The results of a comparative ultrastructural study of spermiogenesis in T/tx, + /tx, +/T, C57BL/6J, BALB/c and randomly breeding Swiss Albino mice are reported. The observations show that aberrant spermiogenesis occurs in males of all strains and genotypes and that the same specific types of abnormal spermatids are found in all of the males examined. No unique morphological defect which could be correlated with the increased transmission frequency of tx-bearing gametes can be found in males heterozygous for the tx allele.

The T-locus in the house mouse is located on chromosome 17 and consists of a wild-type allele ( + ), a dominant mutant allele (T), and a series of recessive alleles (tx). With a few exceptions, males which are heterozygous for these recessive lethal alleles ( + /tx; T/tx) transmit the tx -bearing spermatozoa at a frequency greater than 50 %. Conversely, males heterozygous for the dominant allele (T/ + ) and all heterozygous females (T/tx, + /tx) transmit the alleles in a 1:1 ratio (Dunn & Gluecksohn-Schoenheimer, 1939; Dunn, 1960). As a result of the increased transmission frequency of tx -bearing spermatozoa, litters from + /txinter se matings are composed of more than the expected 25 % homozygous t embryos.

In a light-microscopic study Bryson (1944) found that this increased transmission frequency was not the result of extra post-meiotic mitoses of tx-bearing spermatids and that specific spermatid defects could not account for the increased transmission frequency. Yanagisawa (1965), however, suggested that the increased transmission frequency could result if spermatozoan abnormalities were limited to the + -or T-bearing gametes obtained from heterozygous males (T/tx; +/tx). Because of the hypothesis proposed in this latter study, we have undertaken a comparative ultrastructural analysis of mouse spermiogenesis in ræ-bearing males (tw32, t6, t12) and in different inbred and outbred strains in order to establish: first, if aberrant spermatid development occurs in these diverse males ; second, if these defects are the same or different in the groups of males studied; and third, if any specific spermatid defect can be correlated with the increased transmission frequency of tx -bearing gametes.

The studies were done on T/t6 (breeding stock obtained from Dr M. Lyon), T/t12 (breeding stock obtained from Dr S. Waelsch), T/tw32 (breeding stock obtained from Dr D. Bennett), + /t6, + t12, + /tw32, T6/ +, T12/ +, Tw32/ +, C57BL/6J, BALB/c, and randomly breeding Swiss Albino mice. The inbred BALB/c mating pairs were obtained from Dr G. Wolfe in 1964. These latter animals have been maintained through brother-sister matings. The + /tx and T/ + males were obtained by crossing the specific T/tx males to BALB/c females. The males used in the present studies were tested for their level of fertility according to the protocol of Dunn & Bennett (1969). Using their criteria, all of the males were classified as normal fertile. The averaged transmission frequencies of the tx alleles from the heterozygous males used in this study was 0 · 78 for the t6 allele, and 0 · 75 for both the t12 and tw32 alleles.

Six males of each strain and genotype were sacrificed by cervical dislocation at 6 months of age. This age was chosen in order to eliminate the documented correlations between the animal’s age and the numbers of abnormal spermato-genic cells (Bryson, 1944; Hancock, 1972; Krzanowska, 1972). Testes were removed and placed into 3 % glutaraldehyde in 0 · 1 M-PO4 buffer (pH 7 · 4). The tunica albuginea was removed from each testis and the seminiferous tubules cut into approximately 1 mm segments. These segments were fixed for 2 h in glutaraldehyde, placed into 0 · 1 M-PO4 buffer for 2 h, postfixed in 1 % osmium tetroxide (Millonig’s, pH 7 · 3), dehydrated, and embedded in Epon. Ultrathin sections were stained with either lead citrate (Venable & Coggeshall, 1965), or with both lead citrate and 2 % uranyl acetate (Watson, 1958). The sections were examined with a Philips 300 electron microscope.

A detailed description of normal mouse spermatid development is not included in this report. Several ultrastructural studies of spermiogenesis in various mammalian species are available (Fawcett & Phillips, 1969; Fawcett, Eddy & Phillips, 1970; Fawcett, Anderson & Phillips, 1971). In addition, there are a number of reports describing specific stages of, and the development of specific organelles during, mouse spermiogenesis (Sandoz, 1970; Bennett, Gall, Southard & Sidman, 1971 ; Bryan & Wolosewick, 1973). An overall description of mouse spermatid development, with particular emphasis placed on spermatid head development, has been reported by Dooher & Bennett (1973). The present study includes brief descriptions of the development of only those component structures which exhibit aberrant morphology. The spermatid staging follows that established by Oakberg (1956) as modified by Dooher & Bennett (1973).

General observations

Testes isolated from all strains and genotypes contained abnormal spermatids. The identical types of abnormalities were found in all testes examined. During the course of the study it was noted that aberrant spermatids were frequently clustered in delimited areas of the seminiferous tubules. Consequently, one thin section of a tubule would show few abnormalities whereas sections from a different area of the same tubule or from other tubules would contain numerous aberrant spermatids. This clustering was also reported by Bryson (1944) and Rajasekarasetty (1954) in their light-microscopic studies of mouse spermiogenesis in tx -bearing males.

Because aberrant spermatids were not found in each section of the tubule, and because sections were randomly selected and examined, it was impossible to establish either the total number of abnormal cells or the incidence of a specific type of spermatid abnormality for any individual male. Nevertheless, it was possible to rank the strains and genotypes in order, beginning with those containing the greatest numbers of abnormal cells to those containing the least numbers, utilizing the relative ease by which we could find abnormal spermatids in randomly selected sections. Using this criterion, C57BL/6J and BALB/c males contained the largest number, and the t-bearing (T/tx or + / tx) and Swiss Albino males, the fewest number of aberrant spermatids. Our subjective determination that C57BL/6J contained high numbers of abnormal spermatids agrees with Johnson (1974) who reported that C57BL/6J and A/Gr mouse strains contained more abnormal spermatids (multinucleated spermatids) than did the other inbred (C57BL/Gr; CBA/Gr; AKRfNMRI/Lac) and outbred (p/p mixed; +/p25 mixed; +/ hop mixed) strains which he examined.

Specific abnormalities

Uninuclear spermatid defects

1 Duplicatedproacrosomal vesicles and granules

Normally, during spermatid Stage 2, only one proacrosomal vesicle assumes a juxtanuclear position and becomes closely apposed to the nuclear membrane. During spermatid Stage 3, the convex juxtanuclear membrane of this proacrosomal vesicle becomes situated in a nuclear indentation (Figs. 1 A, B). This delimited area of the nuclear envelope is characterized by both a layer of condensed chromatin subjacent to the inner nuclear membrane and an absence of nuclear pores (Sandoz, 1970).

FIGURE 1–4

A, Acrosome; F, flagellum; IF, implantation fossa; M, mantle; Mc, manchette; N, nucleus; PC, proximal centriole; PG, proacrosomal granule; PR, perinuclear ring; PV, proacrosomal vesicle.

Fig. 1. (A) An electron micrograph of a portion of a normal Stage-3 mouse spermatid. A single proacrosomal vesicle containing a fibrillar proacrosomal granule is situated in a nuclear identation. × 18 900. (B) This insert shows a portion of the proacrosomal vesicle membrane and the subjacent inner and outer membranes of the nuclear envelope. Note the presence of condensed material associated with the inner nuclear membrane and the presence of amorphous material between the outer nuclear membrane and the membrane of the vesicle, × 72000.

FIGURE 1–4

A, Acrosome; F, flagellum; IF, implantation fossa; M, mantle; Mc, manchette; N, nucleus; PC, proximal centriole; PG, proacrosomal granule; PR, perinuclear ring; PV, proacrosomal vesicle.

Fig. 1. (A) An electron micrograph of a portion of a normal Stage-3 mouse spermatid. A single proacrosomal vesicle containing a fibrillar proacrosomal granule is situated in a nuclear identation. × 18 900. (B) This insert shows a portion of the proacrosomal vesicle membrane and the subjacent inner and outer membranes of the nuclear envelope. Note the presence of condensed material associated with the inner nuclear membrane and the presence of amorphous material between the outer nuclear membrane and the membrane of the vesicle, × 72000.

Spermatids in which two proacrosomal vesicles become associated with the presumptive rostral area of the spermatid nucleus are commonly found (Fig. 2). In these nuclei, the region of the nuclear envelope between the two vesicles does not contain condensed chromatin on its inner membrane but usually does contain nuclear pores.

Fig. 2.

A micrograph of a Stage-3 mouse spermatid showing the aberrant association of two proacrosomal vesicles with the nucleus. This defect was found in all of the males examined, × 21000.

Fig. 2.

A micrograph of a Stage-3 mouse spermatid showing the aberrant association of two proacrosomal vesicles with the nucleus. This defect was found in all of the males examined, × 21000.

The consequence(s) of this duplication has not been determined, but the apparent defect has been found in all of the strains and genotypes examined and has also been described in p25H/p25H and p6H/p6H sterile mice (Hunt & Johnson, 1971). It is not known if the two vesicles and their granules ultimately fuse and regulate to form a normal mature acrosome; or if, conversely, there is no regulation and the duplication results in abnormal acrosomal and nuclear development. If regulation does not occur, and if the initial vesicle-nucleus apposition establishes the future rostral portion of the spermatid head, a duplicated acrosomal vesicle may produce two anterior apices. This duplication of apices could produce those defective spermatids which have either bifid or bifurcated heads.

2. Bifid and bifurcated sperm heads

All of the males contained spermatids with bifid or bifurcated heads. This abnormality is also found in hop /hop sterile mice (Johnson & Hunt, 1971). A typical example of this type of spermatid abnormality is shown in Fig. 3. In these spermatids, the head has two rostral parts. Each apex may be enclosed by a separate acrosome or the two may be enclosed by a single acrosome. Both apices, however, are always covered by a single continuous mantle. The extent of bifidity in these spermatids varies and the cleft frequently extends below the level of the perinuclear ring (Fig. 3).

Fig. 3.

An example of a ubiquitous defect, nuclear bifidity, is shown in this micrograph of a Stage-12 spermatid. In this spermatid the bifidity extends below the level of the perinuclear ring. Note the presence of a single mantle, implantation fossa and proximal contriole. × 17800.

Fig. 3.

An example of a ubiquitous defect, nuclear bifidity, is shown in this micrograph of a Stage-12 spermatid. In this spermatid the bifidity extends below the level of the perinuclear ring. Note the presence of a single mantle, implantation fossa and proximal contriole. × 17800.

It is conceivable that most of these aberrant spermatids are derived from those Stage-3 spermatids which contain two proacrosomal vesicles and granules. Alternatively, these bizarre spermatids could result from the partial fusion of the presumptive caudal regions of two nuclei of a binucleated cell. The fact, however, that these aberrant nuclei have only one implantation fossa and one flagellum does not support the latter hypothesis. Binucleated spermatids which develop from a binucleated cell usually have two distinct flagella (cf. Figs. 3 and 20).

Bifurcation is not limited to the rostral area. Spermatids frequently contain nuclei with lateral extensions. These nuclear extensions are always anterior to the perinuclear ring. Although the spermatid membranes are continuous around the outpocketings, the mantle is often discontinuous at these sites (Fig. 4). The causal factor responsible for these lateral nuclear extensions is not known.

Fig. 4.

A longitudinal section of a mature spermatid with a bifurcated nucleus. Although the nuclear and acrosomal membranes are always continuous around these projections, the mantle is often discontinuous (arrow), × 13000.

Fig. 4.

A longitudinal section of a mature spermatid with a bifurcated nucleus. Although the nuclear and acrosomal membranes are always continuous around these projections, the mantle is often discontinuous (arrow), × 13000.

3. Abnormal chromatin condensation

Normally, the condensed chromatin of late-staged spermatids fills the nucleus and is contiguous with the inner membrane of the nuclear envelope. In all of the males, spermatids were found in which the condensed chromatin was completely separated from the inner nuclear membrane or was contiguous with it in some areas and not in others. This retraction of the chromatin occurs most frequently in the postacrosomal region of the nucleus (Fig. 5). This abnormal dehiscence of the nuclear membrane and the condensed chromatin is also found in spermatids from C57BL/6J-qk/qk mice (Bennett et al. 1971).

4. Duplicated implantation fossae and flagella

By spermatid Stage 6 the proximal and distal centrioles, together with the forward basal portion of the single axoneme, have migrated to the caudal portion of the spermatid nucleus. The proximal centriole has become orientated perpendicular to the nucleus, and the circumscribed area of the nuclear membrane above this centriole has indented to form the implantation fossa. The single flagellum projects into the flagellar canal (Fig. 6).

FIGURE 5–9

A, Acrosome; F, flagellum; IF, implantation fossa; M, mantle; PR, perinuclear ring.

Fig. 5. A transverse section of a Stage-14 spermatid in which the condensed chromatin is retracted from the postacrosomal nuclear membrane, × 16000.

FIGURE 5–9

A, Acrosome; F, flagellum; IF, implantation fossa; M, mantle; PR, perinuclear ring.

Fig. 5. A transverse section of a Stage-14 spermatid in which the condensed chromatin is retracted from the postacrosomal nuclear membrane, × 16000.

Fig. 6.

A micrograph of a normal Stage-6 spermatid. Note the presence of a single implantation fossa and flagellum, × 8800.

Fig. 6.

A micrograph of a normal Stage-6 spermatid. Note the presence of a single implantation fossa and flagellum, × 8800.

Spermatid tails which contain two axonemes and their associated flagellar components within a common cytoplasm, bound by a single plasma membrane can be found projecting into the tubule lumen. When such tails are seen in cross-section (Fig. 7) we cannot determine if the two axial filament complexes are from a binucleated spermatid which normally has two flagella (Fig. 20), or if the two complexes are both associated with the nucleus of a uninucleated spermatid. If the axial filament complexes are not traced to their sites of implantation, it is impossible to distinguish between these two alternatives. We have observed longitudinal sections which clearly show that some uninucleated spermatids have two implantation fossae and two axial filament complexes (Fig. 8). These double-tailed spermatids were found in all of the males suggesting that duplicated fossae and flagella are common abnormalities found in all normal fertile mice. This same defect is a phenotypic characteristic of spermiogenesis in sterile hop/hop males (Johnson & Hunt, 1971). However, in these latter males, sperm tail development is either abortive or arrested during the early spermatid stages.

Fig. 7.

A transverse section through a spermatid tail. Two sections of midtails are contained in a common cytoplasm bounded by a plasma membrane. Without tracing the tail it is not possible to determine if these duplicated structures are from a uninucleated or from a binucleated spermatid, × 17400.

Fig. 7.

A transverse section through a spermatid tail. Two sections of midtails are contained in a common cytoplasm bounded by a plasma membrane. Without tracing the tail it is not possible to determine if these duplicated structures are from a uninucleated or from a binucleated spermatid, × 17400.

Fig. 8.

A longitudinal section of a double-tailed Stage-12 uninucleated spermatid. Note the presence of two implantation fossae and two flagellae. × 12000.

Fig. 8.

A longitudinal section of a double-tailed Stage-12 uninucleated spermatid. Note the presence of two implantation fossae and two flagellae. × 12000.

5. Missing doublets and outer dense fibers

Additional aberrations of the spermatid flagella are missing (Fig. 9 A) or extra (Fig. 9B) doublets and/or outer dense fibers and disorganized flagellar components (Fig. 9C). These defects are the least numerous of the specific aberrations found but are present in spermatids of all of the males examined. Flagellar disorders (abnormal distribution and excessive numbers of outer dense fibers and doublets) have also been described in the spermatids of C57BL/6J-qk/qk mice. In this mutant the flagellar components ‘decompose’ after Stage 9 or 10 (Bennett et al. 1971). Excessive numbers of doublets and disorganized axonemal components are also found in the spermatids of hop/hop sterile males (Johnson & Hunt, 1971). Conversely, Olds (1973) found no flagellar abnormalities in her study of spermatids from fertile T/tw18, T/tw32 and sterile tw18/tw32 males. However, the presence of flagellar defects in all of the mutant (including T/tw32) and wild-type males which we studied and the fact that similar defects have been found in the spermatids of other mutant mice suggest that these defects are common spermatid abnormali-ties and argue that some of the axonemal and outer dense fiber defects of the epididymal spermatozoa described by Olds originated during spermiogenesis rather than being effected solely by the epididymal environment.

Fig. 9.

(A) A transverse section through the midpiece of a spermatid tail which lacks doublets 4, 5, 6 and 7. The outer dense fibers appear normal, × 32000. (B) A transverse section through a membrane-limited cytoplasm containing two flagellar endpieces. It is not known if these two tails are from a uninucleated or a binucleated spermatid. Note the presence of an extra doublet in the one endpiece (arrow) and the presence of nine pairs of doublets in the surrounding cytoplasm. These doublets may have come from the second endpiece which appears to have a disrupted plasma membrane and to be devoid of all doublets except for the middle pair, × 36000. (C) This micrograph shows four sections of tails with disorganized and missing axial filament components, × 21600.

Fig. 9.

(A) A transverse section through the midpiece of a spermatid tail which lacks doublets 4, 5, 6 and 7. The outer dense fibers appear normal, × 32000. (B) A transverse section through a membrane-limited cytoplasm containing two flagellar endpieces. It is not known if these two tails are from a uninucleated or a binucleated spermatid. Note the presence of an extra doublet in the one endpiece (arrow) and the presence of nine pairs of doublets in the surrounding cytoplasm. These doublets may have come from the second endpiece which appears to have a disrupted plasma membrane and to be devoid of all doublets except for the middle pair, × 36000. (C) This micrograph shows four sections of tails with disorganized and missing axial filament components, × 21600.

6. Projections of Sertoli cells and manchette into spermatid nuclei

A very common spermatid abnormality in these mice is the projection of Sertoli-cell cytoplasmic extensions which indent the spermatid nucleus. Most frequently the indentations are anterior to the perinuclear ring. In both transverse and longitudinal sections of these nuclei the indented Sertoli cell cytoplasm is circumscribed by the spermatid plasma, acrosomal and nuclear membranes (Figs. 10 A, B). In some cases, portions of the mantle are also included in these projections. These spermatids are similar to those described by Bennett et al. (1971) as being characteristic of sterile C57BL,/6J-qk/qk mice.

FIGURE 10–13

A, Acrosome; M, mantle; m, microtubules; Mc, manchette; N, nucleus; NE, nuclear envelope; PG, proacrosomal granule; PR, perinuclear ring; S, Sertoli cell cytoplasm. Fig. 10. (A) A section through a Stage-14 spermatid. Note the presence of Sertoli cell cytoplasm (arrow) which is projected into the nucleus. This projection has occurred in the rostral area and the cytoplasm is circumscribed by the nuclear, acrosomal and plasma membranes, × 17500. (B) This insert contains a higher magnification of the delimited area of the spermatid head shown in (A), × 36000.

FIGURE 10–13

A, Acrosome; M, mantle; m, microtubules; Mc, manchette; N, nucleus; NE, nuclear envelope; PG, proacrosomal granule; PR, perinuclear ring; S, Sertoli cell cytoplasm. Fig. 10. (A) A section through a Stage-14 spermatid. Note the presence of Sertoli cell cytoplasm (arrow) which is projected into the nucleus. This projection has occurred in the rostral area and the cytoplasm is circumscribed by the nuclear, acrosomal and plasma membranes, × 17500. (B) This insert contains a higher magnification of the delimited area of the spermatid head shown in (A), × 36000.

We have also found spermatid abnormalities in which the microtubules of the manchette are projected into the nucleus. Because of the level of these indentations, the microtubule projections are circumscribed by only the nuclear envelope (Fig. 11). We have not determined if Sertoli cell extensions indent the manchette, which in turn projects into the nucleus, or if the manchette projects independently into the nucleus. This defect has been found in all of the males and has also been described in the spermatids of hop /hop sterile mice (Johnson & Hunt, 1971).

Fig. 11.

A longitudinal section of the head of a Stage-9 spermatid which is normal except for the projection of a portion of the manchette into the nucleus. The indented microtubles are circumscribed by only the nuclear envelope, × 17500.

Fig. 11.

A longitudinal section of the head of a Stage-9 spermatid which is normal except for the projection of a portion of the manchette into the nucleus. The indented microtubles are circumscribed by only the nuclear envelope, × 17500.

7. Nonsequential development of spermatid component parts

Non-sequential spermatid development occurs when one or more component parts of the spermatid fail to reach the stage of development they normally attain prior to the formation of additional component parts. The most frequently observed non-sequential development is the delayed or aberrant development of the acrosome and of the nucleus relative to the temporal appearance of the perinuclear ring and the manchette. The manchette is normally present in stage-8 spermatids and persists through Stage 15 when it disappears (Dooher & Bennett, 1973). In normal Stage-8 spermatids, the nucleus is elongated, patches of densely staining material are found both adjacent to the inner nuclear membrane and scattered inside the nucleus, the nucleoli have disappeared, the acrosomal cap is completely formed and filled with dense material, and the mantle is closely apposed to the acrosome (Fig. 12). Frequently, one finds cells in which the nuclei have the conformation of younger staged spermatids (e.g. Stage 4) but have lost their nucleoli and have a chromatin condensation pattern similar to that of Stage-8 spermatids (Fig. 13). The nuclei of these aberrant spermatids always have defective acrosomal caps (e.g. persistent proacrosomal granules), but they are always encircled by a normal appearing perinuclear ring and manchette. This common abnormality suggests, first, that defective acrosomal formation is associated with, and may be a principle cause for, the abnormal shaping of the spermatid nucleus; second, that the temporal loss of nucleoli and the subsequent pattern of chromatin condensation is not dependent upon the conformation of either the acrosome or the nucleus; and third, that the formation of a normal appearing manchette is not dependent upon the normal conformation of either the acrosome or the nucleus. These findings, in addition to demonstrating that spermatid component parts can develop normally even if the preceding development of other component parts is aberrant, support the hypothesis advanced by Fawcett et al. (1971) that the manchette does not play an active role in the shaping of the postacrosomal region of the mammalian spermatid head.

Fig. 12.

A longitudinal section through a normal, elongated Stage-8 spermatid nucleus. Note the presence of condensed chromatin associated with the inner nuclear membrane and scattered through the nucleoplasm. The perinuclear ring and manchette are present at this stage, × 8000.

Fig. 12.

A longitudinal section through a normal, elongated Stage-8 spermatid nucleus. Note the presence of condensed chromatin associated with the inner nuclear membrane and scattered through the nucleoplasm. The perinuclear ring and manchette are present at this stage, × 8000.

Fig. 13.

A longitudinal section through an aberrant spermatid. The nucleus is shaped like a Stage-4 nucleus but the chromatin condensation pattern is like that of a Stage-8 nucleus (compare with Fig. 12). The acrosome is aberrant and still contains a fibrous proacrosomal granule. Note, however, that the perinuclear ring and the manchette have formed and appear normal, × 8000.

Fig. 13.

A longitudinal section through an aberrant spermatid. The nucleus is shaped like a Stage-4 nucleus but the chromatin condensation pattern is like that of a Stage-8 nucleus (compare with Fig. 12). The acrosome is aberrant and still contains a fibrous proacrosomal granule. Note, however, that the perinuclear ring and the manchette have formed and appear normal, × 8000.

8. Manchette abnormalities

We have found cells which appear to contain excessive numbers of microtubules in all of the testes which we examined. Two classes of these cells have been observed. In the first, the spermatid heads are abnormally shaped and are smaller than normal. In longitudinal sections of these microheaded spermatids the perinuclear ring appears to be greatly extended on either one (Fig. 14) or both sides of the nucleus. As a consequence of this extension, the spermatids appear to contain excessive numbers of microtubules. It is probable, however, that the perinuclear rings and manchettes of spermatids with both normal and abnormal nuclei are of equal size and that the manchettes associated with these nuclei contain the same numbers of microtubules. In cells in which the nuclei are smaller, the normal sized perinuclear rings and manchettes would be disproportionately large; and consequently, the spermatids would appear to contain excessive numbers of microtubules.

FIGURE 14–17

1 C, Condensed chromatin; G, Golgi apparatus; m, microtubules; Mc, manchette; NE, nuclear envelope; PG, proacrosomal granule; PR, perinuclear ring; PV, pro-acrosomal vesicle.

Fig. 14. A longitudinal section through a microheaded Stage-11 spermatid. The perinuclear ring and the associated microtubules of the mantle appear to be extensive on one side of the nucleus (bracket). This type of defect was found in all males, × 16100.

FIGURE 14–17

1 C, Condensed chromatin; G, Golgi apparatus; m, microtubules; Mc, manchette; NE, nuclear envelope; PG, proacrosomal granule; PR, perinuclear ring; PV, pro-acrosomal vesicle.

Fig. 14. A longitudinal section through a microheaded Stage-11 spermatid. The perinuclear ring and the associated microtubules of the mantle appear to be extensive on one side of the nucleus (bracket). This type of defect was found in all males, × 16100.

In the second class the nuclei are completely deformed and their nuclear membranes are disrupted (Fig. 15). Normally, by Stage 14 the chromatin is condensed and is contiguous with the entire inner nuclear membrane. In the defective spermatids at this stage the condensed chromatin is located only in the central area of the nucleus. The microtubules of the manchette protrude into the nucleus at those points which are devoid of nuclear envelope, and they appear to be present in significantly greater numbers than in correspondingly staged, normal spermatids.

Fig. 15.

A longitudinal section through an extremely abnormal spermatid. Note the absence of most of the rostral portion of the head, the disruption and the aberrant reflexion of the nuclear envelope (arrow) and the retraction of the condensed chromatin from the intact portion of the envelope. This type of aberrant spermatid appears to have excessive numbers of microtubules, × 17800.

Fig. 15.

A longitudinal section through an extremely abnormal spermatid. Note the absence of most of the rostral portion of the head, the disruption and the aberrant reflexion of the nuclear envelope (arrow) and the retraction of the condensed chromatin from the intact portion of the envelope. This type of aberrant spermatid appears to have excessive numbers of microtubules, × 17800.

A type of spermatid abnormality similar to the latter class has been described by Dooher & Bennett (1974) in sterile tw2/tw2 males. These authors suggest that the sterility of tw2/tw2 males is related to the formation of spermatids with ‘an unusually large number of disorganized microtubules’ which appear to depolymerize prematurely. The tw2 spermatids seldom survive to maturity, and most are phagocytized by Sertoli cells.

Rattner & Brinkley (1972) have reported that the numbers of microtubules in spermatids are species specific. In order to determine if there is an actual increase in the number of microtubules in those spermatids which appear to have excessive numbers, it will be necessary, therefore, to count the microtubules in both the normal and aberrant cells. Only then will it be possible to state that the aberrant spermatids described in either the present report or in those found in tw2 homozygous males do in fact contain excessive numbers of microtubules.

Binucleated and multinucleated spermatid defects

Testes from males of all genotypes and strains contain both binucleated and multinucleated spermatids. In binucleated cells the two nuclei frequently share a single Golgi apparatus, and subsequently, a single proacrosomal vesicle and granule which forms a single acrosome. Two developmental stages of such binucleated spermatids (Stages 4 and 8) are shown in Figs. 16 and 17. Differing spatial relationships between the forming acrosome and the two nuclei result in a wide distribution of nuclear, and subsequent head, morphologies. These range from spermatids having two heads which are conformationally normal except for sharing a common acrosome to spermatids having two conjoined heads which are highly bizarre (Figs. 18, 19). The aberrant relationship between the acrosome and the two nuclei is often accompanied by other ubiquitous defects. For examples, the two flagella are frequently associated with only one of the two nuclei (Fig. 20); the condensed chromatin is not contiguous with the nuclear membranes (Fig. 21); and the manchette (Fig. 21) or the Sertoli cell cytoplasm projects into either or both of the nuclei and/or into the common acrosome. Similar types of bi nucleated spermatids have been reported in pinkeyed sterile mice (Hunt & Johnson, 1971), in hop /hop sterile mice (Johnson & Hunt, 1971), and in ABP(JAX) mice (Bryan & Wolosewick, 1973).

Fig. 16.

A longitudinal section of a binucleated Stage-4 spermatid. Note that the two nuclei share a common Golgi apparatus and a common proacrosomal vesicle which is spaced equidistant from both nuclei, × 7500.

Fig. 16.

A longitudinal section of a binucleated Stage-4 spermatid. Note that the two nuclei share a common Golgi apparatus and a common proacrosomal vesicle which is spaced equidistant from both nuclei, × 7500.

Fig. 17.

A transverse section of a Stage-8 spermatid which developed from a binucleated cell like that shown in Fig. 16. × 8000.

Fig. 17.

A transverse section of a Stage-8 spermatid which developed from a binucleated cell like that shown in Fig. 16. × 8000.

FIGURE 18–21

A, Acrosome; C, condensed chromatin; IF, implantation fossa; m, microtubules; M, mantle; NE, nuclear envelope.

Fig. 18. A binucleated Stage-9 spermatid. The two nuclei, although joined by a single acrosome, are normally shaped. A single mantle, perinuclear ring and manchette are present. In serial sections of this spermatid it was found that a single flagellum was associated with each nucleus, × 10000.

FIGURE 18–21

A, Acrosome; C, condensed chromatin; IF, implantation fossa; m, microtubules; M, mantle; NE, nuclear envelope.

Fig. 18. A binucleated Stage-9 spermatid. The two nuclei, although joined by a single acrosome, are normally shaped. A single mantle, perinuclear ring and manchette are present. In serial sections of this spermatid it was found that a single flagellum was associated with each nucleus, × 10000.

Fig. 19.

A bizarre Stage-12 spermatid which developed from two nuclei which shared a single proacrosomal vesicle and granule. These aberrant spermatids result when the proacrosomal vesicle is not shared equally by the two nuclei. The pattern of chromatin condensation in each nucleus is normal for Stage-12 spermatids. Note the large vacuoles in the shared acrosome. × 12600.

Fig. 19.

A bizarre Stage-12 spermatid which developed from two nuclei which shared a single proacrosomal vesicle and granule. These aberrant spermatids result when the proacrosomal vesicle is not shared equally by the two nuclei. The pattern of chromatin condensation in each nucleus is normal for Stage-12 spermatids. Note the large vacuoles in the shared acrosome. × 12600.

Fig. 20.

An aberrant Stage-12 binucleated spermatid. In addition to the bizarre shape, one nucleus is lacking a flagellum, while the other nucleus contains two implantation fossae and flagella, × 11900.

Fig. 20.

An aberrant Stage-12 binucleated spermatid. In addition to the bizarre shape, one nucleus is lacking a flagellum, while the other nucleus contains two implantation fossae and flagella, × 11900.

Fig. 21.

This micrograph of a binucleated Stage-14 spermatid shows two additional defects; the retraction of the condensed chromatin from the nuclear membrane, and the projection of the manchette into the nucleus (arrow), × 18200.

Fig. 21.

This micrograph of a binucleated Stage-14 spermatid shows two additional defects; the retraction of the condensed chromatin from the nuclear membrane, and the projection of the manchette into the nucleus (arrow), × 18200.

Not all binucleated cells, however, give rise to aberrant spermatids. In many, the two nuclei develop synchronously but independently of each other. According to Bryan & Wolosewick (1973) this type of bi nucleated cell gives rise to two normal spermatozoa.

Multinucleated spermatids have also been observed in the present study, but they never contain more than four nuclei. Three types of tetranucleated cells are found. In the first, and at the highest frequency, the four nuclei develop synchronously but independently of each other (Fig. 22). In these cells the nuclei are randomly orientated. No structures (e.g. Golgi apparatus, proacrosomal vesicle and granule, centrioles) are shared between or among any of the nuclei. According to Bryan & Wolosewick (1973), this type of multinucleated cell will produce four normal spermatozoa. In the second type, two of the nuclei share a common acrosome while the other two nuclei remain independent (Fig. 23). In these cells the two nuclei joined by a common acrosome subsequently develop aberrantly, showing the same types of abnormalities found in conjoined nuclei of binucleated cells. Finally, two pairs of conjoined nuclei have been observed, each pair sharing a single acrosome. Each joined nuclear pair develops abnormally, showing the same range of spermatid abnormalities found in binucleated spermatids.

FIGURE 22 AND 23

Fig. 22. A section showing three nuclei of a tetranucleated cell. Note that the nuclei are developing synchronously (all are Stage 7) and that the nuclei are randomly orientated, × 7700.

FIGURE 22 AND 23

Fig. 22. A section showing three nuclei of a tetranucleated cell. Note that the nuclei are developing synchronously (all are Stage 7) and that the nuclei are randomly orientated, × 7700.

Fig. 23.

A micrograph of a typical tetranucleated cell in which two nuclei are sharing a common proacrosomal vesicle (arrow), while the other two nuclei are developing independently. Their independent development was established by examining serial sections, × 7600.

Fig. 23.

A micrograph of a typical tetranucleated cell in which two nuclei are sharing a common proacrosomal vesicle (arrow), while the other two nuclei are developing independently. Their independent development was established by examining serial sections, × 7600.

In his light-microscopic studies of developing spermatogenic cells of mice, Bryan (1971) noted that squashed preparations of viable seminiferous tubules contained large numbers of multinucleated cells and that there were as many cells with odd numbers of nuclei (Fig. 11 of his report shows a ‘spermatid containing 13 nuclei’) as with even numbers of nuclei (‘2 to more than 30 nuclei’). The strain(s) of animals used in his study was not reported. Later, using electron microscopic preparations, Bryan & Wolosewick (1973) described the ultrastructure of binucleated and tetranucleated spermatogenic cells in ABP(JAX) mice. In this latter study cells with more then four nuclei were not described. On the basis of their combined observations Bryan & Wolosewick suggested that multinucleated spermatid development was a common phenomenon during spermiogenesis. Hunt & Johnson (1971) described the appearance of multinucleated spermatids in p6H and p25H males using both light and electron microscopy. In the light-microscopic studies they found that the multinucleated spermatids contained as many as 16 nuclei. Later, Johnson (1974), using both light and electron microscopy, examined cells from five inbred and from three outbred mouse strains. He found that all of the males contained multinucleated cells and that the incidence of multinucleated spermatids was strain related. Johnson’s report, therefore, confirmed the hypothesis that binucleated and multinucleated spermatid development was a common phenomenon in mouse spermiogenesis. His observations, however, differed from Bryan’s observations (1971) concerning the incidence of the occurrence of this abnormality. Johnson found that the frequency of this abnormality was significantly lower than that reported by Bryan. Our present observations support Johnson’s findings. While all of the males contained multinucleated spermatids, the frequency of these multinucleated cells was low. Furthermore, we found multinucleated cells more frequently in randomly selected sections of C57BL/6J and BALB/c testes. This also supports Johnson’s observation that the incidence of this abnormality is strain related.

The fact that we found no cells with more than four nuclei while both Bryan (1971) and Hunt & Johnson (1971) reported the presence of cells containing larger numbers of nuclei is not easily resolved. However, since cells with more than four nuclei were found almost exclusively in their light microscopic and not in their ultrastructural studies, it is possible that their preparative procedures for the former studies caused cellular fusion, thus producing cells with artifactual numbers of nuclei. However, in spite of these differences in the maximum numbers of nuclei found in multinucleated spermatozoa, the present observations do support the hypothesis proposed by both Bryan and Johnson: that multinucleate spermatid development is a phenomenon common to all genotypes and strains during mouse spermiogenesis.

  1. All of the males examined in the present study produced abnormal spermatids. The same types of aberrations were found in all of the strains and genotypes studied.

  2. The highest incidences of abnormal spermatids were found in C57BL/6J and BALB/c mice and the lowest, in the randomly breeding Swiss Albino and T/ tx, + /txand T/ + mice.

  3. The defects found in the spermatids of these normal fertile males have also been found by other investigators in other mutant mice and in other inbred and outbred strains of mice.

  4. The ubiquitousness of aberrant spermatid development and the strain-related incidence of aberrant spermiogenesis are obvious from this and other studies. The effects, therefore, of any specific mutant gene on spermiogenesis can be determined only after the distribution and range of aberrant spermiogenesis is established for the inbred or outbred strain(s) in the presence and absence of the mutant allele(s).

  5. The present study shows that the spermatids of + /tx and T/tx have no unique ultrastructural defects which could either result in, or contribute to, the increased transmission frequency of the tx -bearing allele.

This research was supported by United States Public Health Service Grants Nos. HD 00827 and HD 09753. The authors would like to thank Dr Ralph Hillman for his help in the preparation of this manuscript and Marie Morris and Geraldine Wileman for their technical assistance.

Bennett
,
W. I.
,
Gall
,
A. M.
,
Southard
,
J. L.
&
Sidman
,
R. L.
(
1971
).
Abnormal spermiogenesis in Quaking, a myelin-deficient mutant mouse
.
Biol. Reprod
.
5
,
30
58
.
Bryan
,
J. H. D.
(
1971
).
Spermatogenesis revisited. I. On the presence of multinucleate spermatogenic cells in the seminiferous epithelium of the mouse
.
Z. Zellforsch. mikrosk. Anat
.
112
,
333
349
.
Bryan
,
J. H. D.
&
Wolosewick
,
J. J.
(
1973
).
Spermatogenesis revisited. IL Ultrastructural studies of spermiogenesis in multinucleate spermatids of the mouse
.
Z. Zellforsch. mikrosk. Anat
.
138
,
155
169
.
Bryson
,
V.
(
1944
).
Spermatogenesis and fertility in Mus musculus as affected by factors at the TTocus
.
J. Morph
.
74
,
131
186
.
Dooher
,
G. B.
&
Bennett
,
D.
(
1973
).
Fine structural observations on the development of the sperm head in the mouse
.
Am. J. Anat
.
136
,
339
362
.
Dooher
,
G. B.
&
Bennett
,
D.
(
1974
).
Abnormal microtubular systems in mouse spermatids associated with a mutant gene at the T-locus
.
J. Embryol. exp. Morph
.
32
,
749
761
.
Dunn
,
L. C.
(
1960
).
Variations in the transmission ratios of alleles through egg and sperm in Mus musculus
.
Am. Nat
.
94
,
385
393
.
Dunn
,
L. C.
&
Bennett
,
D.
(
1969
).
Studies of effects of β-alleles in the house mouse on spermatozoa. II. Quasi-sterility caused by different combinations of alleles
.
J. Reprod. Fert
.
20
,
239
246
.
Dunn
,
L. C.
&
Gluecksohn-Schoenheimer
,
S.
(
1939
).
The influence of taillessness (anury) in the house mouse. II. Taillessness in a second balanced lethal line
.
Genetics
24
,
587
609
.
Fawcett
,
D. W.
,
Anderson
,
W. A.
&
Phillips
,
D. M.
(
1971
).
Morphogenetic factors influencing the shape of the sperm head
.
Devi Biol
.
26
,
220
251
.
Fawcett
,
D. W.
,
Eddy
,
E. N.
&
Phillips
,
D. M.
(
1970
).
Observations on the fine structure and relationships of the chromatoid body in mammalian spermatogenesis
.
Biol. Reprod
.
2
,
129
153
.
Fawcett
,
D. W.
&
Phillips
,
D. M.
(
1969
).
The fine structure and development of the neck region of the mammalian spermatozoon
.
Anat. Rec
.
165
,
153
184
.
Hancock
,
J. L.
(
1972
).
Spermatogenesis and sperm defects
.
In Edinburgh Symposium on the Genetics of the Spermatozoon
(ed.
R. A.
Beatty
&
S.
Gluecksohn-Waelsch
), pp.
121
130
.
Copenhagen
:
Bogtrykkeriet Forum
.
Hunt
,
D. M.
&
Johnson
,
D. R.
(
1971
).
Abnormal spermiogenesis in two pink-eyed sterile mutants in the mouse
.
J. Embryol. exp. Morph
.
26
,
111
121
.
Johnson
,
D. R.
(
1974
).
Multinucleate spermatids in the mouse. A quantitative electron microscope study
.
Cell Tiss. Res
.
150
,
323
329
.
Johnson
,
D. R.
&
Hunt
,
D. M.
(
1971
).
Hop-sterile, a mutant gene affecting sperm tail development in the mouse
.
J. Embryol. exp. Morph
.
25
,
223
236
.
Krzanowska
,
H.
(
1972
).
Influence of Y chromosome on fertility in mice
.
In Edinburgh Symposium on the Genetics of the Spermatozoon
(ed.
R. A.
Beatty
&
S.
Gluecksohn-Waelsch
), pp.
370
386
.
Copenhagen
:
Bogtrykkeriet Forum
.
Oakberg
,
E. F.
(
1956
).
A description of spermiogenesis in the mouse and its use in analysis of the cycle of the seminiferous epithelium and germ cell renewal
.
Am. J. Anat
.
99
,
391
413
.
Olds
,
P.
(
1973
).
Sperm ultrastructure as an indication of sterility. Schering Workshop on Contraception: The Masculine Gender
(ed.
G.
Raspé
).
Advances in Biosciences
10
,
261
270
.
Rajasekarasetty
,
M. R.
(
1954
).
Studies on a new type of genetically determined quasisterility in the house mouse
.
Fert. Steril
.
5
,
68
97
.
Rattner
,
J. B.
&
Brinkley
,
B. R.
(
1972
).
Ultrastructure of mammalian spermiogenesis. III. The organization and morphogenesis of the manchette during rodent spermiogenesis
.
J. Ultrastruct. Res
.
41
,
209
218
.
Sandoz
,
D.
(
1970
).
Evolution des ultrastructures au cours de la formation de l’acrosome du spermatozoide chez la souris
.
J. Microscopie
9
,
535
558
.
Venable
,
J. H.
&
Coggeshall
,
R.
(
1965
).
A simplified lead citrate stain for use in electron microscopy
.
J. Cell Biol
.
25
,
407
408
.
Yanagisawa
,
K.
(
1965
).
Studies on the mechanism of abnormal transmission ratios at the T-locus in the house mouse. IV. Some morphological studies on the mature sperm in males heterozygous for t-alleles
.
Jap. J. Genet
.
40
,
97
104
.
Watson
,
M. L.
(
1958
).
Staining of tissue sections for electron microscopy with heavy metals
.
J. biophys. biochem. Cytol
.
4
,
475
478
.