Molecular Human Reproduction Vol.7, No.6 pp.

513–520, 2001

Teratozoospermia in mice lacking the transition protein 2

(Tnp2)

Ibrahim M.Adham1, Karim Nayernia1, Elke Burkhardt-Göttges1, Özlem Topaloglu1,

Christa Dixkens1, Adolf F.Holstein2 and Wolfgang Engel1,3
1Instituteof Human Genetics, University of Göttingen, D-37073 Göttingen and 2Department of Anatomy, Eppendorf University
Hospital, D-20251 Hamburg, Germany
3Towhom correspondence should be addressed at: Institute of Human Genetics, University of Göttingen D-37073 Göttingen,
Germany. E-mail: wengel@gwdg.de

It is believed that the transition proteins (Tnp1 and Tnp2) participate in the removal of the nucleohistones and in
the initial condensation of the spermatid nucleus. Later in spermatogenesis, Tnp1 and Tnp2 are replaced by the
protamines 1 and 2. In an effort to elucidate the physiological role of Tnp2, we have disrupted its locus by
homologous recombination. Breeding of the Tnp2–/– males on different genetic backgrounds revealed normal fertility
on the mixed background C57BL/6J⍥129/Sv, but total infertility on the inbred 129/Sv background. Light and
electron microscopy showed that the germ cells were capable of undergoing chromatin condensation, although
many spermatozoa exhibited head abnormalities with acrosomes not attached to the nuclear envelope. Furthermore,
migration of Tnp2–/– spermatozoa from the uterus into the oviduct was reduced. These results suggest that male
infertility of the Tnp2–/– mice is a result of sperm head abnormalities and reduced sperm motility. The increased
level of the Tnp1 transcript in testes of the Tnp2-deficient mice raises the possibility that a deficiency created
through the disruption of the Tnp2 gene can be compensated for by recruitment of the Tnp1.

Key words: acrosome/chromatin condensation/genetic background/teratozoospermia/Tnp2

Introduction maximum concentration. Ultrastructural studies have shown

After the meiotic division, the germ cells enter spermio-
genesis, the haploid phase of spermatogenesis, where round
spermatids differentiate into elongated spermatids and ulti-
mately spermatozoa. One of the morphological changes that
accompany spermatid differentiation is the nuclear organization
of the male germ cell (Fawcett et al., 1971; Dooher and
Bennett, 1973). During this process, various modifications
occur in the nature of proteins associated with the DNA and
the result is the progressive condensation of the chromatin.
This morphological transformation induces the gradual dis-
placement of testis-specific and remaining somatic histones by
transition proteins 1 and 2 (Tnp1 and Tnp2) which are thought
to participate in the initial condensation of the spermatid
nucleus. Shortly thereafter, the transition proteins are replaced
by the protamines Prm1 and Prm2, which are characteristic
for the mature sperm nucleus (Balhorn et al., 1984). Immuno-
staining of rat testis with Tnp1, Tnp2 and Prm1 antisera has
shown that the appearance of the Tnp2 in the spermatid nucleus
precedes that of Tnp1 and Prm1. Tnp2 is found diffusely
distributed over the anterior tip of the nuclei in step 10
spermatids and remains localized over the more anterior portion
of the nucleus even in step 13 spermatids where it is at its
© European Society of Human Reproduction and Embryology
that chromatin condensation occurs between steps 12 and 14,
starting at the anterior portion of the nucleus and then spreading
gradually towards the posterior region (Dooher and Bennett,
1973). Thus, the first reactivity of Tnp2 appears at that time
when the chromatin still has a fibrillar and lightly stained
structure at steps 10–11 (Kistler et al., 1996).
The nucleoprotein genes Tnp2, Prm1 and Prm2 are closely
linked in a stretch of DNA, 13-15 kb long, on human
chromosome 16p13.3 and on mouse chromosome 16 (Schlüter
et al., 1992; Nelson and Krawetz, 1994). In this cluster, a new
member of the protamine family (Prm3) has been identified
and characterized (Schlüter and Engel, 1995; Schlüter et al.,
1996). Tnp1 is the only gene encoding germ cell-specific
nucleoproteins which is localized on a separate chromosome.
The Tnp2 protein, a 117 amino acid long molecule, contains
a basic domain and two proposed zinc finger motifs at the
amino and carboxyl regions, respectively (Baskaran and Rao,
1991). These two domains may be responsible for the inter-
action of the Tnp2 with the DNA. The considerable sequence
variation in the primary structure of Tnp2 between species
leads one to believe that Tnp2 is involved in the establishment
of species-specific sperm nucleus morphology (Fawcett et al.,
513
I.M.Adham et al.
1971; Kleene and Flynn, 1987; Luerssen et al., 1989; Reinhardt
et al, 1991; Keime et al., 1992; Alfons and Kistler, 1993).
However, the first appearance of Tnp2 in nuclei of elongated
spermatids which have essentially completed the morphological
changes of the nuclear shaping and which are undergoing
chromosomal condensation rules out the role of Tnp2 in
determination of the nuclear morphology (Fawcett et al., 1971;
Dooher and Bennett, 1973; Alfons and Kistler, 1993; Oko
et al., 1996).
To investigate the role of Tnp2 in the differentiation and
function of the male germ cell, we have generated mice
containing a targeted disruption of the Tnp2 gene. Male
infertility was associated with the homozygous mutation on
an inbred (129/Sv) genetic background, but fertility was not
affected in Tnp2-deficient mice on a mixed (C57 BL/6J⫻129/
Sv) genetic background. To determine the underlying cause
for male infertility, we have examined several parameters of
sperm function. The cumulative results presented here showed
that the deficiency of Tnp2 leads to sperm head abnormalities
which are most probably due to malformations in the
attachment of the acrosome to the nuclear involvement. The
acrosomal defects appeared to influence the acrosome reaction
and the ability of the spermatozoa to penetrate the zona
pellucida of the oocyte. In addition, the migration of the
spermatozoa through the female genital tract was found to be
impaired.
Materials and methods
Generation of the Tnp2-mutant mice
A P2 clone carrying the mouse Tnp2 gene was isolated from
the C129/ES cell library (Genome Systems, Cambridge, UK) by
polymerase chain reaction (PCR) screening (Schlüter et al., 1996). A
6.3 kb EcoRI fragment and a 3.6 kb EcoRI/XbaI fragment, together
containing the closely linked Prm2, Prm3 and Tnp2 genes, were
subcloned into pBluescript vector (Stratagene, La Jolla, USA) and
mapped with restriction enzymes (Figure 1A). A targeting vector was
designed for insertion of a neomycin-resistance gene driven by a
PGK promoter (pgk-neo) into the SstII site of exon1. A herpes
simplex virus thymidine kinase gene (tk) cassette was attached to the
3⬘ end for negative selection (Figure 1A). Linearized plasmid DNA
(30 µg) was electroporated into R ES cells (Joyner, 1993). Colonies
resistant to G418 (400 µg/ml) and gancyclovir (GANC) (2 µmol/l)
were selected.
Genomic DNA was extracted from ES cells, digested with EcoRI,
electrophoresed and blotted onto Hybond N membranes (Amersham,
Braunshweig, Germany). The blots were hybridized with a 32P-
labelled 1.8 kb XbaI/EcoRI fragment (Figure 1B) at 65°C overnight
and washed twice at 65°C to final stringency at 0.2⫻standard saline
citrate/0.1% sodium dodecyl sulphate (SDS). To confirm a correct
homologous recombination event of the targeted Tnp2 gene and the
absence of additional random integration of the targeting construct,
a neomycin fragment was used to rehybridize the Southern blots.
Among the G418 and GANC resistance colonies, four independent
ES clones were selected and then injected into C57BL/6J blastocysts
to produce chimeric animals (Joyner, 1993). One line yielded germ-
line transmitting chimeras. The chimeric male was mated to C57BL/
6J and 129/Sv females, respectively, and F1 offspring were genotyped
by PCR analyses. Heterozygous animals were crossed to obtain
homozygous mice, which were genotyped by PCR. PCR was per-
514
formed according to standard protocols to discriminate wild-type and
mutant alleles in the DNA from mouse tails. Primer sequences were
as follows: 1 (Tpn2 sense), 5⬘- AACCAGTGCAATCAGTGCACC;
2 (Tpn2 antisense), 5⬘- ATGGACACAGGAACATCCTGG; 3 (Pgk
antisense), 5⬘- TCTGAGCCCAGAAAGCGAAGG.
Thermal cycling was carried out for 35 cycles, denaturation at
94°C for 30 s, annealing at 58°C for 30 min, and extension at 72°C
for 1 min. One-fifth of each reaction mixture was electrophoresed on
2% agarose gels and stained with ethidium bromide. Primer pair 1/2
amplified a 353 bp fragment in the heterozygous and wild-type
samples, whereas the primer pair 1/3 amplified a 616 bp fragment
with the DNA of both heterozygous and homozygous animals.
RNA blot hybridization
Total RNA was extracted from tissues using the RNA Now Kit (ITC
Biotechnologies, Heidelberg, Germany) according to the manufac-
turer’s recommendation. The RNA was size fractionated by electro-
phoresis on a 1% agarose gel containing formaldehyde, transferred
to a nylon membrane, and hybridized with a 32P-labelled cDNA
fragment, under the same conditions as those used for Southern blot
hybridization.
Extraction of basic nuclear proteins and Western blot
Basic nuclear extracts were prepared from mouse testis as described
(Alfonso and Kistler, 1993). Aliquots (10 µg of protein) of nuclear
extract fractions were subjected to 20% polyacrylamide gels con-
taining 0.9 mol/l acidic acid and 6 mol/l urea (Panyim and Chalkley,
1969) and the gels were blotted onto nitrocellulose filters. Membranes
were then incubated with rabbit Tnp2 antiserum or rabbit H.1.1
antiserum as described (Alfonso and Kistler, 1993; Franke et al., 1998).
Electron microscopy
Testes and epididymides were fixed with 5% glutaraldehyde in
0.2 mol/l phosphate buffer, postfixed with 2% osmium tetroxide, and
embedded in epoxy (Epon) resin. Sections at 70 nm were stained
with 1% Toluidine Blue/pyronine.
Analysis of fertility
To assay the fertility of Tnp2–/– males on a mixed (C57BL/6J⫻126/
Sv) and on an inbred (129/Sv) genetic background, sets of 10 Tnp2–/–
and Tnp2⫹/⫹ males of each genetic background from the F2 littermates
were mated, each with two CD1 females for 3 months. Females were
checked for the presence of vaginal plug and/or pregnancy. Pregnant
females were removed to holding cages to allow them to give birth.
We counted the number of litters sired from each group of males in
the 3-month mating period and the size of the litters was determined.
Furthermore, 8 week old CD1 females were superovulated by i.p.
injections of 5 IU pregnant mare serum gonadotrophin (PMSG)
(Intergonan 5 IU; Intervet, Tönisvorst, Germany) followed by 5 IU
human chorionic gonadotrophin (HCG) (Predalon; 5 IU, Organon,
Oberschleißheim, Germany) 46–48 h later, and mated with Tnp2⫹/⫹
or Tnp2–/– males of 129/Sv genetic background. Oocytes from females
with a vaginal plug were isolated. The oviducts were dissected out
and flushed in M2 medium (Sigma). The oocytes were treated with
M2 containing hyaluronidase (300 ng/ml) to remove the cumulus,
washed in M2 and then maintained in M16 (Sigma, Taufkirchen,
Germany) for 2–8 h for assessment of the presence of male and
female pronuclei. The oocytes were then cultured for a further 48 h
in M16 covered with mineral oil to check for progressive development.
Sperm analysis
Epididymides were collected from 3 month old Tnp2⫹/⫹ and Tnp2–/–
males of 129/Sv genetic background and dissected in Tyrode’s
 Teratozoospermia in Tnp2-deficient mice

Figure 1. Targeted disruption of the Tnp2 gene. (A) Restriction map of the genomic fragment containing the closely linked Prm2, Prm3,
Tnp2 and Socs-1 genes, the targeted vector and the predicted restriction map after the homologous recombination. The probe used and the
predicted length of restriction fragments in the Southern blot analysis are shown. Primers 1, 2 and 3 used to amplify the wild-type and
mutant alleles by polymerase chain reaction are indicated. A pgk-neo cassette was inserted in exon 1 of the Tnp2 gene. TK, Thymidine
kinase cassette, E, EcoRI; S, SstII; S*, disturbed SstII; X, XbaI. (B) Southern blot analysis of the transfected ES clones. Genomic DNA
extracted from ES clones was digested with EcoRI and probed with the 3⬘ probe indicated in (A). The wild-type Tnp2 allele generates a
5.4 kb EcoRI fragment, whereas the targeted allele yields a 7 kb EcoRI fragment, as indicated in (A) and (B). (C) Testicular RNA of the
three genotypes was analysed using the Tnp2 cDNA and the neomycin gene as probes. (D) Western blot analysis of basic nuclear proteins
from wild-type (⫹/⫹), heterozygous (⫹/-) and knockout (–/–) mice were incubated with an anti-Tnp2 and anti-histone H1.1 antiserum. The
immunoreactive Tnp2 protein was detectable in wild-type and heterozygous, but not in knockout, mouse samples.

medium. Sperm number and motility were determined by light
microscopy. To examine sperm transport in the female reproductive
tract, males were mated with mature CD1 females. Six hours after
mating, uteri and oviducts from females with a vaginal plug were
flushed with M2 medium, and sperm numbers were counted. To
determine the replacement of the somatic histones by protamines
in the sperm nucleus, spermatozoa were recovered from cauda
epididymidis, centrifuged at 250 g for 5 min, fixed in 3:1 (vol:vol)
methanol:acetic acid and stained with Aniline Blue (Dadoune et al.,
1988). At least 200 spermatozoa from each male were assayed for
staining. To examine the acrosome reaction, epididymal spermatozoa
were capacitated for 15 h in Tyrode’s medium and then incubated
for 5 min at 37°C in 5% CO2 in Tyrode’s medium plus the calcium
ionophore A23187 (20 µmol/l; Sigma). To determine the percentage
of spermatozoa that had undergone acrosome reaction, spermatozoa
were fixed and stained with Coomassie Brilliant Blue R250 as
previously described (Thaler and Cardullo, 1995). At least 200
spermatozoa from each male were examined for the presence or
absence of the characteristic dark blue acrosomal crescent.
IVF assays
Sexually mature Tnp2⫹/⫹ and Tnp2–/– males of 129/Sv genetic
background were used for the experiments. Female CD1 mice
were superovulated. Oocytes were collected 10–12 h after HCG
administration and cumulus cells were removed by treatment with
hyaluronidase. To remove the zona pellucida, oocytes were treated
with acidic Tyrode and washed three times with phosphate-buffered
saline as described (Hogan et al., 1986). Spermatozoa were isolated
from the vas deferens and the cauda epididymis of each male group
and capacitated in Tyrode’s medium at 37°C for 1.5 h. Spermatozoa
(105–106) were added to the oocytes in 400 µl drops of fertilization
medium and incubated for 6 h at 37°C in 5% CO2. Using a large
bore micropipette, oocytes were washed in M16 and the oocytes were
then cultured in M16 as described.
Results
Targeted disruption of the Tnp2 gene in mice
Two genomic fragments, 6.3 and 3.5 kb, containing the closely
linked Prm2, Prm3 and Tnp2 genes, were used to construct
the Tnp2 targeting vector. A replacement-targeting vector was
designed for insertion of the Pgk-neo cassette into exon 1
upstream of the sequence coding for the arginine- and lysine-
rich domain of the Tnp2. The Herpes simplex virus thymidine
kinase (tk) gene, at the 3⬘-end of the construct, enabled us to
use positive and negative selection (Figure 1A) (Mansour
et al., 1988). R1 ES cells (Joyner, 1993) were transfected with
the targeting vector and selected for homologous recombination
515
I.M.Adham et al.
events. Drug resistance clones were selected and DNA was
isolated and screened by Southern blot analysis using an
external probe. A probe upstream of the targeting construct
detected a 5.4 kb EcoRI wild-type fragment and a 7 kb EcoRI
recombinant fragment (Figure 1A, B). One of four Tnp2⫹/–
ES clones injected into C57BL/6J blastocystes gave rise to
chimeric mice that transmitted the Tnp2 mutation into germ-
line. Chimeric mice were intercrossed to C57BL/6J and 129/
Sv females, respectively, to establish the Tnp2-disrupted allele
on a C57BL/6J⫻129/Sv hybrid and on a 129/Sv inbred genetic
background. In both backgrounds, male and female mice
heterozygous for the Tnp2 mutation appeared normal and
fertile. Heterozygous animals were mated, and ~25% (56 of
218) of the offspring were homozygous for the mutant allele.
Increased level of Tnp1 expression in Tnp2–/– testis
To confirm that the engineered disruption of Tnp2 by insertion
of the Pgk-neo cassette had generated a null mutation, we
examined the expression of Tnp2 at the mRNA and the protein
level by Northern blot and immunoblot analysis, respectively.
Northern blot analysis of RNA derived from testes of different
genotypes revealed that the Tnp2–/– mice failed to produce a
detectable 0.6 kb Tnp2 mRNA transcript (Figure 1C). We then
determined whether Tnp2 protein is synthesized in mutant
mice. Basic nuclear proteins were prepared from testes and
subjected to SDS/polyacrylamide gel electrophoresis and blot-
ted onto nitrocellulose filter. Western blot analysis revealed
that polyclonal anti-Tnp2 antibodies detected the protein in
wild-type and heterozygous mice. In contrast, no band corres-
ponding to Tnp2 was found in testes of homozygous Tnp2–/–
mice. Probing the Western blot with the polyclonal H1.1
antiserum revealed equal amounts of the loaded proteins
(Figure 1D).
The Prm3 and Socs-1 genes are located 1.8 kb upstream
and 2.4 kb downstream of the Tnp2 locus respectively. To
investigate whether the insertion of the Pgk-neo cassette
influenced the expression of these two genes, Northern blot
hybridization was performed using Prm3 and Socs-1 cDNA
probes. The results showed that the insertion of the Pgk-neo
cassette in the Tnp2 locus had no influence on the expression
of the closely linked Prm3 and Socs-1 genes (Figure 2A, B).
To verify the Tnp1 expression in testis of Tnp2–/– mice, a
Northern blot with RNA isolated from testes of different
postnatal stages was hybridized with Tnp1 cDNA probes. The
Tnp1 mRNA was first detectable in testis of wild-type and
Tnp2–/– day mice at day 23 of postnatal life. At subsequent
developmental stages, the level of Tnp1 mRNA was signific-
antly increased in testis of Tnp2 –/– mice compared to wild-
type littermates (Figure 2C).
Infertility of the Tnp2–/– males on the 129/Sv background
To investigate the consequences of the Tnp2 gene disruption
on male fertility, we intercrossed 10 Tnp2–/– males on the
C57BL/6J⫻129/Sv mixed and 129/Sv inbred background each
with two wild-type females for 3 months. All matings of
Tnp2–/– males on the hybrid background were productive and
the average litter size was not significantly altered as compared
to the breeding of wild-type littermates with wild-type females
516
Figure 2. Expression of the Prm3, Socs-1 and Tnp1 gene in testis
of the Tnp2–/– mice. (A, B) Northern blots with testicular RNA
isolated from the three genotypes were hybridized with the Prm3
and Socs-1 cDNA probes respectively (C). Total testicular RNA
isolated from 23, 24, 27 and 30 day old Tnp2–/– and Tnp2⫹/⫹ mice
was hybridized with a Tnp1 cDNA fragment. Rehybridization was
performed with the human EF-2 cDNA.
(Table I). In contrast, all Tnp2–/– males on 129/Sv background
were infertile despite normal sexual behaviour towards female
mice and production of copulation plugs. To further evaluate
male fertility, wild-type females were mated with wild-type
and Tnp2–/– males, and oocytes were collected 12 h after
mating and scored for fertilization. Eighty-one per cent of
oocytes harvested from females inseminated by wild-type
males had male pronuclei, and 78% developed to the 4-cell
stage after 48 h culture. In contrast, all 225 oocytes from
the Tnp2–/– matings lacked male pronuclei and failed to
develop further.
To address the question of whether the infertility of the
Tnp2-deficient mice on the 129/Sv background is caused by
failure of spermatozoa to penetrate the zona pellucida and
fertilize the oocyte, we have performed IVF assays. Spermato-
zoa were recovered from Tnp2–/– and wild-type animals and
tested for their ability to fertilize in-vitro zona-intact and
zona-free oocytes. Figure 3 summarizes the data of these
experiments. In in-vitro assays with zona-intact oocytes, only
18.5% of oocytes were fertilized with the spermatozoa of
Tnp2–/– mice and 17% then developed to the 4-cell stage,
whereas 79.5% of oocytes were fertilized with spermatozoa
of wild-type mice. In contrast, insemination of zona-free
oocytes by spermatozoa from Tnp2–/– and wild-type mice did
not result in significant differences in the fertilization rates.
Thus, the lack of Tnp2 prevents the spermatozoon from
penetrating the zona pellucida, but does not affect fertilization
once the oocyte is reached.
Spermatozoa from Tnp2–/– show abnormal acrosomes and
defects in transport
To study further the basis of the infertility of the Tnp2–/–
mice on the 129/Sv background, we investigated the sperm
morphology and the sperm transport through the female genital
 Teratozoospermia in Tnp2-deficient mice

Table I. Fertility of Tnp2⫹/⫹, Tnp2⫹/– and Tnp–/– mice on the genetic background C57BL/6J⫻129/Sv and
129/Sv

Genotype of male

57BL/6J⫻129/Sv 129/Sv

⫹/⫹ ⫹/⫺ ⫺/⫺ ⫹/⫹ ⫹/⫺ ⫺/⫺

No. of males mated 10 10 10 10 10 10

No. of females mated 20 20 20 20 20 20
No. of mice born 312 328 304 204 192 0
Average litter size 7.8 8.2 7.6 5.1 4.8 0
Fertility ratea 100.0 105.0 97.0 100.0 94.1 0
aDetermined as the percentage of average litter size in ⫹/⫹ mice.

Figure 3. Analysis of IVF experiments. (A, B). IVF and early development of zona-intact (A) and zona-free (B) oocytes incubated with
spermatozoa of Tnp2⫹/⫹ and Tnp2–/– mice on the 129/Sv background. The percentage of oocytes (no) incubated with spermatozoa is given
as 100%. The percentage reaching the pronucleus stage (pn) or 4-cell embryo stage (4c) is shown. The numbers in each category appear
above the columns. The results of IVF assay reveal that spermatozoa from Tnp2–/– are generally unable to penetrate the zona pellucida.

Table II. Sperm analysis in Tnp2⫹/⫹ and Tnp2–/– mice (background 129/Sv)

Parameter Genotype

⫹/⫹ ⫺/⫺

No. of spermatozoa in cauda epididymis (⫻107) 2.1 ⫾ 0.5 (5) 1.8 ⫾ 0.6 (5)
No. of spermatozoa in uterus (⫻103) 0.8 ⫾ 0.2 (5) 0.6 ⫾ 0.3 (5)
No. of spermatozoa in oviduct 301 ⫾ 15 (5) 16a ⫾ 5 (5)
Spermatozoa with head abnormalities (%) 6 ⫾ 2 (4) 23a ⫾ 3 (4)
Aniline Blue staining of sperm head (%) 3 ⫾ 1 (4) 44a ⫾ 3 (4)
Spermatozoa undergoing acrosome reaction (%) 76 ⫾ 7 (4) 31a ⫾ 7 (4)

Data from sperm analysis represent the mean ⫾ SEM of the number of individual measurements indicated in
parentheses.
aValues for these paramenters in Tnp2–/– mice are significantly different from those in Tnp2⫹/⫹ mice
(P ⬍ 0.01, Student’s t-test).

tract. Although there were no significant differences in the
mean number of spermatozoa collected either from the cauda
epididymis of Tnp2–/– and wild-type males or from the uterus
of wild-type females inseminated by spermatozoa of both
genotypes, the mean number of Tnp2–/– spermatozoa counted
in oviducts of inseminated females was found to be much
lower than the mean number of spermatozoa from wild-type
males in the oviducts (Table II).
Examination of spermatozoa by light microscopy revealed
that 24% of the Tnp2–/– cauda epididymis spermatozoa exhib-
ited abnormal sperm head morphology, an abnormality that
was only seen in 6% of wild-type spermatozoa (Figure 4A).
To determine whether the gross abnormality associated with
the sperm head was due to aberrant disposition of somatic
lysine-rich histones by cystein- and arginin-rich protamines
during nuclear condensation, we stained spermatozoa collected
from the cauda epididymis by Aniline Blue, which is known
to specifically stain lysine residues of histones in nuclei of
early spermatid stages but not in nuclei of normal mature
spermatozoa where the protamines are the prominent proteins
in the condensed chromatin (Dadoune and Alfonsi, 1986). The
percentage of stained heads was much higher in epididymal
spermatozoa of Tnp2–/– than in that of wild-type mice (Table II).
The sperm head abnormalities were also observed by
electron-microscopical examination of Tnp2–/– epididymal
spermatozoa. However, the abnormal sperm head was not
caused by the disruption of the chromatin condensation in the
sperm nucleus but due to malformations of the acrosomes.
517
I.M.Adham et al.

Figure 4. Morphological analysis of germ cells from Tnp2⫹/⫹ and Tnp2–/– males. (A) Light microscopy of spermatozoa from Tnp2–/– cauda
epididymis showed normal spermatozoa (w) and spermatozoa with abnormal heads (b). (B–E) Thin section illustrating the ultrastructure of
mature spermatids in testis of wild-type (B) and Tnp2–/– (C), and spermatozoa in cauda epididymis of wild-type (D) and Tnp2–/– mice (E).
The chromatin condensation proceeds as the germ cells mature, but the acrosome has become detached from the nucleus of the Tnp2–/–
spermatids (C) and spermatozoa (E) and is underlain by a large subacrosomal space. a ⫽ acrosome; m ⫽ mitochondrial sheath;
n ⫽ nucleus. Original magnification: (B–D) ⫻13 000; (E) ⫻22 000.

Many acrosomes (30%) appeared indented and/or partially
detached from the nuclear envelope (Figure 4E). Indented
acrosomes were also found in 4% of wild-type spermatozoa
(Figure 4D). This abnormal acrosome was also seen in
spermatids within the testis (Figure 4C). Thus, the high
frequency of the indented acrosomes strongly suggests that
the attachment of the acrosomal membrane to the nucleus is
impaired in spermatozoa of Tnp2 null. In addition to the
acrosomal defect, local presence of the sperm tail adjacent to
the sperm head was found in epididymal spermatozoa of
mutant mice. However, the axon of the spermatozoa were found
to possess the normal 9⫹2 arrangement of microtubule pairs.
To determine whether the acrosomal defect of the sperm-
atozoa of Tnp2–/– mice has an influence on the acrosomal
exocytosis, we examined the response of spermatozoa from
518
Tnp2–/– and wild-type mice to the calcium ionophore A23187.
As shown in Table II, spermatozoa of the Tnp2–/– mice differ
significantly in numbers undergoing acrosome reaction as
compared to wild-type spermatozoa.
These data indicate that teratozoospermia is responsible for
the infertility of the Tnp2–/– males of the 129/Sv background.
Discussion
In this study, we have generated mice carrying a null mutation
in the Tnp2 locus and determined the effect of the Tnp2
mutation on male fertility. Breeding of the Tnp2–/– animals
revealed that the loss of the Tnp2 gene in mice causes male
infertility depending on the genetic background. Similar results
have been obtained in mice carrying targeted null mutations

for the Pou-homeodomain gene (Sprm-1), the transition
protein-1 gene (Tnp-1) and the desert hedgehog gene (Dhh).
For these genes, highly variable penetrance of male infertility
on different genetic backgrounds have been described (Bitgood
et al., 1996; Pearse et al., 1997; Yu et al., 2000). The full
penetrance of the Tnp–/– phenotype on the isogenic 129
background and the observation of normal fertility of the
Tnp2–/– mice on the mixed background could indicate that
the interaction of the Tnp2 mutation with the 129 genetic
background involves modifier genes. Preliminary observations
of a normal number of spermatozoa produced and of normal
fertility in most Tnp–/– male mice were cited in one study (Yu
et al., 2000), but the genetic background of the studied Tnp–/–
mice has not been mentioned.
In Tnp2-deficient mice, the sperm heads become severely
deformed coincident with malformation of the acrosome. The
misplacement of the acrosome in spermatozoa may likewise
reflect disruption of the acrosomal reaction, as observed in
spermatozoa from the Tnp2 null mice. In addition, the inability
of the spermatozoa from the Tnp2–/– mice to penetrate the
zona pellucida in in-vitro assays is also a possible explanation
for their infertility. Furthermore, Tnp2–/– spermatozoa in the
oviducts were found to be reduced in number, indicative of
poor motility. Taken together, our results suggest that the
infertility of the Tnp2-deficient males is most likely a result
of abnormal sperm head morphology and poor motility.
It is interesting that the reduced fertility of the Tnp1-deficient
males is also due to abnormalities of the sperm head and
defects in sperm motility (Yu et al., 2000). The similarity of
the phenotypes of spermatozoa from Tnp1 and Tnp2-deficient
mice, and the elevated level of Tnp1 and Tnp2 in testis of
Tnp2- and Tnp1-deficient mice, respectively (these results and
Yu et al., 2000) raises the possibility that a deficiency created
through the disruption of the Tnp2 can be compensated for by
recruitment of Tnp1 and vice versa.
Our data concerning the pathology of the acrosomes of the
Tnp2–/– mice are, for the moment, essentially descriptive. It
has long been suggested that DNA binding properties of Tnp2
serve specific loci for initiation of chromatin condensation
during the later stages of spermatogenesis (Kundu and Rao,
1996). Immunocytochemical distribution of the Tnp2 has
shown that the first immunoreaction is detected at the tip of
the nucleus of step 11 spermatids and this reaction remains
localized over the anterior tip of the nucleus even in step 13
spermatids, in which chromatin condensation begins (Kistler
et al., 1996). The morphological changes of the acrosomal
vesicle during spermatogenesis are concomitant with chromatin
condensation. During spermatid steps 11–13, the acrosomal
cap is flattened, shifted caudally to the anterior tip of the
nucleus and attached finally to the nuclear involvement
(Oakberg, 1956). Therefore, it could be suggested that the
failure of the initial chromatin condensation during spermato-
genesis in Tnp2-deficient mice leads to impairment of acrosome
attachment to the nucleus involvement or to dehiscence of
acrosome from the nucleus.
Alterations in the fine structure of the acrosome have been
noted in spermatozoa of casein kinase II (Ck2)-deficient
mice, in which males are infertile and possess ‘round-head’
Teratozoospermia in Tnp2-deficient mice
spermatozoa with detached acrosome (Xu et al., 1999). Similar
alterations have also been noted in retinoid X receptor β
(RXRβ)-deficient mice and in mice carrying a T/t haplotype
(Dooher and Bennett, 1977; Kastner et al., 1996). It was found
that the acrosome is not firmly attached to the nuclear envelope
and that the spermatozoa of these genotypes also have
reduced motility.
There have been a number of gene disruption experiments
that have generated animals with more subtle abnormalities
than might have been expected. The normal expression of the
closely linked Prm3 and Socs-1 genes in the testis of the
Tnp2 –/– mice can rule out the possibility that the disrupted
Tnp2 locus in some way influences these neighbouring genes.
The phenotype of the Tnp2–/– mouse is clearly due to the Tnp2
gene deletion.
Acknowledgements
We would like to thank M.Schindler and H.Riedesel for assistance
with the generation of knock-out mice; C.Müller and S.Wolf for help
with particular experiments; A.Winkler for secretarial help; and
W.S.Kistler and B.Drabent for providing anti-Tnp2 and anti-H1.1
antiserum, respectively. This work was supported by a grant from
the Deutsche Forschungsgemeinschaft (through SFB 271) to W.E.
References
Alfonso, P.T. and Kistler, W.S. (1993) Immunohistochemical localization of
spermatid nuclear transition protein 2 in the testis of rats and mice. Biol.
Reprod., 48, 522–529.
Balhorn, R., Weston, S., Thomas, C. et al. (1984) DNA packaging in mouse
spermatids. Synthesis of protamine variant and four transition proteins.
Exp. Cell. Res., 150, 298–308.
Baskaran, R. and Rao, M.R.S. (1991) Mammalian spermatid specific protein,
TP2, is a zinc metalloprotein with two finger motifs. Biochem. Biophys.
Res. Commun., 179, 1491–1499.
Bitgood, M.L., Shen, L. and McMahon, A.P. (1996) Sertoli cell signaling by
Desert hedgehog regulates the male germline. Curr. Biol., 6, 298–304.
Dadoune, J.P. and Alfonsi, M.F. (1986) Ultrastructural and cytochemical
changes of the head components of human spermatids and spermatozoa.
Gamete Res., 14, 33–46.
Dadoune, J.P., Mayaux, M.J. and Guihard-Moscato, M.L. (1988) Correlation
between defects in chromatin condensation of human spermatozoa stained
by aniline blue and semen characteristics. Andrologia, 20, 211–217.
Dooher, G.B. and Bennett, D. (1973) Fine structural observations on the
development of the sperm head in the mouse. Am. J. Anat., 136, 339–261.
Dooher, G.B. and Bennett, D. (1977) Spermiogenesis and spermatozoa in
sterile mice carrying different lethal T/t locus haplotypes: A transmission
and scanning electron microscopic study. Biol. Reprod., 17, 269–288.
Fawcett, D.W., Anderson, W.A. and Phillips, D.M. (1971) Morphogenetic
factors influencing the shape of the sperm head. Dev. Biol., 26, 220–251.
Franke, K., Drabent, B. and Doenecke, D. (1998) Testicular expression of the
mouse histone H1.1 gene. Histochem. Cell Biol., 109, 383–390.
Hogan, B., Costantini, F. and Lacy, E. (eds) (1986) Manipulating the Mouse
Embryo. Cold Spring Harbor Laboratory Press, New York, 332 pp.
Joyner A.L. (ed.) (1993) Gene Targeting: A Practical Approach. IRL Press,
Oxford, 230 pp.
Kastner, P., Mark, M., Leid, M. et al. (1996) Abnormal spermatogenesis in
RXRβ mutant mice. Genes Dev., 10, 80–92.
Keime, S., Heitland, K., Kumm, S. et al. (1992) Characterization of four
genes encoding basic proteins of the porcine spermatid nucleus and close
linkage of three of them. Biol. Chem. Hoppe-Seyler, 373, 261–270.
Kistler, W.S., Henriksen, K., Mali, P. et al. (1996) Sequential expression of
nucleoproteins during rat spermatogenesis. Exp. Cell. Res., 225, 374–381.
Kleene, K.C. and Flynn, J.F. (1987) Characterization of a cDNA clone
encoding a basic protein, TP2, involved in chromatin condensation during
spermatogenesis in the mouse. J. Biol. Chem., 262, 17272–17277.
519
I.M.Adham et al.
Kundu, T.K. and Rao, M.R. (1996) Zinc dependent recognition of a human
CpG island sequence by the mammalian spermatidal protein TP2.
Biochemistry, 35, 15626–15632.
Luerssen, H., Maier, W.M., Hoyer-Fender, S. et al. (1989) The nucleotide
sequence of rat transition protein 2 (TP2) cDNA. Nucleic Acids Res.,
17, 3585.
Mansour, S.L., Thomas, K.R. and Capecchi, M.R. (1988) Disruption of the
proto-oncogene int-2 in mouse embryo-derived stem cells: a general strategy
for targeting mutation to non-selectable gene. Nature, 336, 348–352.
Nelson, J.E. and Krawetz, S.A. (1994) Characterization of a human locus in
transition. J. Biol. Chem., 369, 31067–31073.
Oakberg, E.F. (1956) Duration of spermatogenesis in the mouse and timing
of stages of the cycle of the seminiferous epithelium. Am. J. Anat., 99,
507–516.
Oko, R.J., Jando, V., Wagner, C.L. et al. (1996) Chromatin reorganization in
rat spermatids during the disappearance of testis-specific histone, H1t, and
the appearance of transition proteins TP1 and TP2. Biol. Reprod., 54,
1141–1157.
Panyim, S. and Chalkley, R. (1969) High resolution acrylamide gel
electrophoresis of histones. Arch. Biochem. Biophys., 130, 337–346.
Pearse, R.V., Drolet, D.W., Kalla, K.A. et al. (1997) Reduced fertility in mice
deficient for the POU protein sperm-1. Proc. Natl. Acad. Sci. USA, 94,
7555–7560.
520
Reinhardt, N., Kremling, H., Luerssen, H. et al. (1991) Characterization of a
gene encoding a basic protein of the spermatid nucleus, TNP2, and its
close linkage to the protamine gene in the bull. Biol. Chem. Hoppe-Seyler,
372, 431–436.
Schlüter, G. and Engel, W. (1995) The rat Prm3 gene is an intronless member
of the protamine gene cluster and is expressed in haploid male germ cells.
Cytogenet. Cell Genet., 71, 352–355.
Schlüter, G., Kremling, H. and Engel, W. (1992) The gene for human transition
protein 2: Nucleotide sequence, assignment to the protamine gene cluster,
an evidence for its low expression. Genomics, 14, 377–383.
Schlüter, G., Celik, A., Obata, R. et al. (1996) Sequence analysis of concerved
protamine gene cluster shows that it contains a fourth expressed gene. Mol.
Reprod. Dev., 43, 1–6.
Thaler, C.D. and Cardullo, R.A. (1995) Biochemical characterization of
a glycosyl-phosphatidylinositol-linked hyaluronidase on mouse sperm.
Biochemistry, 34, 7788–7795.
Xu, X., Toselli, P.A., Russel, L.D. et al. (1999) Globozoospermia in mice
lacking the casein kinase IIα catalytic subunit. Nature Genet., 23, 118-121.
Yu, Y.E., Zhang, Y., Unni, E., et al. (2000) Abnormal spermatogenesis and
reduced fertility in transition nuclear protein 1-deficient mice. Proc. Natl.
Acad. Sci. USA, 79, 4683–4688.
Received on December 20, 2000; accepted on March 28, 2001