Specific mRNAs in Sertoli and
Germinal Cells of Testes from Stage
Synchronized Rats
Carlos R. Morales, Acacia A. Alcivar, Norman B. Hecht, and
Michael D. Griswold
Program in Biochemistry and Biophysics (M.D.G.)
Washington State University
Pullman, Washington 99164-4660
Department of Biology (A.A.A., N.B.H.)
Tufts University
Medford, Massachusetts 02155
Department of Anatomy (C.R.M.)
McGill University
Montreal, Quebec, Canada
A treatment which used vitamin A depletion followed
by vitamin A repletion was used to synchronize
seminiferous tubules to a few related stages of the
cycle of the seminiferous epithelium. The success
of the synchronization procedure was dependent on
the age and size of the rat at the initiation of the
experiment (20 days of age and 35-40 g) and the
extent to which the vitamin A deficiency had progressed. Administration of retinol was done when
the only viable germinal cells in the testis were
preleptotene spermatocytes and type A spermatogonia but if the deficiency was prolonged spermatogenesis did not recover. Once established synchrony appeared to be sustained at least through
several consecutive cycles. A combination of molecular probes was used to determine if the synchronized testes displayed stage specific variations in
Sertoli cell and germinal cell mRNA levels as has
been reported for normal asynchronized rats. Sertoli
cells in the synchronized testes were shown by
quantitative in situ hybridization and by Northern blot
analysis to have stage specific variations in the
levels of mRNA for transferrin, sulfated glycoprotein1, and sulfated glycoprotein-2. The mRNA levels in
the different stages were qualitatively similar to
those in equivalent stages previously reported for
testes from asynchronous rats. The germinal cell
content of the synchronized testes were examined
with Northern blots probed with nick-translated protamine 1 and transition protein 1 cDNAs. The results
showed that testes synchronized in stages I-VI contained spermatid steps 1-6 and 15-18 and had little
or no protamine 1 or transition protein 1 mRNA but
that testes synchronized to later stages VIII to XIV
had both high levels and the predicted nonpolysomal
and polysomal sizes of these two mRNAs. Since
these germinal cell mRNAs are found in step 7 to 14
spermatids in normal rats, these results confirm the
extent of synchrony in these testes. Thus, the present investigation confirms the relationship of the
synchronized testes to the corresponding stages in
normal rats and establishes the value of this model
in future studies on the nature of the cycle of the
seminiferous epithelium. (Molecular Endocrinology
3: 725-733, 1989)
INTRODUCTION
Morphological and histochemical observations have resulted in a detailed description of the cyclic nature of
germinal cell development in the seminiferous epithelium. The cycle is defined by the recurring presence of
specific germinal cell types within given segments of
seminiferous tubules and by variations in Sertoli cell
synthetic capabilities (1 -6).
Several biochemical and molecular studies on the rat
have described the variable secretion of Sertoli cellspecific products such as androgen binding protein (7),
plasminogen activator (8), and cyclic protein-2 (9) in
relation to the stages of the cycle. The cDNAs for three
specific Sertoli cell products [transferrin, sulfated glycoprotein-1 (SGP-1), and SGP-2] have been cloned and
sequenced (10-13). The availability and use of these
recombinant cDNA probes has allowed the determination of the level of transferrin and SGP-2 mRNA in
Sertoli cells associated with different stages of the cycle
(14). It was shown by quantitative in situ hybridization
that transferrin mRNA was most abundant in Sertoli
cells associated with stages XIII and XIV of the cycle
of the seminiferous epithelium and that SGP-2 mRNA
levels were highest at stages VII—VIII (14,15).
0888-8809/89/0725-0733502.00/0
Molecular Endocrinology
Copyright © 1989 by The Endocrine Society
725
MOL ENDO-1989
726
In addition, cDNA probes for germinal cell specific
products such as protamine 1 and 2 and transition
protein-1 (TP-1) have been isolated and characterized
(16-18). These proteins appear in mid to late spermatid
stages of germinal cell development and are transcription products of haploid cells (19, 20).1 The availability
of cell specific molecular markers such as the cDNAs
described above allow for an unambiguous biochemical
analysis of spermatogenesis without the need for cell
separations.
It has recently been established that the seminiferous
tubules of the rat can be synchronized to the extent
that the testis which ordinarily is comprised of 14 stages
of the cycle of the seminiferous epithelium contains only
a few related stages. These stages traverse through
the cycle in a normal manner and spermiation occurs
synchronously in all tubules once every 12.8 days (21).
The synchrony is induced by feeding a vitamin A-deficient diet to young rats. The vitamin A deficiency (VAD)
produces in turn, a progressive germ cell depletion and
cessation of spermatogenesis resulting in seminiferous
tubules which only contain Sertoli cells, spermatogonia,
and a small number of preleptotene spermatocytes.
The spermatogenic arrest appears to be at the preleptotene stage of meiosis concomitant with a still uncharacterized spermatogonial arrest (21, 22). Under these
conditions spermatogenesis can be rapidly restored by
the administration of retinol, and in the majority of
cases, retinol induces the development of stage synchronized seminiferous tubules (21).
The following report is an extension of those initial
studies which used the synchronized testis model. It
was important to determine if synchrony declined over
time and if the tubules in the synchronized testes were
analogous at the molecular level to stage specific tubules in the normal rat. In addition, markers of germinal
cell development were used to illustrate the extent of
synchrony at the molecular level and to verify the presence of specific germinal cell populations which first
give rise to these products.
RESULTS
Stage Synchronization of Testes
The growth of the animals on the VAD diet (not shown)
increased linearly during the first 8 to 9 weeks. At this
time the weight stabilized for several days and was
followed by a decline in body weight. This decline in
weight is consistent with the total depletion of liver and
body vitamin A stores (23). Simultaneously, several
other symptoms became evident, such as conjunctivitis, blefaritis, and the appearance of the chronic cough
signalling the metaplasia of the tracheal epithelium typical of this condition (24). During the days after vitamin
1
Yelick, P. C , Y. Kwon, J. G. Flynn, A. Borzorgzodeh, K.
C. Kleene, and N. B. Hecht, submitted.
Vol 3 No. 4
A depletion, palpation of the testes was performed daily
in order to detect the testicular atrophy that was a
consequence of cessation of spermatogenesis and
germ cell loss. Reduction of testes size generally occurred between 1 - 2 weeks after the vitamin A depletion
and this was considered as the proper time to initiate
the retinol treatment. Sections of atrophic testes
showed seminiferous tubules lined by Sertoli cells, spermatogonia, and a few preleptotene spermatocytes. The
precise time of depletion of vitamin A stores and testicular atrophy was subject to variation among rats. We
found that if the VAD diet was initiated before the rats
reached 20 days of age there was a high mortality and
if the diet was initiated after the rats had passed 20
days of age it was sometimes difficult to achieve the
deficient state within the normal time period. The optimal conditions were achieved with 20-day-old rats
weighing between 35 and 40 g. In addition, if the
deficiency was extended much beyond the time after
the atrophy of the testes was complete the tubules did
not reinitiate spermatogenesis.
If the protocol described above was followed, reinitiation of spermatogenesis occurred almost immediately
after retinol administration and the first spermiation was
completed 39-41 days later (21, 23). Although the
development of the spermatids proceeded in a synchronous manner (Fig. 1), the number of germ cells per
Sertoli cell appeared to be considerably lower than in
the homologous stages of normal seminiferous tubules.
This was probably due to a reduction in the number of
preleptotene spermatocytes because of the vitamin A
deficiency (22). However, after the second round of
spermiation (day 52 subsequent to the retinol administration) the number of germ cells per Sertoli cell seemed
to be fully recovered and no evident morphological
differences were found between the synchronized tubules and the homologous stages of normal seminiferous tubules. Thus, the synchronized testes used in this
investigation for in situ hybridization and Northern blot
analysis were from animals killed between the second
and third round of spermiation which occurred between
days 52 and 65 after retinol replacement. The stages
as determined by histological sections in some of the
synchronized testes obtained during the course of this
investigation are shown in Fig. 1. The data confirmed
that synchronization persists after more than one round
of spermiation and suggest that a spermatogonial arrest
must coexist with the arrest at preleptotene spermatocyte in order to have a persistent synchronization
(22). As the time after Vitamin A increased the synchronization deviated from the predicted values (Fig. 1).
In Situ Hybridization in Sections of Synchronized
Testes
We have previously used biotinylated cDNAs to localize
transferrin, SGP-1, and SGP-2 mRNA exclusively to
Sertoli cells in the testes (12-14). We obtained the
same results in the testes from synchronized rats. The
appearance of the biotinylated SGP-1 cDNA in Sertoli
mRNA in Sertoli and Germinal Cells
727
in
Normal
13
m rv v vi vn vra ix x xi xn xm xrv
6
3 4
9
9
19
10
NS
10
Fig. 1. Representation of the Stage Frequency of Seminiferous Tubules Found in Histological Sections of Testes from Rats after
Vitamin A Deficiency and Retinol Repletion
The Roman numerals indicate the stages of the cycle according to the classification of Leblond and Clermont (1). The numbers
show the frequency appearance of various stages of seminiferous tubules in percentages. The normal represents the frequency of
appearance of stages in testicular sections of normal Sprague-Dawley rats. The remaining data give the frequency of stage
appearance observed in a single synchronized testes and the shaded areas indicate the predicted stages. Note that as the time
after recovery was extended the synchronized stages drifted beyond the range of predicted stages. The stage frequency was
determined from observations of 200 tubules from at least two different cross-sections on one testis. The value for nonsynchronized
or undeveloped tubules is given as (NS).
cells is shown in Fig. 2a and is representative of the
results obtained for transferrin and SGP-2. When the
3
H-RNA probes for these three mRNAs were detected
after in situ hybridization the radioautographic reaction
was observed exclusively over the seminiferous tubules. The data for the 3H-SGP-1 probe is shown in
Fig. 2b and again is representative of the results obtained with transferrin and SGP-2 probes. As expected,
the silver grain distribution over the seminiferous epithelium presented a scattered appearance. However, a
high proportion of silver grains were observed overlying
those areas where the cytoplasm of Sertoli cells was
present (Fig. 2b). Treatment of the tissue with RNase
before hybridization, in situ hybridization at 0 C, or in
situ hybridization with 3H-RNA strands homologous to
their respective mRNAs eliminated either the biotinylated or the radioautographed reaction (data not
shown).
The number of silver grains associated with crosssections perpendicular to the longitudinal axis of seminiferous tubules was determined in the normal and
synchronized testes using transferrin, SGP-1, and SGP2 3H-RNA probes. For all three probes, no statistically
significant differences were found among the tubules in
a particular stage. For the transferrin probe, grain
counts showed that the lowest level of mRNA was
registered at stage IX (737 ± 131 SD) and XI (704 ± 51
SD) whereas the highest values occurred during stages
XIII (2248 ± 266 SD) and XIV (1929 ± 127 SD) (see Fig.
3). Analysis by t test demonstrated statistically significant differences between stages IX-XI and XIII—XIV
with any of the other quantified stages (P < 0.005). For
the SGP-2 probe the grain counts showed that the level
of SGP-2 mRNA increased from stage II to stage VIII.
At stage IX the level of transcripts decreased sharply,
remaining low until stage II of the cycle (Fig. 4). The
highest values were associated with stage VI, VII, and
VIII. Analysis by t test demonstrated that the peak of
stage VIII (1700 ± 393 SD) and the high values of stages
VI and VII were significantly different from the rest of
the other, stages analyzed (P < 0.005). Finally, the
quantitative analysis of the SGP-1 3H-RNA probe also
showed significant differences in the level of mRNA
from tubules in different stages (P < 0.05). However,
the stage variations were less pronounced for SGP-1
than those observed for transferrin and SGP-2 mRNAs
(Fig. 5). The t test analysis suggested that the relatively
higher level of transcripts of stages VII and VIII were
statistically significant as was the low level of SGP-1
mRNA registered at stage XIV (P < 0.05).
Northern Blot Analyses
The expression of transferrin, SGP-1, and SGP-2
mRNAs in synchronized testes were also examined by
Northern blot analysis. Total RNA was isolated from
the synchronized testes, and poly(A+) RNA was isolated
from Sertoli cells in culture. For the cultured Sertoli
Vol 3 No. 4
MOL ENDO-1989
728
jo
2000 -
IV
VI
VII
VIII
IX
XI
XIII
XIV
Stages
Fig. 4. Variations in the Level of SGP-2 mRNA per CrossSection of Synchronized Seminiferous Tubules
Each column represents the mean number (±SD) of silver
grains counted in four tubule cross sections. Data was obtained from synchronized testes described in Fig. 1.
Fig. 2. Cross-Sections of Synchronized Seminiferous Tubules
Hybridized In Situ with SGP-1 cRNA Probes Transcribed from
pTZ18U
a, Portion of a tubule hybridized with a biotinylated probe
and visualized with the avidin-glucose oxidase/tetranitroblue
reaction. Black deposits of insoluble formazan indicate that
SGP-1 mRNA is localized in Sertoli cells but not in germ cells,
b, Radioautograph of a portion of seminiferous tubule hybridized with a tritiated cRNA probe. Note that the majority of
silver grains overlay the basal pole of Sertoli cells (large arrows)
or the paler stained areas which correspond to the cytoplasm
of Sertoli cells (small arrows).
2000 -|
VI
VII
VIII
IX
XI
XIII
XIV
Stages
Fig. 5. Level of SGP-1 mRNA per Cross-Section of Synchronized Seminiferous Tubules
Each column represents the mean number (±SD) of silver
grains counted in three cross-sections of tubules. Data was
obtained from synchronized testes described in Fig. 1.
3000-1
II
IV
VII
VIII
IX
XI
XIII
XIV
Stages
Fig. 3. Variations in the Level of Transferrin mRNA per CrossSection of Synchronized Seminiferous Tubules
Each column represents the mean number (±SD) of silver
grains counted in four tubule cross-sections. Data was obtained from synchronized testes described in Fig. 1.
cells, 3 fig poly(A+) RNA was used on the blots. For the
synchronized testes, 5 n\ final volume (of 500 ^l) containing the total isolated RNA from one testis was used.
The distribution and proportion of stages as well as the
amount of micrograms per n\ RNA obtained from each
testes were specified in Table 1. The recovery of total
RNA from the testes was approximately 70% in each
sample.
Figure 6, A, B, and C show the autoradiogram of the
nitrocellulose blot and indicate the relative levels of
transferrin [2.7 kilobases (kb)], SGP-2, (2.1 kb), and
SGP-1 (2.6 kb) mRNAs. Densitometric scanning of the
autoradiogram demonstrated that the level of transferrin, SGP-1, and SGP-2 mRNAs in the synchronized
testes were consistent with the variations obtained by
in situ hybridization (data not shown). In addition, mRNA
blots from these same samples were hybridized to nicktranslated cDNAs which represented transcripts from
specific germinal cells (Fig. 7, A and B). The changes in
the length of protamine and TP-1 mRNAs precisely
reflect the cellular composition of the extracts. For
instance, the 580 nucleotide protamine 1 mRNA is
predominant in stages Vll-X (Fig. 7A, lanes 4-6)
whereas the partially deadenylated 450 nucleotide protamine 1 mRNA is the transcript in stages III-VI (Fig.
7A, lanes 2 and 3). A combination of stages Xlll-ll (Fig.
7A, lanes 1 and 7) reflect the cell types containing both
size classes of protamine 1 mRNAs. The TP-1 transcripts also mirror stage dependent size changes with
both size classes of mRNAs predominant around
mRNA in Sertoli and Germinal Cells
729
Table 1. Isolation of RNA from Synchronized Testes
Yield
Sample no.
Stage Status
Wt. of Testis
(9)
(mg Total RNA/
Testis)
I, 88%; XIV, 6%; XIII, 6 %
VII, 22%; VI, 27%; V, 27%; IV, 2 4 %
VII, 14%; VI, 13%; V, 22%; IV, 43%; III, 8 %
X, 4%; XI, 3%; VIII, 57%; VII, 3 1 %
XII, 2%; XI, 5%; X, 10%; IX, 28%; VIII, 4 2 %
XII, 7%; XI, 19%; X, 27%; IX, 30%
II, 15%; I, 37%; XIV, 13%; XIII, 30%
1.48
1.47
1.43
1.20
1.35
1.02
1.42
1.2
1.3
2.1
2.0
2.0
1.5
1.5
Synchronized testes were obtained from retinol-treated vitamin A-deficient rats and the mRNA was isolated as described. Yields
are not corrected for recovery. Asynchronous and/or unrecovered seminiferous tubules were sometimes present in small proportion
(less than 5%) among the synchronized tubules.
1
2
3
4
5
6
7 SC
12
3 4 5 6 7
2.6 k b
A
1
2
3 4
5
6 7 SC
2X)kb
12
4 5 6
B
1 2
3 4
5
6
7 SC
2.6 kb
B
Fig. 6. Northern Blot Analysis of Sertoli Cell mRNA
Lanes 1 to 7 correspond to the stages of the testes specified
in Table 2 and represent the total RNA isolated from the
different synchronized testes. In addition, lane SC corresponds
to RNA isolated from rat Sertoli cells in culture. For a, b, and
c equal volumes of mRNA were added (5 n\) so that the
intensity is relative to the total amount of a specific mRNA per
testis. The lanes represent autoradiograms of blots probed
with nick translated transferrin cDNA (a), SGP-2 cDNA (b), and
SGP-1 cDNA (c).
stages VIII to XII of spermatogenesis and the initial
longer transcription product in stage VII (Fig. 7B). The
stronger TP-1 hybridization signals seen in lanes 4 to 6
reflect the increased accumulation of TP-1 mRNA during steps 7 to 13 in spermatid development which are
found in stages VIII to XIII of the spermatogenic cycle.
A similar temporal pattern of expression was found for
protamine 2 mRNA (data not shown).
Fig. 7. Northern Blot Analysis of Germinal Cell mRNA
Lanes 1 - 7 correspond to the stages of the testes specified
in Table 2 and analyzed in Fig. 5. For A and B, 8-MQ, aliquots
of RNA per lane were loaded. The lanes represent autoradiograms of blots probed with nick-translated protamine-1 cDNA
(A), and TP-1 cDNA (B). In (B) samples 3 and 7 were not
analyzed. The tick marks in (A) correspond to 580 and 450 bp
and in (B) to 600 and 480 bp.
DISCUSSION
Stage Synchronization of Seminiferous Tubules
The present study is an extension of our initial paper
describing retinol-induced stage synchronization of the
seminiferous tubules of the rat (21). Administration of
retinol to vitamin A deficient rats can rapidly reinitiate
spermatogenesis in a synchronous manner from the
most advanced germinal cell present, i.e. the prelepto-
MOL ENDO-1989
730
tene spermatocyte (21). Since the duration of the cycle
is well known in this particular strain of rat, the stages
of the tubules of the synchronized testes were predicted by calculating the rate of differentiation of the
preleptotene spermatocyte during the period from the
first administration of retinol to the day of killing. However, the preleptotene phase of the primary spermatocytes has a duration of approximately 88 h and we do
not know exactly the timing of the preleptotene block.
Therefore, the predicted stages were calculated from
the beginning to the end of the preleptotene stages (VII
and VIII) and had a window of 88 h (21).
In order to achieve satisfactory synchronization the
precise time of vitamin A depletion and the testicular
atrophy that follows this condition must be carefully
monitored. For example, when the exposure to vitamin
A deficiency is prolonged for several weeks the administration of retinol does not reinitiate spermatogenesis
indicating that irreversible damage results. Conversely,
if the retinol treatment was started immediately after
vitamin A depletion when the testes were not fully
atrophied, the synchronization was frequently incomplete because of the presence of a mixed population of
germ cells at the moment of the retinol administration.
An additional biological variable is the age and nutritional status of the weanlings which are first put on the
VAD diet.
The long term study showed that the full recovery of
the normal germinal cell numbers appeared to occur
after the second spermiation. Therefore, the collection
of testicular tissue in the synchronous model may be
optimal after 52 days from the first injection of retinol.
Synchronization could be maintained for at least 78
days after the initiation of retinol administration and
after this time there was no apparent diminution of the
synchronized condition. It appears that once the testes
become synchronized the condition persists. All of the
evidence suggests a spermatogonial arrest in addition
to the arrest at the preleptotene spermatocyte stage.
Although the type A1 spermatogonia appear to be the
logical candidate for this second blockage, more studies
are needed to determine the identity of the arrested
cell. Finally, from day 62 after retinol administration, the
stages of the synchronized testes drifted beyond the
predicted range of stages indicating either that the
duration of the cycle was shorter in our Sprague-Dawley strain or that the initial estimations were longer than
the real values (see Fig. 1). The duration of one cycle
as reported by Leblond and Clermont (1) is 308 h
(12.83) days). Our data as shown in Table 1 would all
fall within predicted values if the duration of one cycle
in our rats was approximately 290 h. We do not know
if this discrepancy is related to the vitamin A manipulation or to differences in rat strains.
Stage Specific Levels of mRNA in the Synchronous
Testes
One of the main purposes of this investigation was to
compare some aspects of the molecular behavior of
Vol 3 No. 4
Sertoli cells and germ cells in relation to the stages of
the cycle using the synchronous model and to compare
the results with published data on the normal testis.
In situ hybridization of sections from the synchronized
testes with the biotinylated and radioactive probes confirmed previous studies which showed that Sertoli cells
expressed transferrin, SGP-1, and SGP-2 mRNA and
proteins (13, 14). Furthermore, the use of 3H-probes
and subsequent radioautography allowed the quantitative determination of mRNA variations in Sertoli cells
among the different synchronized stages. The calculations were dependent on the observations of Wing and
Christensen (25) that the number of Sertoli cells per
unit length of seminiferous tubules remains constant
throughout the 14 stages of the cycle. These data were
consistent with those presented by Bustos-Obregon
(26) from Sertoli cell counts in whole mounts of seminiferous tubules at stages II, IV, V, and VII. Thus, the
number of Sertoli cells per cross-section of tubules is
not subject to cyclical changes. Therefore, in a section
which is perpendicular to the longitudinal axis of a
seminiferous tubule, the number of silver grains reflects
differences in the amount of specific hybridizable mRNA
per Sertoli cell associated with a given stage (14).
The results from the quantitative in situ hybridization
using the transferrin and SGP-2 probes agreed qualitatively with the results we have previously reported on
normal rats. In the studies in this manuscript the stages
analyzed in the synchronized testes came from different
animals while in the previous report a single normal rat
could be used to obtain data for all stages. It has been
shown that transferrin is important in the delivery of
iron to developing germinal cells (15) and that SGP-2
binds to spermatozoa and is also made by the epididymis (12, 22). The stage specific variation in SGP-1
mRNA has not been described for the normal rat but in
the synchronized rat stages VII and VIII expressed
relatively higher amounts of SGP-1 mRNA and stage
XIV relatively lower levels. It has been shown that SGP1 is similar to a human protein designated as sphingolipid activator protein precursor. This protein contains
four domains which can act as sphingolipid binding
regions and activate the metabolic degradation of these
lipids (13). One interpretation of these results is that
transferrin may be more important during the meiotic
division of germinal cells (stages XIII—XIV) while SGP-1
and SGP-2 which interact with spermatozoa are required during the spermiation event (stage VIII).
Northern blot analysis confirmed the presence of
transferrin, SGP-1, and SGP-2 mRNAs in all stages and
the variations in mRNA levels using this semi-quantitative technique were consistent with the data obtained
in situ hybridization. Since Sertoli cells were the only
testicular cells that expressed these mRNAs and assuming that the synchronized testes contained approximately the same number of Sertoli cells per testis, any
variation in the level of transcripts will reflect stagespecific levels related to the presence of a certain stage
or group of stages in a given synchronized testis.
The changes in mRNA length of several germ cell-
mRNA in Sertoli and Germinal Cells
specific proteins during spermiogenesis provided a second means to monitor cell stage and function of the
synchronized rat testis. Several DNA binding proteins,
the transition proteins and protamines, that are involved
in the nucleohistone-nucleoprotamine transition in
mammals are well characterized temporally. All appear
to be under translational regulation since transcription
terminates during mid-spermiogenesis, in some cases
up to a week before the protein is first synthesized.
Coincident with the storage requirements of the mRNAs
coding for TP1 and protamines 1 and 2 are marked
length differences between the stored and translated
mRNAs (19, 20). These changes provide a precise
molecular marker for stage of spermiogenesis and allow
the molecular differentiation of a group of stages from
a synchronized testis to be critically compared to a
control testis. In both mouse and rat, protamine 1 is
transcribed during mid-spermiogenesis as an approximately 580 nucleotide long molecule which is deadenylated to a 450 nucleotide mRNA at its time of translation (18, 27). Similar reductions in length occur for
protamine 2 (from about 830 to about 700 nucleotides)
and TP-1 (from about 600 to about 480 nucleotides)
(see Footnote1). The temporal appearance of the protamine 1 and TP-1 mRNAs in the synchronized rat
testes matches the cellular composition of each sample
very closely and suggests that TP-1 mRNA is transcribed before protamine 1 (Fig. 6, a and b, lane 2).
Moreover, these studies are in good agreement with a
recent in situ hybridization analysis of protamine 1
expression in the rat (27).
Because of the nature of these studies much of the
data that was presented was obtained on a single
sample, i.e. a single synchronized testis. Since each
testis varied somewhat in composition of stages the
data could not be summarized from a number of samples. However, the results are indicative of many more
samples than were presented.
In conclusion, the seminiferous tubules of the synchronous testis model appear to cycle in the same
manner as the homologous seminiferous tubules of the
normal testes. The present study showed that the
model may be useful in resolving complicated problems
such as the analysis of stage-specific variations of
Sertoli or germinal cell mRNAs and in studies of stagespecific transcriptional and translational regulation of
either germ cells or Sertoli cells.
MATERIALS AND METHODS
Preparation of Synchronized Testis
Male Sprague-Dawley rats (20 days old) were fed a vitamin Adeficient diet (United States Biochemical Corporation) for 8 10 weeks. After this period the animals were given by ip
injection 7 mg, and 24 h later, 2.5 mg retinol acetate suspended in sesame oil. In both cases the retinol was dissolved
in 200 A*' ethanol and mixed with 200 n\ distilled water. Subsequently, the rats received a daily oral administration of 0.5
mg retinol in the normal balanced rat diet during the first week
after the first injection. The animals were used between 38
731
and 78 days after the initial retinol injection. In all cases the
rats were anesthetized with sodium pentobarbital and after a
scrotal incision and ligation of the testicular artery, the right
testis was removed, weighed, and immediately frozen on dry
ice. After this, the left testis was fixed by perfusion through
the abdominal aorta with either Bouins' fixative or with 4%
paraformaldehyde and 0.5% glutaraldehyde in PBS. Since
both testes synchronized simultaneously and posed almost
identical stage distribution of seminiferous tubules (21) one
section of each of the fixed testes was stained with periodic
acid Schiff. This procedure allowed the identification of both
the nuclear morphology and the acrosome shape of evolving
spermatids, permitting the precise stage determination of the
seminiferous tubules in a testicular section (1). Stage frequency was calculated by determination of stages in 200
tubules from at least two sections of each testis.
Recombinant DNA Probes
From the initial transferrin cDNA clone that consisted of a 1kb sequence, which coded for the carboxy terminal half of the
protein (11), a 688 base pairs fragment was generated by
digestion of the cDNA with the restriction enzymes Pst\ and
Hinctt, and was subcloned into the HincW and Pst\ sites of the
SP65plasmid(14).
The original SGP-2 cDNA was cloned into the Pst\ site of
the pUC-13 vector which represented a full-length cDNA that
coded for a precursor to SGP-1 (12). The cloned SGP-2 cDNA
was digested with the restriction enzymes Xba\ and PvuW to
generate an 829 base pair (bp) fragment, which was subcloned
into the Xbal and Smal sites of the SP65 plasmid (14).
A cDNA which represented the entire coding region of the
SGP-1 protein (13) was subcloned into the HincW and BamHI
sites of the pTZ 18U vector (designated IAZ 10.3). Thus,
transcription originating from the SP6 promoter of the SP65
plasmid and from the T7 promoter of the pTZ 18U plasmid
produced a cRNA strand complementary to transferrin and
SGP-2 and the SGP-1 mRNAs respectively.
A mouse protamine 1 cDNA containing the complete coding
region and 3'-untranslated region and most of the 5'-untranslated region was cloned into a pGEM vector. Digestion with
the enzymes, EcoRI and Psfl, generated a 400 nucleotide
fragment for hybridization (16).
A mouse TP-1 cDNA containing the coding and 3'-untranslated regions was cloned into a Bluescript vector. Digestion
with EcoRI generated a 435-bp sequence which was used for
hybridization.
Synthesis of cRNA Probes
For in situ hybridization the testicular sections were incubated
with 3H-labeled cRNA probes (5 x10 7 cpm/^g SP65 transferrin, SP65 SGP-2 and IAZ 10.3 SGP-1 cRNA). RNA probes
complementary to transferrin and SGP-2 mRNAs were synthesized from either 1 ^g linearized SP6 plasmid in transcription
buffer: 40 ITIM Tris (pH 7.5), 6 ITIM MgCI2, 2 rtiM spermidine,
0.5 mM each ATP, CTP, and GTP, either biotinylated UTP (1
IDM Bio-11-UTP from Bethesda Research Laboratories, Gaithersburg, MD) or 8 nC\ 3H-UTP (32.2 Ci/mol from New England
Nuclear, Boston, MA), 0.2 M dithiothreitol with the addition of
50 U RNasin and 10 U SP6 RNA polymerase/reaction. In the
case of SGP-1 the riboprobe was synthesized from 1 /xg
linearized pTZ 18U (IAZ 10.3) plasmid in T7 transcription buffer
(United States Biochemical Corporation, Cleveland, OH); 0.5
mM each ATP, CTP, and GTP, 0.1 mM UTP, either 1 mM Bio11-UTP or 8 fiC\ 3H-UTP and 0.2 M dithiothreitol with the
addition of 20 U T7 polymerase.
In Situ Hybridization
Deparaffinized sections of synchronized and normal testes
fixed with 4% paraformaldehyde and 0.5% glutaraldehyde in
Vol 3 No. 4
MOL ENDO-1989
732
PBS were pretreated with 0.2 N HCI for 20 min at room
temperature, followed by a digestion with 2 ^g/ml proteinase
K in 20 miwi Tris-HCI and 2 mM CaCI2 at 37 C and subsequently
immersed in 300 mM glycine during 5 min at room temperature.
The sections were then dehydrated in a graded series of
alcohols and air dried.
In situ hybridization was performed according to the procedure of Guelin et al. (28) as modified by Morales ef al. (14).
Briefly, the modification consisted of the pretreatment of the
sections with a prehybridization mixture containing yeast tRNA
(Sigma, St. Louis, MO), and denatured salmon sperm DNA
(Sigma). The cRNA concentration was approximately 0.35 ng/
fi\ for SGP-2 and SGP-1 and 0.75 ng/^l for transferrin. The
specific activity of the 3H-labeled RNA probes were 2.7 x 106
cpm/fig for SGP-2 and SGP-1 and 6 x 106 cpm/^g for transferrin. All hybridization and wash buffers contained 4x STE
(STE: 150 mM NaCI; 2.5 mM Tris-HCI; 0.25 mM EDTA; pH 7.4)
and 50% formamide. After the final wash, the tissue sections
were dehydrated in alcohol and air dried. For specificity controls, some tissue sections were digested with ribonuclease A
or incubated with the same probes at 4 C. In the case of
transferrin and SGP-2 additional controls were included using
cRNA strands homologous to their mRNAs generated from
the same cDNAs subcloned into the Hinc\\-Pst\ and Xba\Sma\
sites, respectively, of the Sp64 plasmid
Tissue sections were coated with Kodak NTB-2 nuclear
emulsion, and 3H-cRNA: mRNA hybrids were visualized by
radioautography (30). After 2 days or 7 days exposure, the
slides were developed with a Kodak D-170 developer and post
stained in hematoxylin and eosin. For the identification of the
biotinylated probes, a biotinylated glucose oxidase -avidin
system (Vectastain ABC-GO; Vector Labs, Burlingame, CA)
was used, followed by a treatment with phenazine methosulfate as an intermediate electron carrier and tetranitroblue
tetrazolium, which upon reduction formed a black-colored,
insoluble formazan in the site of the reaction (14).
Quantitative Analysis of In Situ Hybridization
One paraffin section of the synchronized testicular tissue was
obtained from each of the experimental animals and subjected
to hybridization as described above. Radioautographs were
examined under the light microscope to select and stage
seminiferous tubules (1). The total number of grains per crosssection of seminiferous tubules was counted in those tubules
that were cut in perfect transverse sections. Cross-sections
from four synchronized testes which contained a closely related distribution of stages were used to obtain each data
point.
The grain counts were submitted to the following statistical
analysis: 1) analysis of variance to determine if the counts
varied significantly among the different stages and to determine if there were significant differences among the four
sections counted for each stage and 2) analysis by f test in
order to find possible significant differences between specific
stages.
presence of ammonium acetate and ethanol. The testis homogenate was treated with 50 ^/ml proteinase K (Sigma), for
45 min at 45 C and extracted with phenol:chloroform (1:1).
The nucleic acid in the aqueous phase was precipitated with
2.2 vol ethanol at - 2 0 C and the pellet was resuspended in
H2O. An equal volume of 4 M LiCI was added to the nucleic
acids and precipitated at 0 C. The RNA concentration was
determined from the absorbance at 260 nm wavelength and
an aliquot was removed and used to determine the percentage
of 3H-lambda cRNA in each sample and the micrograms RNA
per testis and the percentage recovery was calculated.
Northern Blot Analysis
The design of this experiment was based on the observation
that Sertoli cells cease to proliferate by postnatal day 20 (30)
and that the number of Sertoli cells in the adult SpragueDawley testis is approximately 21.9 x 106 with small variations
(<5%) from animal to animal (25). The recovery of poly(A+)
RNA isolated from the synchronized testes was similar in each
of the seven testes, so equal volumes of samples (5 ^l) were
incubated in 6% formaldehyde, 20 mM HEPES (pH 7.8), 1 mM
EDTA, and 50% formamide for 5 min at 68 C. Samples were
rapidly cooled and subjected to electrophoresis in 1.2% agarose gel for 15 h at 20 V. The RNA was transferred to
nitrocellulose and hybridized with 200 ng 32P-labeled nick
translated cDNA probe (108 cpm/Mg plasmid).
For the analysis of protamine 1 and TP-1 mRNA, 8-^g
aliquots of RNA were denatured with formaldehyde and electrophoresed for 15-20 h at 20 V in 1.0% agarose gels containing formaldehyde and were then transferred to nitrocellulose filters. The blots were baked, rehydrated in distilled water,
and prehybridized for 2 h at 65 C in a solution containing 0.40
M phosphate buffer (pH 6.5), 1 x Denhardt's solution, 1 % SDS,
and 50-100 ^g/ml salmon sperm DNA. Hybridization was
performed overnight at 65 C by incubating the filters in fresh
prehybridization solution (~15 ml) with the addition of radiolabeled probe (1-10 x 107 cpm/mg plasmid DNA. Filters were
washed twice at room temperature for 30 min in a solution
containing 0.25 M phosphate and 1 % SDS, and then at 65 C
for 30 min in a solution containing 0.40 M phosphate and 1 %
SDS. Filters were exposed to Kodak XAR film at - 7 0 C for
1-5 days.
Acknowledgments
Received December 12, 1988. Revision received January 18,
1989. Accepted January 18,1989.
Address requests for reprints to: M. D. Griswold, Program
in Biochemistry and Biophysics, Washington State University,
Pullman, Washington 99164-4660.
Supported in part by NIH Grants HD-10808 (to M.D.G.),
HD-21523 (to M.D.G.), GM-29224 (to N.B.H.), and a grant
from MRC (to CM.).
Isolation of Nucleic Acids form Synchronized Testes
REFERENCES
The frozen right testis was thawed and homogenized in buffer
150 mM NaCI, 1 mM Tris, pH 7.1, 0.75% sodium dodecyl
sulfate (SDS), 5 mM ethylediaminetetraacetate (EDTA) plus 1 x
10" cpm 3H-lambda cRNA for 1 min, using a bio-homogenizer
(Biospec Products, Bartlesville, OK). 3H-labeled lambda cRNA
was synthesized from 1 ng control template (Promega Biotec,
Madison, Wl). Synthesis was carried out in transcription buffer:
40 mM Tris (pH 7.5), 6 mM MgCI2, 2 mM spermidine, 0.5 mM
each ATP, CTP, and 0.1 mM uridine 5'-triphosphate (UTP), 8
nC\ 3H-UTP (32.2 Ci/mol), 10 nM dithiothreitol, with the addition
of 50 U RNasin (Promega Biotec) and 10 U SP6 RNA polymerase (Promega Biotec). The reaction was carried out at 37 C
for 1 h and cRNA was then extracted from the unincorporated
nucleotides by successive precipitations at - 2 0 C in the
1. Leblond CP, Clermont Y 1952 Spermiogenesis of rat,
mouse, hamster and guinea pig as revealed by the
"periodic acid-fuchsin sulfurous acid" technique. Am J
Anat 90:167-216
2. Neimi M, Kormano M 1965 Cyclical changes and significance of lipid and acid phosphatase activity in the seminiferous tubules of the rat testis. Anat Rec 151:159-170
3. Kerr JB, DeKretser DM 1975 Cyclic variations in Sertoli
cell lipid throughout the spermatogenic cycle in the rat. J
Reprod Fertil 43:1-8
4. Parvinen M, Ruokonen A 1982 Endogenous steroids in
rat seminiferous tubules. Comparison of different sper-
mRNA in Sertoli and Germinal Cells
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
matogenic stages isolated by transillumination-assisted
microdissection. J Androl 3:211-220
Hilscher B, Passia D, Hilscher W 1979 Kinetics of the
enzymatic pattern in the testis. I. Stage dependence of
enzymatic activity and its relation to cellular interactions
in the testis of the Wistar rat. Andrologia 11:169-181
Morales C, Clermont Y, Nadler NJ 1986 Cyclic endocytic
activity and kinetics of lysosomes in Sertoli cells of the
rat: a morphometric analysis. Biol Reprod 34:207-218
Ritzen EM, Boitani C, Parvinen M, French FC, Feldman
M 1982 Stage-dependent secretion of ABP by rat seminiferous tubules. Mol Cell Endocrinol 25:25-33
Lacroix M, Parvinen M, Fritz IB 1981 Localization of
plasminogen activator in discrete portion (stages VII and
VIII) of the seminiferous tubules. Biol Reprod 25:143-146
Wright WW, Parvinen M, Musto NA, Gunsalus GL, Phillips
DM, Mather JP, Bardin CW 1983 Identification of stagespecific proteins synthesized by rat seminiferous tubules.
Biol Reprod 29:257-270
Griswold MD, Collard M, Hugly S, Huggenvik J 1986 The
use of specific cDNA probes to assay Sertoli cell functions. In: Dhindsa 0, Bahl O (eds) Molecular and Cellular
Aspects of Reproduction. Plenum Press, New York, pp
301-317
Huggenvik Jl, Idzerda RL, Haywood L, Lee DC, McKnight
GS, Griswold MD 1987 Transferrin mRNA: molecular
cloning and hormonal regulation in rat Sertoli cells. Endocrinology 120:332-340
Collard MW, Griswold MD 1987 Biosynthesis and molecular cloning of sulfated glycoprotein-2 secreted by rat
Sertoli cells. Biochemistry 26:3297-3303
Collard M, Sylvester SR, Tsuruta JK, Griswold MD 1988
Biosynthesis and molecular cloning of sulfated glycoprotein-1 secreted by rat Sertoli cells: sequence similarity
with the 70-kilodalton precursor to sulfatide/GMi activator.
Biochem 27:4557-4564
Morales C, Hugly S, Griswold M 1987 Stage-dependent
levels of specific mRNA transcripts in Sertoli cells. Biol
Reprod 36:1035-1046
Morales C, Sylvester SR, Griswold MD 1987 Transport
of iron and transferrin synthesis by the seminiferous epithelium of the rat in vivo. Biol Reprod 37:995-1005
Kleene KC, Distel RJ, Hecht NB 1985 The nucleotide
sequence of a cDNA clone encoding mouse protamine I.
Biochemistry 24:719-722
733
17. Yelick PC, Balhorn R, Johnson PA, Corzett M, Mazrimas
JA, Kleene KC, Hecht NB 1987 Mouse protamine 2 is
synthesized as a precursor whereas mouse protamine 1
is not. Mol Cell Biol 7:2173-2179
18. Kleene KC, Bozorgzadeh A, Flynn JF, Yelick PC, Hecht
NB 1988 Nucletoide sequence of a cDNA clone encoding
mouse transition protein 1. Biochim Biophys Acta
950:215-220
19. Kleene KC, Distel RJ, Hecht NB 1984 Translational regulation and coordinate deadenylation of a haploid mRNA
during spermiogenesis in the mouse. Dev Biol 105:71 -79
20. Hecht NB, Bower PA, Waters SH, Yelick PC, Distel RJ
1986 Evidence for haploid expression of mouse testicular
genes. Exp Cell Res 164:183-190
21. Morales C, Griswold MD 1987 Retinol-induced stage synchronization in seminiferous tubules of the rat. Endocrinology 121:432-434
22. Griswold MD, Morales C, Sylvester S 1988 Molecular
biology of the Sertoli cell. Oxf Rev Reprod Biol 10:124161
23. Huang HFS, Hembree WC 1979 Spermatogenic response
to vitamin A in vitamin A deficient rats. Biol Reprod
21:891-904
24. Mason KE 1933 Differences in testes injury and repair
after vitamin A deficiency, vitamin E deficiency and inanition. Am J Anat 52:153-239
25. Wing TY, Christensen AK 1982 Morphometric studies on
rat seminiferous tubules. Am J Anat 165:13-25
26. Bustos-Obregon E 1970 On Sertoli cell number and distribution in rat testis. Arc Biol 81:99-108
27. Mali P, Sandberg M, Vuorio E, Yelick PC, Hecht NB,
Parvinen M 1988 Localization of protamine 1 mRNA in
different stages of the cycle of the rat seminiferous epithelium. J Cell Biol 107:407-412
28. Guelin M, Kejzlarova-Lepesant J, Lepesant JA 1985 In
situ hybridization: a routine method for parallel localization
of cDNA sequences and their transcripts in consecutive
paraffin sections with the use of 3H-labeled nick translated
cloned DNA probes. Biol Cell 53:1-12
29. Kopriwa BM, Leblond CP 1962 Improvements in the
coating technique of radioautography. J Histochem Cytochem 10:269-284
30. Steinberger A, Steinberger E 1970 In vivo growth and
development of mammalian testes. In: Johnson AD,
Gomes WR, Van Denmark VL (eds) The Testis. Academic
Press, New York, vol 11:363-391