|
|
|
|||||||
|
|||||||||
(Received for publication, September 1,
1995; and in revised form, October 31, 1995) From the
H1t is synthesized in mid to late pachytene spermatocytes of the
male germ line and is the only tissue-specific member of the mammalian
H1 histone family. As a step toward identifying DNA sequences that
confer its tissue-specific expression, we have produced transgenic mice
containing the intact rat H1t gene as well as a H1t-lacZ
fusion gene. Transgenic mice carrying a 6.8-kilobase fragment of rat
genomic DNA encompassing the H1t gene expressed rat H1t at high levels
in the testis and in no other organ examined. H1t fragments truncated
to within 141 base pairs (bp) of the gene in the 5` direction or within
837 bp in the 3` direction retained testis specificity. Expression of
rat H1t protein was also evident in the testes of the transgenic mice,
and in some lines the level of rat H1t exceeded that of the mouse
protein. The stage of spermatogenesis of transgene expression was
assessed by following appearance of transgenic mRNA in developing mice
and by immunohistochemistry using an antiserum to rat H1t. In lines
from three different constructs, expression was restricted to germinal
cells, although in two strongly expressing lines the transgenes were
expressed somewhat prematurely in preleptotene spermatocytes. An
H1t(-948/+71)-lacZ fusion was also expressed
specifically in the spermatocytes and round spermatids of a transgenic
line, confirming that sequences sufficient for correct tissue and
developmental expression lie within this 1,019-bp segment of the gene.
Spermatogenesis involves a complex developmental program in
which relatively undifferentiated cells of the male germ lineage
transform into spermatozoa. A substantial list of testis-specific genes
that are expressed in a particular temporal order during mammalian
spermatogenesis is known (reviewed in Wolgemuth and Watrin(1991) and
Hecht(1993)), and a current challenge is to understand the factors
responsible for this pattern of gene regulation. The testis-specific
histone variants are particularly interesting because they belong to
families that each have many members expressed in somatic cells. As the
fundamental packaging proteins of chromatin (Wolffe, 1992), the five
histone classes are found in all tissues. However, in mammals the male
germ line is unique in having specific histone variants not expressed
elsewhere. These include the variants H1t (Seyedin and Kistler, 1980;
Cole et al., 1986), H2A (Trostle-Weige et al., 1982;
Huh et al., 1991), and H2B (Shires et al., 1976; Kim et al., 1987), which have been characterized both as proteins
and cloned genes, whereas the variant TH3 has been identified at the
protein level (Trostle-Weige et al., 1984). An H4 gene linked
to the H1t gene and also expressed at high levels in the testis has
been designated H4t, although it encodes the standard somatic form of
the H4 protein (Grimes et al., 1987). In rats these testis
histones are expressed according to different schedules, with TH3
appearing in late spermatogonia, TH2A and TH2B appearing early in cells
undergoing meiosis (spermatocytes), and H1t appearing only later in mid
to late pachytene spermatocytes (Bucci et al., 1982; Meistrich et al., 1985; Kremer and Kistler, 1991). These variants
replace their somatic counterparts to varying degrees and remain in
developing germ cells until the haploid nucleus elongates and
condenses, when they disappear and are replaced by
spermatid/sperm-specific nuclear proteins. Unlike the situation in
various invertebrates such as sea urchins (Poccia, 1986), the
testis-specific histone variants of mammals are not retained in the
nuclei of mature sperm. H1 histones are believed to be responsible
for condensing adjacent nucleosomes into the compact, solenoidal, 30-nm
chromatin fiber. H1t, while bearing an unmistakable resemblance to
members of the common somatic H1 family, is sufficiently different in
amino acid sequence to constitute a recognizable variant class
(Kistler, 1989), and a recent study suggests that H1t is unable to
bring about a condensed chromatin state (De Lucia et al.,
1994). Isolation of the H1t gene (Cole et al., 1986; Grimes et al., 1987; Drabent et al., 1991) revealed that its
immediate 5`-flanking region was extremely similar to the promoter
regions for standard H1 variants expressed in somatic cells. This
similarity includes four nucleotide regions found just upstream of all
standard H1 genes (Coles and Wells, 1985; Heintz, 1991). A study of the
role of H1t upstream sequences fused to a reporter gene in a
transfected cell line revealed some variation in expression depending
on the length of upstream sequences present but could not address the
issue of testis specificity (Kremer and Kistler, 1992). Transgenic
mice have proved useful in defining DNA regions necessary for testis
expression of such genes as protamine 1 (Peschon et al., 1987;
Zambrowicz et al., 1993), protamine 2 (Stewart et
al., 1988), the testis-specific form of angiotensin-converting
enzyme (Langford et al., 1991), phosphoglycerate kinase-2
(Robinson et al., 1989), and proenkephalin (Galcheva-Gargova et al., 1993). The present report describes analysis of the
expression of the rat H1t gene in transgenic mice in order to establish
the boundaries of the DNA region necessary for developmental and
tissue-specific expression. Using genomic fragments containing the
natural gene, we found that constructs with as little as 141 bp (
Figure 1:
Diagram of cloned rat genomic DNA
fragments used as transgenes and probes. Nucleotides are numbered
relative to the transcriptional initiation point. TG1
(-2383/+4488) is bounded by EcoRI sites. TG2
(-948/+4488) is truncated upstream at the single PvuII site. TG3 (-141/+4488) is truncated upstream
at a PstI site. TG4 (-2149/+1578) is truncated
upstream at a SacI site and downstream at a StuI
site. TG-Lac consists of the -948/+71 fragment of H1t fused
to the lacZ expression module of placI as described under
``Experimental Procedures.'' Probe A (PstI
-141/PstI +90) includes the 5`-UTR as well as the
promoter region. Probe B (PstI +91/SalI
+804) includes all of the gene except for the
5`-UTR.
The number of copies of integrated transgenes in members of
established lines was estimated by immobilizing 5 µg of genomic DNA
as dots on nitrocellulose (Schleicher & Schuell, BA85) and
hybridization to randomly primed
For S1
nuclease mapping, the 231-bp PstI fragment
(-141/+90) that overlaps the cap site (see Fig. 1, probe A) was ligated to the PstI site of M13 mp18,
and a single stranded uniformly
Figure 2:
Transgene expression analyzed by Northern
blots. RNA samples were prepared, separated, and blotted to nylon
membranes as described under ``Experimental Procedures.''
Blots were hybridized to
Figure 3:
Transgene expression analyzed by S1
nuclease digestion. RNA samples were prepared and hybridized to a
uniformly
Encouraged by this result, we truncated this fragment
about 1 kb upstream of the gene at a PvuII site to yield the
TG2 construct. Three founder animals were identified for this
construct, of which two (TG2-97 and TG2-99) were bred to generate
lines. Each TG2 line was found to display testis-specific expression of
the transgene by both S1 assays (not shown) and Northern blots (see Fig. 2, TG2-99). We then truncated the upstream
region at a PstI site 141 bp from the cap site (Fig. 1, TG3) and also investigated the region downstream of the gene
by shortening a fourth construct at a StuI site at +1578
relative to the cap site (Fig. 1, TG4). Five founder
animals were identified for each of these constructs (see Table 1). Three of the TG3 constructs passed the transgene and
generated lines (TG3-4301, TG3-5592, and TG3-5597), whereas two of the
founders failed to generate lines. Testis-specific expression of the
transgene was determined by Northern blotting for all three lines, for
example TG3-4301 (Fig. 2). Lines were derived from all five of
the TG4 founders. Four of them gave testis-specific expression detected
either by S1 analysis or Northern blotting, for example TG4-4616 (Fig. 2). The fifth TG4 founder, TG4-308, was a male that passed
the transgene to each of 15 daughters but to none of 18 sons,
suggesting that the transgene had integrated on the X chromosome. His
daughters passed the transgene to approximately 50% of their sons, as
expected, but none of those tested expressed rat H1t. This may be
explained by the general inactivity of the X chromosome during meiosis
in male mammals (Gordon and Ruddle, 1981).
Figure 4:
Transgene expression in the testes of
animals from different lines analyzed by Northern blotting. RNA was
prepared, separated, and hybridized to uniformly
Mouse and rat H1t could be resolved
from one another by SDS gel electrophoresis, and it was therefore
possible to examine the relative concentrations of H1t protein
generated by both the endogenous mouse gene and the transgene. Protein
samples were prepared from seven of the transgenic lines and examined
by SDS-polyacrylamide gel electrophoresis (Fig. 5). In general,
those lines with a high level of transgene mRNA also expressed high
levels of rat H1t protein with a concomitant reduction of endogenous
mouse H1t. In some cases the level of rat H1t exceeded that of mouse
H1t (Fig. 5, TG1-688, TG2-97, and TG2-99). Line TG4-296, which did not show transgene expression
by Northern blot, did show a faint level of transgene protein (Fig. 5). TG1-555 was the only exception to the generalization
regarding protein expression, because this line had a high level of rat
mRNA (Fig. 3, lane 3) but only a low level of protein
expression (Fig. 5). This discrepancy remains unexplained.
Figure 5:
Transgene expression in the testes of
animals from different lines analyzed by SDS-polyacrylamide gel
electrophoresis. A protein fraction consisting primarily of H1 histones
was prepared from the testes of individual animals by differential
trichloroacetic acid precipitation and stepwise elution from BioRex 70
as described under ``Experimental Procedures.'' This fraction
(about 40 µg for the majority of samples) was separated on a 15%
SDS gel and stained with Coomassie Blue. The lanes for rat and mouse
show samples from control animals. Rat H1t, which migrates slightly
more rapidly than mouse H1t, is indicated by the pointer where present.
Protein extracts were prepared from different animals than used for RNA
extraction (lines TG1 and TG2) or from one testis, whereas the other
was used for RNA extraction (lines TG3 and
TG4).
Figure 6:
Developmental analysis of transgene
expression by Northern blotting. A, RNA samples were prepared
from the testes of animals at various times after birth, separated
electrophoretically, blotted to nylon membranes, and hybridized with
the
Although
the precise location of transgene expression might best be determined
by in situ hybridization, an initial attempt to accomplish
this with a
Figure 7:
H1t
expression in testes of normal and transgenic animals detected by
immunohistochemistry. Testes were fixed in Bouin's fluid, and
routine paraffin-embedded sections were prepared. These were treated
with a polyclonal antiserum raised to rat H1t, and the immune complexes
were visualized by exposure to a peroxidase-conjugated second antibody
and diaminobenzidine, as described under ``Experimental
Procedures.'' The sections shown were not counterstained. A, low power view of nontransgenic mouse testis. The antiserum
reacts with nuclei of mid to late pachytene spermatocytes (pc)
as well as nuclei of round (rt) and elongating (et)
spermatids. No reaction is seen over spermatogonia, early
spermatocytes, elongated spermatids, or interstitial cells outside the
tubules. B, low power view of transgenic testis from line
TG1-688. Note that in contrast to A, nuclei are prominently
labeled in some tubules around the periphery of the tubules. C, control slide in which the primary antiserum was omitted. D, E, and F are higher power views of
sections from TG1-688. D, a tubule with unlabeled
spermatogonia (sg), unlabeled early pachytene spermatocytes (pc), and heavily labeled round spermatids (rt). E, a tubule slightly more advanced than that of D, in
which the spermatogonia have largely progressed to preleptotene
spermatocytes (plc). The pachytene spermatocytes (pc)
remain unlabeled, whereas round spermatids (rt) are heavily
labeled. F, leptotene spermatocytes (lc) around the
periphery of the tubules are labeled, whereas late pachytene
spermatocytes (pc) are labeled as well as the elongating
spermatids (et) found in tubules of this stage. Bar,
50 µm.
Although the
antiserum does not distinguish between rat and mouse H1t, the most
straightforward interpretation of the immunological results and the
developmental Northern blot (Fig. 6) is that the transgene is
expressed in this line significantly earlier in germ cell development
than the endogenous gene. The loss of immunoreactivity in the zygotene
and early pachytene cells is a mystery. It could reflect a temporary
masking of antigenic determinants or destabilization of the association
of H1t with chromatin and associated degradation of the protein. This
same pattern of immunoreactivity was observed in the highly expressing
TG4-4616 line, whereas the lower expressing TG3-4301 line had a pattern
of immunoreactivity indistinguishable from the nontransgenic mouse
testis (results not shown).
Figure 8:
Expression of lacZ transgene in
seminiferous tubule whole mounts. Teased testis tubules from a mouse of
the TG-LacZ-5371 line (A) or from a nontransgenic control
mouse (B) were stained for 2 days to detect
Figure 9:
To determine the site of
transgene expression within the testes of the 5371 line, histological
cross sections were prepared. Staining was observed over pachytene
spermatocytes and round spermatids but not over other cell populations
in the tubules (Fig. 10), mimicking the location of H1t itself
(compare Fig. 7). It is puzzling that the staining of
spermatocytes and round spermatids was quite variable, with some cells
staining darkly and others in the same cross section remaining
virtually unstained (Fig. 10). This pattern was observed in each
of three animals examined from this line.
Figure 10:
Expression of lacZ transgene in
seminiferous tubule cross sections. Teased testis tubules from a mouse
of the TG-LacZ-5371 line (A and C) or from a
nontransgenic control mouse (B and D) were stained
for 2 days to detect
Whereas all of the genomic fragments yielded testis-specific
expression, the levels of expression in the most 5`- and 3`-truncated
fragments (TG3 and TG4) were generally not so high as with the
full-length fragment (TG1). An earlier study in somatic cells
transfected with the H1t 5`-flanking region fused to the
chloramphenicol acetyltransferase gene indicated that stimulatory
sequences lie between -368 and -693 (Kremer and Kistler,
1992). The lower expression of two of the TG3 lines (TG3-5592 and
TG3-5597) compared with the TG1 and TG2 lines agrees with this
conclusion, although expression from the remaining TG3 line (TG3-4301)
was stronger and comparable with the TG2 lines. The most variable
levels of expression were found with TG4 lines, which deleted
downstream sequences past +1578. The H4t gene lies within the
region deleted (Fig. 1), and it is plausible to suppose that
sequences neighboring H4t could have a stimulatory effect on both the
H1t and H4t genes. Although the H4t gene is expressed in late
pachytene spermatocytes, it is also expressed in the brain and to a
lesser extent in liver (Wolfe et al., 1989). Accordingly, in
the case of H4t, expression in spermatocytes is not coupled to complete
repression in somatic organs. Because the H4t gene encodes a protein
with the same amino acid sequence as somatic H4 (Grimes et
al., 1987), its expression would not have an obvious functional
consequence in somatic cells. Thus, controls on its somatic expression
may not be as stringent as for the other germ cell-specific histone
variants.
The possible
role of distant sequences on the expression of standard H1 genes is not
well explored. However, participation of distant upstream sequences is
documented for expression of the mouse H1
Several explanations for the early expression of
the transgene can be considered. First, it is possible that an
additional cis-acting DNA locus that lies outside the 6.8-kb region we
have investigated is required to fine tune H1t appearance. This seems
unlikely because one of the the three lines did not show premature
transgene expression. A second possibility is that the natural gene
also becomes active in late spermatogonia/early spermatocytes but at
too low a level to be detected by the techniques applied. The multicopy
transgene might simply yield a detectable signal coincident with the
earliest activation of the gene through a gene dosage effect. A third
explanation is that having many copies of the gene, all integrated in
tandom (as is the general case for transgenes (Jaenisch, 1988) but has
not been investigated in the lines under discussion), leads to an
enhanced probability of expression when necessary factors are limiting,
perhaps because the effect of the many tandem repeats is to build a
higher local concentration of the appropriate factors than would
otherwise occur. Although this might lead to a nonproductive
distribution of factors over different promoters, it might also lead
eventually to a productive clustering on one or more of the tandom
repeats. A fourth explanation is that a negative-acting factor in
limiting concentration is titrated out by multiple transgene copies. At
the present time we do not have any reason to support one of these
possibilities more strongly than the others. It is curious, however,
that the various testis-specific histone variants appear at slightly
different developmental times, TH3 in spermatogonia, TH2A and TH2B in
early spermatocytes, and H1t and H4t in late spermatocytes (Meistrich et al., 1985; Kim et al., 1987; Kremer and Kistler,
1991; Grimes et al., 1987). It is unknown whether any of the
factors that determine male germ cell-specific expression are shared
among these genes, but if so, there could be some sort of titration of
DNA response elements during germ cell development, and the presence of
multicopies may somehow change the outcome of this titration. It may be
relevant that a somewhat comparable situation has been described for
the transgenic expression of the human keratin-1 gene, which is
normally expressed only in the suprabasal layers of the skin. In
transgenic mice, expression of human keratin-1 was restricted to the
skin, but in contrast to the natural gene, the transgene was expressed
prematurely in the mitotically dividing basal cells (Rosenthal et
al., 1991).
In
summary, results presented here indicate that the sequence elements
conferring spermatocyte-specific expression to H1t are restricted to at
most 1 kb of 5`-flanking sequence and may lie within as little as 212
bp. We are thus encouraged to look to this region for the DNA sequences
responsible and their associated binding factors.
Note Added in
Proof-Recently vanWert et al.(1995) described
testis-specific expression of rat H1t in transgenic mice carrying a
genomic fragment identical to TG1 (Fig. 1).
Volume 271,
Number 8,
Issue of February 23, 1996 pp. 4046-4054
©1996 by The American Society for Biochemistry and Molecular Biology, Inc.
ONE KILOBASE OF 5`-FLANKING SEQUENCE MEDIATES CORRECT EXPRESSION OF
A lacZ FUSION GENE (*)
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
)of upstream or 800 bp of downstream sequence result in
testis-specific expression, although elements affecting the level of
expression may lie outside this region. A construct with 0.95 kb of
upstream sequence fused to the Escherichia coli lacZ gene also
expressed uniquely within the germ cells of the testis. Interestingly,
some transgenic mice expressed the rat gene prematurely during
spermatogenesis, and in some lines the level of expression of the
transgene exceeded that of the endogenous H1t gene.
Constructs and Transgenic Mice
Constructs were
derived from a 6.8-kb EcoRI genomic fragment containing the
rat H1t gene (Cole et al., 1986; Grimes et al., 1990)
using standard cloning techniques (Sambrook et al., 1989) and
are diagrammed in Fig. 1. The TG1 and TG2 fragments were excised
directly from this clone, whereas the TG3 and TG4 fragments each
required sequential transfer of selected restriction fragments to
pBluescript SK(+) (Stratagene, La Jolla, CA). For the H1t-lacZ fusion construct, the +70Tth111I site in the H1t
5`-UTR was blunted with Klenow DNA polymerase and joined to an NcoI linker (5`-AGCCATGGCT-3`). A fragment running from this NcoI site upstream to the -948 PvuII site was
then joined to placI, obtained from R. Palmiter. This promoterless lac
expression vector has an NcoI site at the initiating ATG and
is identical to placF (Mercer et al., 1991) except that the
intron and poly(A) regions downstream from lacZ were provided
by the mouse metallothionein II gene fragment from BamHI
(+350) to PstI (+840) (Searle et al.,
1984). Fragments for injection were excised from plasmid vectors,
resolved by agarose gel electrophoresis, transferred
electrophoretically to pieces of NA45 membrane (Schleicher &
Schuell), eluted in 1.5 M NaCl, extracted with n-butanol, and ethanol precipitated. In some cases DNA
fragments were recovered from agarose gels using glass beads (Qiaex,
QIAGEN, Chatsworth, CA) as described by the manufacturer. DNA was then
dissolved in sterile 0.1 mM EDTA, 10 mM Tris-HCl (pH
8.0) at about 2 µg/ml and injected into the pronuclei of (C57BL/6
DBA/2)F2 embryos (Brinster et al., 1985). Transgenic
founder animals were identified by Southern blotting of RsaI-digested DNA obtained from tail samples (Hogan et
al., 1986). For natural gene constructs these blots were
hybridized to a P-labeled PstI-SalI fragment that includes most of the
rat gene and cross-hybridizes to mouse H1t (see Fig. 1, probe B). Transgenic founders were subsequently mated to pure
strain C57BL/6 mice; positive offspring from this cross were likewise
mated and the line thus propagated through successive generations.
P-labeled probe A (see Fig. 1). Dots were excised and counted in a scintillation
counter, and cpm hybridizing to the transgene was calculated by
subtracting cpm hybridizing to control mouse DNA. The cpm hybridizing
to rat genomic DNA were used as a standard, assuming two copies of the
gene per diploid genome.
RNA Analysis
RNA was extracted from tissues by the
method of Chomczynski and Sacchi(1987). For Northern blotting, 20
µg of total RNA was analyzed on a 1.5% agarose/formaldehyde gel
containing 0.5 µg/ml ethidium bromide (Davis et al.,
1986). Uniformity of RNA loading was checked under UV illumination. RNA
was transferred to Hybond N membranes (Amersham Corp.), immobilized by
exposure to UV light, and probed with a P-labeled 231-bp PstI restriction fragment that contains the entire 69 bp of
the 5`-UTR and 21 bp of the coding region of rat H1t mRNA (see Fig. 1, probe A). Hybridization was at 42 °C in
0.2% bovine serum albumin, 0.2% Ficoll, 0.2% polyvinylpyrolidone, 10%
dextran sulfate, 1% SDS, 100 µg/ml sheared denatured calf thymus
DNA, 5 mM EDTA, 50 mM NaH
PO
(pH 7.2), 0.9 M NaCl, 50% formamide. Final washes were
in 0.1% SDS, 75 mM NaCl, 0.5 mM EDTA, 5 mM Tris-HCl (pH 8) at 68 °C. Membranes were exposed to Kodak
X-Omat AR film in the presence of an intensifying screen.P-labeled probe was
generated using Klenow DNA polymerase I with the universal M13 forward
sequencing primer, followed by cleavage with EcoRI and
isolation of the probe by elution from a denaturing polyacrylamide gel.
RNA (10 µg) was incubated with 10
cpm of probe,
hybridized, and digested with S1 nuclease as described (Berk, 1989).
The protected fragment was analyzed on a 8% sequencing type gel, which
was dried and exposed to film.Extraction and Analysis of H1t
A mouse testis was
homogenized in 5 ml of cold 0.2 M HSO,
and then brought to 3% trichloroacetic acid by addition of 0.5 ml of
33% trichloroacetic acid. After 15 min on ice, the mixture was
centrifuged 10,000 g for 15 min, and volume of 100%
trichloroacetic acid was added to the supernatant. After 15 min on ice
and centrifugation as before, the supernatant was discarded and the
pellet was rinsed twice with ethanol:ether (1:1) and allowed to dry.
The H1 proteins were dissolved in 10 mM acetic acid at 1 ml/g
of tissue extracted. This extract also contains high mobility group
(HMG) proteins, of which HMG 1 and 2 migrate in the vicinity of rat H1t
on SDS-polyacrylamide gel electrophoresis. To eliminate HMG
contamination, the H1 extract was diluted to 1 ml with 0.1 M NaHPO (pH 7.3) and applied to a 200-µl
column of BioRex 70 (200-400 mesh, Bio-Rad) equilibrated with the
same buffer in a 1-ml disposable syringe. HMGs were eluted from the
column by 1 ml of 7% guanidine HCl, 0.1 M NaHPO (pH 7.3), and the H1 fraction was
then eluted by 300 µl of 15% guanidine HCl, 0.1 M NaHPO (pH 7.3). The H1 fraction was
desalted with a 2-ml column of Sephadex G-25 medium, developed with 10
mM acetic acid, and lyophilized. H1 proteins were analyzed on
a 15% SDS-polyacrylamide gel prepared as described by Laemmli(1970)
except that the stacking gel was made with the same buffer as the
resolving gel. This modification enhanced the separation of mouse and
rat H1t. Protein bands were visualized by staining with Coomassie
Brilliant Blue R-250.Immunohistochemistry
Testes were fixed by
immersion in Bouin's solution (glacial acetic acid, formalin,
saturated picric acid, 1:5:15), embedded in paraffin, sectioned,
mounted, and hydrated by standard methods. Slides were immersed
successively for 5 min each in 1% LiCO
/70% ethanol, 1%
HO/70% ethanol, and 0.3 M glycine.
Using a humidified chamber, sections were then incubated in TBS (150
mM NaCl, 20 mM Tris-HCl, pH 7.4) containing 10%
normal goat serum for 15 min followed by three washes in TBST (TBS
containing 0.1% Tween 20). Rabbit anti-rat H1t ()was diluted
1:100 in TBS containing 10% goat serum and 1 mM EDTA and
incubated at 37 °C for 1.5 h. Slides were then washed three times
in TBST. Peroxidase-conjugated goat anti-rabbit IgG (Bio-Rad) diluted
1:1000 in TBS containing 10% goat serum was applied and incubated at 37
°C for 30 min. Slides were washed twice in TBST and once in TBS
before developing with 0.06% diaminobenzidine, 0.03%
HO, 50 mM Tris-HCl (pH 7.6). In some
cases slides were developed with the Immunopure Metal Enhanced DAB
Substrate Kit (Pierce) and stained lightly with Gill's
hematoxylin (Fisher). Histochemical staining for 
-galactosidase
was done essentially as described by Behringer et al.(1993)
using decapsulated testes with the tubules slightly teased apart to
insure complete penetration by the chromogenic substrate. The use of
detergents at the indicated concentrations (Behringer et al.,
1993) largely suppressed endogenous -galactosidase activity
characteristic of interstitial cells.
Tissue extracts were
prepared by disrupting tissues with a glass/glass homogenizer in 5 vol
of 100 mM Na-Galactosidase AssaysHPO (pH 7.3), 0.1 mM MgCl, 2 mM MgSO, 40 mM 2-mercaptoethanol, 0.2 mM phenylmethylsulfonyl fluoride,
0.005% leupeptin, and 0.1% Triton X-100. Extracts were clarified by
centrifugation at 10,000 g for 15 min at 4 °C.
Assays were performed as described (Sambrook et al., 1989)
except they were monitored continuously with a recording
spectrophotometer at 420 nm using an extinction coefficient for o-nitrophenol of 10,650 cm
mol to compensate for incomplete dissociation of the reaction product
at pH 7.3. Protein was determined by the biuret reaction.
Overview of Natural Transgene Constructs and Their
Expression
It was possible to use the entire rat H1t gene in
transgenic experiments because of its small size and because the mRNA
and protein encoded can each be distinguished from the products of the
endogenous mouse H1t gene. For the initial experiment, we used a 6.8-kb EcoRI genomic clone containing 2.4 kb upstream of the
transcriptional initiation site and 3.8 kb of sequence downstream from
the 3` end of the gene (see Fig. 1for diagram of constructs).
The downstream region of this fragment (Fig. 1, TG1)
also contains the H4t gene (Grimes et al., 1987, 1990).
Potential transgenic animals were screened by Southern analysis of tail
biopsy DNA using a 713-bp PstI-SalI restriction
fragment that hybridizes to both rat and mouse H1t genes (Fig. 1, probe B) and identifies a characteristic
1.4-kb RsaI fragment of the rat gene. Two founder animals were
identified for this construct, TG1-555 and TG1-688, and each was bred
to develop lines. Testis RNA from each line was assayed for the
presence of rat-specific H1t mRNA by both S1 nuclease assays and by
Northern blotting. Both lines were found to express the transgene
prominently in an organ-specific fashion by either assay. Fig. 2shows testis-specific expression of the TG1-688 line by
Northern blot, and Fig. 3shows testis-specific expression of
the TG1-555 line by S1 assay. Thus, by either assay strong signals were
evident in RNA samples from transgenic testis and undetectable with RNA
from spleen, small intestine, heart, liver, lung, kidney, or brain. The
S1 assay verified that the transgene mRNA in both TG1 lines protected
the same probe fragment as normal rat testis H1t mRNA, indicating that
transcriptional initiation was occurring correctly (Fig. 3). In
addition, the sensitivity of the S1 assay emphasized that leaky
expression was very low or nonexistent in the somatic organs tested.
Data for these and other transgenic animals are summarized in Table 1.
P-labeled probe A (Fig. 1), which is specific for the 5`-UTR of rat H1t mRNA. The first four lanes contain samples from nontransgenic mice (lanes 1 and 2) and rats (lanes 3 and 4) to demonstrate the specificity of the probe for rat H1t. Lanes 5-12 contain samples from tissues of individual
male animals from the indicated transgenic
lines.
P-labeled single stranded probe overlapping the
5` end of the rat H1t mRNA, digested with S1 nuclease, and resolved on
a denaturing polyacrylamide gel as described under ``Experimental
Procedures.'' Correctly initiated rat H1t mRNA protects 90 bp of
this probe. Lane 1, end-labeled HinfI digest of
pBR322. Lane 2, 10,000 cpm of undigested probe. RNA from
nontransgenic mouse organs (lanes 3 and 4) and rat
organs (lanes 5 and 6) demonstrate the specificity of
the probe. (The full-length probe remaining in the rat liver sample (lane 5) was not reproducible and was apparently due to
incomplete digestion of this sample.) Lanes 7-13 contain
RNA from organs of a single male animal of the TG1-555
line.
Comparative Expression of the Rat H1t Transgene in Testes
of Different Lines
Among different lines the level of transgene
expression was quite variable and presumably reflects the influence of
adjacent sequences in different chromosomal integration sites. However,
the tissue specificity of transgene expression was not influenced by
integration site because the expression of all H1t constructs was
limited to the testis. To compare the relative level of transgene
expression among the lines developed from our constructs we prepared a
Northern blot from representatives of the four families of transgenic
lines (Fig. 4). Control RNA samples from nontransgenic mouse and
rat testis served to verify specificity of the probe (Fig. 4, lanes 1 and 2). Both TG1 lines (Fig. 4, lanes 3 and 4) had higher levels of rat H1t mRNA in
testicular RNA samples than did rat testis RNA itself. The single TG2
line analyzed (Fig. 4, lane 5) had somewhat lower
levels of H1t mRNA, comparable with those seen in the rat testis lane.
Expression among the three TG3 constructs was somewhat variable (Fig. 4, lanes 6-8), ranging from about the same
as rat testis to somewhat lower. Expression among the five TG4 lines
was the most variable (Fig. 4, lanes 9-13), with
expression ranging from undetectable (lanes 12 and 13) to very high and comparable with the TG1 lines (lane
11). Although TG4-296 expression was not detected by this Northern
blot, a low level of expression was seen by S1 nuclease assay (not
shown). Expression levels among these various constructs was not
strictly tied to copy number of the transgene, but the general pattern
was for higher copy number lines to have higher expression,
particularly when comparing members of the same construct. Thus
TG3-4301 (34 copies) was stronger than TG3-5592 (9 copies) or TG3-5597
(11 copies). Among the TG4 lines, TG4-4616 (30 copies) was by far the
strongest expressor, whereas TG4-286 (20 copies) and TG4-4605 (9
copies) were substantially weaker, although TG4-4604 (9 copies) and
TG4-296 (7 copies) were weaker still. As mentioned above, the lack of
detectable expression for TG4-308 (2 copies) may have been due to its
location on the X chromosome.
P-labeled
probe A (Fig. 1) as described under ``Experimental
Procedures.'' The specificity of the probe is shown by its
selective hybridization to rat testis (lane 2) but not mouse
testis (lane 1) RNA. The remaining lanes show results from RNA
samples from the testes of individual male animals of the indicated
transgenic lines. Ribosomal RNA bands stained with ethidium bromide (lower panel) demonstrate lack of degradation and comparable
loading of the various lanes.
Determination of the Developmental Expression of the
Transgenes during Spermatogenesis
The previously described
results documented the organ specificity of transgene expression.
However, those results could not establish that the transgene was
expressing in the correct population of germinal cells, mid to late
pachytene spermatocytes (Meistrich et al., 1985; Grimes et
al., 1990; Kremer and Kistler, 1991). The time when germ cells in
the immature mouse first reach particular stages of development is
known (Bellvéet al., 1977). This makes
it possible to assign initial transgene expression tentatively to a
particular point in spermatogenesis by determining the earliest
transgene expression in developing animals. RNA was extracted from
testes of control mice and from those of the TG1-688 line at 7, 12, and
19 days after birth. A Northern blot was prepared and probed initially
with the -141/+90 rat-specific probe A (Fig. 6A). This blot showed that the transgene was
expressing even in 7-day-old mice, although the first pachytene
spermatocytes are not normally seen in the mouse testis until animals
are 14 days old (Bellvéet al., 1977).
The blot was stripped and rehybridized to the +91/+804
nonspecific H1t probe B in order to show the onset of H1t expression in
the control mouse testes (Fig. 6B). Although a more
intense signal was detectable over the transgenic lanes, the control
mouse showed no endogenous H1t expression in the 7- or 12-day-old
samples, with the first expression occurring in the 19-day-old testes.
This result suggested that the transgene in the TG1-688 line was
expressing significantly earlier than the endogenous gene.
P-labeled -141/+90 rat-specific probe A. B, after stripping, the blot was rehybridized to
P-labeled +90/+804 nonselective H1t probe
B.
S-riboprobe made from the -141/+90
rat H1t fragment was unsuccessful, perhaps because of the relatively
short 90-bp hybridization target. Accordingly we decided to make use of
an H1t-specific polyclonal antiserum that reacts with both mouse and
rat H1t to test for the presence of H1t protein in early germ cells.
With control adult mouse testis, this antiserum identified the nuclei
of late pachytene spermatocytes, round spermatids, and early elongating
spermatids, but no reaction was seen over the nuclei of early germ
cells, elongated spermatids, or somatic cells of the testis (Fig. 7A). In striking contrast, testis sections of a
mouse homozygous for the TG1-688 transgene showed an additional set of
positively staining nuclei superimposed on those observed in the
control mouse. Some tubules exhibited a distinctive pattern of staining
of round nuclei in cells about the periphery of the tubule, which was
never observed with material from nontransgenic animals (Fig. 7B). At higher magnification, it could be
determined that early spermatogonia were generally unreactive with the
antiserum (Fig. 7D) but that the earliest meiotic
cells, preleptotene and leptotene spermatocytes, reacted very strongly (Fig. 7, E and F). Oddly, whereas these early
spermatocytes were distinctively positive, spermatocytes beginning at
about the zygotene phase became unreactive, and reactivity did not
reappear until late pachytene spermatocytes. This accounts for the ring
of unstained early pachytene nuclei seen in some tubules (Fig. 7, B and E). Precise identification of
the earliest expressing cells was not possible with these sections, but
they may be B type spermatogonia. Despite the unexpected detection of
the transgene in early spermatocytes, expression was confined to germ
cells, because no immune reaction was seen in nuclei outside the
seminiferous tubules or within Sertoli cell nuclei.
Testis-specific Expression of H1t-lacZ Fusion
Gene
The previous results leave open the possibility that
testis-directing sequences might lie within the H1t gene. To
demonstrate that this is not the case, we isolated two founder lines
that each contained an integrated fusion gene in which the
-948/+71 fragment of H1t sequence was ligated to the E.
coli lacZ structural gene. When a testis from the TG-LacZ-5371
line was lightly fixed and soaked in a histochemical stain solution for
-galactosidase, it turned a dark blue within a matter of hours,
indicating a positive reaction. Low power microscopic examination of
the seminiferous tubules showed that staining was present in only a
fraction of the cells, whereas no staining was observed in tubules of a
nontransgenic control animal (Fig. 8). The second founder line
did not give a positive histochemical stain for -galactosidase in
the testes or in any other organ tested. To quantitate
-galactosidase activity in the TG-LacZ-5371 line, extracts were
prepared from a number of organs. Only the testes showed an activity
significantly above (6-8-fold) the background levels seen in
nontransgenic controls (Fig. 9).
-galactosidase activity. Intact tubules were photographed using a
deep red filter to enhance the contrast of the blue reaction product.
No blue color was seen in the control tubules, and gray spots visible
on the photograph are due to elongated spermatid nuclei. Bar,
100 µm.
-Galactosidase activity in
homogenates of various organs. Representative organs were assayed for
-galactosidase activity as described under ``Experimental
Procedures.''
-galactosidase activity. Tubules were then
embedded for light microscopy, sectioned, and counterstained with
nuclear fast red. The same cross section was photographed through a
light green filter to identify nuclei (A and B) or
through a deep red filter to enhance the contrast of the indigo
reaction product (C and D). -Galactosidase
activity was observed over round spermatids (thin arrow) and
pachytene spermatocytes (thick arrow) but not over early germ
cells found in the outermost germ cell layer of the tubules or over
elongated spermatids found near the lumen. Bar, 30
µm.
Sequences Regulating Rat H1t Expression Lie within a
Discrete Chromosomal Region
H1t is a tissue-specific member of
the H1 gene family. Its expression is limited to late pachytene
spermatocytes. In this study we have defined the DNA region that
conveys this tissue specificity by introducing both the natural rat H1t
coding sequence as well as an H1t-lacZ fusion gene into
transgenic mice. Genomic fragments with 2.4 kb of upstream sequence and
4 kb of downstream sequence resulted in high level, testis-specific
expression of the natural rat H1t transgene. Fragments truncated to
within 141 bp upstream or 837 bp downstream of the 741-bp gene
maintained testis-specific expression, although with variable levels of
expression. An H1t-lacZ fusion gene with 0.95 kb of
5`-flanking sequence also expressed specifically in the correct germ
cell populations. Thus the sequences sufficient to activate expression
in the male germ line yet prevent expression in somatic cells must lie
within the 1019-bp fragment running from -952 to +71.
Correct expression may indeed be driven by merely the
-141/+71 fragment, although it is also possible that
redundant control sequences are found in the -952 to -141
and downstream regions so that correct expression is compatible with
deletion of either region individually but not in combination. Sequences Responsible for Testis-specific Expression
Remain Speculative
Most H1 genes code for members of the
standard somatic H1 family (H1a-e), and analysis of their
transcriptional control has focused on cell cycle regulation in
cultured cells (reviewed in Heintz(1991)) rather than expression in
tissues. These studies have identified a set of four conserved
sequences lying within 100 bp upstream of the transcriptional
initiation site of all standard H1 genes (Coles and Wells, 1985). Two
of these are binding sites for novel, H1-specific factors, the H1 box
(Dalton and Wells, 1988) and the CCAAT region (van Wijnen et
al., 1988; Gallinari et al., 1989; Martinelli et
al., 1994). The remaining two DNA elements (GC box and TATA box)
are presumed to interact with common transcription factors. As H1t has
each of these conserved promoter elements, they do not seem to account
for its tissue specificity. Recently Grimes and colleagues (Grimes et al., 1992a, 1992b) have identified a binding activity in
testis nuclear extracts for a palindrome found between the GC and CCAAT
boxes in the promoter region of the H1t genes of several species but
absent in standard somatic H1 promoters. Whether this factor plays a
role in H1t expression remains to be established. In this context it is
interesting to note that Chae and Lim(1992) have reported a possible
repressor of the testis-specific H2B gene in extracts of immature rat
testes that bound to a site just 3` of the TATA box.
gene. H1 is a minor linker histone whose expression is associated with the
onset of a nonproliferative, differentiated phenotype. Unlike a
standard H1 gene, the H1 gene encodes a polyadenylated
mRNA. Although its promoter shares three of the conserved elements of
the standard H1 genes, an apparent histone H4 promoter element is
substituted for the usual H1-specific CCAAT sequence (Breuer et
al., 1989, 1993; Khochbin and Lawrence, 1994). Studied by
transfection in mouse F9 embryonal carcinoma cells, basal expression of
the mouse H1 5`-flanking region was shown to depend on an
80-bp region at about -500 that bound several nuclear proteins
(Breuer et al., 1993). The H1 gene is induced by
vitamin A in F9 cells, and this induction is associated with a pair of
tandem retinoic acid response elements located just 3` to this 80-bp
region necessary for basal expression (Breuer et al., 1989,
1993). Although the relevance of this extreme H1 variant to H1t is
uncertain, it provides a precedent for influence of relatively distant
sequences on mammalian H1 gene expression. Roles for sequences upstream
of the proximal promoter region have also been proposed through
extensive study of a cell cycle-regulated human H4 gene (reviewed in
Stein et al.(1992)).Why Is the H1t Gene Expressed Prematurely in Some
Transgenic Lines
The natural H1t gene is expressed only in mid
to late pachytene primary spermatocytes beginning about stage VII in
the rat as determined by appearance of the protein in young animals
(Seyedin and Kistler, 1980), by analysis of H1t synthesis in purified
germ cell populations (Meistrich et al., 1985), and by
detection of H1t mRNA via in situ hybridization (Kremer and
Kistler, 1991). The results presented here for the developmental onset
of H1t mRNA expression in normal mouse testis as well as the
immunohistological detection of H1t protein confirm this mid/late
pachytene expression. It was therefore surprising to find that two of
three transgenic lines examined expressed the transgene prematurely,
clearly in preleptotene spermatocytes, and perhaps even in B type
spermatogonia. Although we have not checked for subtle effects, the
early expression of rat H1t in these mice had no obvious effects on
their fertility.Competition between Transgene and Endogenous H1t
Genes
In some transgenic lines, e.g. TG1-688 and
TG2-97, levels of the endogenous mouse H1t protein were markedly
reduced. This result could be explained by competition among excess
protein molecules attempting to bind to limited places available on
chromatin. In a recent study, Brown and Sittman(1993) observed that
overproduction of H1e or H1c in somatic cell lines led to
correspondingly reduced levels of other H1 variants, although the
mechanism for the reduction of the endogenous H1 variants was not
experimentally addressed. Competition for H1(t)-specific transcription
factors could also account for a reduced level of mouse H1t protein if
the mouse H1t mRNA was also low. Some lines of transgenic mice
expressing a p53 promoter-chloramphenicol acetyltransferase fusion,
which was expressed at high levels in the testes, displayed a reduced
level of p53 mRNA in the testes, associated with a failure of cells to
complete meiosis (Rotter et al., 1993). One explanation
suggested for this effect was a competition of the transgene and
endogenous p53 promoters for limiting transcription factors.
))
We thank Linda Bade for help with preparation of
histological sections, Clark Millette for assistance in identifying H1t
immunoreactive germ cells in transgenic animals, and Richard Showman
for the use of microscopic facilities. We are grateful to Richard
Palmiter for providing placI.
©1996 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. M. Kehoe, M. Oka, K. E. Hankowski, N. Reichert, S. Garcia, J. R. McCarrey, S. Gaubatz, and N. Terada A Conserved E2F6-Binding Element in Murine Meiosis-Specific Gene Promoters Biol Reprod, November 1, 2008; 79(5): 921 - 930. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Ma, G. C. Horvath, M. K. Kistler, and W. S. Kistler Expression Patterns of SP1 and SP3 During Mouse Spermatogenesis: SP1 Down-Regulation Correlates with Two Successive Promoter Changes and Translationally Compromised Transcripts Biol Reprod, August 1, 2008; 79(2): 289 - 300. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. C. Horvath, W. S. Kistler, and M. K. Kistler RFX2 Is a Potential Transcriptional Regulatory Factor for Histone H1t and Other Genes Expressed During the Meiotic Phase of Spermatogenesis Biol Reprod, November 1, 2004; 71(5): 1551 - 1559. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Han, W. Xie, S. H. Kim, L. Yue, and J. DeJong A Short Core Promoter Drives Expression of the ALF Transcription Factor in Reproductive Tissues of Male and Female Mice Biol Reprod, September 1, 2004; 71(3): 933 - 941. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Wolfe and S. R. Grimes Specific Binding of Nuclear Proteins to a Bifunctional Promoter Element Upstream of the H1/AC Box of the Testis-Specific Histone H1t Gene Biol Reprod, June 1, 2003; 68(6): 2267 - 2273. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. C. Horvath, S. E. Clare, M. K. Kistler, and W. Stephen Kistler Characterization of the H1t Promoter: Role of Conserved Histone 1 AC and TG Elements and Dominance of the Cap-Proximal Silencer Biol Reprod, October 1, 2001; 65(4): 1074 - 1081. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Bartell, D. A. Fantz, T. Davis, M. J. Dewey, M. K. Kistler, and W. S. Kistler Elimination of Male Germ Cells in Transgenic Mice by the Diphtheria Toxin A Chain Gene Directed by the Histone H1t Promoter Biol Reprod, August 1, 2000; 63(2): 409 - 416. [Abstract] [Full Text]
|
|
|
|
|
|
P. P. Reddi, C. J. Flickinger, and J. C. Herr Round Spermatid-Specific Transcription of the Mouse SP-10 Gene Is Mediated by a 294-Base Pair Proximal Promoter Biol Reprod, November 1, 1999; 61(5): 1256 - 1266. [Abstract] [Full Text]
|
|
|
|
|
|
S. A. Wolfe, P. J. Mottram, J. M. vanWert, and S. R. Grimes Localization of Upstream Elements Involved in Transcriptional Regulation of the Rat Testis-Specific Histone H1t Gene in Somatic Cells Biol Reprod, October 1, 1999; 61(4): 1005 - 1011. [Abstract] [Full Text]
|
|
|
|
|
|
J. M. vanWert, H. R. Panek, S. A. Wolfe, and S. R. Grimes The TE Promoter Element of the Histone H1t Gene Is Essential for Transcription in Transgenic Mouse Primary Spermatocytes Biol Reprod, July 1, 1998; 59(3): 704 - 710. [Abstract] [Full Text]
|
|
|
|
 |
|
 S. E. Clare, D. A. Fantz, W. S. Kistler, and M. K. Kistler The Testis-specific Histone H1t Gene Is Strongly Repressed by a G/C-rich Region Just Downstream of the TATA Box J. Biol. Chem., December 26, 1997; 272(52): 33028 - 33036. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |