ABSTRACT
Many proteins of the SNF2 family, which share a similar DNA-dependent ATPase/putative helicase domain, are involved in global
transcriptional control and processing of DNA damage. We report here the partial cloning and characterization of
89B helicase
, a gene encoding a new
Drosophila
melanogaster
member of the SNF2 family. 89B Helicase protein shows a high degree of homology
in its ATPase/helicase domain to the global transcriptional activators SNF2 and
Brahma and to the DNA repair proteins ERCC6 and RAD54. It is, however, most strikingly similar to the
Saccharomyces cerevisiae
protein Mot1, a transcriptional repressor with many target genes for which no
homologue has yet been described.
89B helicase
is expressed throughout fly development and its large transcript encodes a >200 kDa protein. Staining with anti-89B Helicase antibodies reveals that the protein is present uniformly in
early embryos and then becomes localized to the ventral nerve cord and brain.
On the polytene chromosomes, 89B Helicase is bound to several hundred specific
sites that are randomly distributed. The homology of 89B Helicase to Mot1, its
widespread developmental expression and its large number of targets on the
polytene chromosomes of larval salivary gland cells suggest that 89B Helicase
may play a role in chromosomal metabolism, particularly global transcriptional
regulation.
The
Saccharomyces cerevisiae
transcriptional activator SNF2 (
1
) has been implicated in control of expression of a broad range of diversely
regulated genes. SNF2 is part of a complex of proteins, including four other
SNF proteins and six non-SNF proteins (
2
-
5
). The SNF2-containing complex is thought to carry out its role in global regulation of transcription by remodelling chromatin structure and counteracting the non-specific repressive effects of histones and other proteins involved
in gene packaging (reviewed in
6
). Chromatin restructuring could affect such processes as nucleosome packing (
7
), DNA looping or attachment to the nuclear matrix (
8
). Whether the chromatin alterations then facilitate binding of general
transcription factors, assist a variety of gene-specific DNA binding activators to gain access to their DNA binding sites,
promote interactions between the general and specific classes of proteins or
act in some other way to bring about transcription, is presently under
investigation (
9
).
Recently, many eukaryotic and several prokaryotic and viral (
10
,
11
; reviewed in
12
,
13
) regulatory proteins with similarity to SNF2 have been discovered. Many of the
SNF2-related proteins are also likely to function in regulation of gene
transcription as part of multiprotein complexes (
14
,
15
). These include
Drosophila melanogaster
Brahma (Brm;
16
), an activator of the homeotic genes, and its vertebrate homologues (
17
-
20
) and the
Drosophila
ISWI protein, which is required to perturb nucleosome structure and generate an
accessible heat shock promoter (
21
,
22
). In contrast to the activators SNF2, Brm and ISWI, another SNF2-related protein, Mot1, acts as a transcriptional repressor in
S.cerevisiae
(
23
,
24
). Additional members of the SNF2 family function in other aspects of
chromosomal metabolism, such as DNA repair. These include human ERCC6, which is
involved in excision repair of transcriptionally active DNA (
25
), and its yeast homologue RAD26 (
26
,
27
).
We have identified a gene encoding a new
Drosophila
member of the SNF2 family and tentatively designated it
89B helicase
, according to its chromosomal location and the presence within it of the
ATPase/presumptive helicase domain
which characterizes the family. This domain includes seven motifs (I, Ia and II-VI) which have been defined for two related superfamilies of DNA- and RNA-dependent ATPases and putative helicases (
28
-
30
). Motifs I and II are responsible for binding of the Mg-nucleoside triphosphate moiety. Among members of the SNF2 family, motifs V and VI have the largest differences from the corresponding motifs of
other helicase families (
10
,
11
). The nucleic acid-stimulatable ATPase/helicase domain is functionally significant. For both SNF2 (
31
) and the mammaliam Brm homologue Brg1 (
17
), mutations in this domain have been shown to impair transcriptional activation
in vivo
. However, additional experiments have shown that this region is not sufficient
(
17
,
31
-
33
) and that other domains participate in transcriptional regulation.
The sequence identity of the ATPase/helicase region of the 89B Helicase protein is shown to be greatest to the global transcriptional repressor Mot1, which until now has been the only representative within its own
SNF2 subfamily (
13
). 89B Helicase also shares homology with Mot1 outside the ATPase signature
motifs. The expression of
89B helicase
throughout
Drosophila
development and the tissue distribution of its protein product in embryos is
presented. We have examined the possibility that 89B Helicase protein binds to
the polytene chromosomes of larval salivary glands and observed targetting of
the protein to a large number of chromosomal sites. Together these data suggest
that 89B Helicase may be a global transcriptional regulator of many target
genes that functions throughout
Drosophila
development.
Drosophila melanogaster
insertion strain P282 (
34
), carrying a P element insertion in position 89B on the third chromosome, was
obtained from the
Drosophila
Stock Center (Bloomington, IN). We identified this line as carrying a mutation in the morphogenetic locus
serpent
, which we were interested in cloning, by virtue of its non-complementarity to a known allele of
srp
. While embryos homozygous for this insertion die, their cuticular phenotype is
almost normal. DNA was extracted from adult heterozygous flies, restricted with
either
Eco
RI,
Bam
HI or
Sal
I and plasmid rescue was carried out as described (
35
).
A 3.15 kb
Eco
RI fragment from the plasmid-rescued DNA (Fig.
1
A) was labelled with the Multiprime DNA Labelling System (Amersham) and used to
probe a random-primed 0-16 h embryonic cDNA library (
36
). Two of the ~200 000 recombinant phage hybridized to the genomic
Eco
RI probe. These were called RP2 and RP3. The same probe was used to screen a
poly(A)
+
-primed 0-14 h embryonic cDNA library (
37
), yielding another two hybridizing phage, MN0.9kb and MN1.4kb. Finally,
sequences from cDNA clones RP3, MN0.9kb and MN1.4kb were applied to a poly(A)
+
-primed 9-12 h size-selected embryonic library (
38
). This probe detected two phage, YKZ12 and YKZ9.
Poly(A)
+
RNA was isolated from flies of different developmental stages using Dynabeads
Oligo(dT)
25
(Dynal). Frozen, ground tissues were homogenized in lysis buffer as per the manufacturer's recommendations and the supernatant applied to beads. Approximately 4 [mu]g of poly(A)
+
RNA was loaded per well in a formaldehyde-containing gel. A mixed probe with sequences from cDNA clones RP3, MN0.9kb
and MN1.4kb was applied to nylon filters (Nytran; Schleicher & Schuell). The following day, the blots were washed in 0.2* SSC (0.3 M sodium chloride, 0.03 M sodium citrate, pH 7), 1% SDS at 65oC.
Antibodies were generated against a bacterial fusion protein containing 140
amino acids from 89B Helicase by inserting RP2 into the pGEX 2T expression vector (Pharmacia LKB Biotechnology Inc.). The 89B
Helicase fusion protein was induced, harvested (
39
) and then absorbed onto glutathione-agarose beads (Sigma). For further purification prior to immunization
into rabbits, it was released from the beads and isolated by preparative SDS-PAGE, followed by light staining with Coomassie blue.
Schneider L2 cells were grown in Schneider medium (Biological Industries,
Kibbutz Beit HaEmek) supplemented with 10% fetal calf serum (FCS) and two Kc
derivatives, 167 and 7E10 (
40
), were grown in D22 (Sigma), 5% FCS. For Western analysis, the cells pelleted
from 1 ml medium were dissolved in 20 [mu]l Laemmli buffer and the supernatant applied to a 7.5% polyacrylamide
minigel (BioRad mini Protean gel apparatus). After transfer to a Nytran filter,
the immunoblots were blocked with 5% low fat milk and 2% bovine serum albumen
(BSA) overnight in the cold. They were treated with a 1:100 dilution of primary
antibody in 5% normal goat serum (NGS) for 1 h and then alkaline phosphatase-conjugated affinity-purified goat anti-rabbit antibody (Jackson) at a 1:2000 dilution in 2% NGS, also
for 1 h. Detection of bound antibody was carried out by adding nitroblue
tetrazolium chloride (Boehringer) and X-phosphate (Boehringer). The reaction was stopped with 20 mM EDTA.
For immunohistochemistry of cells, these were pelleted, fixed in 4% formaldehyde
and spread onto poly-L-lysine-treated slides. They were then stained by previously described
procedures (
41
) with some slight modifications. The primary antibodies were applied at a 1:200 dilution and the peroxidase-conjugated goat anti-rabbit antibody (Jackson) was used at a 1:300 dilution. The staining
solution contained 200 [mu]l DAB [1 mg/ml 3,3'-diaminobenzidine tetrahydrochloride (Sigma) in PBS with 0.1%
Triton (PBT)], 100 [mu]l phosphate-buffered saline (PBS) and 3 [mu]l 3% H
2
O
2
. The cells were viewed and photographed with Nomarski optics on a Zeiss
Axioskop.
For examination of the localization of 89B Helicase in embryos, embryos of mixed
ages were fixed, stained and viewed as for the cells, except that the secondary
antibody was conjugated to alkaline phosphate and an appropriate substrate was
applied as in Western analysis. In these experiments anti-89B Helicase or pre-immune serum that had been pre-absorbed on fixed Schneider L2 cells and GST were employed.
Salivary glands of wandering third instar larvae that had been growing at 18oC on molasses-containing medium were dissected into PBT. They were fixed for 10 s
in 3.7% formaldehyde in PBT followed by 2-3 min in 3.7% formaldehyde in 50% acetic acid, spread onto 0.1% poly-L-lysine (Sigma)-coated slides and covered with Sigmacote (Sigma)-treated coverslips. The slides were frozen in dry
ice and the coverslips flipped off. The chromosomes were washed twice in PBS
and then kept in 100% ethanol for up to 1 week.
For staining, the chromosomes were rehydrated in PBS, incubated for 1 h in a
moist chamber covered with blocking solution [PBS containing 3% BSA, 10%
powdered non-fat milk, 0.2% Nonidet P40 (Sigma) and 0.2% Tween 20 (Baker)] and rinsed.
The anti-89B Helicase antibody was diluted 1:50 in blocking solution containing 2%
NGS and applied under a coverslip at 4oC. The following day, the slides were washed in solution 1 (300 mM NaCl,
0.2% Nonidet P40, 0.2% Tween 20 in PBS), solution 2 (like solution 1 but 400 mM
NaCl) and PBS. Affinity-purified biotinylated goat anti-rabbit antibody (Jackson) was applied for 1 h at a 1:200 dilution in
blocking solution containing 2% NGS followed by extra-avidin peroxidase (Sigma) at a 1:40 dilution in PBS for 30 min at 37oC. The slides were washed in 3% H
2
O
2
and transferred to staining solution. The chromosomes were mounted in
gelatin/glycerol (Sigma) for observation at 100* (oil immersion) without counterstaining.
The
89B helicase
RNA was identified by Northern blot analysis of poly(A)
+
RNA prepared from embryos, third instar larvae, pupae and adult males. Using
stringent washing conditions, we detected a prominent transcript >7.4 kb in
size that is large enough to encompass the partial cDNA of
89B helicase
. This transcript was present at all the developmental stages examined (Fig.
4
, upper panel), including staged embryos aged 2-5 and 5-9 h and adult females (not shown). Upon longer exposures of the
same blot, we also observed a number of smaller RNA species at all the examined
stages (Fig.
4
, lower panel). The most visible of these are two very small species of 1.3 and
1.1 kb and another two of 2.2 and 1.7 kb. Some of these multiple mRNA
components might be the products of alternative RNA processing of the
89B helicase
gene, since under conditions of reduced stringency we saw no evidence for an additional cross-hybridizing gene in the
Drosophila
genome (data not shown). However, since the cDNA that we have in hand is larger
than these smaller species, their relevance is unclear. From these blots it can
be concluded that
89B helicase
is expressed at all stages of
Drosophila
development.
In order to visualize the protein encoded by
89B helicase
, we generated anti-89B Helicase antibodies as described in Materials and Methods and employed
them in Western analyses of protein extracts from three
Drosophila
tissue culture cells lines. Based on the size of the predominant
89B helicase
transcript and the partial cDNA, we expect the encoded protein to be >90 kDa in
size. Indeed, the anti-89B Helicase antibodies recognized a common slowly migrating (>200 kDa)
polypeptide in extracts from two Kc-derived lines, 167 and 7E10 (Fig.
5
, upper panel, left, arrow). As anticipated, the antibodies, although not pre-absorbed in this case (see Materials and Methods) on Schneider L2, which
do not express
89B helicase
transcripts (not shown), did not detect a large protein in these cells (Fig.
5
, upper panel, left). The ~210 kDa polypeptide was also absent from all the cell extracts in parallel Western
blots treated with pre-immune serum (Fig.
5
, upper panel, right).
We examined whether the anti-89B Helicase antibodies can be employed for immunohistochemistry by
staining the same three
Drosophila
cell lines (again with antibodies that had not been pre-absorbed). The immunohistochemical results confirmed what we observed in
Western analysis; while Kc167 and Kc7E10 cells did exhibit an immunoreactive
signal (Fig.
5
A and C), the anti-89B antibodies did not stain Schneider L2 cells (Fig.
5
E). There was no reaction with the pre-immune serum (Fig.
5
B, D and F). Thus, the >200 kDa protein observed in Western blots of extracts
from Kc cells is likely to represent a polypeptide encoded by
89B helicase
. Furthermore, the anti-89B Helicase antibodies appear to give a specific signal when employed for
immunohistochemistry.
In order to determine the tissue distribution of 89B Helicase throughout
embryonic development, whole mount embryos of different ages were treated with
anti-89B Helicase antibodies that had been pre-absorbed as described in Materials and Methods. While pre-immune serum (Fig.
6
A and B) or antiserum against GST (not shown) gave no specific
immunohistochemical signal, the anti-89B Helicase antibodies revealed expression of the protein in unfertilized
eggs and early embryos. 89B Helicase is ubiquitously distributed throughout the
embryo (but not in pole cells) during the first part of embryogenesis,
including the blastoderm stage (Fig.
6
C), gastrulation and germband extension (not shown). At ~8 h of embryogenesis, during germband retraction, the protein becomes
highly localized to the ventral nerve cord and brain (Fig.
6
D). In the CNS, it is preferentially found in the longitudinal connectives
rather than in the horizontal commissures of the scaffold (Fig.
6
E). The same pattern of expression was observed using anti-89B Helicase antibodies that had been pre-absorbed on fixed embryos (not shown). These results demonstrate
that, despite the widespread developmental expression of
89B helicase
, there are differential levels of the protein in different developing tissues.
Several transcriptional regulator proteins that control a large number of target
genes in
Drosophila
bind to multiple sites on polytene chromosomes, including their known targets of regulation (see for
example
45
,
46
). After ascertaining immunohistochemically that 89B Helicase is found in the nuclei of third instar larvae salivary glands
(data not shown, but see Discussion), we examined the possibility that 89B
Helicase is also associated with the chromosomes. Spread polytene chromosomes
from third instar larval salivary glands were treated with anti-89B Helicase followed by a biotinylated secondary antibody and extra-avidin-conjugated peroxidase. Following colour development, we
observed brown bands of immunoreactivity distributed over all the chromosomes
(Fig.
7
) and estimate that 89B Helicase is associated with several hundred sites along
the polytene chromosomes. It is absent from puffed regions of the chromosomes (Fig.
7
and inserts at the top). While different samples of the same chromosome
exhibited identical staining patterns (Fig.
7
, inserts), we did not detect any periodicity or pattern in the distribution of
89B Helicase immunoreactivity on the polytene chromosomes. This implies that
89B Helicase is associated with particular target genes that are randomly
dispersed in the genome.
The number of proteins assigned to the SNF2-related family of DNA-dependent ATPases/putative helicases that function in various
aspects of DNA maintenance and processing has increased rapidly over the last
few years. Among these proteins, Mot1 is one of the few that is the only
representative within its proposed subfamily (
13
). In addition to the absence of any known homologue, Mot1 is unique among the
previously known SNF2 proteins because it acts as a global transcriptional
repressor. Here we describe the isolation, partial cloning and characterization
of
89B helicase
, a new
Drosophila
gene encoding a product with sequence homology to the SNF2 family. In the
sequence and positioning within the protein of the ATPase/helicase domain, the
spacing of the individual helicase motifs and the absence of a C-terminal bromodomain, 89B Helicase is most similar to yeast Mot1. Thus,
89B Helicase identifies a new member of the Mot1 subfamily of proteins within
the SNF2-related family. The particularly high overall degree of similarity between
89B Helicase and Mot1 argues in favour of 89B Helicase, like Mot1, playing a
role in transcriptional control (
13
). The widespread expression of the gene product of
89B helicase
, as observed in developmental Northern blots and immunohistochemical analysis of embryonic tissues probed with anti-89B Helicase antibodies, implies that it may play a prominent role throughout development. This is further supported by the prevalent distribution of 89B Helicase
at several hundred specific sites on larval salivary gland polytene
chromosomes.
mot1
encodes an essential protein required for repression of basal transcription of
many genes in yeast (
23
,
24
). Mot1 is thought to function in negative control of transcription by acting as a polymerase II-specific TATA binding protein (TBP)-associated factor (TAF) that complexes with TBP. It may catalyse the removal
of TBP from DNA in an ATP-dependent manner (
47
-
49
), thus destabilizing the binding of TFIID, which consists of TBP and TAFs, to
DNA. It has been suggested that the function of Mot1 in ATP-dependent removal of TBP from DNA may be analogous to the proposed role of the SNF complex in displacement of histones from DNA. Whether 89B Helicase functions in a manner similar
to that proposed for other SNF proteins or at the level of the basal
transcriptional machinery, as demonstrated for Mot1, remains to be examined.
89B helicase
is expressed throughout the
Drosophila
life cycle. In Northern analysis there was prominent expression of a >7.4 kb
transcript at all stages of fly development and in adult males and females.
This transcript encodes a >200 kDa protein that we have visualized in
Drosophila
tissue culture cell extracts. The large size of the protein is characteristic of
many members of the SNF2 family, such as SNF2 (
1
), Brg1 (
17
), hBrm (
19
) and Mot1 (
48
). The early appearance of the 7.4 kb transcript in poly(A)
+
RNA derived from staged embryos and the immunohistochemical observation of 89B Helicase in pre-blastoderm stage embryos suggest that there may be a maternal contribution of
89B helicase
to the embryo.
We have observed differential levels of expression of 89B Helicase in embryonic
tissues by indirect immunohistochemical analysis with anti-89B Helicase antibodies. While the results from Northern analysis using an
89B helicase
probe are indicative of a very general pattern of expression over time and early
embryos do exhibit uniform staining, in the latter part of embryogenesis there
is a concentration of protein in the ventral nerve cord and brain, two tissues
where, presumably, the complex cell differentiation events that occur require
many levels of gene regulation. Although we did not observe nuclear expression
for 89B Helicase in embryos or tissue culture cells, we have detected the
protein localized to the nuclei of cells from the salivary glands of wandering
third instar larvae (from which the polytene chromosomes were removed to examine binding of 89B Helicase). Interestingly, its subnuclear distribution is strikingly non-homogeneous in these cells. Furthermore, when we examined salivary glands
of slightly older animals (very early prepupae prior to the stage of salivary
gland histolysis at 15 h after puparium formation;
50
), the nuclear distribution was altered and we also detected 89B Helicase in the
cytoplasm (Goldman-Levi and Zak, in preparation). These observations indicate that 89B
Helicase can be found either in the nucleus or the cytoplasm. Alterations in
the subcellular distribution of this protein may be one mode of regulating its
level of activity.
Using our anti-89B Helicase antibodies to examine the sites of 89B Helicase binding on
spread polytene chromosomes, we have demonstrated a distinct pattern of bands that implies an association of the protein with particular genes. 89B Helicase is thus, to our knowledge,
the first SNF2-related protein shown to bind to polytene chromosomes. How is 89B Helicase bound to chromatin? While several of the SNF proteins do have motifs characteristic of
activators, such as glutamine- and proline-rich regions in SNF5 (
51
) and acidic regions in SNF6 (
52
), they do not have clear DNA binding motifs. They were thus, until recently,
thought to be directed to DNA in general, and specifically to particular
promoters, via interactions with other proteins that do bind DNA, such as as-yet-uncharacterized members of the SNF complex or the DNA binding gene-specific activators whose activation they assist, such as
yeast GAL4 (
53
,
54
),
Drosophila
Bicoid (
53
) and fushi tarazu (
54
) and mammalian glucocorticoid (
55
), estrogen and retinoic acid receptors (
18
). Now, however, it has been shown that the purified SNF complex does have a
high affinity for DNA (
56
) and also that the
SNF
gene products are integral components of the yeast RNA polymerase II holoenzyme
(
57
). It remains to be seen whether a DNA binding motif is present in 89B Helicase,
as in CHD1, the only helicase domain-containing protein that has been demonstrated to possess DNA binding
capability (
58
,
59
), or whether the binding of 89B Helicase too is mediated by protein-protein interactions.
The large but discrete number of anti-89B Helicase immunoreactive bands on the polytene chromosomes, together
with the widespread developmental pattern of expression of the
89B helicase
gene, suggests that the protein plays a global role in some aspect of
chromosomal metabolism. This does not, however, appear to be associated with
basic housekeeping functions that are required for viability of tissue culture
cells, since the loss of 89B Helicase from
Drosophila
Schneider L2 cells is not lethal. The absence of 89B Helicase from chromosomal
puffs implies that the protein is not targetted to all areas which are
transcriptionally active. It will be interesting to determine whether there is
a correlation between the presence or absence of this protein and the
transcriptional state of the DNA.
This research was supported by a Research Career Development Award of the Israel
Research Cancer Fund to NBZ, a research grant from the Israel Academy of Arts
and Sciences to NBZ and pre-doctoral support by the Kraut Endowment Fund to RG-L. We would like to thank Jeremy Thorner and Karin Hansen for
helpful discussions about
mot1
, Allen Shearn and Dennis LaJeunesse for advice in immunohistochemical staining
of the polytene chromosomes and Ze'ev Paroush for careful reading of the
manuscript.
REFERENCES
Return