doi:10.1006/jmbi.2000.4027 available online at http://www.idealibrary.com on
J. Mol. Biol. (2000) 302, 17±25
Upstream Flanking Sequences and Transcription
of SINEs
Astrid M. Roy1, Neva C. West1,2, Aravinda Rao1, Prateek Adhikari3,
Claudina AlemaÂn1,3, Anthony P. Barnes1,2 and Prescott L. Deininger1,4*
1
Tulane Cancer Center, SL-66
and, Department of
Environmental Health Sciences
Tulane University Medical
Center, 1430 Tulane Avenue
New Orleans, LA 70112, USA
2
Neuroscience Center of
Excellence, Louisiana State
University Health Sciences
Center, 1901 Perdido Street
New Orleans, LA 70112, USA
3
Biochemistry Department,
Louisiana State University
Health Sciences Center, New
Orleans, LA 70112, USA
4
Laboratory of Molecular
Genetics, Alton Oschner
Medical Foundation, New
Orleans, LA 70121, USA
SINEs, short interspersed repeated DNA elements, undergo ampli®cation
through retroposition and subsequent integration into a new location in
the genome. Each new SINE insertion will be located in a new chromosomal environment, with different ¯anking sequences. Modulation of transcription by different ¯anking sequences may play an important role in
determining which SINE elements are preferentially active in a genome.
We evaluated the ability of upstream ¯anking sequences to regulate the
transcription of three different SINEs (Alu, B2 and ID) by constructing
chimeric constructs with known 50 ¯anking sequences of RNA polymerase III-transcribed genes. Upstream sequences from the 7SL RNA gene,
U6 RNA gene, vault RNA gene, and BC1 gene increase transcription of
Alu, B2 and BC1 in transient transfections of NIH3T3, HeLa, Neuro2a
and C6 glioma cell lines. The 7SL sequence proved most ef®cient in
increasing SINE transcription. The 7SL upstream fused to the BC1 RNA
gene (an ID element) was used to create a transgenic mouse line. In contrast to the tissue-speci®c endogenous BC1 transcription, BC1 transgene
transcripts were detected in all tissues tested. However, expression was
much higher in those tissues that express the endogenous gene, demonstrating both transcriptional and post-transcriptional regulation. The BC1
RNA was detected in a similar ribonucleoprotein complex in the different
tissues.
# 2000 Academic Press
*Corresponding author
Keywords: Alu; BC1; polymerase III; SINE; transgenic mice
Introduction
SINEs are short interspersed repeated DNA
elements commonly found in introns, 30 untranslated regions of genes, and intergenic regions. Alu
repeats are the major class of SINEs in the human
genome. Rodents have an Alu-related element, B1,
and two unrelated elements, B2 and ID (Weiner
et al., 1986; Rogers, 1985).
The SINE ampli®cation model includes the transcription of the element by RNA polymerase III
(pol III), reverse transcription of the transcript, followed by integration of the cDNA into a new site
in the genome (Rogers & Willison, 1983; Weiner
et al., 1986; Boeke et al., 1985). SINEs contain an
internal RNA pol III promoter that consists of an A
box located just downstream from the transcription
start site, and a B box located approximately 50 bp
downstream from the A box. In vitro transcription
studies demonstrated that the internal promoter is
E-mail address of the corresponding author:
[email protected]
0022-2836/00/010017±9 $35.00/0
required for transcription (Fuhrman et al., 1981;
Sutcliffe et al., 1984; Kleinert et al., 1988; Willis,
1993). Only a limited number of SINEs, termed
master or source genes, appear to be capable of
actively undergoing retroposition. In fact, very
small amounts of SINE transcripts from limited
individual elements are observed in vivo (Paulson
& Schmid, 1986; Matera et al., 1990; Shaikh &
Deininger, 1996; Kim et al., 1995). The lack of in vivo
transcriptional activity of most copies of the SINE
elements has been attributed to a multitude of factors. Some of these factors include the stability of
the RNA structure of the transcript (Labuda &
Zietkiewicz, 1994), mutations within the SINE
element in¯uencing the internal promoter (Daniels
& Deininger, 1991; Chu et al., 1995), methylation
(Liu & Schmid, 1993; Englander et al., 1993; Liu
et al., 1994; Vorce et al., 1994), and chromatin context (Slagel & Deininger, 1989; Englander &
Howard, 1995). Furthermore, very few SINE
elements are actually capable of undergoing the
retroposition process. Transcription has been
# 2000 Academic Press
18
suggested to be a major factor in selecting active
SINEs from pseudogene copies of the SINEs
(Schmid & Maraia, 1992).
The in¯uence of upstream ¯anking regions on
various RNA pol III transcribed genes (Schmid &
Maraia, 1992) including Alu elements (Chesnokov
& Schmid, 1996) and ID elements (Martignetti &
Brosius, 1995) has been observed previously.
Chesnokov et al. (1996) reported that wild-type p53
represses pol III transcription of Alu in a promoterdependent manner. However, when they used the
7SL RNA gene upstream activating sequences or
an Alu source gene with its own ¯anking sequence,
both relieved the p53 repression, suggesting an
additional regulatory level on Alu transcription.
Our goal is to further explore the potential of
upstream ¯anking regions to control and enhance
RNA pol III transcription of SINES. To analyze the
effect of upstream ¯anking regions on SINE RNA
pol III transcription, we created chimeric constructs
with SINEs and different ¯anking sequences. We
chose the upstream sequences of four known RNA
pol III transcribed genes: human 7SL RNA gene
(Kleinert et al., 1988), mouse U6 RNA gene (Das
et al., 1988), rat vault RNA gene (vRNA)
Figure 1. Schematic of the plasmid constructs. The
constructs consist of three parts: a 50 upstream region, a
SINE body, and a unique 30 end. Upstream ¯anking
sequences were obtained from four known RNA polymerase III transcribed genes: (1) human 7SL RNA gene
(accession number M20910); (2) mouse U6 RNA gene
(accession number X06980); (3) BC1 gene; and (4) rat
vault RNA gene (vRNA). Flanking vector provided the
sequence for the no upstream control. The SINEs BC1,
B2, and the Alu with their own internal pol III promoters (A and B box) were used as bodies of the construct.
The Alu body is the consensus sequence for the PS (Sx)
subfamily. The ID body used matches the ID portion of
the BC1 master gene (Kim et al., 1994) and for the B2
body sequence matches the B2 element within intron 4
of the murine b-glucuronidase gene (Gus-s) (Roy et al.,
1998). All constructs are ¯anked at the 30 end by the
unique sequence of rat BC1 that provides the RNA pol
III terminator (Kim et al., 1994). Transcripts start at the
‡1 site of the SINE and end within the unique sequence
of the BC1 30 end as indicated by the arrow. The plasmids are named according to the upstream sequence
(superscript) and the SINE body and the 30 end (superscript), for example: p7SLAluSxBC1 (7SL upstream - Alu
Sx body - BC1 30 unique sequence). The oligo ``unique1`` (broken line) complementary to the BC1 unique
sequence of the RNA was used as probe.
Modulation of SINE Transcription
(Kickhoefer et al., 1993), and the rat BC1 RNA gene
(Kim et al., 1994) (Figure 1). The consensus
sequence or a close match, all with perfect internal
pol III promoters of three SINES: Alu, ID and B2
were used for the body of the chimera. Different
SINEs were explored, since Alu is derived from
7SL, while ID and B2 are derived from tRNAs and
their internal promoters vary. In addition, the ID
elements have one particular copy, BC1, which has
the unusual property of tissue speci®city, exhibiting extremely robust expression in brain and to a
lesser degree in testes. To eliminate the in¯uence of
the 30 end on transcript termination and stability,
all constructs contain the rat 30 unique end from
the BC1 RNA gene (Kim et al., 1994).
Results
Enhancement of SINE transcription by
upstream sequences
Transcription of the three SINES from chimeric
constructs ¯anked by different upstream sequences
(Figure 1) was evaluated by transient transfections
of the human cell line HeLa (epitheloid carcinoma
from cervix) and the three rodent cell lines,
NIH3T3 (®broblasts), C6 (rat glial tumor), and
Neuro2a (mouse neuroblastoma). Since BC1 transcription is almost exclusive to the brain, two
brain-derived cell lines were chosen for transfections to evaluate any potential in¯uencing factors
that may be present or absent in cells from this
tissue.
The transcripts of B2 (267 bp), ID (199 bp) and
Alu (370 bp) were quanti®ed from Northern blots
probed with an end-labeled oligonucleotide
(unique-1) complementary to the unique portion of
the BC1 30 end (Figure 1). A constant amount of
either p7SLAluSxBC1 (for B2 and ID constructs) or
pBC1BC1 (for Alu constructs) was cotransfected as
an internal control for transfection ef®ciency and
used to standardize the quanti®cation (Figure 2(a)).
Transcription rates are reported as normalized
values relative to the constructs carrying the BC1
upstream.
All four upstream sequences increased SINE
transcription (p < 0.05 two-sided ANOVA). The
SINEs without an added upstream sequence
(¯anked by vector sequence) are hardly detectable
above background (Figure 2(a) and (b)). Transcription of the various SINEs varied slightly between
the four cell lines, but the same general pattern of
enhancement is observed. The 7SL-upstream
sequence proved to be the most ef®cient, with the
highest increased transcription of Alu and ID by
several fold (p < 0.05). The B2 transcription from
both 7SL and RVG upstream sequences appears to
be comparable in strength and better than the BC1
and U6 upstream sequences. However, only the
7SL and RVG effect on B2 transcription observed
in the Neuro2a cell line is signi®cantly different
(p < 0.05) from the other upstream regions.
Upstream enhancement strength can be indirectly
19
Modulation of SINE Transcription
Figure 2. Effect of different upstream regions on SINE transcription in transiently transfected cell lines. (a) Alu transcripts in C6 glioma cells, BC1 transcripts in Neuro2a cells and B2 transcripts in NIH3T3 cells were detected using
end-labeled unique-1 probe. The lanes 1-5 are constructs with the following upstream regions: 1, 7SL; 2, U6; 3, BC1;
4, RVG; 5, no upstream; lanes 6-8 are controls: 6, vector, 7, only internal control pBC1BC1BC1 (for Alu) and
p7SLAluSxBC1 (for BC1), and 8, no DNA transfected. Individual transcripts are indicated on the left; C, denotes the
internal cotransfected control. (b) SINE constructs were transfected into four cell lines; C6 glioma (striped bar), Neuro2a (white bar), NIH3T3 (gray bar), and HeLa (black bar). Bars represent the mean the standard error of the
mean. Note the different scales (x-axis) between Alu and BC1/B2. The upstream regions of the constructs are indicated on the x-axis (only the vector control group is shown). The expression levels were normalized to the internal
controls. The ratio of the evaluated SINE transcript/control transcript was normalized to the pBC1 (SINE) transcription
ratio (y-axis), which was assigned the arbitrary value of 1 (#). Statistical signi®cance calculated by the Tukey twosided Analysis of Variance (ANOVA) was used to determine p values with n > 3. * ˆ p < 0.05 statistically different
from all of the groups. ** ˆ p < 0.05 statistically different from control groups.
evaluated by the competition between the construct and the control plasmid, also a RNA pol IIItranscribed gene. In particular, the p7SLAluSxBC1
transcription's ef®ciency can be observed by
the competition with the pBC1BC1BC1 control plasmid (Figure 2(a), C6 glioma panel, lane 1). In
addition, SINE transcription from the different constructs was tested in vitro using HeLa extracts. The
results obtained paralleled those observed using
transient transfections, where the 7SL-upstream
sequence proved to be the most ef®cient (data not
shown).
Transcription of BC1 in a
mouse line
7SL
BC1 transgenic
A transgenic mouse line (G) carrying the 7SLBC1
sequence was created. Progeny of the G line-founder mouse were screened by PCR to determine the
presence of the transgene, 7SLBC1. Although no
detailed analyses have been made, the transgenic
line appears to be normal in behavior, development and reproductive capability. In addition, we
have analyzed 14 F2 generation mice (ten male and
four female transgenics produced from mating of
20
F1 heterozygotes) by multiple breeding to detect if
any were homozygous for the BC1 transgene
homozygotes. We would expect one quarter of the
progeny to be homozygous, but all 14 were found
to be heterozygous. This may re¯ect either lethality
due to higher BC1 expression levels in the homozygotes, a recessive mutation caused by the integration of the transgene, or simple chance
assortment.
RNA extracts of brain, liver, kidney, spleen,
intestine, testes, heart and muscle (not shown)
were evaluated by Northern blot analysis with the
unique-1 probe. BC1 transcripts are observed in all
the transgenic mouse tissues tested (Figure 3), in
contrast to the wild-type control where BC1 is
detected only in brain and testes. In order to minimize RNA degradation from some of the tissues,
RNA was processed rapidly and not quanti®ed
before analysis. Therefore, BC1 transcripts were
quanti®ed relative to the endogenous 7SL RNA
(Table 1). However, the same general tissuespeci®c transcription pattern is observed, highest
in brain (approximately ®ve times the level of 7SL
RNA), followed by testes and in much smaller
amounts in the rest of the tissues evaluated.
BC1-RNP complex formation in transgenic
mouse tissues
BC1 RNA interacts with protein(s) to form an
RNP (ribonucleoprotein) complex (Anzai & Goto,
1988; Kobayashi et al., 1991; Cheng et al., 1996).
Although the RNP proteins must be abundant in
brain, the abundance or presence of these proteins
in other tissues was previously unknown. We evaluated the formation of the BC1 RNP complex in
several different tissues from the transgenic mouse.
``Native mobility shift assays'' for the endogenous
RNP from extracts of brain, liver, kidney, heart,
and testes were probed for BC1 RNA (Figure 4).
BC1 RNP is detected in large amounts in brain, in
smaller amounts in testes, liver, heart and spleen
(not shown), and barely detectable in kidney when
samples are equalized by protein concentration
(Figure 4(a)). Free BC1 RNA is not detected. However analysis of samples with approximately equal
concentrations of BC1 RNA demonstrate the presence of BC1 RNP in the tissues evaluated
Modulation of SINE Transcription
Table 1. BC1 tissue expression in wildtype and
transgenic mice tissues
Tissue
Brain
Liver
Kidney
Spleen
Intestine
Muscle
Testes
Wild-type
a
6.6(1.3)
ND
ND
ND
ND
ND
*1.0
7SL
BC1
Transgenic
2.54(0.89)a
0.53(0.20)b
0.53(0.17)b
0.33(0.10)b
0.40(0.19)b
0.49(0.14)b
*1.0
Relative amounts of BC1 RNA (mean standard deviation)
expressed in different tissues of wild-type and 7SLBC1 transgenic mice as determined by Northern blot analysis. *BC1 RNA
from testes from each mouse was arbitrarily designated as 1.0.
Statistical signi®cance was calculated by t-test analysis with
n 5 6.
a
Signi®cantly different from testes p < 0.001.
b
Signi®cantly different from testes and brain p < 0.002, but
not signi®cantly different from each other p ˆ 0.235.
ND, not detected.
(Figure 4(b)). All the BC1 RNPs detected present
the same electrophoretic mobility as the one
observed in brain, suggesting it is either the same
complex, or at least one involving closely related
proteins. The presence of this RNP in the different
tissues indicates that the RNA binding proteins
involved in this SINE RNP are ubiquitous. No
comparisons between RNA levels (Figure 3) and
levels of RNP complexes (Figure 4) in the different
tissues can be made because sample standardization for each experiment is different.
Discussion
The vast majority of Alu elements are incapable
of making copies (Paulson & Schmid, 1986; Liu &
Schmid, 1993). The same could be true for other
SINE families as well (Kim et al., 1995; Deragon
et al., 1996). Since most SINE elements are very
similar in sequence, it seems likely that major in¯uences on whether a SINE is active or not reside in
the ¯anking sequences of speci®c elements, which
will differ for each individual element insertion.
The BC1 RNA gene, an ID SINE element, is the
only identi®ed SINE that has served as an ef®cient
``master'' element for SINE ampli®cation over a
long period of time. Since transcription is required
for retroposition, and the BC1 RNA gene is highly
Figure 3. BC1 tissue expression
pattern in 7SLBC1 transgenic and
wild-type mice. Total RNA from
different tissues was hybridized
with BC1 (unique-1) and 7SL
(a7SL-1) speci®c oligos: 1, brain; 2,
liver; 3, kidney; 4, spleen; 5, intestine; 6, testes; and 7, heart.
21
Modulation of SINE Transcription
1985). Our analysis, using several chimeric constructs, further corroborates the importance of the
upstream ¯anking sequence in the control and
in¯uence of RNA pol III transcription of SINEs.
These results showed that these SINEs, regardless
of tRNA or 7SL origin, responded approximately
the same to the various upstream sequences. In
addition, similar activation occurred, regardless of
the cell type transfected. Thus, the activator
sequences appear to be fairly generic and lead to
ubiquitous expression patterns. Previous analyses
of these four upstream sequences suggested the
presence of cis-acting upstream elements affecting
their transcription i.e. PSE, SP1, ATF, OCT or DSE
(Bredow et al., 1990; Martignetti & Brosius, 1995;
Carbon et al., 1987; Danzeiser et al., 1993;
Kickhoefer et al., 1993). In addition, TATA-like
elements are present in three of the upstream
sequences used. The 7SL upstream sequence has
no real TATA box, but has a TAGTA that may be
a functional binding site (Bredow et al., 1990).
These upstream cis-acting elements present in the
constructs may provide an advantage to the RNA
poly III internal promoter by recruiting transcription factors to the site.
The upstream sequence regulates BC1
transcription in vivo
Figure 4. BC1-RNP complex formation in different tissues of 7SLBC1 transgenic mice. Mobility shift assays of
extracts from different tissues of transgenic mice were
probed with unique-1 oligo to detect BC1 RNA: 1,
brain; 2, liver; 3, kidney; 4, heart; and 5, testes. Tissue
extracts were loaded to contain (a) equal amounts of
protein or (b) similar amounts of BC1 RNA as previously determined by protein detection assays or
Northern blot analyses. Arrows indicate the following:
BC1 RNP (®lled black arrow), expected position of free
BC1 RNA (open arrow), and the loading wells (small
arrow).
regulated transcriptionally, we decided to evaluate
the in¯uence of upstream ¯anking sequences in
SINE transcription.
Specific sequences containing RNA pol III
activating elements can upregulate
SINE transcription
Although most SINEs have the internal RNA pol
III promoters (A and B boxes), very few are
actively transcribed. Previous work indicates the
importance upstream ¯anking regions have on
RNA pol III transcription of SINEs (Martignetti &
Brosius, 1995; Chesnokov & Schmid, 1996;
Kobayashi & Anzai, 1998) and a number of other
RNA pol III-transcribed genes (Ullu & Weiner,
RNA poly III-transcribed RNAs are generally
ubiquitously expressed, making the tissue-speci®city of the BC1 RNA quite unusual. We considered the possibility that RNA poly III
transcription of the BC1 RNA gene was ubiquitous,
but that the RNA expression was primarily regulated post-transcriptionally. In our 7SLBC1 transgenic mice, the expression of the rat BC1 transgene
in all tissues con®rms that the RNA is at least
reasonably stable in other tissues. Although we
obtained only one founder mouse, it is doubtful
that our results are in¯uenced greatly by a ``position'' effect. Our goal was to eliminate the BC1
tissue-speci®c regulation. Therefore, we utilized an
upstream region from a constitutively expressing
gene, 7SL, to try to disrupt the normal regulation.
Because we obtained expression in every tissue
tested, this allowed us to demonstrate that BC1 is
stable, and makes a similar protein complex in all
tissues tested. Although the expression is now constitutive, we continue to see the same neuronal and
testes-speci®c expression preference in the transgene as for the endogenous gene. This would be
extremely unlikely to be caused by a fortuitous
integration site that mimics the endogenous gene
and almost certainly demonstrates a post-transcriptional component to the regulation. Finally, it is
worth considering the potential in¯uence of multiple copies of the transgene. As there are already
more than 10,000 BC1 closely related elements in
the mouse genome, copy number of the repetitive
portion of the element is very unlikely to have any
in¯uence. In general, multiple copies of the
upstream portion could be considered to titrate out
22
speci®c factors. However, since we are using the
upstream region to drive constitutive expression, it
seems unlikely that we could mimic the endogenous regulation. Therefore the primary repression of
the endogenous BC1 gene in those tissues must be
predominantly at the level of transcription.
These data con®rm previous work utilizing
in vitro transcription that suggested transcriptional
speci®city (Martignetti & Brosius, 1995). Work on
another BC1 transgenic mouse line (containing the
BC1 RNA gene 1.4 kb SacI-BamHI fragment)
suggests that its endogenous upstream sequence
contains suf®cient information to direct tissuespeci®c transcription (Martignetti & Brosius, 1995).
In addition, one of the ®ve E-boxes, (pol II enhancer element and binding site for myogenic factors
of the MyoD family) apparently necessary for
effective transcription of BC1, was recognized
speci®cally by brain nuclear protein(s) (Kobayashi
& Anzai, 1998). The authors suggest that the
upstream E-box and its binding protein may be
involved in the regulation of the preferential RNA
pol III BC1 expression in brain.
Although the rat BC1 transgene is expressed in
all tissues, a tissue-speci®c pattern is maintained.
Stability of the transcript may be the cause for the
differential expression of BC1 in the tissues.
Because the endogenous BC1 and the rat BC1
transgene transcripts are almost identical (only two
bases different) something other than the basic primary sequence and structure may be involved in
BC1 RNA stability. Increased BC1 stability in brain
may be due to the compartmentalization of the
BC1 RNA to the dendrites (Tiedge et al., 1991,
1991; Muslimov et al., 1997); thus removing the
BC1 RNA from the normal degradation pathway.
Alternatively, the protein interaction with RNA
may provide protection from degradation. Several
proteins have been suggested as candidates
through in vitro analyses, but have not been
directly linked to the endogenous complex
(Kobayashi et al., 1992a,b; Muramatsu et al., 1998;
Kremerskothen et al., 1998). Our ``native mobility
shift assays'' demonstrate that the BC1 RNA in
some of the tissues of the transgenic mouse tested
was present as an RNP complex. All the RNP complexes observed appear to be the same as the one
present in the brain of the wild-type mouse.
Although we cannot rule out that different tissues
have different members of a closely related family
of RNA binding proteins, it seems likely that the
RNA binding protein(s) may be of a ubiquitous
nature and have a function other than forming a
complex with BC1 RNA. In addition, the proteins
are unlikely to be involved in a purely neural function. It is interesting that no free BC1 RNA was
observed in our native mobility shift assays, even
when overexpressed to very large amounts in
brain (Figure 4(a)). There are several potential
explanations for this observation. First, the protein
normally may be expressed in large amounts, making it available to form the complex upon the transcription of BC1. Alternatively, the increased
Modulation of SINE Transcription
expression of BC1 may indirectly force the cell to
upregulate the proteins' expression by sequestering
them from their normal function. Finally, the
amount of the protein may dictate the amount of
RNP complex made, with the prompt degradation
of any free BC1 RNA.
Previous reports suggest that the autoantigen,
La, may be part of this RNP complex
(Kremerskothen et al., 1998). We have been unable
to precipitate the BC1 RNA with anti-La antibodies
(data not shown). It is possible that the appropriate
epitopes are not exposed in the RNP complex. La
is ubiquitously expressed with nuclear localization
(Keech et al., 1993; Bachmann et al., 1997). Such
published data seem inconsistent with the extremely high levels of complex that are present in the
brain cytoplasm. Thus, although we believe that La
probably associates with the BC1 RNA transiently,
it seems unlikely that it is part of the cytoplasmic
RNP complex that we observe.
Upstream modulation of general SINE
transcription and potential role in retroposition
Optimal ¯anking sequences that allow SINEs to
express strongly are probably rare. We do not yet
know enough about RNA poly III upstream activating sequences to know how position and orientation affect transcription ef®ciency. In addition to
a paucity of genomic activating sites, it is also
likely that there is a strong selection against
expression of SINEs, because a strongly transcribing element may result in a high level of deleterious mutations through retroposition. In addition,
high levels of expression may be disruptive to cell
function by competing for RNA binding proteins,
or in¯uencing processes such as translation, as has
been suggested for Alu (Chu et al., 1998).
The origin and evolution of BC1
RNA expression
The BC1 RNA gene shows sequence and
expression conservation consistent with a gene
under functional selection (Deininger et al., 1996).
Thus, the BC1 RNA gene has ``exapted'' (Brosius &
Gould, 1992) to some function following its formation in the rodent genome. It may be the fortuitous formation or insertion of the BC1 element
adjacent to upstream activating sequences that
allowed it to express and both initiate the ID
family of SINEs and continue to serve as a master
element for ampli®cation throughout the rodent
lineage (Kim et al., 1994). In a case like this, where
an element begins to express to high levels, it is
important that the expression does not compete
with normal cellular functions (i.e. sequester key
RNA binding proteins that are needed for other
RNAs). In the case of BC1, this may have been
helped by the tissue speci®city of the expression,
directing the RNA to tissues that could best accommodate the high level of the RNA. It seems likely
that BC1 RNA expression was originally fairly ubi-
23
Modulation of SINE Transcription
quitous, but that its regulation quickly evolved to
allow expression where the RNA would have the
least negative impact and that it gradually exapted
to having a positive role in those tissues.
Materials and Methods
Construction of plasmids
Constructs containing different upstream sequences of
known RNA poly III transcribed genes (7SL, U6, RVG,
and BC1) were cloned upstream of three different SINEs
(B2, rat BC1, and Alu) all with the unique 30 end from
the mouse BC1 master gene, see Figure 1. The plasmids
are named according to the upstream sequence (superscript), the SINE body, and the 30 end (superscript), for
example: p7SLAluSxBC1 or pBC1BC1BC1. The particular Alu
body is the consensus sequence for the PS (Sx) subfamily. The ID body used exactly matches the ID portion of
the BC1 master gene (Kim et al., 1994) and for the B2
body, the sequence used (a close match to the consensus)
is from the B2 element within intron 4 of the murine bglucuronidase gene (Gus-s) (Roy et al., 1998). Upstream
sequences, individual SINEs and the BC1 30 downstream
sequence were individually ampli®ed and joined by PCR
using overlapping primers (Ling & Robinson, 1997). The
®nal PCR product of the complete construct was cloned
into the pCRII or pCRII.1 vector of the TA cloning kit
(InVitrogen). Evaluation of both strands by sequencing
con®rmed the correct sequences of the constructs. Puri®ed plasmids were prepared by alkaline lysis of bacterial
cells followed by banding twice in a CsCl gradient. DNA
concentrations were determined spectrophotometrically
by using A260 and veri®ed by visual examination of ethidium bromide-stained agarose gels.
Transcription in cell lines
Transient transfections were carried out in the human
cell line HeLa (ATCC CCL2), and the rodent cell lines
Neuro2a (ATCC CCL131), NIH3T3 (ATCC CRL1658),
and C6 glioma (ATCC CCL107). Monolayers were
grown to 40-60 % con¯uency and transfected with 10 mg
of the construct-containing plasmid and 2 mg of control
plasmid using LipofectAmine1 (Gibco Life Sciences) following the manufacturer's recommended protocol. Total
RNA was isolated 16-20 hours post-transfection.
DNA and RNA analysis
Genomic DNA was isolated from mouse tails using
the Easy DNA2 kit (InVitrogen) following the manufacturer's recommended protocol. Selected tissues were
removed from euthanized mice and immediately processed. RNA was extracted from cell lines and mouse tissues utilizing the Trizol2 Reagent (Life Technologies,
Inc.) according to the manufacturer's protocol. Equal
amounts of RNA were electrophoresed on a 2 % agaroseformaldehyde gel and then transferred to a nylon membrane, Hybond-N (Amersham). Northern blots were
hybridized utilizing one or both of the following endlabeled DNA oligos: unique-1 50 -TGTGTGTGCCAGTTACCTTG-30 (complementary to the unique region of
the BC1 RNA) and a7SL 50 -CCGATCGGCATAGCGCACTA-30 (complementary to a region of the 7SL RNA
that does not share sequence similarity to Alu RNA) in
5 SSC, 5 Denhardt's, 1 % (w/v) SDS and 100 mg/ml
herring sperm DNA. Oligonucleotides were end-labeled
with [g-32P]ATP (Amersham) using T4 polynucleotide
kinase (NEB), and puri®ed by ®ltration through a Sephadex G-50 column. Blots were washed thrice at 42 C with
a low stringency buffer (2 SSC and 1 % (w/v) SDS)
and subjected to autoradiography or quanti®ed using a
Phosphorimager screen and ImageQuant Software (Molecular Dynamics). Statistical analysis was performed
using the Jandel SigmaStat Statistical Software Version 2,
Jandel Corporation.
Transgenic mouse production
Transgenic mice were produced by Chrysalis DNX
Transgenic Sciences Corporation (Princeton, NJ, USA).
A 385 bp EcoRI fragment containing the complete 7SLBC1
unit was microinjected into C57BL/6 SJLF2 hybrid
mouse eggs. Transgenic mice were identi®ed by PCR
analysis of tail DNA with transgene-speci®c primers.
The primers used were 50 -CCTCCAGACCGCCCAGTGTGGGTGT-30 (for the upstream 7SL region), and
50 -GATACTGCGATATCCTAAAGGGCAG-30 (for the
BC1 30 end). Ampli®cation was performed in an MJ thermal cycler using 50 pmol of each of the primers for ®ve
cycles as follows: 94 C, 58 C, and 72 C each for
30 seconds, followed by 25 cycles at 94 C, 46 C, and
72 C for 30 seconds each. The ®nal chain extension step
was at 72 C for ®ve minutes. Only one line, referred to
as the G line, was established containing the transgene.
Transcription of the transgene in the selected tissues was
detected by Northern blot analysis and RT-PCR with
subsequent sequencing of the cDNA.
Native mobility shift assay
Cytoplasmic extracts were obtained by differential
centrifugation as previously described (Cheng et al.,
1996; Kobayashi et al., 1992a) with minor modi®cations.
In brief, tissue (0.5 g/ml of buffer) was homogenized in
buffer: 100 mM KCl, 15 mM MgCl2, 10 mM Tris (pH 7.5),
0.32 M sucrose, 0.2 mM PMSF, 1 mM DTT, 0.1 % (w/v)
aprotinin and 1.5 ml/g tissue of protease inhibitor cocktail (SIGMA). Samples were centrifuged at 2000 g for ®ve
minutes, followed by a second spin of the supernatant at
15,000 g for ten minutes. The supernatant was further
cleared by centrifugation at 105,000 g for two hours.
Finally, glycerol was added to the supernatant to a ®nal
concentration of 20 % (v/v). Samples of the tissue
extracts were electrophoresed in high ionic non-denaturing polyacrylamide gels as described previously
(Ausubel et al., 1996). The gels were transferred to zetaprobe blotting membrane (Bio-Rad) in Tris-glycine buffer
(0.05 M Tris, 0.375 M glycine, 2 mM EDTA (pH 8.5)) for
one hour at 350 mA. UV-crosslinked membranes were
hybridized with end-labeled unique-1 oligo under the
same conditions described above.
Acknowledgments
A.M.R. was supported by a Brown Foundation fellowship from the Tulane Cancer Center. This research was
supported by National Institutes of Health RO1
GM45668 to P.L.D. We thank Dr Valerie Kickhoefer at
the Department of Biological Chemistry, UCLA School
of Medicine for generously supplying the plasmid containing the RVG upstream region and Dr David H. Kass
for his assistance with the B2 sequence.
24
References
Anzai, K. & Goto, S. (1988). ID sequence-binding protein
factor complexed in ribonucleoprotein particles.
Neurosci. Letters, 89, 246-250.
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D.,
Seidman, J. G., Smith, J. A. & Struhl, K. (1996). Current Protocols In Molecular Biology, John Wiley &
Sons, Inc., Canada.
Bachmann, M., Grolz, D., Bartsch, H., Klein, R. R. &
Troster, H. (1997). Analysis of expression of an
alternative La (SS-B) cDNA and localization of the
encoded N- and C-terminal peptides. Biochim. Biophys. Acta, 1356, 53-63.
Boeke, J., Gar®nkel, D. J., Styles, C. A. & Fink, C. R.
(1985). Ty elements transpose through an RNA
intermediate. Cell, 40, 491-500.
Bredow, S., Surig, D., Muller, J., Kleinert, H. & Benecke,
B. J. (1990). Activating-transcription-factor (ATF)
regulates human 7S L RNA transcription by RNA
polymerase III in vivo and in vitro. Nucl. Acids Res.
18, 6779-6784.
Brosius, J. & Gould, S. J. (1992). On ``genomenclature'': a
comprehensive (and respectful) taxonomy for pseudogenes and other ``junk DNA''. Proc. Natl Acad.
Sci. USA, 89, 10706-10710.
Carbon, P., Murgo, S., Ebel, J. P., Krol, A., Tebb, G. &
Mattaj, I. W. (1987). A common octamer motif binding protein is involved in the transcription of U6
snRNA by RNA polymerase III and U2 snRNA by
RNA polymerase II. Cell, 51, 71-79.
Cheng, J. G., Tiedge, H. & Brosius, J. (1996). Identi®cation and characterization of BC1 RNP particles.
DNA Cell Biol. 15, 549-559.
Chesnokov, I. & Schmid, C. W. (1996). Flanking
sequences of an Alu source stimulate transcription
in vitro by interacting with sequence-speci®c transcription factors. J. Mol. Evol. 42, 30-36.
Chesnokov, I., Chu, W. M., Botchan, M. R. & Schmid,
C. W. (1996). p53 inhibits RNA polymerase IIIdirected transcription in a promoter-dependent
manner. Mol. Cell Biol. 16, 7084-7088.
Chu, W. M., Liu, W. M. & Schmid, C. W. (1995). RNA
polymerase III promoter and terminator elements
affect Alu RNA expression. Nucl. Acids Res. 23,
1750-1757.
Chu, W. M., Ballard, R., Carpick, B. W., Williams, B. R.
& Schmid, C. W. (1998). Potential Alu function:
regulation of the activity of double-stranded RNAactivated kinase PKR. Mol. Cell. Biol. 18, 58-68.
Daniels, G. & Deininger, P. (1991). Characterization of a
third major SINE family of repetitive sequences in
the galago genome. Nucl. Acids Res. 19, 1649-1656.
Danzeiser, D. A., Urso, O. & Kunkel, G. R. (1993). Functional characterization of elements in a human U6
small nuclear RNA gene distal control region. Mol.
Cell. Biol. 13, 4670-4678.
Das, G., Henning, D., Wright, D. & Reddy, R. (1988).
Upstream regulatory elements are necessary and
suf®cient for transcription of a U6 RNA gene by
RNA polymerase III. EMBO J. 7, 503-512.
Deininger, P. L., Tiedge, H., Kim, J. & Brosius, J. (1996).
Evolution, expression, and possible function of a
master gene for ampli®cation of an interspersed
repeated DNA family in rodents. In Progress in
Nucleic Acid Research and Molecular Biology (Cohn,
W. E. & Moldave, K., eds), pp. 67-88, Academic
Press, San Diego.
Modulation of SINE Transcription
Deragon, J. M., Gilbert, N., Rouquet, L., Lenoir, A.,
Arnaud, P. & Picard, G. (1996). A transcriptional
analysis of the S1Bn (Brassica napus) family of SINE
retroposons. Plant Mol. Biol. 32, 869-878.
Englander, E. W. & Howard, B. H. (1995). Nucleosome
positioning by human Alu elements in chromatin.
J. Biol. Chem. 270, 10091-10096.
Englander, E. W., Wolffe, A. P. & Howard, B. H. (1993).
Nucleosome interactions with a human Alu
element. Transcriptional repression and effects of
template methylation. J. Biol. Chem. 268, 1956519573.
Fuhrman, S., Deininger, P. L., LaPorte, P., Friedmann, T.
& Geiduschek, P. (1981). Analysis of transcription
of the human Alu family of ubiquitous repeating
element by eukaryotic RNA polymerase III. Nucl.
Acid. Res. 9, 6439-6455.
Keech, C. L., Gordon, T. P., Reynolds, P. & McCluskey,
J. (1993). Expression and functional conservation of
the human La (SS-B) nuclear autoantigen in murine
cell lines. J. Autoimmun. 6, 543-555.
Kickhoefer, V. A., Searles, R. P., Kedersha, N. L.,
Garber, M. E., Johnson, D. L. & Rome, L. H. (1993).
Vault ribonucleoprotein particles from rat and bullfrog contain a related small RNA that is transcribed
by RNA polymerase III. J. Biol. Chem. 268, 78687873.
Kim, J., Martignetti, J., Shen, M. R., Brosius, J. &
Deininger, P. (1994). The Rodent BC1 RNA gene is
a master gene for ID element ampli®cation. Proc.
Natl Acad. Sci. USA, 91, 3607-3611.
Kim, J., Kass, D. H. & Deininger, P. L. (1995). Transcription and processing of the rodent ID repeat family
in germline and somatic cells. Nucl. Acids Res. 23,
2245-2251.
Kleinert, H., Gladen, A., Geisler, M. & Benecke, B. J.
(1988). Differential regulation of transcription of
human 7 S K and 7 S L RNA genes. J. Biol. Chem.
263, 11511-11515.
Kobayashi, S. & Anzai, K. (1998). An E-box sequence
acts as a transcriptional activator for BC1 RNA
expression by RNA polymerase III in the brain. Biochem. Biophys. Res. Commun. 245, 59-63.
Kobayashi, S., Goto, S. & Anzai, K. (1991). Brain-speci®c
small RNA transcript of the identi®er sequences is
present as a 10 S ribonucleoprotein particle. J. Biol.
Chem. 266, 4726-4730.
Kobayashi, S., Higashi, N., Suzuki, K., Goto, S., Yumoto,
K. & Anzai, K. (1992a). The 10 S BC-1 ribonucleoprotein particle contains identi®er sequence-binding
proteins that interact with an array of GCAAG/
CTTGC motifs between split promoter sequences
for RNA polymerase III. J. Biol. Chem. 267, 1829118297.
Kobayashi, S., Tokuno, T., Suzuki, K., Higashi, N., Goto,
S. & Anzai, K. (1992b). Developmental change in
subcellular location of Bp-1 protein with an ability
to interact with both identi®er sequence and its
brain-speci®c transcript, BC-1 RNA. Biochem. Biophys. Res. Commun. 189, 53-58.
Kremerskothen, J., Nettermann, M., Op, D. B.,
Bachmann, M. & Brosius, J. (1998). Identi®cation of
human autoantigen La/SS-B as BC1/BC200 RNAbinding protein. DNA Cell. Biol. 17, 751-759.
Labuda, D. & Zietkiewicz, E. (1994). Evolution of secondary structure in the family of 7SL-like RNAs.
J. Mol. Evol. 39, 506-518.
Modulation of SINE Transcription
Ling, M. M. & Robinson, B. H. (1997). Approaches to
DNA mutagenesis: an overview. Anal. Biochem. 254,
157-178.
Liu, W. M. & Schmid, C. W. (1993). Proposed roles for
DNA methylation in Alu transcriptional repression
and mutational inactivation. Nucl. Acids Res. 21,
1351-1359.
Liu, W. M., Maraia, R. J., Rubin, C. M. & Schmid, C. W.
(1994). Alu transcripts: cytoplasmic localisation and
regulation by DNA methylation. Nucl. Acids Res. 22,
1087-1095.
Martignetti, J. A. & Brosius, J. (1995). BC1 RNA: transcriptional analysis of a neural cell-speci®c RNA
polymerase III transcript. Mol. Cell. Biol. 15, 16421650.
Matera, A. G., Hellmann, U. & Schmid, C. W. (1990). A
transpositionally and transcriptionally competent
Alu subfamily. Mol. Cell. Biol. 10, 5424-5432.
Muramatsu, T., Ohmae, A. & Anzai, K. (1998). BC1
RNA protein particles in mouse brain contain two
y-, h-element-binding proteins, translin and a
37 kDa protein. Biochem. Biophys. Res. Commun. 247,
7-11.
Muslimov, I. A., Santi, E., Homel, P., Perini, S., Higgins,
D. & Tiedge, H. (1997). RNA transport in dendrites:
a cis-acting targeting element is contained within
neuronal BC1 RNA. J. Neurosci. 17, 4722-4733.
Paulson, K. E. & Schmid, C. W. (1986). Transcriptional
inactivity of Alu repeats in HeLa cells. Nucl. Acids
Res. 14, 6145-6158.
Rogers, J. H. (1985). The origin and evolution of retroposons. Int. Rev. Cytol. 93, 187-279.
Rogers, J. H. & Willison, K. R. (1983). A major
rearrangement in the H-2 complex of mouse t haplotypes. Nature, 304, 549-552.
25
Roy, A. M., Gong, C., Kass, D. H. & Deininger, P. L.
(1998). Recent B2 element insertions in the mouse
genome. DNA Sequence, 8, 343-348.
Schmid, C. W. & Maraia, R. (1992). Transcriptional regulation and transpositional selection of active SINE
sequences. Curr. Opin. Genet. Dev. 2, 874-882.
Shaikh, T. H. & Deininger, P. L. (1996). The role and
ampli®cation of the HS Alu subfamily founder
gene. J. Mol. Evol. 42, 15-21.
Slagel, V. & Deininger, P. (1989). In vivo transcription of
a cloned prosimian primate SINE sequence. Nucl.
Acids Res. 17, 8669-8682.
Sutcliffe, J. G., Milner, R. J., Gottesfeld, J. M. &
Reynolds, W. (1984). Control of neuronal gene
expression. Science, 225, 1308-1315.
Tiedge, H., Fremeau, R. T. J., Weinstock, P. H., Arancio,
O. & Brosius, J. (1991). Dendritic location of neural
BC1 RNA. Proc. Natl Acad. Sci. USA, 88, 2093-2097.
Tiedge, H., Chen, W. & Brosius, J. (1993). Primary structure, neural-speci®c expression, and dendritic
location of human BC200 RNA. J. Neurosci. 13,
2382-2390.
Ullu, E. & Weiner, A. M. (1985). Upstream sequences
modulate the internal promoter of the human 7SL
RNA gene. Nature, 318, 371-374.
Vorce, R. L., Lee, B. & Howard, B. H. (1994). Methylation- and mutation-dependent stimulation of Alu
transcription in vitro. Biochem. Biophys. Res. Commun.
203, 845-851.
Weiner, A., Deininger, P. & Efstradiatis, A. (1986). The
reverse ¯ow of genetic information: pseudogenes
and transposable elements derived from nonviral
cellular RNA. Annu. Rev. Biochem. 55, 631-661.
Willis, I. M. (1993). RNA polymerase III. Genes, factors
and transcriptional speci®city. Eur. J. Biochem. 212,
1-11.
Edited by M. Gottesman
(Received 28 February 2000; received in revised form 16 June 2000; accepted 12 July 2000)