Thyroid hormone influences conditional transcript elongation of the apolipoprotein A-I gene in rat liver

Abstract

Chronic administration of thyroid hormone (T3) increases apoA-I gene expression in rat liver by enhancing mRNA maturation, but reduces apoA-I mRNA synthesis to 50% of control. To gain insight into the inverse relation of mRNA maturation and mRNA synthesis, we measured transcription in livers of control and T3-treated rats (50 micrograms/100 g body weight for 7 days) by nuclear run-on assays using overlapping antisense RNA probes encompassing the apoA-I gene. In control rats, after normalization for hybridization efficiency and probe length, the hybridization signals with intron 3 probes were reduced to 45% of those obtained with exon 1 to exon 3 probes (P < 0.01) indicating transcriptional arrest or pausing close to the exon 3-intron 3 border or 450 to 650 nucleotides downstream of the transcription start site. In T3-treated rats, the elongation block was nearly twice as effective, while the rate of transcription initiation was similar to control. In contrast, the distribution of nascent transcripts across the apoA-IV gene was symmetric, and T3-treatment suppressed apoA-IV mRNA synthesis by processes operating in the 5' region such as transcription initiation. Thus, conditional transcript elongation contributes to the regulation of apoA-I gene expression in rat liver.

Full text

Thyroid hormone influences conditional transcript
elongation
of
the apolipoprotein
A4
gene
in
rat liver
Yen-Chiu Lin-Lee, Selma
M.
Soyal, Andrei
Surguchov,
Sandra Sanders, Wolfgang Strob1,l
and Wolfgang Patsch‘
Department
of
Medicine, Baylor College
of
Medicine, Houston,
TX
77030
Abstract
Chronic administration
of
thyroid hormone (T3)
increases apoA-I gene expression in rat liver by enhancing
mRNA
maturation, but reduces apoA-I
mRNA
synthesis
to
50% of control. To gain insight into the inverse relation of
mRNA
maturation and
mRNA
synthesis, we measured tran-
scription in livers of control and T3-treated rats (50 pg/lOO g
body weight for
7
days) by nuclear run-on assays using over-
lapping antisense
RNA
probes encompassing the apoA-I
gene. In control rats, after normalization for hybridization
efficiency and probe length, the hybridization signals with
intron 3 probes were reduced to 45% of those obtained with
exon
1
to exon 3 probes
(P
<
0.01) indicating transcriptional
arrest or pausing close to the exon 3-intron
3
border or 450
to 650 nucleotides downstream
of
the transcription start site.
In Ts-treated rats, the elongation block was nearly twice as
effective, while the rate of transcription initiation was similar
to control.
In
contrast, the distribution of nascent transcripts
across the apoA-IV gene was symmetric, and T3-treatment
suppressed apoA-IV
mRNA
synthesis by processes operating
in the
5’
region such as transcription initiation. Thus,
conditional transcript elongation contributes to the regula-
tion
of
apoA-I gene expression in rat liver.-Lh-Lee,
Y-C.,
S.
M.
Soyal,
A.
Surguchov,
S.
Sanders,
W.
Strobl,
and
W.
Patsch.
Thyroid hormone influences conditional transcript
elongation
of
the apolipoprotein
A-I
gene in rat 1iver.J.
Lipid
RH.
1995.
36:
1586-1594.
Supplementary key
words
gene expression transcript arrest
Apolipoprotein (apo) A-I is the main apolipoprotein
of HDL and its plasma concentration is inversely associ-
ated with the incidence of coronary artery disease
(CAD)
(1).
By binding to cell surface proteins, opera-
tionally termed HDL-receptors
(2),
apoA-I promotes the
translocation of cholesterol from intracellular pools to
the cell membrane, facilitates the transfer of cholesterol
from cell membranes
to
nascent HDL, and traps choles-
terol via 1ecithin:cholesterol acyltransferase-mediated
esterification in the core of HDL particles
(3-5).
Hence,
apoA-I is critically involved in the initial phase of reverse
cholesterol transport, a function which may, at least in
part, explain its antiatherogenic role.
In most mammalian species, the apoA-I gene is
pri-
marily expressed in liver and intestine (6-8), but the
mechanisms controlling changes in apoA-I mRNA ex-
pression in response to metabolic signals differ between
the two tissues (8-10). In some animal models, changes
in plasma apoA-I levels resulting from dietary or hormo-
nal perturbations correlate with changes in hepatic, but
not intestinal apoA-I mRNA concentrations (10-15).
Experiments in transgenic mice as well as in vitro trans-
fection studies have identified the apoA-I gene elements
mediating hepatocyte-specific expression (8,9, 16). Sev-
eral nuclear proteins converge at three distinct sites in
the
5’
flanking region
of
the apoA-I gene and govern its
expression through synergistic interactions
(
1’7,18).
The
frequency of transcription initiation may therefore be
an
important control point in hepatic apoA-I gene ex-
pression, but the significance of transcriptional regula-
tion for changes in apoA-I gene expression in vivo has
been addressed only in very few studies
(15,
19-22).
Even less is known whether transcription initiation is the
sole control point in apoA-I mRNA synthesis or whether
other mechanisms such as conditional transcript elon-
gation contribute to changes in apoA-I gene expression
in vivo.
Among physiological perturbations, changes in thy-
roid hormone status are associated with distinct changes
of
hepatic apoA-I gene expression (10, 15, 19,
23).
A
single receptor-saturating dose of triiodothyronine (T3)
increases hepatic apoA-I gene transcription, abundance
levels of nuclear and total hepatic apoA-I mFWA, and
plasma apoA-I levels
(15,
19). After repeated daily injec-
Abbreviations: apo, apolipoprotein;
Ts,
triiodothyronine.
‘Present address: Department
of
Pediatrics, University
of
Vienna,
Austria.
2To
whom correspondence should be addressed at: Department
of
Laboratory Medicine, Landeskrankenanstalten Salzburg, A-5020
Salzburg, Austria.
1586
Journal of
Lipid
Research
Volume
36,
1995
by guest, on July 14, 2011www.jlr.orgDownloaded from

tions of T3, plasma apoA-I levels
as
well
as
nuclear and
total hepatic apoA-I mRNA levels remain elevated, but
transcription from the apoA-I gene is reduced to
30-50% of control animals (15). Such a relation between
apoA-I mRNA abundance and mRNA synthesis implies
that posttranscriptional events control apoA-I gene ex-
pression and raises the possibility of inhibition of tran-
scription by processes related to mRNA maturation.
Characterization of the nuclear apoA-I mRNA matu-
ration showed that the decrease in apoA-I gene tran-
scription in Tytreated rats was associated with a com-
mensurate decrease in the abundance of primary
transcripts (24). However, the abundance of mature
nuclear and cytoplasmic mRNA was 3-fold higher in
Ts-treated than in control rats. Compartmental model-
ing
of
apoA-I mRNA processing suggested that chronic
T3 treatment enhances mRNA maturation 7-fold by
protecting the mRNA precursor devoid of intron 2, but
containing introns
1
and
3
from degradation and/or
facilitating the splicing
of
intron
1
from this precursor
(24).
To gain insight into the inverse association between
rates of apoA-I mRNA maturation and gene transcrip-
tion,
we
measured transcription across
the
apoA-I gene
with overlapping single-stranded probes. We report
here that transcription
of
the apoA-I gene is hindered
by an elongation block in the basal state. In chronically
hyperthyroid rats, this elongation block is twice as effec-
tive as under basal conditions and accounts for the
decreased apoA-I mRNA synthesis rate.3
MATERIALS AND METHODS
[5' y32PIdATP (4500 Ci/mM) and [5' a-32PIdCTP
(3000 Ci/mM) were purchased from ICN Radiochemi-
cal~ (Irvine, CA); [a--S5S]dATP (600 Ci/mM) was from
Amersham Corp. (Arlington Heights, IL); [5' a-32P]UTP
(3000 ci/mM) was from New England Nuclear Re-
search/DuPont (Boston, MA). T4 DNA ligase, T4
polynucleotide kinase, calf intestine alkaline phos-
phatase, proteinase
K,
placental ribonuclease inhibitor,
RNase-free DNase I, DNase-free RNase, restriction en-
zymes, and a-amanitin were obtained from Boehringer
Mannheim (Indianapolis, IN). DNA polymerase I and
Klenow-large fragment were from GIBCO BRL Life
Technologies, Inc. (Gaithersburg, MD). Amplitaq@
DNA polymerase was from Perkin-Elmer Cetus (Nor-
walk,
CT),
Qiaex from Qiagen Inc. (Chatsworth, CA),
and Chroma Spin-100 columns from Clontech Labora-
tories, Inc. (Palo Alto, CA). The Sequenase@ Version 2.0
SPart
of
this
research appeared
in
abstract
form
(25).
kit was from United States Biochemical Corp. (Cleve-
land,
OH)
and the MEGAscriptTM in
vitro
Transcrip-
tion Kit was from Ambion Inc. (Austin,
TX).
Nitrocellu-
lose membranes were obtained from Stratagene
Cloning Systems (La Jolla, CA) and
ISS
PromPP" was
from Integrated Separation Systems (Natick, MA).
Experimental
animals
and
isolation
of
nuclei
Adult male Sprague-Dawley rats (Texas Animal Spe-
cialties, Humble, TX) weighing about 250 g were housed
in a room with a 12-h light cycle (7-19 h). Animals were
fed normal rat chow ad libitum. T3 was dissolved in 0.15
N NaCl, pH
11.
Animals were injected with T3 (50
pg/100 g body weight) subcutaneously for
7
days. Rats
serving as injection controls received the alkaline 0.15
N NaCl solution only. Food was removed at
9
AM,
and
2-4 h later animals were anesthetized with pentobarbital
(5
mg/100 g). Rat livers were removed and liver cell
nuclei were prepared by the method of Northemann et
al.
(26) as described (27). The DNA content of the nuclei
was determined by a fluorimetric assay (28) using
salmon sperm DNA as a standard.
Cell-free
transcription
Nascent 3*P-labeled RNA transcripts were obtained
from isolated hepatocyte nuclei by the method
of
Birch
and Schreiber (29) as described previously (27). Nuclei
(0.5-1
x
lo8)
were incubated in a total volume of 350 p1
containing 50 mM HEPES, pH 7.5,50 mM NaC1,2.5 mM
MgC12, 0.05 mM EDTA,
5
mM dithiothreitol,
1
mM of
each ATP, CTP, GTP, 2 mM creatine phosphate, 2 pg
creatine phosphokinase, 25% glycerol,
20
pg heparin,
1
mM spermine,
1
mM EGTA, 0.1 mM phenylmethylsulfone
fluoride,
60
units of human placental ribonuclease in-
hibitor
and
100
pCi of [32P]UTP at 26°C for 30 min.
After incubation, the reaction mixture was treated with
30 units of DNase1 for 10 min and digested with 140
pg,"l proteinase
K
and
0.5%
SDS for 30 min at 37°C.
RNA was extracted twice with phenol-chloro-
form-isoamyl alcohol 25:24:
1
(v/v/v)
and precipitated
from the aqueous phase with ethanol. Unincorporated
[3*P]UTP was removed by Chroma Spin-100 columns.
Nascent 32P-labeled transcripts were partially hydro-
lyzed with 0.2 N NaOH, 10 mM EDTA, 0.2% SDS at
4°C
for
15
min and neutralized with
0.5
M
HEPES (30) prior
to ethanol precipitation with
1
pl
ISS
promPPT. Total
[32P]UTP incorporation ranged from 0.15 to 40
pmol/mg DNA per min. Under these conditions, tran-
scription was DNA-dependent, and RNA polymerase
activity amounted to
55%
of total transcription. Tran-
scription from either the apoA-I or apoA-IV gene was
completely abolished by 2.5 pg,"l a-amanitin (15).
RNA was synthesized by in vitro transcription of
pGEMSZf clones containing various rat apoA-I and
Lin-Lee
et
al.
ApoA-I
gene
transcription
and
thyroid
hormone
1587
by guest, on July 14, 2011www.jlr.orgDownloaded from

apoA-IV gene inserts. ApoA-I and apoA-IV gene seg-
ments were obtained by the amplification of genomic
DNA using the polymerase chain reaction (31). Primers
shown in
Table
1
were synthesized using a Cyclone Plus
DNA synthesizer and reagents from Milligen Biosearch
Division (Burlington, MA). PCR assays contained
1
pg
DNA isolated from rat liver (32), 0.2 p~ of each up-
stream and downstream primer, 200 p~ of each dNTP,
units of AmplitaqB in a 100 pl reaction volume that was
overlaid with mineral oil. Samples were subjected to
initial denaturation for 5 min at 94°C; 30 cycles of
amplification each consisting of
1
min at 60°C (anneal-
ing),
1
min at 72°C (extension), and
1
min at 94°C
(denaturation); and a final extension at 72°C for 10 min.
PCR products were separated by electrophoresis in
1.2% agarose, purified using Qiaex, repaired using
Klenow large fragment DNA polymerase I, and inserted
into the SmaI site of pGEMSZ by blunt-end ligation as
described (33). The orientation of inserts was deter-
mined by sequencing (34) which
also
served to verify the
identity of cloned DNA segments.
After linearization of
DNA,
sense and antisense
RNA
was transcribed from the SP6 or T7 promoter using
MEGAscript, the respective polymerases, and
m7G( 5’)ppp( 5’)G according to the instructions of the
manufacturer. 3H-labeled RNA was synthesized by in-
cluding 5 pCi [3H]UTP in in vitro transcription mix-
tures. To determine hybridization background, RNA
was transcribed using T7 RNA polymerase and pGEM-
32
linearized with HaeIII as a template. RNA was dis-
solved in 7.5
x
SSC, 37.5% formaldehyde, incubated at
65°C for 2 h and applied to nitrocellulose filters (2
pg/dot) by dot blotting (35). Filters were air-dried,
baked at 80°C for 2 h, and prehybridized overnight with
0.15 ml
of
20
mM PIPES, pH 6.4,
0.8
N NaC1, 2 mM
EDTA, pH
8.0,
2
x
Denhardt solution, 0.2% SDS, 50%
formamide, 200 pg/ml tRNA,
1
pg/ml poly
(A).
Hy-
10
mM Tris-HC1, pH
8.3,
50 mM KCl, 2.5 mM MgC12, 2.5
bridization was carried out in polypropylene tubes for
60 h at 46°C in a total volume of 0.15 ml prehybridiza-
tion solution with 1-10
x
lo6
cpm
of
extracted nuclear
[32P]RNA. To monitor hybridization efficiency, 3H-la-
beled sense RNA, transcribed from clones containing
the respective apoA-I or apoA-IV gene inserts, was in-
cluded in the hybridization reaction. After hybridiza-
tion, filters were washed three times with 2
x
SSC, 0.1%
SDS
for 30 min at room temperature, then twice with
0.1
x
SSC, 0.1% SDS at 49°C for 15 min. After incubation
with 0.125 pg/ml RNase at room temperature for 10
min, filters were incubated with 100 pg/ml proteinase
K
at 37°C for 30 min (30, 35). Nascent 32P-labeled
transcripts bound to filters were quantified by using a
Betascope 603 Blot Analyzer (Betagen Corp., Waltham,
MA). In addition, filters were subjected to autoradiog-
raphy using Kodak X-OMAV’AR film (New Haven,
CT). Relative rates of apoA-I and apoA-IV mRNA
syn-
thesis were calculated by subtracting the counts per
minute of 32P bound to filters containing RNA tran-
scribed from nonrecombinant pGEM-3Z from the
counts per minute of 32P bound to filters with RNA
transcribed from clones containing apoA-1 or apoA-IV
gene inserts. Counts per minute bound were divided
by
the 32P-labeled RNA input. Values were corrected for
hybridization efficiency and divided
by
the number of
nucleotides per probe
to
compensate for probe length,
RESULTS AND DISCUSSION
Three antisense and three sense RNA transcribed
from cloned DNA containing near contiguous se-
quences spanning all four exons
of
the apoA-I gene were
used to measure transcriptional activities across the
apoA-I gene
and
to
distinguish between mRNA synthe-
sis from the coding and non-coding strands
(Fig.
1).
In
control rats, transcriptional activity was clearly detect-
TABLE
1.
Oligonucleotides used for amplification of DNA fragments of the apoA-I
and
apoA-IV genes
Fragment Upstream Primer Downstream Primer
APOA-I
a
b
d
e
f
g
h
C
5’-GACTG’ITGGAGAGCTCCGS’
(-3, +14)
5’-CGGCAGAGACTATGTGTCCCA-3’
(+535, +555)
5’-CATGCGTGTGAATGCAG-3’ (+1453, +1469)
5’-C’ITCAGGATGAAAGCTGCA-3‘
(+233, +251)
5’-ATGATCCTGTAACTGAGCTG-3’
(+667, +686)
5’-GGATCCGCC’ITGCAACTGGCACCAC-3‘
(+899, +9 18)
5’-GCTGCTCTC’ITCCCCTCTAG-3’
(+1120, +1139)
5’-GAGIITCTGGCAGCAACATGAGCS’
(+449, +470)
5’-TCATC”‘GCTGCCAGAAC-3’
(+468, +451)
5’-GTCGACTAGCCCAGAACTCCTGAGT-3’
(+123i +1214)
5’GTGTCGACGTCTCATACTCTAAACCS’
(+1942, +1925)
5’-GG’ITCCTCTGCCCACCCT-3’
(+657,+640)
5’CCACGATCACAGATGTGG’IT-3’
(+935, +916)
5’-GTCTGCAGATCCATGCACATG-3’
(+1086, +1066)
5’-CTCCTCGTTCCAC’ITCTCCT-3’
(+1345, +1326)
5’-CCTTCCAGGC’ITCCAGCA-3’
(+1673, +1656)
ApoA-IV
a
5’-TCCTCACAGCGACACAGTGA-3’
(+2,
+2
1)
5’-AG’ITG’lTCCACAGCCTCCTT-3‘
(+5 12, +493)
b
5’-GACATCAGAGTClTGCCTCT-3’
(+62
1,
+640)
5’-CCTCCATCTTGTCCCTGTAG-3’
(+1113, +1094)
C
5’-CTGGAAGACCXGCGCAGCAG3’
(+1687, +1706)
5’-CTCCTGGACCTGTl’CCTGAA-3’
(+2 187, +2168)
Numbers in parentheses are relative to the major transcription
start
sites (1
1).
1588
Journal
of
Lipid
Research
Volume
36,1995
by guest, on July 14, 2011www.jlr.orgDownloaded from

a
I
b
I
E
(-3
(+M)
(+sn
(+1232) (+1453) (+194t)
Fig.
1.
Elongation block in apoA-I gene transcription in livers of control and TJ-treated rats. A, Map of the rat
apoA-I gene; A,
B,
C,
D,
refer to exons
1,
2,
3,
and
4;
a,
b, and c represent the portions
of
the apoA-I gene
encompassed by in vitro synthesized
RNA
probes. Numbers in parentheses are relative to the major transcription
start site
(1
1).
B,
Autoradiograph of dot blot hybridization
of
32P-labeled nuclear RNA to excess antisense
(-)
or sense
(+)
RNA
probes a, b, and c. Nuclei were isolated from the livers
of
control
rats
or
rats
injected for
7
days
with
50
pg/lOO
g
body weight of TJ (T3) and then allowed to incorporate [q2P]UTP in vitro. The number
of
labeled
U
residues in the sequence, complementary to each
(-)
probe is
122
for probe a,
144
for probe b,
and
97
for probe c. P represents background hybridization to an RNA transcribed from nonrecombinant
pGEM-3Z which contains an irrelevant sequence.
able with the antisense
RNA
probe
a,
extending from
the transcription start site into exon
3
of the apoA-I
gene. The signal intensity considerably decreased with
the antisense
RNA
probe
b
which extended from exon
3
into exon
4.
Hybridization signals obtained with probe
c,
which was shorter than probe
b
and extended from
exon
4
into the
3'
untranslated region of the apoA-I
gene, showed intensities comparable to those with
probe
b.
The drop in transcriptional activity in going
from probe
a
to probe
b (Table
2)
was confirmed in two
other sets of animals and was a consistent finding in
more than
10
independent liver samples. Southern blots
of genomic
DNA
restricted with BamHI or SstI and
probed with the
DNA
inserts of clones
a, b,
and
c
demonstrated only one band (not shown). Thus, the
decrease in signal intensity was not confounded by
repetitive sequences of probe
a.
Furthermore, the nu-
cleotide composition of probes
a
and
b
provided no
explanation for the drop in incorporation. In addition,
hybridization signals with probes
a, b,
and
c
were de-
creased by more than
95%
when
2.5
pg/ml a-amanitin
was included during nuclear run-on experiments (data
not shown). The newly synthesized
32P-RNA
transcripts
were therefore likely to result from RNA polymerase I1
activity.
In T3-treated rats, the signal intensity with probe
a
was
similar to that of control rats (Fig.
1
and Table
2).
However, the drop in signal-intensity was more pro-
nounced in going from probe
a
to probe
b
(Table
2).
Virtually no hybridization signal was detected using
sense
RNA
as
a hybridization probe in both control and
Ts-treated rats indicating that transcription from the
non-coding strand
was
negligible. The decrease in the
levels
of
RNA
polymerase I1 activity downstream of the
sequence encompassed by probe
a
was consistent with
RNA polymerase I1 pausing, attenuation, or premature
Lin-Lee et
al.
ApoA-I gene transcription and thyroid hormone
1589
by guest, on July 14, 2011www.jlr.orgDownloaded from

TABLE
2.
Effect of chronic
Ts
administration
on
transcriptional activity across the apoA-I gene in
rat
liver
mRNA
Synthesis
Hybridization Probe
Control
-r:*
pm/nucleotide
%
pm/nucleotide
v/o
~
a
(-3
to
+468) 0.148
f
0.018b,C
100
f
12 0.147
f
0.024d,e
99
f
16
b
(+535
to
+1232) 0.095
f
0.014h,f 64
f
9
0.038
It
0.003d,'
26
f
2
c
(+1453
to
+1942) 0.073
f
0.039
49
f
24
0.055
f
0.015'' 37
f
10
ApoA-I mRNA synthesis rates were measured by nuclear
run-on
assays using nuclei from four or five individual
rat livers per group. Nuclei were isolated from livers of control rats or rats injected with
50
pg of
TS
per 100 g
body weight for
7
days. mRNA synthesis rates are expressed
as
parts per million (ppm) of input 3*P-labeled RNA
(3
x
lo6
cpm) and are means
f
SD.
Results are corrected for the number of nucleotides per probe and for
hybridization efficiency which averaged
16, 13,
and
13%
for probes
a,
b,
and
c,
respectively. mRNA synthesis
rates are also expressed
as
percentage values of probe
a
in control.
dAntisense RNA obtained by in vitro transcription; numbers in parentheses refer to nucleotides relative to
the major transcription start site
(1 1).
b.cJ€'
<
0.05;
d,eP
<
0.001,
two-way analysis of variance.
termination. The difference in probe
b
hybridization
signals between control and Ts-treated rats, despite
similar signal-intensities with probe
a,
was likely
to
re-
flect specific processes that occurred during transcrip-
tion of the apoA-I gene as a result
of
T3 administration.
To ascertain the specificity of hormonal effects on
apoA-I transcript elongation, we measured RNA polym-
erase I1 densities across the apoA-IV gene. Using full-
length apoA-IV cDNA as a hybridization probe, we have
previously shown that chronic T3 administration de-
creases apoA-IV gene transcription
to
50%
of control
(36).
In T3-treated rats, apoA-IV gene transcription de-
creased to
51%
of control when measured with the
5'
antisense RNA probe
a
extending from position
2
to
5
12
relative to the transcription start site
(Table
3,
Fig.
2).
Furthermore, signal intensities obtained with probes
b
and
c
encompassing intron
2
and exon
3
showed the
increases expected for the
3'
nascent transcripts in the
absence of elongation blocks or pauses. The weak hy-
bridization signals obtained with the
3'
sense RNA were
a consistent finding in several liver samples and sug-
gested transcription from the non-coding strand of the
apoA-IV gene, but do not affect the conclusion that
chronic TS administration suppresses apoA-IV gene
transcription by processes operating in the
5'
region of
the gene such as transcription initiation.
To more closely map the site at which elongation
of
apoA-I nascent transcripts is inhibited, overlapping
RNA probes containing about
500
nucleotides each and
encompassing the apoA-I gene region from exon
2
to
exon
4
were used to quantify 32P-transcripts synthesized
in run-on assays
(Table
4,
Fig.
3).
Corrected hybridiza-
tion signals with probe
d
were similar to those obtained
with probe
a
in both control and experimental rats. A
decline in hybridization signals was observed with
probes
e
and
f.
These findings in normal rats are con-
TABLE
3.
Effect of chronic T3 administration
on
transcriptional activity across the apoA-IV gene in rat liver
mRNA
Sytithesis
Hybridization Probe"
Control
T3
pm/nucleotide
%
pm/nucleotide
%
a
(+2
to
+512)
b
(+535
to
+1232)
c
(+1453
to
+1942)
0.117
f
0.020b 100
f
17 0.060
f
0.012bc 51
*
10
0.137
f
0.040d 117
f
34 0.094
f
0.012'.d 80
f
10
0.134
f
0.011 114f9 0.117
f
0.021 100
f
18
ApoA-IV mRNA synthesis rates were measured by nuclear
run-on
assays using nuclei from four
or
five
individual rat livers per group. Nuclei were isolated from livers of control rats or rats injected with
50
pg of
TS
per
100
g
body weight for
7
days. mRNA synthesis rates are expressed
as
parts per million (ppm) of input
S'P-labeled RNA
(3
x
lo6
cpm) and are means
f
SD.
Results are corrected for the number of nucleotides per
probe and for hybridization efficiency, which averaged
15, 18,
and
18%
for probes
a,
b,
and
c,
respectively.
mRNA synthesis rates are also expressed
as
percentage values of signals obtained with probe
a
in control.
XAntisense RNA obtained by in vitro transcription; numbers in parentheses refer to nucleotides relative to
the transcription start site
(1 1).
br3dData pairs significantly different
(P
<
0.05)
using two-way analysis of variance.
1590
Journal
of
Lipid
Research
Volume
36,
1995
by guest, on July 14, 2011www.jlr.orgDownloaded from

A.
B.
a
ib
(+2)
(+512)
(+62l)
(+1113)
(+1687)
(+2187)
Fig.
2.
Transcriptional activity across the apoA-IV gene in livers of control and Ts-treated rats. A, Map of the
rat apoA-IV gene;
A,
B,
and
C
refer to exons
1.2,
and
3;
a, b, and c represent the portions of the apoA-IV gene
encompassed by in vitro synthesized RNA probes. Numbers in parentheses are relative to the transcription start
site
(1
1).
B, Autoradiograph of dot blot hybridiiration of szP-labeled nuclear RNA to excess antisense
(-)
or
sense
(+)
RNA probes a, b, and c. Nuclei were isolated from the livers of control rats
(Co)
or
rats injected with
TS
(T3)
and then allowed to incorporate [J2P]UTP in vitro. The number of labeled U residues in the sequence,
complementary to each
(-)
probe is
110
for probe a,
101
for probe b, and
78
for probe c. P represents background
hybridization to an RNA transcribed from nonrecombinant pGEM-3Z.
TABLE
4.
Mapping of transcriptional activity across apoA-I gene
in livers of control and chronically hyperthyroid rats
Hybridization
Probe"
mRNA
Synthesis
Control
T.?
d
(+233
to
+657)
e
(+449
to
+935)
f
(+667
to
+1086)
g
(+899
to
+1345)
h
(+1120
to
+167?;)
%
100
f
17h.cd.e
99
2
1f.g.h.i
45
f
6h 36
f
6'
44
?
4c-i 25
f
5~d
49
f
4d 41 f6h
45
f
5c 42
f
7'
ApoA-I mRNA synthesis rates were measured in three independent
nuclear run-on assays usingnuclei from three to six individual rat livers
per group. Nuclei were isolated from livers of control rats
or
rats
injected with
50
pg of Ts per
100
g body weight for
7
days. mRNA
synthesis rates are expressed
as
percent of signals obtained with probe
d
in control. Input of 8zP-labeled RNA was
3.5
x
lo6
cpm. Results are
means
f
SD, and are corrected for the number of nucleotides per
probe and for hybridization efficiency, which averaged
19.22, 15, 18,
and
18%
for probes
d,
e,
f,
g,
and
h,
respectively.
'Antisense RNA obtained by in vitro transcription; numbers in
parentheses refer to nucleotides relative to transcription start site
(1 1).
h.c.def.R.hjJData pairs significantly different
(P
<
0.01)
using two-way
analysis of variance.
sistent with a transcriptional elongation block or pause
site between nucleotide position
450
and
650
of the
apoA-I gene relative to its major transcription start site.
As
Ts-treated rats showed a more pronounced drop in
signal intensity with probes
e
and
f,
it
is
possible that
chronic hormone administration augments the proc-
esses hindering transcription elongation at this site.
However, the use of nuclear run-on experiments pre-
cludes the precise location of the elongation block site
because of the limitations on the size of probes required
for hybridization. Moreover, hybridization signals
downstream
of
the main block tended to increase in
Ts-treated rats, but remained at the reduced level or
tended to decrease in control rats (Fig.
3,
Tables
2
and
4).
Hence, the possibility that an additional elongation
block site distinct from but adjacent to the site detected
in control rats was induced by Ts-treatment, cannot be
excluded.
To our knowledge, this is the first report on expres-
sion control by transcriptional arrest or pausing among
members
of
the apolipoprotein multigene family. Be-
cause there is extensive homology within and between
apolipoprotein genes and among different species
(37,
Lin-Lee et
al.
ApoA-I gene transcription and thyroid hormone
1591
by guest, on July 14, 2011www.jlr.orgDownloaded from

k
c
h
(+1120),(+1673)
B.
T3
d
e
I
I
I
h
Fig.
3.
Mapping of transcriptional arrest or pausing in the apoA-I gene in livers of control and chronically
hyperthyroid rats. A, Map of the rat apoA-I gene; A,
B,
C,
D,
refer to exons
1,
2,
3, and
4;
d, e, f,
g,
and h
represent the portions of the apoA-I gene encompassed by in vitro synthesized RNA probes. Numbers in
parentheses are relative to the major transcription start site
(11).
B,
Autoradiograph of dot blot hybridization
of
3zP-labeled nuclear RNA to excess antisense
(-)
RNA probes d-h. Nuclei were isolated from the livers of
control rats
(Co)
or Ts-injected rats (T3) and then allowed to incorporate [s2P]vTp in vitro. The number of
labeled
U
residues in the sequence, complementary to each
(-)
probe is
91
for probe d,
96
for probe e, and
85
for probe f,
88
for probe
g,
and
90
for probe h. P represents background hybridization
to
an RNA transcribed
from nonrecombinant pGEM-32.
38), control of the elongation phase of transcription may
be involved in the expression level
of
other apolipopro-
teins.
The expression of several prokaryotic, eukaryotic,
and viral genes may be controlled by transcriptional
arrest. Examples
of
eukaryotic genes include several
cellular proto-oncogenes, the murine and human
adenosine deaminase genes, the human histone H3.3
gene, and the Drosophila genes, specifically the heat-
shock genes hsp
70
and hsp 26 (reviewed in refs. 39-41).
Examples of viral transcription units include the human
immunodeficiency virus (42), simian virus 40 (43), ade-
novirus type
2
(44), and polyomavirus (45).
To
our
knowledge, transcriptional arrest sites, characterized
previously in eukaryotic genes and viral transcription
units, were mapped to the first exon or intron of the
respective genes. None of them occurred in the context
of possible feedback inhibition of transcription. Hence,
the premature transcription arrest occurring close to
the exon 3-intron
3
border
of
the
apoA-I gene and 450
to
650
nucleotides downstream
of
the major transcrip
tion start site is unique. Conditional transcript elonga-
tion may be associated with disruption
of
the ternary
polymerase complex and release of prematurely termi-
nated transcripts (42,46) or with pausing of the polym-
erase complex without release of immature transcripts
(41, 47). Our current experiments do not allow us to
distinguish between these two possibilities.
Several mechanisms have been identified which may
contribute to transcriptional attenuation. Specific se-
quences or structures in DNA or RNA may impede the
progress of the polymerase complex (46,48), events at
the promoter and protein factors may exert effects on
the processivity of
RNA
polymerase (49), and proteins
bound to the template may block the progress of tran-
scription (50). In preliminary experiments, we detected
several in vivo and in vitro footprints within exon
3
and
intron
3
of the apoA-I gene. Furthermore, the elonga-
1592
Journal
of
Lipid
Research
Volume
36,
1995
by guest, on July 14, 2011www.jlr.orgDownloaded from

tion block was competitively relieved with increasing
concentrations of in
vitro
synthesized intron
3-RNA
fragments in nuclear
run-on
assays
of
both
control
and
Ts-treated rats
(25).
While these preliminary
data
await
more rigorous testing in
other
experimental systems,
they
could link mRNA maturation with transcriptional
activity. More effective mRNA maturation would result
in decreased levels of mRNA
degradation
products
which
in
turn would
augment
transcriptional arrest
or
pausing. Such an autoregulatory mechanism could also
serve to amplify changes in transcription initiation. With
more frequent transcript initiation,
the
level
of
mRNA
degradation products would increase
and,
as
a
conse-
quence, transcriptional pausing
or
arrest would
be
re-
duced.
I
This work was supported by National Institutes of Health
Grants
R01
HL34457 and HL27341.
Manuscript received
3
Januar
1995
and
in
revised
fonn
4
April
1995.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
Maciejko,
J.
J.,
D. R. Holmes,
B.
A. Kottke, A.
R.
Zinsmeis-
ter, D.
M.
Dinh, and
S.
J.
T. Mao. 1983. Apolipoprotein
A-I
as
a marker of angiographically assessed coronary
artery disease.
N.
Eng1.J.
Med
309:
385-389.
Graham, D. L., and
J.
F.
Oram.
1987. Identification and
characterization of a high density lipoprotein-binding
protein in cell membranes by ligand blotting.
J.
Biol. Chem.
Nowicka, G., T. Briining,
A.
Bottcher,
G.
Kahl,
and
G.
Schmitz. 1990. Macrophage interaction of HDL sub-
classes separated by free flow isotachophoresis.
J.
Lipid
Jackson, R.
L.,
A.
M.
Gotto,
0.
Stein, andY. Stein. 1975.
A comparative study on the removal
of
cellular lipides
from Landschutz ascites cells by human plasma apolipo-
proteins.
J.
Biol. Chem.
250:
7204-7209.
Fielding, C.
J.,
V.
G. Shore, and P.
E.
Fielding. 1972. A
protein cofactor of 1ecithin:cholesterol acyltransferase.
Biochem. Bqhys.
Res.
Commun.
46:
1493-1498.
Wu, A-L., and
H.
G. Windmueller. 1979. Relative contri-
butions by liver and intestine to individual plasma
apolipoproteins in the rat.J.
Biol.
Chem.
254
7316-7322.
Zannis,
V.
I.,
F.
S.
Cole, C. L. Jackson, D.
M.
Kurnit, and
S.
K.
Karathanasis. 1985. Distribution of apolipoprotein
A-I, C-11, C-111, and E mRNA in fetal human tissues. Time
dependent induction
of
apolipoprotein
E
mRNA by cul-
tures
of
human monocyte macrophages.
Biochemistry.
24:
Sastry,
K.
N.,
U.
Seedorf, and
S.
K.
Karathanasis. 1988.
Different cis-acting DNA elements control expression of
the human apolipoprotein A-I gene in different cell types.
Mol.
Cell. Biol.
8:
605-614.
Walsh, A.,
Y.
Ito, and
J.
L.
Breslow. 1989. High levels of
human apolipoprotein A-I in transgenic mice result in
increased plasma levels of small high density lipoprotein
262:
7439-7442.
Res.
31:
1947-1963.
4450-4454.
(HDL) particles comparable to human HDLs.
J.
Biol.
10.
Apostolopoulos,
J.
J.,
G.
J.
Howlett, and N. Fidge. 1987.
Effects
of
dietary cholesterol and hypothyroidism on rat
apolipoprotein mRNA metabolism.
J.
Li$d
Res.
28:
11. Haddad,
I.
A.,
J.
M.
Ordovas, T. Fitzpatrick, and
S.
K.
Karathanasis. 1986. Linkage, evolution, and expression of
the rat apolipoprotein A-I, C-111, and
A-IV
genes.J.
Biol.
12. Ertel Miller,
J.
C., R.
K.
Barth, P. H. Shaw, R.
W.
Elliott,
and
N.
D. Hastie. 1983. Identification of a cDNA clone
for mouse apolipoprotein A-I (apoA-I) and its use in
characterization of apoA-I mRNA expression in liver and
small intestine.
hoc.
Nutl.
Acud. Sci.
USA.
80
1511-1515.
13. Sorci-Thomas, M.,
M.
M.
Prack,
N.
Dashti, F. Johnson, L.
L. Rudel, and D. L. Williams. 1989. Differential effects of
dietary fat on the tissue-specific expression of the apolipo-
protein A-I gene: relationship to plasma concentration of
high density lipoproteins.
J.
Lipid
Res.
30
1397-1403.
14. Sorci-Thomas,
M.,
M. M.
Prack,
N.
Dashti, F. Johnson, L.
L. Rudel, and D. L. Williams. 1988. Apolipoprotein (apo)
A-I production and mRNA abundance explain plasma
apoA-I and high density lipoprotein differences between
two nonhuman primate species with high and low suscep-
tibilities to diet-induced hypercholesterolemia.
J.
Biol.
15. Strobl, W.,
N.
L. Corder, Y-C. Lin-Lee, A.
M.
Gotto, Jr.,
and
W.
Patsch. 1990. Role of thyroid hormones in
apolipoprotein A-I gene expression in rat liver.
J.
Clin.
Invest.
85:
659-667.
16. Papazafiri, P.,
K.
Ogami,
D.
P.
Ramji, A. Nicosia, P.
Monaci, C. Cladaras, and
V.
I.
Zannis. 1991. Promoter
elements and factors involved in hepatic transcription of
the human apoA-I gene positive and negative regulators
bind to overlapping sites.
J.
Biol.
Chem.
266:
5790-5797.
17. Ladias, A. A., and
S.
K.
Karathanasis. 1991. Regulation
of
the apolipoprotein A-I gene by
ARP-1,
a novel member of
the steroid receptor superfamily.
Science.
251:
561-565.
18. Widom, R. L.,
J.
A. A. Ladias,
S.
Kouidou, and
S.
K.
Karathanasis. 1991. Synergistic interactions between tran-
scription factors control expression of the apolipoprotein
A-I
gene in liver cells.
Mol.
Cell. Biol.
11:
677-687.
19. Apostolopoulos,
J.
J.,
M.
J.
La
Scala, and G.
J.
Howlett.
1988. The effect of triiodothyronine on rat apolipopro-
tein A-I and A-IV gene transcription.
Biochem. Biophys.
Res.
Commun.
154
997-1002.
20. Panduro,
A,,
F.
Shalaby, and
D.
A.
Shafritz. 1987. Chang-
ing patterns of transcriptional and post-transcriptional
control
of
liver-specific gene expression during
rat
devel-
opment.
Genes
Dev.
1:
1172-1182.
21. Panduro, A., Y-C. Lin-Lee, L. Chan, andA. Shafritz. 1990.
Transcriptional and posttranscriptional regulation of
apolipoprotein
E,
A-I and A-I1 gene expression in normal
rat liver and during several pathophysiologic states.
Bio-
22. Strobl,
W.,
B.
Knerer, R. Cratzl,
K.
Arbeiter,Y-C. fin-Lee,
and
W.
Patsch. 1993. Altered regulation of apolipopro-
tein A-IV gene expression in the liver of the genetically
obese Zucker rat.J.
Clin.
Invest.
92
1766-1773.
23. Davidson,
N.
O.,
R.
C. Carlos,
M.
J.
Drewek, and T. G.
Parmer. 1988. Apolipoprotein gene expression in the rat
is regulated in a tissue-specific manner by thyroid hor-
mone.
J.
Lipid
Res.
29
1511-1522.
Ch
264:
6488-6494.
642-648.
Chm
261:
13268-13277.
Chem.
263:
5183-5189.
chemist?.
29:
8430-8435.
Lin-Lee et
al.
ApoA-I
gene
transcription
and
thyroid
hormone
1593
by guest, on July 14, 2011www.jlr.orgDownloaded from

24.
Soyal,
S.
M., C. Seelos, Y-C. Lin-Lee,
S.
Sanders, A. M.
Gotto,
Jr.,
D. L. Hachey,
and
W.
Patsch.
1995.
Thyroid
hormone influences
the
maturation of apolipoprotein A-I
messenger
RNA
in rat 1iver.J.
Biol.
Chem.
270
3996-4004.
25.
Lin-Lee,Y-C.,
S.
M.
Soyal,
S.
R. Sanders, A. Sourgoutchev,
and
W.
Patsch.
1993.
Thyroid hormone alters apolipopro-
tein A-I gene transcription in rat liver by transcriptional
pausing.
Circulation.
88:
1-421.
26.
Northemann,
W.,
M.
Heisig, D. Kunz, and P. C. Heinrich.
1985.
Molecular cloning of cDNA sequences for rat
a2macroglobulin and measurement of its transcription
during experimental inflammation.
J.
Biol. Chem.
260:
27.
Strobl,
W.,
N. L. Gorder, G. A. Fienup, Y-C. Lin-Lee, A.
M. Gotto, Jr., and
W.
Patsch.
1989.
Effect of sucrose diet
on apolipoprotein biosynthesis in rat liver: increase in
apolipoprotein E gene transcription.
J.
Biol. Chem.
264
28.
Kapuscinski,
J.,
and B. Skoczylas.
1977.
Simple and rapid
fluorimetric method for DNA microassay.
Anal. Biochem.
29.
Birch, H. E., and G. Schreiber.
1986.
Transcriptional
regulation of plasma protein synthesis during inflamma-
ti0n.J.
Biol. Chem.
261:
8077-8080.
30.
Chen-Kiang,
S.,
and D.
J.
Lavery.
1989.
Pulse labeling of
heterogeneous nuclear RNA in isolated nuclei.
Methods
Enzymol.
180
82-96.
31.
Saiki, R.,
S.
Scharf, F. Faloona, K.
G.
Mullins, G. T. Horn,
H. A. Erlich, and N. Arnheim.
1985.
Enzymatic amplifica-
tion of Pglobin genomic sequences and restriction site
analysis for diagnosis of sickle cell anemia.
Science.
230:
32.
Sambrook,
J.,
E. F. Fritsch,
andT.
Maniatis.
1989.
Molecu-
lar Cloning: A Laboratory Manual, rev.
ed.,
Vols.
1-3.
Cold Spring Harbor Laboratory, Cold Spring Harbor,
NY.
33.
Radosavljevic, M., Y-C. Lin-Lee,
S.
M.
Soyal,
W.
Strobl,
C.
Seelos, A. M. Gotto, Jr., and
W.
Patsch.
1992.
Effect of
sucrose diet on expression
of
apolipoprotein genes A-I,
34.
Sanger,
F.,
S.
Nicklen,
and
A. R. Coulson.
1977.
DNA
sequencing with chain-terminating inhibitors.
Proc. Natl.
Acad.
Sci.
USA.
74:
5463-5467.
35.
Flaspohler,
J.
A., and C. Milcarek.
1992.
Use of in vitro
transcribed RNAs
as
immobilized targets during
RNARNA hybridization.
BioTechniques.
13:
68-72.
36.
Lin-Lee, Y-C.,
W.
Strobl,
S.
Soyal, M. Radosavljevic, M.
Song, A. M. Gotto, Jr., and
W.
Patsch.
1993.
Role of
thyroid hormone in the expression of apolipoprotein
A-IV and C-111 genes in rat liver.
J.
Lipid
Res.
34:
249-259.
6200-6205.
1190-1 194.
83:
252-257.
1350- 1354.
C-111 and A-IV.
Athmosclmosk.
95:
147-156.
37.
Karathanasis,
S.
K.
1985.
Apolipoprotein multigene fam-
ily and tandem organization of human aDoliDoDrotein
A-I,
C-111, and A-Ifgenes.
Proc. Natl. Acad.'Sci; &A.
82:
6374-6378.
38.
Luo, C. C.,
W.
H. Li,
M.
N. Moore, and L. Chan.
1986.
Structure and evolution of
the
apolipoprotein multigene
fami1y.J
Mol.
Biol.
187:
325-340.
39.
Spencer,
C.
A., and
M.
Groudine.
1990.
Transcription
elongation and eukaryotic gene regulation.
Oncog&e.
5:
777-785.
40.
Lis,
J.,
and
C.
Wu.
1993.
Protein traffic on the heat shock
promoter: parking, stalling, and trucking along.
Cell.
74:
41.
Kash,
S.
F.,
J.
W.
Innis, A. U. Jackson, and R.
E.
Kellems.
1993.
Functional analysis of a stable transcription arrest
site in the first intron of the murine adenosine deaminase
gene.
Mol. Cell. Biol.
13:
2718-2729.
42.
Kao, S-Y., A.
F.
Calman, P. A. Luciw, and B.
M.
Peterlin.
1987.
Anti-termination of transcription within the long
terminal repeat of HIV-1 by tat gene product.
Nature.
330:
43.
Hay, N., H. Skolnik-David, and
Y.
Aloni.
1982.
Attenu-
ation in the control of SV40 gene expression.
Cell.
29:
44.
Maderious, A., and
S.
Chen-Kiang.
1984.
Pausing and
premature termination of human RNA polymerase I1
during transcription of adenovirus in vivo and in vitro.
Proc.
Natl. Acad. Sci. USA.
81:
5931-5935.
45.
Skarnes,
W.
C.,
D.
C. Tessier, and N. H. Acheson.
1988.
RNA polymerases stall and/or prematurely terminate
near both early and late promoters
on
polyoma virus
DNA.J.
Mol. Biol.
203:
153-171.
46.
Bentley,
D.
L., and M. Groudine.
1988.
Sequence require-
ments for premature termination of transcription in the
human c-myc gene.
Cell.
53:
245-256.
47.
Rougvie, A.
E.,
and
J.
T.
Lis.
1988.
The RNA polymerase
I1 at the
5'
end
of
the uninduced hsp70 gene of
D.
melanogaster
is
transcriptionally engaged.
Cell.
54
48.
Reines,
D.
1992.
Elongation factordependent transcript
shortening by templateengaged RNA polymerase 11.
J.
Biol. Chem.
267:
3795-3800.
49.
Hair, A,, and G. T. Morgan.
1993.
Premature termination
of tubulin gene transcription in Xenopus oocytes is due
to
promoterdependent disruption of elongation.
Mol.
Cell. Biol.
13:
7925-7934.
50.
Connelly,
S.,
and
J.
L.
Manely.
1989.
A CCAAT box
sequence in the adenovirus major late promoter func-
tions
as
part of an RNA polymerase I1 termination signal.
Cell.
57:
655-663.
1-4.
489-493.
183-193.
795-804.
1594
Jod
of
Lipid
Research
Volume
36, 1995
by guest, on July 14, 2011www.jlr.orgDownloaded from