ABSTRACT
Transcription factor IIIA (TFIIIA)
,
a cysteine-rich regulatory protein
,
is the prototype for the largest known superfamily of eukaryotic transcription
factors. Members of the TFIIIA superfamily contain Cys
2
His
2
zinc finger domains responsible for nucleic acid binding. Xenobiotic metal ions
,
which lack known biological function
,
were previously used as probes for the structure and function of steroid hormone receptors which contain Cys
2
Cys
2
zinc finger domains. Structural alterations in cysteine-rich regulatory proteins by such ions
in vivo
might potentiate carcinogenesis and other disease processes. In the present
study cadmium
and other xenobiotic metal ions were used to probe the structure and function of
TFIIIA. The specific interaction of TFIIIA with the internal control region (ICR) of the 5S RNA gene, as assayed by DNase I protection, was inhibited by Cd
2+
ion concentrations of
>=
0.1
[mu]
M. Aluminum ions were also found to inhibit the TFIIIA-5S RNA gene interaction
,
albeit at higher concentrations (
>=
5
[mu]
M). Inhibition by either metal ion was not readily reversible. Other xenobiotic
metal ions, such as mercury or cesium, were not found to be inhibitory under
these conditions. None of these ions at the concentrations used in this study
affected the ability of DNase I to digest DNA or restriction enzymes to
specifically cleave DNA. Preincubation of TFIIIA bound to 5S RNA with either Cd
2+
or Al
3+
resulted in subsequent DNA binding upon dilution and RNA removal, whereas
preincubation of free TFIIIA with the metal ions resulted in inhibition of
subsequent DNA binding. Because 5S rRNA also binds the TFIIIA zinc finger
domains
,
these results indicate that the 5S RNA bound to TFIIIA protects the protein
from metal inhibition and implicates the zinc fingers in the inhibition
mechanism. The nature of the footprint inhibition indicates that the N-terminal fingers of TFIIIA are affected by the metal ions. Cd
2+
and Al
3+
ions also inhibited the ability of TFIIIA to bind complementary single-stranded DNA and promote renaturation, as measured by Tris-phosphate agarose gel electrophoresis. This gel assay is sensitive to DNA conformation and Al
3+
ions were found to alter the conformation of single- and double-stranded DNA in this assay. The inhibition of TFIIIA function
in vitro
by xenobiotic metals offers new insights into the structure and function of TFIIIA and TFIIIA-type zinc finger proteins. Inhibition by Cd
2+
occurs at much lower concentrations than previously observed with steroid
hormone receptors and suggests that Cys
2
His
2
zinc finger proteins may be especially sensitive to such agents
in vivo.
Cysteine-rich zinc finger proteins constitute a large portion of the transcription
factors that regulate initiation of RNA synthesis in higher organisms. Two
prominent superfamilies of zinc finger proteins are the TFIIIA-type proteins (
1
,
2
) and the steroid hormone receptors (
3
). TFIIIA regulates eukaryotic 5S RNA synthesis by RNA polymerase III (
4
) and was the first cysteine-rich regulatory protein identified to contain zinc and require the metal
for function (
5
). This protein binds to the 50 bp internal control region (ICR) of the 5S RNA
gene and also binds to the 5S RNA gene product competitively with DNA (
6
-
9
). The zinc in TFIIIA is not tightly bound, since mild treatment with metal
chelators stripped the metal from the protein and inhibited DNA binding; zinc
addtion to the apoprotein restored specific DNA binding (
5
). The amino acid sequence of TFIIIA revealed a repetitive pattern of nine 30
amino acid domains (`fingers'), each containing two Cys and two His residues
capable of metal coordination (
10
). The Zn
2+
ions hold the structure together, since their removal results in unfolding of
TFIIIA and concomitant loss of specific DNA binding ability (
11
). Crystallographic analysis of TFIIIA-type Cys
2
His
2
zinc finger domains bound to DNA revealed compact finger domains wrapped around
the major groove (
12
). The centrally located Zn
2+
ion in each finger was coordinated by the two Cys residues in an antiparallel [beta]-sheet and by the two His residues located on the same face of an [alpha]-helix. Residues in the [alpha]-helix interact specifically via hydrogen
bonds with base pairs in the DNA, whereas other amino acids throughout the
domain make ionic contacts on DNA phosphates (
12
). Mutagenesis of TFIIIA revealed the necessity for all four metal coordinating
residues, as well as the integrity of interfinger linker regions, for specific
DNA binding (
13
). Proteins containing TFIIIA-type zinc finger domains (Cys
2
His
2
) number in the hundreds in vertebrates and constitute the largest known
superfamily of such proteins in higher organisms (
1
,
2
). TFIIIA-type zinc finger proteins regulate a multitude of processes, including
embryogenesis and oncogenesis (
14
-
16
).
The steroid hormone receptor superfamily comprises another large group of
cysteine-rich zinc finger transcription factors which translocate into the nucleus
upon hormone binding. These proteins activate expression from enhancer regions
of a number of hormone responsive genes. Unlike the TFIIIA superfamily, the DNA
binding domains of hormone receptors always contain just two Cys
2
Cys
2
zinc fingers (
3
). The first finger and linker region comprise an [alpha]-helix and make specific DNA contacts in the DNA major groove,
whereas the second finger is involved in protein-protein interactions forming the active receptor dimer (
17
). Because of the cysteine-rich nature of the DNA binding domains of steroid hormone receptors,
studies have examined the effects of metals other than zinc on the structure
and function of hormone receptors. Such studies have added significance since a number of metal ions,
including xenobiotic ions, are believed to have etiological roles in
carcinogenesis and other disease processes (
18
,
19
). Elucidating mechanisms of metal ion-finger protein interactions will provide molecular insights into the
structure and function of these proteins, as well as insights into potential
disease processes. With respect to the estrogen receptor (ER), one study
indicated that 1 mM Cd
2+
could inhibit DNA binding by activated (hormone-bound) receptors and 0.1 mM Cd
2+
could inhibit initial binding of hormone to receptor (
18
). Another study demonstrated that Cd
2+
could replace zinc in the ER and suffice for DNA binding, although some
functional variation was observed (
19
). In the present study, the effects of Cd
2+
and other xenobiotic metals on the DNA binding behavior of a Cys
2
His
2
zinc finger protein,TFIIIA, were examined in order to gain understanding of the
structure and function of Cys
2
His
2
zinc finger proteins and to elucidate potenial toxicity or regulatory
mechanisms by such cations or other small molecules.
Immature ovarian tissue was dissected from 4-5 cm female
X.laevis
frogs (Nasco, Fort Atkinson, WI) and homogenized briefly in 50 mM Tris-HCl, pH 7.5, 50 mM KCl, 5 mM MgCl
2
, 0.5 mM dithiothreitol (DTT) and 0.2 mM phenylmethylsulfonyl fluoride. The
homogenate was centrifuged at 10 000
g
for 20 min and aliquots of the supernatant were layered on 15-30% glycerol gradients in homogenization buffer and centrifuged for 24 h at 34 000 r.p.m. in a SW41 rotor. All manipulations were performed at 0-4oC. The 7S fractions containing TFIIIA bound to 5S rRNA were
identified by UV absorption of gradient fractions and further purified to 90-95% purity (judged by SDS-PAGE of 39 kDa TFIIIA) by DEAE-cellulose chromatography essentially as described
previously (
5
). 5S RNA was removed from TFIIIA by digestion with RNase A (10 [mu]g/ml) in 20 mM Tris-HCl, pH 7.6, 320 mM KCl, 2 mM MgCl
2
, 0.4 mM DTT and 0.1% NP-40 for 30 min at room temperature and then placed on ice. Protein
concentration was determined by the method of Bradford using bovine serum
albumin as the standard (
20
).
A plasmid containing the
Xenopus borealis
somatic 5S RNA gene was purified by CsCl/ethidium bromide equilibrium gradient
centrifugation (
13
). The 303 bp DNA fragment containing the 120 bp
Xenopus
5S RNA gene was end-labeled on the coding strand by digesting the plasmid first with
Eco
RI and incorporating [[alpha]-
32
P]dATP with reverse transcriptase (
13
). The end-labeled fragment was ethanol precipitated, redigested with
Bam
HI and the smaller fragment was purified by PAGE. Specific activity of the
fragment was determined by absorbance at 260 nm and Cerenkov counting. To
examine the effects of various metals on the TFIIIA-5S RNA gene interaction TFIIIA (10 nM) was incubated with the various
metals (chloride salts of all metals were ultrapure grade and purchased from
Aldrich; concentrations are indicated in the figure legends) in 20 mM Tris-HCl, pH 7.6, 70 mM NH
4
Cl, 7 mM MgCl
2
, 0.4 mM DTT and 0.1% NP-40 for 10 min prior to addition of the DNA fragement containing the end-labeled 5S RNA gene (1 nM). After addition of DNA the reaction (20 [mu]l) was incubated at room temperature for an additional 15 min.
The reactions were then incubated for an additional 1 min in DNase I (2 [mu]g/ml final concentration) and digestion was terminated by addition of 100 [mu]l stop buffer (20 mM Tris-HCl, pH 7.6, 1 mM EDTA, 0.1% SDS, 30 [mu]g/ml sonicated salmon sperm DNA). The samples were ethanol precipitated, dried, resuspended in 4 [mu]l formamide soution (20 mM Tris-HCl, pH 7.6, 95% deionized formamide, 1 mM EDTA,
0.01% xylene cyanol and bromphenol blue), heated to 95oC for 5 min and electrophoresed on a 7 M urea-7% polyacrylamide gel until the xylene cyanol marker had migrated two
thirds the distance of the gel. Gel autoradiography was performed overnight at -70oC with Kodak XAR-5 film and a Dupont Cronex intensifying screen.
Complementary single-stranded DNA template for renaturation assays was obtained by denaturing
the 303 bp
32
P-end-labeled fragment (in 20 mM Tris-HCl, pH 7.6, 50 mM KCl, 1.5 mM MgCl
2
) at 100oC for 5 min, followed by quick cooling in an ice/water bath. The non-denatured 303 bp end-labeled fragment was used as the double-stranded DNA control. TFIIIA-dependent DNA renaturation was performed under the
same TFIIIA binding conditions as described in the DNase I protection assay
above. TFIIIA, DNA and metal ion concentrations are given in the figure
legends. The renaturation reactions (20 [mu]l) were quenched by addition of 4 [mu]l electrophoresis sample buffer (20 mM Tris-HCl, pH 7.6, 10 mM EDTA, 0.5% SDS, 25% glycerol, 0.01% bromphenol
blue). Agarose gel (0.6%) electrophoresis utilizing a 20 mM Tris-phosphate buffer, pH 8.3, was used to resolve single- and double-stranded DNA in the renaturation assays (
21
). Gels were exposed to Kodak XAR-5 film overnight at 4oC.
Cd
2+
ions (0.1-1.0 mM) were found to inhibit steroid hormone receptors and have been
useful probes of receptor structure and function (
18
). It was important to determine the effects of various concentrations of Cd
2+
and other xenobiotic metal ions on the structure and function of the other
major class of zinc finger transcription factors, the TFIIIA-type, with their Cys
2
His
2
DNA binding domains. Figure
1
is an autoradiogram exhibiting the effects of Cd
2+
and Hg
2+
ions on the interaction of
Xenopus
TFIIIA with the internal control region (ICR) of the
Xenopus
5S RNA gene as assayed by DNase I protection. In Figure
1
A all concentrations of Cd
2+
tested (5-20 [mu]M, lanes 3-6) were found to be inhibitory, as evidenced by the loss of
TFIIIA-dependent DNAse I protection from nucleotides +43 to +96 on the coding
strand of the 5S RNA gene (cf. minus and plus TFIIIA controls, lanes 1 and 2).
This inhibitory concentration for Cd
2+
on TFIIIA function (5 [mu]M) is significantly lower than that observed with steroid hormone receptors.
In contrast, Hg
2+
over this same concentration range had no effect on the TFIIIA-5S RNA gene interaction (Fig.
1
B, DNAse I protection observed between nucleotides +43 and +96, lanes 3-6). Because Cd
2+
but not Hg
2+
ions were able to inhibit TFIIIA binding to the 5S RNA gene, it was necessary
to examine potential effects of other metal ions on this interaction. In order
to determine how specific this metal ion inhibition was Al
3+
, Zn
2+
, Mg
2+
, Mn
2+
and Cs
2+
(all at 10 [mu]M final concentration) were examined for inhibitory effects on TFIIIA-5S RNA gene interactions as determined by DNase I digestion (Fig.
2
). At this concentration only Al
3+
(Fig.
2
, lane 3) was able to inhibit this interaction to any appreciable extent, as
evidenced by loss of DNase I protection from nucleotides +43 to +96. Lead ions
were also found to inhibit TFIIIA-DNA interactions, albeit at higher concentrations (75-100 [mu]M range; data not shown). It is noted that although Cd
2+
and Al
3+
inhibit TFIIIA-DNA interactions at low concentrations they are not general inhibitors of
DNA-protein interactions, as no effect on DNase I digestion is observed (Figs
1
and
2
, cf. digestion pattern in lane 1, no TFIIIA, with lanes exhibiting TFIIIA
binding inhibition). In addition, these metal ions did not inhibit a variety of
restriction enzymes tested (data not shown).
Besides binding to the 5S RNA gene, TFIIIA also has the ability to promote DNA
renaturation and was the first transcription factor shown to possess such an
activity (
21
,
22
). More recently the transcription factor/tumor suppressor p53 was also shown to
promote DNA renaturation (
23
). TFIIIA renatures DNA by binding initially to complementary single-stranded DNAs, accelerating their reassociation and then remaining
associated with the renatured double-stranded DNA (
21
,
22
). Unlike the DNase I protection assay of the 5S gene, TFIIIA depleted of Zn
2+
ions retains the ability to promote DNA renaturation (
22
). Because TFIIIA-dependent DNA renaturation is a DNA binding mechanism distinct from that assayed by DNase I protection, it was of interest
to determine whether xenobiotic metal ions also inhibit this process. DNA
migration in the Tris-phosphate agarose gel electrophoresis system used in the DNA renaturation
assay is sensitive to some types of DNA structural changes (
22
). This assay is useful in this study because it can detect structural changes
in the nanomolar DNA concentration range. Before assaying for xenobiotic metal
effects on TFIIIA-dependent DNA renaturation we discovered that Al
3+
ions by themselves altered the electrophoretic mobility of both single- and double-stranded DNA in this Tris-phosphate gel system (Fig.
4
). Lanes 1 and 2 in the Figure
4
autoradiogram show migration of the native double-stranded and heat-denatured 303 bp end-labeled fragment in this gel system; the denatured single-stranded DNA migrates substantially slower (lane 2) than
the double-stranded DNA (lane 2). Lanes 3 and 4 contain the same DNA samples as lanes
1 and 2, but incubated with 10 [mu]M Al
3+
ions prior to electrophoresis. The migration of both DNA forms is retarded by
Al
3+
ions to about the same extent and the bands are also broadened. The binding of
Al
3+
ions could reduce the negative charge density of DNA to a variable degree and
slow migration. However, simple electrostatic binding to DNA is expected to
occur with other metal cations, yet this migration effect is not observed with
10 [mu]M Cd
2+
(lanes 5 and 6), Hg
2+
(lanes 7 and 8) or Zn
2+
(not shown). We conclude that Al
3+
ions induce a unique structural change in both single- and double-stranded DNA under these conditions. Zn
2+
and Cd
2+
ions do cause DNA structural changes (as deduced by chemical modification), but
at significantly higher concentrations than used in the present study (
24
,
25
).
How do Cd
2+
and Al
3+
ions inhibit TFIIIA binding to DNA? One possibility is that the ions interact
with the DNA template and alter the structure necessary for TFIIIA binding. Another possibility is that
the ions interact directly with TFIIIA and disrupt the protein structure,
resulting in binding inhibition. Experiments involving Al
3+
ion inhibition of TFIIIA-dependent DNA renaturation (Fig.
5
) suggested the latter. However, the previous DNA binding experiments did not
clearly distinguish between these two possibilities, since protein, DNA and
metal ions were incubated together in the same reaction tube. To examine for
metal inhibition on TFIIIA directly Cd
2+
or Al
3+
ions were added to the free protein at higher concentration and then diluted ~25-fold in the final DNase I protection assay, out of the inhibitory
range for the metals deduced from Figure
3
. In addition, TFIIIA bound to 5S RNA was analyzed in the same way, because 5S
RNA and the 5S RNA gene compete for the same binding region on TFIIIA (
8
,
9
). Importantly, 5S RNA bound to TFIIIA protects the zinc fingers from denaturation by EDTA,
which occurs rapidly in the absence of bound 5S RNA (
5
). This observation indicates that the finger domains are not as accessible to
solutes when complexed with 5S RNA.
When free TFIIIA is exposed to an inhibitory Cd
2+
concentration and then diluted out of the inhibitory range in the DNase I protection reaction
with the 5S RNA gene binding inhibition is maintained, as evidenced by the loss
of DNase I protection from nucleotides +43 to +96 (Fig.
7
A, lane 3). Significant binding inhibtion is not observed when TFIIIA bound to
5S RNA is initially exposed to Cd
2+
followed by dilution and RNA removal in the DNase I assay (lane 5). Lanes 1 and
2 are the minus and plus TFIIIA controls and lane 4 is a control in which 5S
RNA is removed from TFIIIA (by RNase digestion) in the DNase I protection
reaction. As a control for potential binding of Cd
2+
to RNA, a mixture of TFIIIA and tRNA was exposed to the metal ions and
inhibition was still observed (not shown). These results indicate that metal
inhibition of TFIIIA can occur by direct ion interaction with the protein, that the interaction is not readily reversible by
dilution (or by a 200-fold molar excess of added Zn
2+
; not shown) and that the zinc finger domains (protected by 5S RNA) are likely
targets for the inhibition. A similar experiment was also conducted with Al
3+
(Fig.
7
B). Lanes 3 and 4 exhibit the DNase I digestion patterns when free TFIIIA and TFIIIA bound to 5S RNA are initially exposed to Al
3+
ions and then diluted ~25-fold before assaying for DNA binding activity (along with RNA removal
in the lane 4 reaction). Binding inhibition is only observed with the free
TFIIIA sample (lane 3), again indicating a direct, irreversible effect of Al
3+
on TFIIIA. Lanes 7 and 8 in Figure
7
B exhibit the DNase I protection patterns when free TFIIIA and TFIIIA bound to
5S RNA in the 7S particle are exposed to Hg
2+
ions followed by dilution, RNase treatment for the lane 8 sample and DNase I
digestion of the binding reaction. Protection is seen in both lanes, indicating
that Hg
2+
ions have no effect on TFIIIA under either of these conditions. Lanes 1 and 5 are minus
TFIIIA controls and lanes 2 and 6 are controls containing TFIIIA liberated from
the 7S particle by digestion with RNase prior to the binding assay (lane 2) or
in the binding assay (lane 6).
This work was supported by grants from the National Institute of General Medical
Sciences and the Oklahoma Center for the Advancement of Science and Technology. The authors thank J. Smith for excellent technical assistance.
Chesterfield Gunn, M.D., (co-author) died on July 30, 1995. The medical and scientific community is
poorer in his absence.
REFERENCES
Return