Chromium Induced Neoplasia Using p53, Gadd45, NF-kB, and AP-2 Response Elements
As Models For Binding Affinity
Julia Michelle Kotler
The University of Montana IBS-CORE Program
Mentor- Dr. Kent
Sugden
INTRODUCTION- p53 is
a tumor suppressor protein that has come to the forefront of cancer
research. It is well proven that the
p53 protein acts as the “guardian” of the genome by controlling progression
through the cell cycle and by provoking an arrest in G1 and G2
phases in response to DNA damage (1). p53
exerts its function by binding to a specific DNA response element (p53RE) that
can trigger several biochemical pathways and lead to cell cycle arrest or
alternatively apoptosis (1,2). Apoptosis is a programmed cell death mechanism
mediated by p53 and the activation of the Bax gene to eliminate individual
cells that may lead to disease states (2,5)
The importance of p53 is most evident from the finding that
approximately 50 percent of all human cancers contain cells with point
mutations or deletions in both alleles of the p53 gene (2).
Chromium is a well established
metal carcinogen that is present in both occupational and environmental
settings (3). Singh J, Carlisle DL, et
al. reported that at the cellular level, chromium exposure may lead to cell cycle
arrest, premature terminal growth arrest, or neoplastic transformation
(3). There are two stable forms of
chromium found in nature, these are in the oxidation states of Cr(III) and
Cr(VI) (3). Cr(III) is unable to enter cells but Cr(VI) enters through a
nonspecific anionic transporter (3).
After uptake, Cr(VI) is converted to the reduced oxidation states of
Cr(V), Cr(IV), and finally down to the stable form of Cr(III) (4). The
intracellular reduction of Cr(VI) to lower valence states is a necessary step
in the formation of DNA lesions (5).
Cr(VI)
activates p53 by reactive oxygen species(ROS) mediated free radical reactions
that occur during the oxidative reduction of hexavalent chromium within the
cell (6). Oxidative DNA damage is
considered to be an important mechanism in the genotoxicity of Cr(VI) (5).
The
strongest p53 binding consensus sequence has been reported by several sources
to be the palindrome 5’-GGACATGCCCGGGCATGTCC-3’ (7,8).
Use of Cr(VI) in vitro with
DNA showed damage that was non-random
with guanine bases being the most heavily damaged (6). Another investigation showed again that the
frequency of base damage varied along the DNA, with guanine being the most
commonly damaged base (9). With a total of 16 guanine bases along the p53
double stranded consensus binding sequence, it seems evident that an
investigation into the mutations caused by chromium on this sequence could
prove an essential key to discovering the etiology of chromium induced
neoplasia.
More
insights come from investigating information from the structures deposited in
The Protein Data Bank (pdb). Clear interactions between specific bases and
amino acids as well bonding distances can be obtained (10).
figure
1a figure 1b
Figure 1- (a)- p53 cocrystal
structure highlighting interaction of Lysine120B and Arginine280B with
Guanine1109 (10). (b)- cocrystal
structure of p53 bound to its consensus oligonucleotide (10).
The
bond distance between the guanine shown in Figure 1a and the two corresponding
amino acids is well within the hydrogen bonding distances commonly
characterized in protein-DNA interactions (10). Oxidative damage to DNA commonly results in the formation of
oxidized bases which offer altered structures of the nucleic acid guanine (11).
Ž
Ž
By
affecting the structure, donor-acceptor relationships for protein recognition
and DNA binding are postulated to be changed. Therefore resulting in loss of
recognition and binding. This theory
also carries over to another transcription factor which has a similar binding
motif.
p50 was the first DNA-binding subunit of the NF-kappaB transcription factor to be identified (12) and is a product of its p105 precursor (13, 14). Both in vitro and in vivo, p50 has been shown to form transcriptionally active heterodimers with another NF-kB subunit p65(15, 16). p50 is also shown to have high transcriptional activity alone as a homodimer (16). The NF-kB transcription factor is known to participate in the apoptotic signal that regulates the TNF-a pathway in response to certain cellular assaults(17). Recently however, NF-kB has been associated with a p53 dependent apoptotic pathway(18).
Crystal structure investigations into p50 (19) show similar binding motifs with p53 in guanine, lysine and arginine interactions. Similar hydrogen bonding distances between these amino acids and guanine are observed thus linking a connection between activity and interplay.
figure 2a figure 2b
Figure 2- (a)- p50 bound to consensus oligo as homodimer. (b)- p50 structure highlighting interactions between guanine4d, Lysine 241a, and Arginine54a.
The hypothesis rests on the theory that the interactions between p53 and its consensus oligonucleotide and NF-kB and its consensus sequence would be altered after exposure to chromium and binding would not occur. As a consequence of binding failure a loss in cell cycle regulation would occur.
Methods and
Materials-
Oligonucleotide
sequences and preparations-
p53
consensus Oligonucleotide preparation- The 20
base pair palindromic consensus sequence 5’-GGACATGCCCGGGATGTCC-3’ was
purchased from Sigma Genosys in a 1m mole synthesis. HPLC anion exchange purification was carried out and
self-annealing was performed using a temperature gradient program on the PCR
machine. This involved heating the
oligo in deionized water from 25*C to 95*C for 190 seconds and allowing the
denaturing to occur for 6 minutes with a gradual decrease from 95- 4*C for 120
minutes. For confirmation that
annealing was successful, a non-denaturing 20% polyacrylamide gel was run (data
not shown).
Gadd45-
p53 binding site Oligonucleotide preparation- The
34 base pair sequence 5’- TGGTACAGAACATGTCTAAGCATGCTGGGGACTG-3’ and it’s
complimentary strand were purchased from Sigma
Genosys in a 1m mole synthesis.. HPLC purification and a temperature gradient
PCR annealing were performed as indicated above.
NF-kappaB binding
site Oligonucleotide preparation- The 22 base pair sequence 5’-AGTTGAGGGGACTTTCCCAGGC-3’ was
purchased in the double stranded form from Promega
(E3291).
Oligo
was supplied in TE buffer.
AP-2 binding site
Oligonucleotide preparation- The 26 base pair sequence 5’-GATCGAACTGACCGCCCGCGGCCCGT-3’ was
purchased in the double stranded version from Promega supplied in TE buffer.
SP-1 (competitor)
binding site Oligonucleotide preparation- The 22 base pair sequence
5’-ATTCGATCGGGGCGGGGCGAG-3’ was
purchased in the double stranded version from Promega supplied in TE buffer.
Chromium(V)-salen
Preparation-
A
25mM 0.5mL stock solution of Cr-V-salen was prepared fresh for every
experiment. Added 6.25mg of
Cr-III-salen and 0.5mL CH3CN (dry) vortexed to dissolve and 5mg of
Iodosyl benzene added, vortexed and incubated at room temperature for ten
minutes or until color change from dark red to dark brown was observed. 4mL of stock solution equals
500mM, 2mL = 250mM, 1mL = 100mM.
Cr(V)-salen and
p53RE piperdine damage experiments-
A
denaturing 20% polyacrylamide gel was prepared as follows- 31.5g Urea was added to 37.5mL 40%, 19:1
polyacrylamide, 7.5mL 10X TBE buffer, 3.3mL 1.6% Ammonium Persulfate, and 50mL TEMED.
P53RE
oligo’s were 5’ 32P-g-ATP labeled using
PolyNucleotideKinase. Cr-V-salen reactions were incubated for 30 minutes and
speed-vaced until dry. Concentrations
of 500mM, 250mM, and 100mM Cr-V-salen were used to
measure concentration dependant damage.
Piperdine cleavage reactions were done using 1.0mL 1M solution of
piperdine. 100mL of solution was added to each sample and
incubated at 94*C for 30 minutes. The
samples were then dried using the speed vac and an additional 20mL of deionized water and dried again. Maxam-Gilbert G/A reactions were performed
as follows- 10mL deionized water added to
10mL labeled DNA and incubated with 2mL of formic acid at 37*C for 30 minutes,
dried and redissolved in 20mL deionized water and dried
again, strand cleavage (piperdine reaction) was then carried out. The final
procedure was the addition of 2mL formamide loading buffer
which was heated at 94*C for 10 minutes
and then flash frozen for crisper bands prior to loading. Gel was pre-electrophoresed at 2000V and
45*C for 1 hour prior to loading.
Running buffer was 1X TBE. Gels
were run for approximately 2 hours or until indicator bands had traveled 15cm
from the bottom of the wells. Results
were analyzed using autoradiography.
Gel Shifts(EMSA)
Experiments
Protein Products
and Antibodies-
Recombinant
Human p53 Protein- 10mg Recombinant p53 protein was purchased from BD PharMingen (cat.# 556439) suspended in PBS (pH=7.4) no sodium azide.
Recombinant
Human Wildtype p53 Baculovirus Lysate- 500mL lysate purchased from Pierce Chemicals (product # 29963) buffered in 20mM Tris-HCl, pH=
6.8, 10mM EDTA, 2% SDS, 10% glycerol, 0.3% beta-mercaptothanol, and 0.3%
Bromophenol blue
Purified
Mouse Anti-Human p53 Monoclonal Antibody-
0.1mg purified monoclonal antibody was purchased from BD PharMingen (cat.# 554293) suspended
in 150mM Tris-HCl (pH=8.0).
HeLa
Nuclear Extract- 40mL of
5mg/mL total protein was purchased from Promega (cat. # E352A) suspended in 20mM HEPES, pH=7.9, 0.1M KCl,
0.2mM EDTA, 0.5mM PMSF, 0.5mM DTT, and 20% glycerol.
AP-2
Extract- 20mL of 1.4mg/mL extract
purchased from Promega (cat. # E354A)
suspended in 50mM Tris-HCl, pH=7.7, 130mM KCl, 1mM EDTA, 10mM MgCl2,
10mM ZnSO4, and 20% glycerol.
NF-kB (p50, human) Extract- 50 gsu (gel shift units)
purchased from Promega (cat.# E3770)
suspended in 50mM NaCl, 5mM DTT, 0.5mM PMSF, 20mM HEPES (pH= 7.9), 10mM zinc acetate, 0.1% NP-40 (w/v), and 10%
glycerol.
NF-kB (p50) (Ab-1)- 100mL of polyclonal antisera
purchased from Oncogene Research Products
(cat.# PC136) suspended in 0.1% sodium azide.
Gel Shift reactions- Gels were run on a 4% non-denaturing polyacrylamide matrix
containing the following- 3.0mL 10X TBE buffer (pH=7.9), 6.0mL 40% 19:1
acrylamide, 2.0mL 80% glycerol, 48.5mL
deionized water, 30mL TEMED, and 450mL 10% Ammonium Persulfate. Gels were pre-electrophoresed at 4*C and
250V for 1 hour prior to loading.
Reactions were carried out in 5X binding buffer containing 5mM MgCl2,
2.5mMEDTA, 2.5mM DTT, 250mM NaCl, 50mM Tris-HCl (pH=7.5), 0.25mg/mL poly dIdC in 20%
glycerol. 1.75pmoles of 5’ 32P-g-ATP labeled oligonucleotide was divided
according to experiment and incubated with indicated proteins and/or antibodies
for 30 minutes at room temperature.
Ficoll loading buffer was added to the lanes with only naked DNA to
allow for indicator bands while the gels were running. Gels were run in 0.5X TBE for 45 minutes at 4C
and 250V and then analyzed by autoradiography.
RESULTS/
DISCUSSION
The results
of these experiments show that Cr(V)-salen does affect the p53 binding site as
evidenced by the Cr(V)-salen piperdine experiments. The results from Figures 3a and 3b show that damage can be seen
but results are inconclusive as to the extent of this damage. Further and repeated experiments using the
p53 consensus sequence after treatment with Cr(V)-salen will be undertaken to
elucidate these results.
figure 3a figure 3b
FIGURE 3-(a) Lane 1-p53RE (20
bp). Lane 2- p53RE + 125mM Cr(V)-salen.
Lane 3-p53RE + 62.5 mM Cr(V)-salen.
Lane 4- p53RE + 25mM Cr(V)-salen.
Lane 5- p53RE + piperdine. Lane
6-p53RE + 125mM Cr(V)-salen + piperdine. Lane 7- p53RE+
62.5 mM Cr(V)-salen + piperdine. Lane 8- p53RE + 25mM Cr(V)-salen + piperdine. Lane 9- Maxam-Gilbert G/A reaction. Lane 10- Maxam-Gilbert C/T reaction.
(b)-Lane1- p53RE. Lane 2-p53RE + 500mM Cr(V)-salen.
Lane 3- p53RE + 250mM Cr(V)-salen. Lane 4- p53RE +100 mM Cr(V)-salen.
Lane 5- p53RE + piperdine. Lane
6- p53RE + 500mM Cr(V)-salen + piperdine. Lane 7- p53RE + 250mM Cr(V)-salen + piperdine. Lane 8- p53RE +100 mM Cr(V)-salen + piperdine. Lane 9- Maxam-Gilbert G/A.
The gel shift experiments with the
p53 binding site (20 bp) showed that the recombinant p53 protein purchased from
BD PharMingen was ineffective at causing a shift. The p53 baculovirus lysate also showed no signs of binding
activity to the 20 base pair p53RE. The
gadd45 p53RE oligonucleotide was used after finding that Fornance, AJ, et al
(20) were successful in getting it to bind to p53. It however showed no binding activity with either the recombinant
p53 protein or the lysate (data not shown).
After finding that HeLa cell extracts
contained low levels of p53 and a nuclear cofactor HMG-1(high mobility group-1
protein) (21) that stimulates p53 binding by bending the DNA it was used in
conjunction with the recombinant protein and/or lysate or alone with the p53
consensus sequences. It was determined
that the HeLa extracts were responsible for the gel shifts that were produced
by eliminating all other protein products from the reactions (figure 4c). Specific binding of p53 to the consensus
sequences could not be determine by lack of “supershifts” in the lanes that
contained DNA, protein and p53 monoclonal antibodies.
The gadd45 p53RE (34 bp) showed
multiple bands or “multiple shifts” in the lanes containing it and the HeLa
extract (figure 4b) leading to the conclusion that it was recognized by two
different proteins in the extract.
Again, specificity could not be determined because the p53 monoclonal
antibody failed to produce a supershift.
The NF-kB (p50) consensus
oligonucleotide showed no binding to the HeLa extract (figure 4a) but did show
binding to the p50 extract. Further
experiments using the p50 polyclonal antibody will be done to show binding
specificity. The AP-2 consensus
oligonucleotide did show binding to the HeLa extract but not to the NF-kB
extract (figure 4a).
figure 4a figure
4b figure 4c
FIGURE 4- (a)EMSA Lane 1- NF-kB oligo. Lane 2- NF-kB oligo + p50 extract. Lane 3- NF-kB oligo + HeLa extract. Lane 4- AP-2 oligo. Lane 5- Ap-2 oligo + p50 extract. Lane 6- AP-2 oligo + HeLa extract.
(b)- EMSA Lane 1- gadd45 p53RE. Lane 2- gadd45 p53RE + HeLa extract. Lane 3- gadd45 p53RE + HeLa extract + p53 mAb. Lane 4- gadd45 p53RE + p53 lysate. Lane 5- gadd45 p53RE + p53 lysate + p53 mAb. Lane 6- gadd45 p53RE + HeLa extract + p53 lysate. Lane 7- gadd45 p53RE + HeLa extract + p53 lysate + p53 mAb.
(c)- EMSA Lane 1-p53RE. Lane 2- p53RE + recombinant p53. Lane 3- - p53RE + recombinant p53 + p53 mAb. Lane 4- p53RE + recombinant p53 + HeLa extract. Lane 5- p53RE + HeLa extract. Lane 6- p53RE + recombinant p53 + HeLa extract + p53 mAb. Lane 7- p53RE + recombinant p53 + HeLa extract + excess p53 mAb.
Figure 5a shows an inconclusive effect of Cr(V)-salen treatment on the gadd45 p53RE that shows a general distortion of binding. The effects of Cr(V)-salen treatment will be studied further using the more direct method of inserting altered base products in the consensus sequences to measure binding affinity. Further oxidation beyond 8-oxo-dG will be assayed by treating the altered sequences with Cr(V)-salen to produce the guanidinohydantoin and spiroaminodihydantoin configurations. These further experiments should provide clues to the protein-DNA interactions that may occur due to chromium exposure.
Figure 5
Figure 5- Lane 1- gadd45 p53RE. Lane 2- gadd45 p53RE + 1mL HeLa extract. Lane 3- gadd45 p53RE + 2mL HeLa extract. Lane 4- gadd45 p53RE + 3mL HeLa extract. Lane 5- gadd45 p53RE + 3mL HeLa extract + p53 mAb. Lane 6- gadd45 p53RE + 500mM Cr(V)-salen. Lane 7- - gadd45 p53RE + 500mM Cr(V)-salen + 3mL HeLa extract. Lane 8- - gadd45 p53RE + 500mM Cr(V)-salen + 3mL HeLa extract + p53 mAb.
ACKNOWLEDGEMENTS-
This work was funded in part by an IBS-CORE grant given to the University of Montana by the Howard Hughes Medical Institute, and National Institute of Environmental Health Sciences Grant No. ES10437.
I would also like to thank Dr. Brooke Martin for her
support on this project.
REFERENCES CITED-
(1). Bourdon, J.C., Deguin-Chambon, V., et
al. (1997) Oncogene 14, 85-94.
(2). Karp, K.
(1999) Cell and Molecular Biology 2nd
Edition, 701-732.
(3). Singh, J.,
Carlisle, D.L., et al.
(1998) Oncology Report 6, 1307-18.
(4). Sugden, K.D., Wetterhahn, K.E.
(1997) Chemical Research In Toxicology 10, 1397-1406.
(5). Sugden, K.D. (1999) Journal of Inorganic Biochemistry 77,
177-183.
(6). Ye, J., Wang, S., et al. (1999)
Journal of Biological Chemistry 49,
34974-80.
(7). Rodriguez, H., Akman, S.A. (1998) Free
Radical Research 6,
499-510.
(8). Funk, W.D.,
Pak, D.T., et al. (1992) Molecular
and Cellular Biology 6, 2866-2871.
(9). Tokino, T., Thiagalingani, S., et al. (1994)
Human Molecular Genetics 3,
1537-1542.
(10) Cho, Y., Gorina,
S., Jeffrey, P., Pavletich, N. (1994)
Science 265, 245-265.
(11) Duarte, V., Muller,
JG., Burrows, CJ. (1999)
Nucleic Acids Research 27,
496-502.
(12) Baeuerle, PA.,
Baltimore, D. (1989) Genes
Dev 3, 1689-1695.
(13) Ghosh, S., et
al. (1990) Cell 62
1019-1023.
(14) Kieran, M., et
al. (1990) Cell 62
1007-1011.
(15) Schmid, RM., et
al. (1991) Nature 352 733-738.
(16) Fujita, T., et al. (1992) Genes Dev 6 775-782
(17) Baeuerle, PA., Henkel, T. (1994)
Annu. Rev. Immunol. 12
141-179.
(18) Esche, H., et al. (1999) Oncogene 17 2728-2738.
(19) Sigler, PB., et al. (1995) Nature
373 303-310.
(20) Fornance, AJ., et al. (1998) Molecular and Cellular Biology 5
2768-2778.
(21) Prives, C., et al. (1998) Genes Dev 12 462-471.