ã
Oncogene (2001) 20, 748 ± 758
2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00
www.nature.com/onc
Expression of homologues for p53 and p73 in the softshell clam (Mya
arenaria), a naturally-occurring model for human cancer
Melissa L Kelley2, Per Winge3, Jason D Heaney1, Raymond E Stephens4,5, Julianne H Farell1,
Rebecca J Van Beneden2, Carol L Reinisch4, Michael P Lesser1 and Charles W Walker*,1
1
Department of Zoology and Marine Biomedical Research Group, Rudman Hall, University of New Hampshire, Durham, New
Hampshire, NH 03824, USA; 2Department of Biochemistry, Microbiology and Molecular Biology and School of Marine Science,
University of Maine, Orono, Maine, ME 04469-5751, USA; 3UNIGEN Center for Molecular Biology, University of Trondheim
N-7005, Trondheim, Norway; 4Marine Biological Laboratory, Woods Hole, Massachusetts, MA 02543, USA; 5Department of
Physiology/L720, Boston University School of Medicine, Boston, Massachusetts, MA 02118-2394, USA
Homologues for human p53 (Hsp53) and p73 (Hsp73)
genes were cloned and expression patterns for their
corresponding proteins analysed in tissues from normal
and leukemic softshell clams (Mya arenaria). These are
the ®rst structural and functional data for p53 and p73
cDNAs and gene products in a naturally occurring, nonmammalian disease model. Core sequence of the predicted
clam p53 (Map53) and p73 (Map73) proteins is virtually
identical and includes the following highly conserved
regions: the transcriptional activation domain (TAD),
MDM2 binding site, ATM phosphorylation site, proline
rich domain, DNA binding domains (DBDs) II-V, nuclear
import and export signals and the tetramerization domain.
The core sequence is a structural mosaic of the
corresponding human proteins, with the TAD and DBDs
resembling Hsp53 and Hsp73, respectively. This suggests
that Map53 and Map73 proteins may function similarly to
human proteins. Clam proteins have either a short
(Map53) or long (Map73) C-terminal extension. These
features suggest that Map53 and Map73 may be alternate
splice variants of a p63/p73-like ancestral gene. Map73 is
signi®cantly upregulated in hemocytes and adductor
muscle from leukemic clams. In leukemic hemocytes, both
proteins are absent from the nucleus and sequestered in the
cytoplasm. This observation suggests that a non-mutational p53/p73-dependent mechanism may be involved in
the clam disease. Further studies of these gene products in
clams may reveal p53/p73-related molecular mechanisms
that are held in common with Burkitt's lymphoma or other
human cancers. Oncogene (2001) 20, 748 ± 758.
Keywords: p53; p73; Burkitt's lymphoma; cancer; Mya
arenaria clam; nuclear exclusion
Introduction
The tumor suppressor gene p53 is one of the best
studied genes in human cancer research. In normal
*Correspondence: CW Walker
Received 7 August 2000; revised 22 November 2000; accepted 29
November 2000
cells, p53 plays a pivotal role in regulating the cell cycle
and in preventing neoplastic transformation (May and
May, 1999; Burns and El-Deiry, 1999). The p53 protein
functions as a sequence-speci®c transcription factor
that promotes expression of genes involved in cell cycle
arrest, DNA editing and repair or DNA damageinduced apoptosis (Kaelin, 1999). The intense interest
in this gene results from its high frequency of mutation
in human cancers (de Fromentel et al., 1992; May and
May, 1999). Mutations in human p53 [Homo sapiens
p53 (Hsp53)] have been found in more than 50% of
human cancers and are especially numerous in colorectal, lung and breast cancers and leukemias (Greenblatt et al., 1994). Alternatively, abnormal p53 activity
may result from di€erential expression and degradation
(Shieh et al., 1997) and phosphorylation and acetylation (Burns and El-Diery, 1999). In some cancers,
exclusion from the nucleus blocks p53-mediated
inhibition of cellular proliferation (Moll et al., 1992,
1996; Lu et al., 2000).
For more than 20 years, p53 was thought to be a
unique gene in vertebrate genomes. Recently, however,
two additional members of this expanding gene family
were found in humans (Kaghad et al., 1997; Schmale
and Bamberger, 1997; Osada et al., 1998; Senoo et al.,
1998; Trink et al., 1998; Yang et al., 1998). Human p63
(Hsp63) and p73 (Hsp73) genes encode proteins that
share signi®cant amino acid identity with p53 (Levrero
et al., 1999). Both Hsp63 and Hsp73 have multiple
splice variants that produce proteins with di€ering
functions dependent upon the particular tissue type in
which they are found (Marin and Kaelin, 2000). The a
splice variants of Hsp63 and Hsp73 proteins possess a
variable C-terminal extension beyond the core sequence
(Kaghad et al., 1997; Schmale and Bamburger, 1997).
These splice variants have reduced functional similarity
to p53 and contain a conserved sterile alpha motif
(SAM domain) that interacts with proteins involved in
signal transduction and transcriptional repression (Jost
et al., 1997; Kaghad et al., 1997; De Laurenzi et al.,
1998; Yang et al., 1998; Thanos and Bowie, 1999).
While p53 family members exist throughout the
mammals, homologues for p73 in other vertebrates
have only been reported for Barbus barbus (a cyprinid
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
®sh: Accession #AF043641; Bhaskaran et al., 2000). In
the invertebrates, a p73-like gene has been identi®ed
for Loligo forbesi (squid; Accession #U43595), a p53like gene has been identi®ed and its role in apoptosis
characterized for Drosophila melanogaster (fruit ¯y;
Brodsky et al., 2000; Ollmann et al., 2000; Steller,
2000) and previously, partial p53 sequences were
identi®ed for M. arenaria (Barker et al., 1997;
Accession #U45237 and #U45238). Here we report
the full-length p53 and p73 cDNA homologues for M.
arenaria (Accession #AF253323 and #AF253324,
respectively).
M. arenaria provides a non-traditional model for
studies of p53 and p73 protein structure and function
in a naturally-occurring leukemia. Bivalve mollusks
have an open circulatory system in which a tubular
heart weakly pumps hemolymph to major vessels and
then to endothelium-lined sinuses that permeate the
gonad, gills, foot, mantle, digestive gland and other
organs. In normal clams, agranular and granular
hemocytes have a broad range of functions, including
defense against pathogens, digestion and excretion
(Au€ret, 1988). The origin of these hemocytes remains
unknown (Cheng, 1981 and references therein). M.
arenaria develops a fatal leukemia at high incidence (up
to 60%) in natural populations along the northeastern
coastal US (Reinisch et al., 1984; Elston et al., 1992).
Morphologically, diseased hemocytes resemble human
Burkitt's lymphoma cells (Baine, 1999; Walker CW,
unpublished observations).
In this study, we present cDNA sequences for the
clam homologues of the Hsp53 and Hsp73 genes. The
highly conserved structure of the clam genes relative
to the corresponding human genes suggests that they
may have similar functions. Protein expression
patterns for both homologues, M. arenaria p53
(Map53) and p73 (Map73) are compared for normal
and leukemic clam tissues. Only Map73 is di€erentially expressed in this clam disease background. Both
Map53 and Map73 proteins are sequestered in the
cytoplasm and thus absent from the nuclei of leukemic
clam hemocytes.
Results
Comparison of p53 and p73 from M. arenaria with other
p53 homologues
The full-length cDNAs for Map53 and Map73 were
identi®ed from normal gill tissue. The cDNA for
Map53 is 2537 base pairs (bp) long with an open
reading frame (ORF) of 1305 bp; Map73 is 2451 bp
long with an ORF of 1839 bp (Figure 1). There are
two possible start codons (nucleotides (nt) 79 ± 81 and
103 ± 105) that initiate ORFs. Map53 encodes a protein
of 435 amino acids (relative to the second ATG) with a
calculated molecular mass of 52.8 kDa; Map73 encodes
a protein of 613 amino acids with a calculated
molecular mass of 74.8 kDa. The core sequence for
both Map53 and Map73 is essentially identical through
749
Figure 1 Nucleotide and predicted amino acid sequence including 5' and 3' untranslated regions (UTRs) for Map53 and Map73.
Core sequence (nucleotide (nt) 1 ± 1317) for Map53 and Map73 is
nearly identical. Five nt di€erences were identi®ed in the core
sequence between Map53 and Map73 with two resulting in amino
acid changes (nt 131 and 323=C, both amino acids change to P;
nt 330=T, 768 and 824=C, no amino acid changes; indicated by
italics). PolyA(+) tail (not shown) immediately follows end of 3'
UTR for both Map53 and Map73. Two possible start codons
(ATG) are shown in bold, and cytoplasmic and nuclear
polyadenylation signals are underlined; stop codons are indicated
by asterisks (*). Numbers correspond to nt sequence
amino acid 405, after which the sequences diverge
(Figures 1 and 2). Map53 has a short C-terminal
extension, while Map73 contains an elongated Cterminal extension including a SAM domain. Map53
di€ers signi®cantly in structure from D. melanogaster
p53. Map73 is more similar to squid (L. forbesi) Sqp73
than to either Hsp53 or Hsp73 (Figure 2).
Transcriptional activation domain (TAD) In Hsp53,
amino acids 12 ± 23 constitute a TAD that interacts
with the basal transcriptional machinery to promote
expression of genes related to: (a) negative regulation
of p53 (Honda et al., 1997); (b) DNA editing and
repair (Tanaka et al., 2000) and (c) apoptosis
(Miyashita et al., 1994; O'Connor et al., 1997;
Avramis et al., 1998; Tanaka et al., 2000). The TAD
of Map53/73 is 67% identical to that of Hsp53, while
only 33 and 42% identical to Hsp63 and Hsp73,
respectively. Amino acid residues F19, W23 and L26
Oncogene
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
750
Figure 2 Alignment of amino acid sequences for p53 and p73 homologues from M. arenaria (Map53 and Map73) with human p53
(Hsp53), p63a (Hsp63a) and p73a (Hsp73a), the p73 homologue from L. forbesi (Sqp73) and the p53 homologue from D.
melanogaster (Dmp53). Abbreviations: PxxP, proline rich domain; NLS I & II, nuclear localization domains; NES, nuclear export
domain, SAM, sterile alpha motif; DNA binding domains, DBDs I ± V; *, putative conserved phosphorylation sites; #, conserved
hydrophobic residues which in Hsp53 interact with MDM2; }, residues in Hsp53 that bind DNA; ¨, Hsp53 mutational hot-spots;
Zn, residues involved in zinc binding
present in Hsp53 and essential for binding, MDM2
(HDM2 in humans) are conserved in the clam
homologues (Figure 2; Kussie et al., 1996; Dobbelstein
et al., 1998).
DNA-binding domain (DBD) The DBD in Map53/73
is localized between amino acid residues 102 and 292
and contains four highly conserved DNA-binding
motifs (II-V; Figure 2). Overall predicted protein
Oncogene
conservation of Map53/73 to Hsp53 is 40% in the
DBD. Speci®cally within this region, conservation of
Map53/73 to Hsp53 is 63% in domain II, 67% in
domain III, 73% in domain IV and 82% in domain V.
Amino acid conservation between Map53/73 and
Hsp73 is 77% in domain II, 75% in domain III,
78% in domain IV and 88% in domain V. Thus, the
TAD and DBDs of clam p53/73 seem to be a mosaic
of Hsp53 and Hsp73a sequences, with the TAD more
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
closely related to Hsp53 and the remaining DBDs more
closely related to Hsp73.
Other conserved regions Following the TAD in Hsp53
is a proline rich domain (amino acids 63 and 97)
containing ®ve PXXP motifs that are involved in
apoptosis and may bind SH3-domain containing
proteins (Walker and Levine, 1996; Venot et al.,
1998). In Map53/73, two PXXP motifs are present
towards the C-terminal end of the corresponding
proline rich domain; one is conserved with Hsp53
and the other is not conserved (Figure 2).
Map53/73 has a bipartite nuclear localization signal
(NLS), while the NLS of Hsp53 is tripartite (Dang and
Lee, 1989; Shaulsky et al., 1991; Liang and Clarke,
1999). NLS I of M. arenaria is conserved with all p53/
63/73 homologues in vertebrates and invertebrates
while NLS II is weakly similar to the NLS of
vertebrate p53 (Figure 2). The Map53/73 NLS I
precedes the tetramerization domain and also contains
the lysine residue that is acetylated in Hsp53 by PCAF,
which enhances p53 sequence speci®c DNA binding
(Nishimori et al., 2000; Liu et al., 2000).
Hsp53 functions as a tetramer, with residues 323 ±
356 required for tetramerization (Figure 2). Overall,
the tetramerization domain of Map53/73 is 59%
similar to Hsp53 and 81% similar to Hsp73. A nuclear
export signal exists within the tetramerization domain.
In Hsp53, the nuclear export signal is a leucine rich
sequence, MFXXLXXXLXL (Stommel et al., 1999); a
similar sequence (MLXXIXXXLXI) is also present in
the tetramerization domain of Map53/73.
Map53 has a unique short C-terminal region that is
not conserved relative to Hsp53 and only slightly
conserved relative to Hsp63 and Hsp73. Map73 has a
much longer C-terminal region than Map53 and
contains a SAM domain that is greater than 50%
conserved with those of Hsp63 and Hsp73.
Expression of p53 and p73 in normal and leukemic clams
Six clam tissues (gill, gonad, digestive gland, mantle,
adductor muscle and hemocytes) were examined by
Western blot analysis for expression of Map53 and
Map73. Map53 was expressed at higher levels than
Map73 in all tissues examined with the exception of the
mantle where Map73 levels were relatively higher than
Map53 (Figure 3). Map73 was not detected in normal
clam hemocytes and may be present at very low levels
in adductor muscle (Figure 3). Comparison of Map53
and Map73 expression patterns between normal and
leukemic clams showed that Map73 expression was
elevated in both adductor muscle and hemocytes of
leukemic clams (Figure 3). Map53 and Map73 were
expressed at similar levels in leukemic clam hemocytes.
Coomassie stained gels con®rmed that there was no
detectable di€erence in protein loading (data not
shown). The unidenti®ed band (*70 kDa) detected in
adductor muscle and hemocytes was inconsistently
observed in di€erent tissues in both normal and
leukemic clams (Figure 3). Western blot analysis of in
vitro expressed Map53 and Map73 proteins con®rms
that the Map53/73 antibody recognizes both proteins
(Figure 3).
751
Subcellular localization of Map53/73
In normal clam hemocytes, immunocytochemistry using
the Map53/73 antibody demonstrated that Map53
protein is predominantly present within the nucleus
(Figure 4a). In leukemic hemocytes, Map53 and Map73
were exclusively present within the cytoplasm and absent
from the nucleus (Figure 4b). As a positive control, the
Map53/73 polyclonal antibody recognized human proteins where they are normally present within most nuclei
of patient-derived, human monocytic leukemia cells
(AML-M5; Figure 4c).
Cytospins of clam and human leukemia cells
Morphologically, normal clam hemocytes (Figure 5a,
inset) are very di€erent from leukemic clam hemocytes
(Figure 5a). In some ways, leukemic clam hemocytes
resemble cells from a Burkitt's lymphoma cell line
(Ramos CRL-1596; ATCC; Figure 5b) and cells from a
Burkitt's lymphoma patient (Figure 5c). Like human
Burkitt's lymphoma cells, clam leukemia cells have a
monotonous appearance (Figure 5a). A striking
similarity exists between clam leukemia and human
Burkitt's lymphoma cells. Within the cytoplasm of
these cancer cells which are derived from radically
di€erent sources (mollusk and mammal), there are
numerous vacuoles (5 ± 15/cell) that stain intensely with
oil red O. In humans, these vacuoles are diagnostic
features for Burkitt's lymphoma (Grogan et al., 1982;
Baine, 1999) (Figure 5b,c). For comparison, monocytic
leukemia cells from an AML-M5 patient were also
stained with oil red O; oil red O staining vacuoles are
absent from these cells (Figure 5d).
Discussion
To our knowledge, this study provides the ®rst fulllength cDNA sequences for both p53 and p73 and the
only existing data on the expression and subcellular
localization of the corresponding proteins for any nonvertebrate organism. The highly conserved structural
domains that de®ne the p53 gene family members in
humans are present in Map53 and Map73 suggesting
that their functional roles may also be conserved in
clams. Di€erential patterns of expression for Map53
and Map73 are also documented in this naturally
occurring, non-mammalian cancer. This disease in the
softshell clam has morphological similarity to human
diseases like Burkitt's lymphoma (Baine, 1999; Walker,
unpublished observations). Furthermore, elevated expression of the Map73 protein and absence of Map53/
73 proteins from the nuclei of leukemic clam
hemocytes, suggest a biological role for Map53 and
Map73 in this clam disease, which appears similar in
many respects to several human cancers.
Oncogene
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
752
Figure 3 Western blot analyses showing in vitro expressed proteins and protein expression patterns for Map53 and Map73 in six
clam tissues. Representative normal and leukemic individuals are shown for gonad, gill, digestive gland, mantle and adductor
muscle. Hemocytes from three normal and three leukemic individuals are shown. Lane 1=Map53, Lane 2=Map73, N=normal,
L=leukemic
Figure 4 Immunocytochemistry using the Map53/73 polyclonal
antibody on cytospins of: (a) normal clam hemocytes showing
primarily nuclear localization of Map53 (arrows); (b) leukemic
clam hemocytes showing localization of Map53 and Map73 in the
cytoplasm and absence from the nucleus; (c) human AML-M5
leukemia cells showing positive control of the Map53/73 antibody; insets in a,b are negative controls without primary
antibody. Scale bars: a,b=10 mm; c=20 mm
Structural aspects of Map53 and Map73: comparison to
Hsp53
In Map53/73 there are two possible start codons
(ATG), both of which initiate an ORF (Figure 1).
With the exception of some protooncogenes and
growth-control genes, `non-functional' initiating codons are rare in vertebrates, but appear to be more
common in D. melanogaster (Kozak, 1987, 1989).
Based on the conserved Kozak sequence (Kozak,
Oncogene
1989) and alignment with other p53 homologues, we
predict that the second ATG is the most likely start site
for Map53/73 translation. However, if translation
begins at the ®rst ATG, then an N-terminal tag may
result on Map 53/73 that is similar to Hsp63 but not
found in other Hsp53/p73s. In the core sequence of
Map53 and Map73, ®ve nucleotide di€erences were
identi®ed that likely re¯ect polymorphisms. It is
possible that the di€erences may have resulted from
polymerase errors in the PCR or sequencing reactions,
although clones were sequenced in both directions at
least once using a high ®delity polymerase in the PCR
(see Materials and methods).
Signi®cant sequence identity to the Hsp53 TAD
suggests that Map53/73 may be responsible for the
transactivation of genes involved in: (a) negative
regulation of p53 (Honda et al., 1997); (b) DNA editing
and repair (Tanaka et al., 2000) and (c) apoptosis
(Miyashita et al., 1994; O'Connor et al., 1997; Avramis
et al., 1998; Tanaka et al., 2000). Of particular interest
is MDM2, a ring ®nger, E3 ubiquitin-protein ligase that
blocks transcriptional activity of Hsp53 in addition to
targeting it for proteolysis. The binding site for MDM2
(QETFSDLWKLLP) is contained within the TAD of
Hsp53 and has the conserved motif xxxFxxxWxxLx.
Amino acids F, W and L are important in stabilization
of hydrophobic interactions between MDM2 and
Hsp53 (Kussie et al., 1996). The binding site for
MDM2 in Map53/73 is 58% identical to Hsp53.
Residues 16 ± 27 (QETFEYLWHTLE) in Map53/73
include the critical amino acids F, W and L. Potentially
important regulatory phosphorylation sites are also
present within the TAD of Map53/73, e.g. S15 (Hsp53
S15) and T18 (Hsp53 T18). In humans, phosphorylation of S15 is sucient to block MDM2 binding and
leads to accumulation of p53 protein (Banin et al.,
1998; Canman et al., 1998). Hsp73 and Hsp63 also
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
753
Figure 5 Cytospins stained with oil red O for each of the following: (a) leukemic clam hemocytes; notice mitotic division (arrows)
(inset shows normal clam hemocytes at the same magni®cation); (b) Ramos RA 1 Burkitt's lymphoma cell line (CRL-1596; ATCC);
(c) Burkitt's lymphoma cells (patient-derived); (d) AML M5 cells (patient-derived). Scale bars: a=10 mm; b ± d=20 mm
contain the MDM2 binding motif with 42 and 33%
identity to Hsp53, respectively. Although Hsp73 can
bind MDM2, it is not targeted for degradation by
MDM2 (Balint et al., 1999; Zeng et al., 1999; Momand
et al., 2000). The MDM2 binding sites in Map53 and
Map73 are identical to each other and more similar to
the same site in Hsp53 than either Hsp63 or Hsp73,
suggesting that both clam proteins may be negatively
regulated and tagged for destruction by an MDM2
homologue. In order to test this hypothesis, studies are
underway to identify an MDM2 homologue in M.
arenaria.
The DBDs are highly conserved in vertebrate p53
family members. The DBDs of Hsp53 contain 80 ± 90%
of the known mutations that have been described in a
wide variety of human cancers (Cho et al., 1994;
Greenblatt et al., 1994). Eight amino acid residues that
are most frequently mutated in Hsp53 (V173, R175,
C243, G245, R248, R249, V273 and D282) are conserved
in Map53/73 (V204, R206, C273, G275, R278, R279,
V302 and D311; May and May, 1999). These include
four residues that bind zinc and two residues that directly
contact DNA (May and May, 1999). Conservation of
these amino acids suggests that Map53/73 binds DNA at
the canonical p53 DNA-binding sites and that these
residues should be analysed for mutations. Mutations
have rarely been demonstrated in Hsp73 (Corn et al.,
1999) or in Hsp63 (Sunahara et al., 1999). Mutations in
p53 are also rare in non-mammalian organisms,
although a point mutation has been identi®ed for
Map53/73 (Barker et al., 1997).
Many SAM-containing variants of Hsp63 and Hsp73
genes result from alternative splicing and do not
possess p53-like functions such as oligomerization,
activation of promoters containing p53-binding sites
and induction of apoptosis (Jost et al., 1997; Kaghad et
al., 1997; De Laurenzi et al., 1998; Yang et al., 1998;
Thanos and Bowie, 1999; Kaelin, 1999). Because of the
almost complete identity of the core sequences of
Map53 and Map73, these gene products may result
from alternative splicing and may have variable
functions related to the nature and timing of
interactions between other proteins and the SAM
domain of Map73. This unique feature of Map53 and
Map73 proteins provides an excellent opportunity for
study of SAM domain functions in general.
Oncogene
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
754
Protein expression
Hsp53 is important in regulating cell cycle and apoptosis
and is ubiquitously expressed at low levels in normal
tissues (Marin and Kaelin, 2000). Expression patterns for
Hsp73 and Hsp63 are less well characterized, but these
proteins are usually restricted to speci®c tissues (Yang et
al., 1998; Kaelin, 1999). Since Hsp53 is widely expressed
in human tissues, it is not surprising that the Map53
protein was encountered in all six clam tissues (Figure 3).
Map73 was not detected by this method in normal clam
hemocytes, although it might be present at very low
levels in adductor muscle from normal clams (Figure 3).
Western blots also revealed a third band of *70 kDa
that may be an unidenti®ed p53-related protein or
another splice variant of Map53/73 (Figure 3).
Map53 and Map73 in clam leukemia
In humans, expression of p53 is depressed in fresh
tissues derived from patients with acute myelogenous
leukemia (AML) and in myeloproliferative diseases
such as chronic myeloid leukemia (CML), myelo®brosis and polycythemia vera (Prokocimer et al., 1986).
In contrast, expression is often elevated in Burkitt's
lymphoma (Prokocimer et al., 1986).
Although elevated expression of Map53 was not
observed in leukemic clam hemocytes, the increase in
Map73 expression seen in hemocytes and adductor
muscle is consistent with an increase in p73 transcripts
and protein in a variety of leukemic cell lines and
leukemia patients (Peters et al., 1999). However, Corn et
al. (1999) found that the promoter region of the p73 gene
is methylated in Burkitt's lymphoma cells and concluded
that p73 is not likely to be expressed in these cancers.
The Map53/73 antibody does not distinguish between
the two clam proteins in in situ assays. Immunocytochemistry using this antibody demonstrated that clam
p53 family members were localized predominantly
within the nucleus in normal clam hemocytes (Figure
4a) and exclusively within the cytoplasm of leukemic
clam hemocytes (Figure 4b). Western blot analysis of
normal clam hemocytes showed expression of only
Map53; Map73 was not detected in normal hemocytes
by this method (Figure 3a). Therefore, it is likely that in
normal clam hemocytes, nuclear staining is only from
p53 localization (Figure 4a). In clam leukemia cells, the
antibody may simultaneously be demonstrating localization of p53 and p73 in the cytoplasm (Figure 4b). The
clam p53/73 antibody also recognized p53 family
members in the nuclei of human AML-M5 cells derived
from a patient (Figure 4c).
The precise mechanisms of involvement of p53
family members in human leukemias, especially
Burkitt's lymphoma, are only now being determined.
An explanation of our results may be provided by the
recently discovered anti-apoptotic activity of mammalian p73 found in the developing neurons of mice
(Pozniak et al., 2000). Relative levels of Map53 and
Map73 might determine whether clam leukemia cells
proliferate or undergo apoptosis and die.
Oncogene
Finally, the cytoplasmic sequestration, or nuclear
exclusion, of p53/p73 proteins in clam leukemia cells
re¯ects a similar situation seen in several human
cancers (Moll et al., 1992, 1996; Schlamp et al., 1997;
Nagai et al., 1998; Lu et al., 2000). A number of
mechanisms may be responsible for this phenomenon.
It is possible that the nuclear import mechanism
mediated by the NLS of Map53/73 is malfunctioning.
In mammalian cells, nuclear import normally involves
recognition of the NLS by import proteins like
importin a (Liang and Clarke, 1999; Kim et al.,
2000) and subsequent transport of p53 to the nucleus
on microtubules using the molecular motor dynein
(Giannakakou et al., 2000). Alternatively, the nuclear
export mechanism mediated by the NES of Map53/73
may be overactive, removing Map53/73 as soon as the
proteins enter the nucleus (Giannakakou et al., 2000).
Studies are underway to discriminate between these
alternative means for excluding Map53/p73 from the
nucleus, using antibodies to dynein for the former and
leucomycin B for the latter (Giannakakou et al., 2000).
Regardless of which mechanism is involved, the result
would be inhibition of any anti-proliferative activities
of Map53/73. Such mechanisms could be very
important in understanding this disease in clams, since
leukemia cells in clams divide slowly, do not die during
the life-time of the diseased clam and accumulate to
approximately 600 times the numbers of normal
hemocytes (Walker CW, unpublished observations).
Clam leukemia o€ers signi®cant advantages over
currently available models for studying the normal and
abnormal functions of p53 family members. The clam
model provides an in vitro and in vivo alternative to the
relatively few human leukemia cell lines. Also, naturally
occurring populations of clams (including leukemic
individuals) are more similar to an outbreeding, human
clinical population than are those generated from inbred
mouse strains or by intentional exposure to known
tumor viruses. Currently M. arenaria is already being
used to study a gonadal neoplasm (germinoma; Van
Beneden et al., 1998). Further studies of these gene
products in clams may reveal p53/p73-related molecular
mechanisms that are held in common with Burkitt's
lymphoma or other human cancers.
Evolution of p53 homologues
p53 family members have been found throughout the
vertebrates and in some invertebrates, like mollusks (L.
forbesi and M. arenaria) and arthropods (D. melanogaster). The existence and structure of these genes,
particularly in invertebrates, suggests that a p53/p63like gene is ancient and already existed in organisms
550 ± 600 million years ago (Figure 6). Both deuterostomes and protostomes have p63/p73-like proteins
with a C-terminal SAM domain, suggesting that this
may be the ancestral form of the protein. Similarly, the
alternatively spliced forms of the p63 and p73 proteins
are evolutionary conserved and probably existed in the
ancestral gene. Map53 and Map73 share a nearly
identical core sequence, diverging at the 3' end. This
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
Figure 6 Phylogenetic relationship between members of the p53
family. The unrooted, neighbor-joined tree, based on alignment of
vertebrate and invertebrate p53, p63 and p73 proteins, suggests that
overall the invertebrate p53/p73s have a closer relationship to the
p63/p73 proteins than to the p53s. Numbers indicate branches with
signi®cant bootstrap values. Five hundred bootstrap replicas were
analysed. The scale bar indicates the number of substitutions per
site. The p53s from the ®sh superorder Ostariophysi and the
unranked group Neoteleostei are indicated. Abbreviations: Bb,
Barbus barbus; Bt, Bos taurus; Ca, Chlorocebus aethiops. Cf, Canis
familiaris; Cg, Cricetulus griseus; Cp, Cavia porcellus; Dm,
Drosophila melanogaster; Dr, Danio reiro; Fc, Felis catus; Gg,
Gallus gallus; Hs, Homo sapiens; Ip, Ictalurus punctatus; Ma, mya
arenaria; Mm, Mus musculus; Mn, Mastomys natalensis; Oa, Ovis
aries; Oc, Oryctolagus cuniculus; Ol, Oryzias latipes; Om,
Oncorhynchus mykiss; Pf, Platichthys ¯esus; Rn, Rattus norwegicus;
Sq, Loligo forbesi; SS, Sus scrofa; Tm, Tetraodon miurus; Xh,
Xiphophorus helleri; Xl, Xenopus laevis
leads us to believe that Map53 and Map73 may be
splice variants of the same gene. We are sequencing
genomic DNA to determine the validity of this idea.
The Map53/73 core sequence appears to be a mosaic of
the corresponding human proteins, with the TAD more
like Hsp53 and the DBDs more like Hsp63 and Hsp73.
The high sequence conservation suggests that traditional human cell cycle, negative-regulatory, DNA
editing and repair and apoptosis genes may be
transcriptional targets for clam p53/p73. Since this is
the only non-vertebrate organism for which we
currently have information on the structure, expression
and subcellular localization of both p53 and p73,
further studies of the genomic sequence and gene
products in the clam may elucidate the function and
evolution of this important gene family.
Materials and methods
Cloning and sequencing Map53 and Map73
A cDNA library was constructed in lambda ZAP (Stratagene) using polyA(+)-selected RNA samples from normal
clam tissue (visceral mass). Ampli®ed titer of the library was
2.56108 PFUs/ml with 84% recombinants. Partial sequence
for Map53 was obtained by RT ± PCR using polyA(+)selected RNA as template and degenerate primers designed
after conserved motifs in vertebrate p53 DBD III (VVR/
KRCPHH) and DBD V (EVRVCACP). The Map53 RT ±
PCR product was used as a probe to screen the cDNA
library. Partial sequences for Map53 (Accessions #45237 and
#U45238) were obtained and extended using 5' and 3' RACE
PCR (Gibco/BRL) on cDNA from normal clam hemocytes
using primers designed to the Map53 sequence. Finally,
Map53 and Map73 full-length clones, including 5' and 3'
UTRs, were obtained by Marathon RACE PCR (CLONTECH). Positive RACE PCR products were identi®ed by
Southern blot analysis (Sambrook et al., 1989), cloned into a
pCRII vector (Invitrogen) and sequenced in both directions
at the University of Maine Automated Sequencing Facility.
Contiguous sequences were assembled and primers were
designed to amplify the full-length Map53 and Map73
sequences from single-stranded cDNA synthesized from gill
RNA of the same individual. The same forward primer (5'TCATATTTGGTAATATGAGCCACGAG-3') was paired
with a reverse primer for either Map53 (5'-CCAACTCAAATCGTTGTGAAAATGG-3')
or
Map73
(5'GTTCAGCGTCAAAAGCAGATTAGTGA-3') to amplify
a single band with a high ®delity polymerase (Accutaq LA;
Sigma). The full-length PCR products for Map53 and Map73
were cloned and sequenced as above (Accession #AF253323
and #AF253324, respectively). Nucleotide di€erences in the
resulting sequences probably re¯ect polymorphisms.
755
Antibody synthesis
A polyclonal antibody to Map53/73 was raised against a
peptide synthesized from a highly immunogenic region
including part of DBD V (Map53/73 23-mer CACPGRDRKADERGSLPPMVSGG). The peptide was synthesized by
the Merri®eld solid-phase method on an Applied Biosystems
Model 431A peptide synthesizer, using Fast Moc chemistry
cycles, conjugated to BSA with gluteraldehyde, and the
resulting conjugate was used to immunize three New Zealand
White Rabbits (Millbrook Farm, MA, USA; Harlow and
Lane, 1999). Antibodies were anity-puri®ed on an A-Gel
column (BioRad). The polyclonal antibody obtained was
successfully screened for its ability to recognize both Map53
and Map73 proteins by Western blot analysis of both in vitro
expressed clam proteins and whole cell extracts of clam
proteins from various tissues (Figure 3).
In vitro protein synthesis and Western blots
Six tissues (hemocytes, gill, gonad, digestive gland, mantle
and adductor muscle) were dissected from normal and
leukemic clams and ¯ash frozen in liquid nitrogen. Tissues
were homogenized in 30 mM Tris pH 8.0, 3 mM MgCl2, 0.5%
Trition X-100, 5 mM DTT which contained protease
inhibitors (7.4 mg/ml TLCK, 5 mg/ml leupeptin, 7 mg/ml
pepstatin A, 16.7 mg/ml aprotinin, 1 mM PMSF). Homogenates were centrifuged at 11 000 g at 48C for 10 min.
Concentration of solubilized protein in the supernatants
was determined by Bradford assay (Bio-Rad).
Proteins were denatured at 958C for 3 min in loading
bu€er (50 mM Tris-HCl, pH 6.8, 2% SDS and 10% glycerol)
with 100 mM dithiothreitol (DTT). Fifty to 75 mg of protein
were loaded onto duplicate 10% Tris/glycine SDS ± PAGE
gels. After electrophoresis at 20 mA/gel, one gel was stained
in Coomassie Brilliant Blue R-250 : methanol : acetic acid
(0.1% : 50% : 10%) and destained in methanol : acetic acid
(45% : 10%). Gels were then soaked in methanol : acetic
Oncogene
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
756
acid : glycerol (7% : 7% : 1%) for 30 min and dried under a
vacuum at 808C for 1 h. The proteins on the duplicate gel
were electroblotted at 200 mA for 50 min onto an Immobilon-P membrane (Millipore) using a semi-dry blotting system
(Labconco). Membranes were blocked overnight in TTBS
(100 mM Tris pH 7.5, 0.9% NaCl, 0.1% Tween 20)+1 mg/
ml gelatin at 48C. After three washes in TTBS, membranes
were incubated in a 1 : 150 dilution of Map53/73 antibody for
2 ± 4 h at room temperature. Membranes were washed in
TTBS and incubated 45 ± 60 min with a 1 : 10 000 dilution of
goat-anti-rabbit IgG secondary antibody conjugated to
horseradish peroxidase (Sigma). Membranes were washed in
TTBS, equilibrated in 0.1 M Tris-HCl, pH 8.5, then incubated for 1 min in enhanced chemiluminescence (ECL)
solution (0.1 M Tris pH 8.5, 0.05% hydrogen peroxide,
680 mM p-coumeric acid and 0.13 mg/ml luminol in DMSO
(Sigma)). Membranes were exposed to Hyper®lm MP
(Amersham Pharmacia Biotech) for 30 ± 60 s.
35
S-radiolabeled and unlabeled Map53 and Map73 proteins
were expressed in vitro using the TNT Quick-Coupled
Reticulocyte Lysate System (Promega). Two ml of the TNT
reactions were subjected to SDS-polyacrylamide gel electrophoresis, followed by autoradiography (data not shown) or
Western blot analysis (as described above) to con®rm protein
synthesis.
Collection of clam hemocytes and staging clam leukemia
Clams were collected from intertidal sand ¯ats on Marsh
Island in New Bedford Harbor at Fairhaven, MA, USA. For
biopsy, a small aliquot (2 ± 3 ml) of hemolymph was removed
from the blood sinus surrounding the heart using a syringe
and 21 gauge needle, placed in a ¯at-bottom 96-well
microtiter plate and incubated for 2 ± 24 h at 88C. Depending
upon the microscopic appearance of the hemocytes, clams
were classi®ed as normal (no leukemic cells), incipient or
intermediate (50 and 4100%) or 100% leukemic. When
appropriate, more precise classi®cation of clam hemocytes
was accomplished using the murine monoclonal antibody,
1E10, which is speci®c for a transmembrane protein in
leukemic hemocytes of M. arenaria and capable of di€erentiating one leukemic cell in a ®eld of 100 000 cells (Miosky
et al., 1989).
Immunocytochemistry
Hemocytes from normal clams, 100% leukemic clams and
cells from a human AML-M5 patient were prepared as
cytospins on poly-L-lysine coated slides (Shandon) in a
Shandon Cytofuge3 and air dried. Cells were ®xed and
permeabilized by immersion in cold (48C) acetone for 10 min.
Endogenous peroxidase activity was quenched by 30 min
immersion in 0.3% H2O2 in MeOH at room temperature. In
situ detection of Map53/73 was accomplished by overnight
incubation (48C) with Map53/73 antibody diluted 1 : 50 in
10 mM NaPO4 pH 7.5, 0.9% NaCl, 0.1% Triton X-100. To
visualize the results, slides were developed with the Vectastain
ABC Elite Rabbit IgG kit (Vector Laboratories), mounted in
DPX mountant (EMS) and photographed using a Spot
digital camera on a Zeiss Axiophot microscope. Control
cytospins received identical treatment in the absence of
primary antibody.
Cytospins of clam and human leukemia cells and oil red O
staining
Ramos RA 1 Burkitt's lymphoma cell line (CRL-1596;
ATCC), Burkitt's lymphoma cells from a patient and acute
monocytic leukemia cells from a patient (AML-M5) were
grown at 378C in RPMI 1640 with 10% fetal calf serum and
1% penicillin-streptomycin (Sigma) and subsequently separated in Ficoll (A Freedman, personal communication).
Approximately 500 ml of cells from these two human sources
and also clam leukemia cells were centrifuged at 900 r.p.m.
for 8 min in a Shandon Cytofuge3, treated with oil red O to
stain oil droplets and counter-stained with hematoxylin
(Humason, 1979). After staining, cells were mounted in
DPX mountant (EMS) and photographed using a Spot
digital camera on a Zeiss Axiophot microscope.
Amino acid alignment and phylogenetic analysis
Alignment of the predicted protein sequences for Map53 and
Map73, Hsp53, Hsp63a and Hsp73a splice variants, L. forbesi
p73 and D. melanogaster p53 was accomplished with Clustal
W (Thompson et al., 1994) and manually re®ned with the
GenDoc program (Nicholas and Nicholas, 1997). The
distances between the sequences were calculated and a
neighbor-joining tree produced with the Clustal W program.
Gap penalty and gap extension penalty was set to 10 and 0.1,
respectively. Bootstrap analysis was performed with 500
bootstrap replicas. The phylogenetic tree was made with the
TreeView program (Page, 1996).
Acknowledgments
This study was supported by a grant from NCI (Molecular
Cytology, HD28204-01A1) to CW Walker. CL Reinisch
was supported by NIH grant CA44307. We thank J Levy
and Dr Andy Laudano (UNH, Department of Molecular
Biology and Biochemistry) for synthesizing the peptide
used in generating the Map53/73 antibody. Dr A Freedman of Dana Farber Cancer Institute, Boston, MA, USA,
generously donated the Ramos cell line and the Burkitt's
lymphoma and AML-M5 cells from human patients. T
Johnson performed the cytospins and oil red O staining of
human leukemia and clam leukemia cells. We also thank P
Singer and K Veit of the University of Maine Automated
Sequencing Facility.
References
Au€ret M. (1988). Amer. Fish. Soc. Spec. Pub., 18, 169 ± 177.
Avramis VI, Nandy P, Kwock R, Solorzano MM, Mukherjee
SK, Danenberg P and Cohen LJ. (1998). Anticancer Res.,
18, 2317 ± 2338.
Baine B. (1999). Leukemia Diagnosis. Blackwell Science, 2nd
edn., London, pp. 1 ± 200.
Balint E, Bates S and Vousden KH. (1999). Oncogene, 18,
3923 ± 3929.
Oncogene
Banin S, Moyal L, Shiel SY, Taya Y, Anderson CW, Chessa
L, Smorodinsky NI, Prives C, Reiss Y, Shiloh Y and Ziv
Y. (1998). Science, 281, 1674 ± 1677.
Barker CM, Calvert RJ, Walker CW and Reinisch CL.
(1997). Detection of mutant p53 in clam leukemia cells.
Exp. Cell. Res., 232, 240 ± 245.
Bhaskaran A, May D, Rand-Weaver M and Tyler CR.
(2000). Comp. Biochem. Physiol., 126, 49 ± 57.
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
Brodsky MH, Nordstrom W, Tsang G, Kwan E, Rubin GM
and Abrams JM. (2000). Cell, 101, 103 ± 113.
Burns TF and El-Diery WS. (1999). J. Cell. Physiol., 181,
231 ± 239.
Canman CE, Lim DS, Cimprich KA, Yhaya Y, Tamai K,
Sakaguchi K, Appella E, Kastan M and Siliciano JD.
(1998). Science, 281, 1677 ± 1679.
Cheng TC. (1981). Invertebrate Blood Cells. Academic Press,
London, pp. 233 ± 300.
Cho JJ, Gorina S, Je€rey PD and Pavletich NP. (1994).
Science, 265, 346 ± 355.
Corn PG, Kuerbitz SJ, van Noesel MM, Esteller M,
Compitello N, Baylin SB and Herman JG. (1999). Cancer
Res., 59, 3352 ± 3356.
Dang CV and Lee WM. (1989). J. Biol. Chem., 264, 18019 ±
18023.
de Fromentel C, May-Levin F, Mouriesse H, Lemerle J,
Charandrasekaran K and May P. (1992). Genes Chrom.
Cancer, 4, 1 ± 15.
De Laurenzi V, Cotsanzo A, Barcaroli D, Terrinoni A, Falco
M, Annicchiarico-Petruzzelli M, Levrero M and Melino
G. (1998). J. Exp. Med., 188, 1763 ± 1768.
Dobbelstein M, Wienzek S, Konig C and Roth J. (1998).
Oncogene, 18, 2101 ± 2106.
Elston RA, Moore JD and Brooks K. (1992). Rev. Aquatic
Sci., 6, 405 ± 438.
Giannakakou P, Sackett DL, Ward Y, Webster KR,
Blagosklonny MV and Fojo T. (2000). Nat. Cell. Biol.,
2, 709 ± 717.
Greenblatt MS, Bennett WP, Hollstein M and Harris CC.
(1994). Cancer Res., 54, 4855 ± 4878.
Grogan TM, Warnke RA and Kaplan HS. (1982). Cancer,
49, 1817 ± 1828.
Harlow E and Lane D. (1999). Using Antibodies: A
Laboratory Manual. Cold Spring Harbor Laboratory
Press, Cold Spring Harbor, NY, pp. 1 ± 495.
Honda R, Hirofumi T and Yasuda H. (1997). FEBS, 420,
25 ± 27.
Humason GL. (1979). Animal Tissue Techniques. 4th ed. WH
Freeman and Co., San Francisco, 1 ± 661.
Jost CA, Marin MC and Maelin Jr WG. (1997). Nature, 389,
191 ± 194.
Kaelin WG. (1999). Oncogene, 18, 7701 ± 7705.
Kaghad M, Bonnet H, Yang A, Creancier L, Biscan Y-C,
Valent A, Minty A, Chalon P, Lelias J-M, Dumont X,
Ferrara P, McKeon F and Caput D. (1997). Cell, 90, 809 ±
819.
Kim IS, Kim D-H, Han S-M, Chin MU, Nam H-J, Cho H-P,
Choi SY, Song B-J, Kim S-R, Bae Y-S and Moon Y-H.
(2000). J. Biol. Chem., 275, 23139 ± 23145.
Kozak M. (1987). Nucleic Acids Res., 15, 8125 ± 8148.
Kozak M. (1989). J. Cell. Biol., 108, 229 ± 241.
Kussie PH, Gorina S, Marechal V, Elenbaas B, Moreau J,
Levine AJ and Pavletich NP. (1996). Science, 274, 948 ±
953.
Levrero M, De Laurenzi V, Contanzo A, Gong J, Melino G
and Wang JY. (1999). Cell Death Di€er., 6, 1146 ± 1153.
Liang S-H and Clarke MF. (1999). J. Biol. Chem., 274,
32699 ± 32703.
Liu Y, Colosimo L, Yang XJ and Liao D. (2000). Mol. Cell.
Biol., 20, 5540 ± 5553.
Lu W, Pochampally R, Chen L, Traidej M, Wang Y and
Chen J. (2000). Oncogene, 19, 232 ± 240.
May P and May E. (1999). Oncogene, 18, 7621 ± 7636.
Marin MC and Kaelin WG. (2000). Biochem. Biophys. Acta,
1470, 94 ± 100.
Miosky DL, Smolowitz RM and Reinisch CL. (1989). J.
Invert. Pathol., 53, 32 ± 40.
Moll UM, Riou G and Levine AJ. (1992). Proc. Natl. Acad.
Sci., 89, 7262 ± 7266.
Moll UM, Ostermeyer AG, Haladay R, Wink®eld B, Frazier
M and Zambetti G. (1996). Mol. Cell. Biol., 16, 1126 ±
1137.
Momand J, Wu HH and Dasgupta G. (2000). Gene, 242, 15 ±
29.
Miyashita T, Krajewski S, Krajewska M, Wang HG, Lin
HK, Liebermann DA, Ho€man B and Reed JC. (1994).
Oncogene, 9, 1799 ± 1805.
Nagai J, Kigasawa H, Koga N, Katoh A, Nishihira H and
Nagao T. (1998). Leuk. Lymphoma., 28, 591 ± 597.
Nicholas KB and Nicholas HB. (1997). www.psc.edu/
biomed/genedoc.
Nishimori H, Nishikawa R, Fujimaki T, Nakagomi T,
Matsutani M, Huang HJ and Cavanee WK. (2000). J.
Neurooncol., 46, 17 ± 22.
O'Connor PM, Jackman J, Bae I, Myers TG, Fan S, Mutoh
M, Scudiero DA, Monks A, Sausville EA, Weinstein AN,
Friend S, Fornace AJ and Kohn KW. (1997). Cancer Res.,
57, 4285 ± 4300.
Ollmann M, Young LM, Di Como CJ, Karim F, Belvin M,
Robertson S, Whittaker K, Demsky M, Fisher WW,
Buchman A, Duyk G, Friedman L, Prives C and
Kopczynski C. (2000). Cell, 101, 91 ± 101.
Osada M, Ohba M, Kawahara C, Ishioka C, Kanamaru R,
Katoh I, Ikawa Y, Nimura Y, Nakagawara A, Obinata M
and Ikawa S. (1998). Nature Med., 4, 839 ± 843.
Page R. (1996). Comput. Appl. Biosci., 12, 357 ± 358.
Peters UR, Tschan MP, Kreuzer KA, Baskaynak G, Lass U,
Tobler A, Fey MF and Schmidt CA. (1999). Cancer Res.,
59, 4233 ± 4236.
Pozniak CD, Radinovic S, Yang A, McKean F, Kaplan DR
and Miller FD. (2000). Science, 289, 304 ± 306.
Prokocimer M, Shaklai M, Bassett HB, Wolf D, Gold®nger
N and Rotter V. (1986). Blood, 68, 113 ± 118.
Reinisch CL, Charles AM and Stone AM. (1984). Hazardous
Waste, 1, 73 ± 81.
Sambrook J, Fritsch EF and Maniatus T. (1989). Molecular
Cloning: A Laboratory Manual, 2nd edn. Cold Spring
Harbor, NY: Cold Spring Harbor Press.
Schlamp CL, Poulsen GL, Nork TM and Nickells RW.
(1997). J. Natl. Cancer Inst., 89, 1530 ± 1536.
Schmale H and Bamberger C. (1997). Oncogene, 15, 1363 ±
1367.
Senoo M, Seki N, Ohira M, Sugano S, Watanabe M,
Tachibana T, Shinkai Y and Kato H. (1998). Biochem.
Biophys. Res. Comm., 248, 603 ± 607.
Shaulsky G, Gold®nger N and Rotter V. (1991). Cancer Res.,
51, 5232 ± 5237.
Shieh SY, Ideda M, Taya Y and Prives C. (1997). Cell, 91,
324 ± 334.
Steller H. (2000). Nature Cell Biol., 2, 100 ± 102.
Stommel JM, Marchenko ND, Jimenez GS, Moll UM, Hope
TJ and Wahl GM. (1999). EMBO J., 18, 1660 ± 1672.
Sunahara M, Shishikura T, Takahashi M, Todo S,
Yamamoto N, Kimura H, Kato S, Ishika C, Ikawa S,
Ikawa Y and Nakagawara A. (1999). Oncogene, 18, 3761 ±
3765.
Tanaka H, Arakawa H, Yamaguchi T, Shiraishi K, Matsui
K, Takei Y and Nakamura Y. (2000). Nature, 404, 42 ± 49.
Thanos C and Bowie J. (1999). Protein Sci., 8, 1708 ± 1710.
Thompson JD, Higgins DG and Gibson TJ. (1994). Nucleic
Acids Res., 22, 4673 ± 4680.
757
Oncogene
Homologues for human p53/p73 in a clam cancer model
ML Kelley et al
758
Oncogene
Trink B, Okami K, Wu L, Sriuranpong V, Jen J and
Sidransky D. (1998). Nature Med., 4, 747 ± 748.
Van Beneden RJ, Rhodes LD and Gardner GR. (1998). Mar.
Environ. Res., 46, 209 ± 213.
Venot C, Maratrat M, Dureuil C, Conseiller E, Bracco L and
Debussche L. (1998). EMBO J., 17, 4668 ± 4679.
Walker XW and Levine AJ. (1996). PNAS, 93, 15335 ±
15340.
Yang A, Kaghad M, Wang Y, Gillett E, Fleming MD,
Dotsch V, Andrews NC, Caput D and McKeon F. (1998).
Mol. Cell, 2, 305 ± 316.
Zeng X, Chen L, Jost CA, Maya R, Keller D, Wang X,
Kaelin WG, Oren M, Chen H and Lu H. (1999). Mol. Cell.
Biol., 19, 3257 ± 3266.