healthmedicinet

In order to find novel regulators of p53 and Mdm2, we screened a cDNA library from
the teleost Medaka. The library was prepared from stage eighteen (neurula), stage
twenty-four (beginning of neurogenesis) and stage thirty-two (completion of organogenesis)
Medaka embryos 8]. It contains almost eighteen thousand full lengths genes of which almost fourteen
thousand are annotated. In addition, about three thousand five hundred full length
cDNA clones with only partial sequence information are contained in the library. During
library preparation, only the most complete representative of each gene was selected,
resulting in a unigene cDNA library. In order to simplify the screening procedure,
which involved labor-intensive SDS-PAGE and Western Blot analysis of cellular lysates,
pools of twenty-four clones were prepared as transfection ready plasmid DNA samples.
Over seven hundred such cDNA pools were arrayed in 96-well plates for the primary
transfection-based screening experiments.

Each pool of cDNAs from the library was co-transfected into p53-negative H1299 cells
together with p53 and mdm2. The absolute levels and ratios of overexpressed p53 and Mdm2 were optimized to allow
both increases and decreases in their abundance to be detected. To monitor Mdm2 activity
as well as Mdm2 and p53 abundance in the absence of co-transfected cDNAs from the
library, control transfections using only p53 or p53 and mdm2 were performed. As an additional control to determine whether cDNA library pools
have a general impact on protein turnover we also transfected the Wnt-receptor protein
ror2 (Fig. 1). Twenty-four hours post-transfection total cell lysates were prepared and the amount
of p53, Mdm2 and ROR2 proteins was determined by SDS-PAGE and Western Blotting analysis
(Fig. 1). Each library pool was analyzed at least twice and only pools that specifically
altered the abundance of co-expressed p53 and/or Mdm2 in at least two experiments
were considered as candidates. Using this primary screening approach, we identified
one hundred and six pools that either increased or decreased abundance of p53 or Mdm2
or both (Table 1). The majority of the hits increased the abundance of either p53 alone or of p53
and Mdm2, while a lower number of library pools resulted in decreased p53 and/or Mdm2
abundance.

Fig. 1. Screening of a Medaka cDNA library. a Schematic drawing of the screening process. 24 clones of the cDNA library were pooled
and all pools were arrayed in 96-well plates. P53-negative H1299 cells were transfected
in 96 well-plates with p53 (5 ng), mdm2 (45 ng), one of the cDNA library pools (150 ng) and myc–ror2 (5 ng) as control. To monitor Mdm2 activity, and Mdm2 and p53 abundance in the absence
of co-transfected pools of the cDNA library, 2 out of 14 wells were transfected with
p53 and myc-ror2 (control 2) or with p53 and mdm2 and myc-ror2 (control 1) without library pools. For transfection, total amounts of plasmid DNA
in all samples were adjusted to 205 ng using vector DNA. 24 h after transfection,
cells were lysed and abundance of p53, Mdm2 and Myc-ROR2 were determined by Western
Blotting. Abundance of PCNA was monitored for loading control. b H1299 cells were transfected with plasmids encoding p53, Mdm2 and Myc-ROR2 together
with the indicated pools of the cDNA library, without the library (ctrl 1) or without
the library and without the plasmid encoding Mdm2 (ctrl 2), for control. 24 h after
transfection, cells were lysed and abundance of p53, Mdm2, Myc-ROR2 and PCNA were
determined by Western Blotting

Table 1. Summary of the screening of a Medaka cDNA library for evolutionarily conserved regulators
of p53 and Mdm2. The table shows the number of hits and their activity on the abundance
of p53and/or Mdm2

In order to confirm that the complexity of the primary screening conditions would
not prevent identification of p53 modulators, we spiked a non-modifying (control)
pool from the library with plasmid DNA encoding the known p53 modifiers p14
^ARF
and USP7 (see Additional file 1). P14
^ARF
stabilizes p53 by preventing the Ubiquitin-mediated degradation of p53 by Mdm2 9] and the ubiquitin-specific protease USP7 deubiquitinates both Mdm2 and p53, resulting
in their stabilization 10]. Co-expression of p14 ^arf
and usp7 in amounts similar to those of the individual clones present in the screened pools
(about 6 ng) resulted in the expected stabilization of p53 and Mdm2, confirming the
suitability of the primary screening conditions (see Additional file 1). Indeed, one of the candidate pools identified in the primary screening contains
the Medaka usp7 gene homologue. The Medaka p14 ^arf
homologue was however not present in the library.

In order to identify which genes in the candidate pools were responsible for regulating
p53 and/or Mdm2 abundance, we performed a secondary screen using the twenty-four single
cDNA clones contained within each candidate pool. The individual cDNA clones were
amplified from the arrayed master library and, like the primary screen, were transfected
together with p53, mdm2 and ror2 (Fig. 2a). Each cDNA was analyzed at least twice and only those clones that altered p53 and/or
Mdm2 abundance in at least two experiments were considered as candidates. From the
one hundred and six pool hits that were identified in the first round of screening,
the secondary screening procedure was performed on twenty. The majority (eleven) of
these twenty pools harbored a single regulator of p53 or p53 and Mdm2, two pools harbored
two regulators and, strikingly, one pool harbored three apparent regulators of p53
and Mdm2 (Fig. 2, Table 2). Thus, overall, 70 % of candidate pools identified and selected for secondary screening
were found to contain regulators of p53 and/or Mdm2. Six of the identified regulators
affected only p53 and two of the cDNAs affected only Mdm2 abundance. Four of the identified
novel regulators increased Mdm2 abundance and decreased p53 abundance and six increased
both p53 and Mdm2 abundance (Table 2). Among the seventeen cDNAs that we found to regulate p53 or p53 and Mdm2, twelve
were annotated and five had unknown sequences. The molecular function of the genes
that we found in the sub-screen (Table 2) ranged from a putative MAPK-activating protein (C1ORF144/SZRD11, 11]) to a lysosomal protein that is involved in the catabolism of heparin and keratan
sulphate (GNS; 12]).

Fig. 2. Subscreening of the Medaka cDNA library. a From the cDNA pools that were considered to contain regulators of p53 and/or Mdm2,
the individual bacteria were amplified from the master library. The plasmids were
purified and 12.5 ng of the individual cDNAs were transfected together with p53 (5 ng), mdm2 (45 ng) and myc-ror2 (5 ng) into H1299 cells in a 96-well format. 24 h after transfection, cells were
analyzed by Western Blotting. b H1299 cells were transfected with plasmids encoding p53, Mdm2 and Myc-ROR2 together
with the individual clones of the indicated pool hit or with plasmids encoding p53,
Mdm2 and Myc-ROR2 without cDNA clones, for control (ctrl). 24 h after transfection,
cells were harvested and analyzed as described in the legend to Fig. 1

Table 2. Regulators of p53 and Mdm2 identified by subscreening of the Medaka cDNA library.
The table shows the number of the clone in the library, the identification number
of the Ensembl database (Ensmble ID), the name of the gene, further details about
the protein and its activity within the p53/Mdm2 circuit

Several of the cDNAs identified in the secondary screen were again transfected into
H1299 cells to confirm that they indeed regulate the levels of p53 and/or Mdm2. All
of the identified Medaka clones regulated p53 and/or Mdm2 (Fig. 3), confirming the result from the secondary screen.

Fig. 3. cDNAs identified in a screen using a cDNA library from Medaka induced mammalian p53
and Mdm2. H1299 cells were transfected with the indicated cDNAs together with plasmids
encoding p53, Mdm2 and Myc-ROR2. For control, cells were transfected only with plasmids
encoding p53, Mdm2 and Myc-ROR2. 24 h after transfection, cells were harvested and
analyzed as described in the legend to Fig. 1

We have performed a cell culture-based cross-species expression screen where a cDNA
library from Medaka was used to screen for regulators of human p53 and mouse Mdm2
in human cells. One aim of this screen was to identify evolutionarily conserved modifiers
of the p53 signaling pathway. Considering the evolutionary distance of teleosts and
mammals we tested whether the human orthologues of the identified Medaka genes would
also be able to regulate p53 and Mdm2. As shown in Fig. 4, the human orthologues of some of the genes identified (e.g. c1orf144, trim25, fam83f and rbm15) also regulated p53 and/or Mdm2 abundance (Fig. 4). C1ORF144/SZRD1, increased p53 abundance, TRIM25 increased p53 and Mdm2 abundance,
FAM83F increased p53 abundance and decreased Mdm2 abundance and RBM15 increased Mdm2
abundance and decreased p53 abundance. We also tested the ability of C1ORF144/SZRD1
and FAM83F to regulate endogenous p53 levels in p53-positive cells and in each case
we could see the expected effect (see Additional file 2). Furthermore, we recently demonstrated that TRIM25 also regulates endogenous p53
13]. These results demonstrate that cell culture-based expression screening using a cDNA
library from Medaka can identify evolutionarily conserved regulators of p53 and Mdm2.

Fig. 4. Human homologs of the cDNAs that were identified by screening a Medaka cDNA library
control p53 and Mdm2 abundance. H1299 cells were transfected with increasing amounts
of plasmids encoding human Myc-tagged C1ORF144, human V5-tagged TRIM25, human Flag-tagged
FAM83F and human Flag-tagged RBM15 together with plasmids encoding p53 and Mdm2. For
control, cells were transfected only with plasmids encoding p53 and Mdm2. 24 h after
transfection, cells were harvested and analyzed as described in the legend to Fig. 1