Introduction

p73, the recently discovered p53 family member, is a nuclear protein that shares a remarkable homology, both at sequence and protein levels, with p53.1 In common with p53, the p73 protein shows three key functional domains: (a) the N-terminal transactivation domain, which shares 29% homology with the N-terminal part of p53; (b) the sequence-specific DNA-binding domain, which shares 63% homology with the corresponding p53 domain and (c) the tetramerization domain, which shares 42% homology with the oligomerization domain of p53.2 The three-dimensional structure of the C-terminal tail of p73 has recently been solved by nuclear magnetic resonance spectroscopy.3 It consists of a five-helix bundle characterized by a marked similarity to the structure of sterile α motif (SAM). These domains are known to be protein–protein interaction modules present in several cytoplasmic signaling proteins and in transcription factors.

Unlike p53, the p73 gene encodes several polypeptides. Two p73 polypeptides were originally identified.1 The longer one, named p73α, comprises 636 amino acids. The shorter one, named p73β, derives from an alternative splicing of exon 13. Additional p73 isoforms that arise from diverse alternative splicings at the C-terminus have recently been identified4, 5, 6, 7, 8 (Figure 1).

Figure 1
figure 1

Schematic representation of the p73 gene isoforms α–ɛ and ΔN-p73. Genomic organization of the p73 gene: TAD, transactivation domain; DBD, DNA-binding domain; OD, oligomerization domain; SAM, sterile α motif like. TAp73 C-terminal p73 protein isoforms α, β, γ, δ, ɛ are generated by alternative splicing of exons 10–14. The ΔN-p73 forms lacking the TAD is translated from an alternative exon 3′ located in intron 3. The location of primers used for amplification are indicated: ↱ ↰ primers for α–ɛ isoforms; ↱ ↰ primers for ΔN isoform.

Full size image

Amino-terminally truncated isoforms (ΔN-p73) that lack the transactivation domain and exert dominant-negative function towards p53, p73 and p63 activity have been described9, 10, 11, 12, 13 (Figure 1). These latter isoforms take origin from a cryptic promoter located in the third intron of the p73 gene9 (Figure 1).

Ectopic expression of p73 isoforms in both p53 +/+ and p53−/− recapitulates the well-characterized p53 antitumoral effects, such as growth arrest, apoptosis and differentiation.14, 15, 16 These effects mainly occur through the activation of common and distinct target genes compared to those recruited by wild-type p53.17 Unlike p53, which represents the most frequently mutated gene in human cancers, p73 is rarely mutated.1, 18, 19, 20, 21 Despite its localization in a genomic region frequently altered in neuroblastoma and other cancers, there is still scarce evidence supporting a role of p73 in the pathogenesis of any specific human tumor. As to hematologic neoplasms, no mutations of this gene have been detected in a recent survey including most common myeloproliferative and lymphoproliferative malignancies.21

p73-deficient mice exhibit severe defects, including hydrocephalus, hypocampal dysgenesis, chronic infections and inflammation and abnormalities in the pheromone sensory pathway.9 It has also been reported that p73 mRNA is upregulated during differentiation of muscle, neuronal and hematopoietic cells. Ectopic expression of p73α promotes neuronal and hematopoietic differentiation.16, 22, 23

Here, we have investigated the expression pattern of TAp73α, its spliced isoforms and ΔN-p73 in diagnostic samples derived from patients with acute myelogeneous leukemia (AML), representative of all major morphologic and genetic subsets. We detected p73 expression in all AML types but significantly different ΔN-p73 expression patterns in APL as opposed to other AMLs.

Materials and methods

Patient samples and RNA preparation

Leukemia samples were obtained from peripheral blood or bone marrow specimens collected at diagnosis from 71 AML patients. All patients were diagnosed and treated at the Department of Human Biotechnology and Hematology of the University ‘La Sapienza’ of Rome. Informed consent was obtained from the patients or their parents. The series was representative of the main morphologic subtypes according to the FAB classification system24 and included the following forms: M0 (five samples), M1 (five samples), M2 (five samples), M3 (41 samples), M4 (five samples), M5 (five samples) and M6 (five samples). With concern to genetic characterization, all M3 cases were featured by the presence of the t(15;17) translocation and/or the PML/RARα fusion. Karyotype was available for 28 of 30 non-M3 AML cases. These were subclassified according to the MRC criteria25 as follows: good risk seven cases, intermediate risk 16 cases and poor risk five cases.

A leukemic infiltration >80% was present in all selected samples. Following centrifugation on a Ficoll–Hypaque density gradient, the mononuclear cell fraction was isolated and washed twice in sterile phosphate-buffered saline. Total RNA was extracted from mononuclear cells by the guanidium-isothiocyanate/phenol–chloroform method according to Chomczynsky and Sacchi.26 The quality of RNA was assessed in all cases by agarose gel visualization and amplification of an internal control gene (see below).

All cases were characterized at the genetic level for the presence of major fusion proteins including PML/RARα, AML1-ETO, CBFβ-MYH11 and MLL alterations using standardized methods reported elsewhere.27

Purification and differentiation of hematopoietic progenitor cells (HPC)

HPCs were purified from peripheral blood of healthy donors after informed consent according to the method reported elsewhere.28 Purified HPCs were induced into specific granulopoietic differentiation with interleukin-3 (1 U/ml), granulocyte/monocytes CSF (0.1 ng/ml) and saturating amounts of G-CSF (500 U/ml). Cells were studied at day 14 when they have reached terminal maturation and are fully differentiated.

Normal monocytes were isolated from peripheral blood mononuclear cells by plastic adherence and were used only when >95% cells stained positive for CD14. CD34+ human hemopoietic progenitors cells have been purified from normal cord blood using magnetic beads coated with anti-CD34 antibodies (CD34 Multisort-kit, Miltenyi Biotech GmbH, Bergisch Gladbach, Germany), according to the procedure reported by the manufacturer. At the end of the procedure, purified cells were 97±2% CD34+.

Analysis of the C-terminal p73 isoforms α−ɛ and ΔN-p73 isoform

For the screening of the C-terminal isoforms α–ɛ and for ΔN-p73 isoform, we analyzed total RNA in all cases. RNA (1 μg) was reverse transcribed using random hexamer primers in 20 μl of reaction buffer using MMLV reverse transcriptase and recombinant RNAsin. cDNA (10 μl) were amplified in a total volume of 50 μl of the reaction mixture containing 0.8 mM of each dNTP, 1 × PCR buffer, 1 U of Taq-Gold DNA polymerase (manufactured by Roche) and 10 pmol of each primer. Preheating of the mixture at 94°C for 5 min was followed by 35 cycles of 30 s at 95°C, 2 min at 55°C, 2 min at 72°C. A final extension of 5 min was carried out at 72°C on a Gene Amp PCR system 9700 (Perkin-Elmer, Emeryville, CA, USA). With the aim of simultaneously analyzing the different C-terminal p73 isoforms α−ɛ the following primers, spanning exons 8–14, were used: forward primer, 5′-GACCGAAAAGCTGATGAGGA-3′; and backward primer, 5′-CAGATGGTCATGCGGTACTG-3′.

The following plasmids (pcDNA-HA-α, β, γ, δ, ɛ) were used as templates to detect the length of PCR-amplified fragments corresponding to α–ɛ isoforms. The specificity of the amplified PCR products was confirmed by gel purification and direct sequencing.

Expression of the ΔN-p73 isoform was investigated by nested reverse transcription (RT)-PCR. All cases except two in which no more RNA was available were analyzed. For the first PCR round, 5 μl of cDNA obtained as reported above were amplified in a 50 μl of the reaction mixture containing 0.8 mM of each dNTP, 1 × PCR buffer, 1 U of Taq-Gold DNA polymerase (manufactured by Roche) and 10 pmol of each primer (forward primer, exon 3′: 5′-AAGCGAAAATGCCAACAAAC-3′; backward primer, exon 4: 5′-GGTCCATGGTGCTGCTCAGC-3′). Exon 3′ (Figure 1) is absolutely specific to the ΔN-p73 transcript. Preheating of the mixture at 94°C for 7 min was followed by 30 cycles of 30 s at 95°C, 2 min at 56°C, 2 min at 72°C. A final extension of 5 min was carried out at 72°C on a Gene Amp PCR system 9700. The next PCR was carried out amplifying 10 μl of the first PCR product using the same conditions detailed above, except for the annealing temperature (55°C) and the primers (forward primer: ACTAGCTGCGGAGCCTCTC; backward primer: TGCTCAGCAGATTGAACTGG). The same conditions were used to analyze the isoforms α–ɛ and ΔN-p73 in normal human leukocytes, granulocytes, monocytes, CD34+ progenitors and spleen using a CLONTECH cDNA (CLONTECH Laboratories Inc., Palo Alto, CA 94303-4230, USA). The porphobilinogen deaminase (PBGD) gene served as an internal control for the RT and was amplified from the same cDNA using the following oligonucleotides as primers: forward primer, 5′-CTGGTAACGGCAATGCGGCT-3′; and backward primer, 5′-GCAGATGGCTCCGATGGTGA-3′. PCR products were electrophoresed and visualized on 2% agarose gels stained with ethidium bromide. A schematic representation of the p73 isoforms with the location of the above-described primers used for amplification is shown in Figure 1

ΔN-p73 protein analysis

Whole-cell protein extracts were lysed in lysis buffer (50 mM Tris (pH 6.8), 7% glycerol, 2% SDS, 10 mM DTT, 1mM phenylmethylsulfonyl fluoride and protease inhibitors mixture). The extracts were sonicated for 10 s and centrifuged at 14 000 r.p.m. for 10 min to remove the debris. Protein concentrations were determined by a colorimetric analysis assay (Bio-Rad, Milan, Italy). Total cell lysates (200 μg/lane) were size fractionated on 10% SDS-polyacrylamide gels and blotted into nitrocellulose filters (Bio-Rad). The membranes were blocked in 5% dry milk/TBS-T for 1 h, incubated over night at 4°C with a 1: 1000 dilution of a mixture of anti-p73 monoclonal antibody (Ab-4, NeoMarkers) and visualized using the enhanced chemiluminescence system (Amersham Pharmacia Biotech, Milan, Italy).

Statistical analysis

The analysis was performed using the SAS software. The association between ΔN-p73 expression and other prognostic factors and the differences in the distributions of variable groups of patients were assessed by Kruskal−Wallis, χ2 or the Fisher's exact test.

The probability of disease-free survival was calculated using the Kaplan−Meier method and the prognostic value of potential factors was analyzed using the log-rank test with stratification for risk group.

All analyses were two tailed and were considered statistically significant when P<0.05.

Results and discussion

As shown in Figure 1, our RT-PCR strategy allowed to analyze simultaneously the RNA expression pattern of TA-p73α and related isoforms such as p73β, p73γ, p73δ and p73ɛ in leukemia samples. In agreement with previously reported findings,4, 7, 29, 30 we detected the presence of p73 mRNA transcripts in AML (Figure 2a and b) as compared to the absence in normal human leukocytes, granulocytes, monocytes, CD34+ progenitors and spleen (Figure 2c). This expression was mainly detected for the shorter isoforms and independently from the FAB subtype. No apparent relationship was found between the various short p73 isoforms and karyotypic groups in AML.

Figure 2
figure 2

Expression pattern of p73 isoforms α–ɛ and ΔN-p73 in patient samples representative of different acute leukemia FAB subtypes (M0–M6). (a) Lanes 1–20 show the results obtained in patient samples representative of M0 (lanes 1–5), M1 (lanes 6–10), M2 (lanes 11–15), M3 (lanes 16–20), respectively. (b) Lanes 21–35 show the results obtained in patient samples representative of M4 (lanes 21–25), M5 (lanes 26–30), M6 (lanes 31–35), respectively. (c) Normal human monocytes, granulocytes, spleen, leukocytes and CD34+ selected progenitors. For control size of the indicated p73 isoforms, PCR of corresponding plasmid for each isoform α–β–γ–δ–ɛ were carried out as described in Material and methods. RT-PCR of PBGD gene served as an internal control for the reverse transcription.

Full size image

It has originally been reported that short p73 isoforms are less efficient than p73α in transactivating gene target promoters and promoting growth suppression and apoptosis.5, 6, 31, 32, 33 However, the molecular mechanisms underlying the reduced transcriptional activity of p73γ, p73δ and p73ɛ are still under investigation. A recent report has shown that the potent transcriptional coactivator Yes-associated protein (YAP) can physically associate with p73α and β but not with p73γ and δ, suggesting that the lack of the recruitment of specific cofactors accounts for the impaired transcriptional activity of alternative spliced isoforms of p73.34, 35 Thus, reduced p73 tumor suppressor activity due to the selective presence of alternative spliced p73 isoforms may potentially contribute to both the transformed phenotype and chemoresistance in leukemic cells.

As to ΔN-p73, its expression pattern was rather heterogeneous in distinct AMLs (Figure 2a and b). In fact, 96.4% of the samples (27 out of 28) representative of different AML FAB subtypes (M0, M1, M2, M4, M5, M6) expressed detectable levels of ΔN-p73 mRNA (Table 1 and Figure 2a and b). After a preliminary analysis on five cases that suggested a distinct expression pattern of ΔN-p73 in M3 subset as compared to other AMLs, the series of M3 samples was expanded to 41 cases. The analysis of these 41 APL (M3) samples showed that only 13 (31.7%) of them expressed ΔN-p73 m-RNA (Figure 3). The APL samples were also analyzed for the expression of the short isoforms at C-terminus. No differences in the expression of p73 short isoforms were detected between ΔN-p73-positive and ΔN-p73-negative APL cases, as well as in the non-APL leukemias (data not shown).

Table 1 ΔN-p73 expression in patient samples according to FAB classification
Full size table
Figure 3
figure 3

Expression pattern of ΔN-p73 isoform in 41 samples of APL patients. PBGD as internal control.

Full size image

To demonstrate that the ΔN-p73 protein was present in samples from AMLs other than M3, we carried out a Western blot analysis of representative cases. As shown in Figure 4 the ΔN-p73 protein was detectable by a p73 antibody in all the analyzed samples. These findings correlate well with the results of ΔN-p73 mRNA analysis.

Figure 4
figure 4

Analysis of p73 protein in representative samples of AML FAB subtypes as well as in normal granulocytes and monocytes. Patient samples representative of M0, M1, M2, M4, M5, M6. Normal monocytes, normal granulocytes. Jurkat cells and ΔN-p73 in vitro translated served as negative and positive controls, respectively.

Full size image

ΔN-p73 is a truncated protein that takes origin from a cryptic promoter located in the third intron of the p73 gene and lacks the transcriptional activation domain. Recent studies have clearly shown that ΔN-p73 protein impairs both p53 and p73 transcriptional activity and p53/p73-mediated apoptosis in response to agents inducing DNA damage.10, 11, 12, 13, 36, 37 A rather speculative hypothesis might suggest that lack of the expression of dominant-negative ΔN-p73 protein contributes to the well-known responsiveness of APL to currently adopted treatments, which include anthracycline-based chemotherapy in addition to retinoic acid.38 In line with this hypothesis, it is remarkable that missense mutations of the p53 gene, the most frequent genetic alteration of human cancer that causes the loss of oncosuppressor activities of wild-type p53, have been shown to occur very rarely in APL.39, 40

Recently, it has been shown that the PML gene contains p53-binding sites, which confer responsiveness to p53. Therefore, PML has been proposed as a direct target modulating p53 biological activity.41, 42, 43 The p53/PML crosstalk is likely to be impaired in APL carrying the PML/RARα fusion. It is conceivable to hypothesize that such impairment is balanced by the lack of ΔN-p73 inhibitory effects on p53 activity. p73 upregulation has been observed during myeloid differentiation of the promyelocytic NB4 and the AML-M2 HL-60 cell lines.22, 23, 29 The transcriptional regulation of p73 transcript at the crossroad between proliferation and differentiation has been shown to be regulated by factors that are distinct from those controlling p53 gene expression.17 It has been reported that human tumor-derived p53 mutants can physically associate with and strongly impair the antitumoral effects of the diverse p73 isoforms.31, 44, 45, 46 This interaction involves the specific DNA-binding domain of p73 and the core domain of mutant p53.31, 46 A recent report has shown that in cells carrying mutant p53, the in vivo recruitment of p63 and p73 to the regulatory regions of specific target genes is largely impaired.47

As to the analysis of presenting features and disease outcome according to ΔN-p73 expression, 41 APL patients treated uniformly with retinoic acid and anthracycline chemotherapy were evaluated. As shown in Table 2, no significant differences were detected concerning median age, sex, FAB, median WBC, platelet count and DFS between the two groups of patients positive or negative for ΔN-p73 expression, thus suggesting that ΔN-p73 expression cannot be considered as a prognostic marker.

Table 2 Clinical characteristics of APL patients according to ΔN-p73 expression pattern
Full size table

In conclusion, our study identifies a novel biological feature of APL that further distinguishes this peculiar leukemia subset from other AMLs and may contribute to explain its favorable response to treatment. Additional laboratory studies are needed to elucidate the genetic mechanism underlying the infrequent ΔN-p73 expression in APL and its biologic consequences on cell survival.