Targeting the p53 Pathway in Ewing Sarcoma.

The p53 tumour suppressor plays a pivotal role in the prevention of oncogenic transformation. Cancers frequently evade the potent antitumour surveillance mechanisms of p53 through mutation of the TP53 gene, with approximately 50% of all human malignancies expressing dysfunctional, mutated p53 proteins. Interestingly, genetic lesions in the TP53 gene are only observed in 10% of Ewing Sarcomas, with the majority of these sarcomas expressing a functional wild-type p53. In addition, the p53 downstream signaling pathways and DNA-damage cell cycle checkpoints remain functionally intact in these sarcomas. This paper summarizes recent insights into the functional capabilities and regulation of p53 in Ewing Sarcoma, with a particular focus on the cross-talk between p53 and the EWS-FLI1 gene rearrangement frequently associated with this disease. The development of several activators of p53 is discussed, with recent evidence demonstrating the potential of small molecule p53 activators as a promising systemic therapeutic approach for the treatment of Ewing Sarcomas with wild-type p53.

1. Introduction

The p53 protein, known as the “guardian of the genome” [1] and voted Science magazine's “Molecule of the Year” in 1993, plays a pivotal role in the cellular defense against transformation of cells in the presence of oncogenic or genotoxic stress [2]. This is achieved through the ability of the p53 transcription factor to drive the expression of downstream target genes to evoke cellular responses such as cell cycle arrest, apoptosis, DNA damage repair, and senescence [3]. The development of a malignant neoplasm generally requires attenuation of these p53 responses, and this can occur via mutation of the p53 protein. TP53 is the most frequently altered gene in cancer, with p53 mutations observed in approximately half of all tumours [4]. In contrast, TP53 mutations are infrequent in the Ewing Sarcoma Family of Tumours (ESFTs) with the majority of these sarcomas expressing a functional wild-type p53 [5-14]. Such features are rarely seen in cancers and are suggestive that ESFTs will be sensitive to p53-based targeted therapeutic strategies.

2. Genomic Integrity of TP53 Is Preserved in Ewing Sarcomas

Ewing Sarcomas arise in the bones of children and young adolescents and are the most lethal of all bone tumours [15, 16]. These sarcomas are infrequent neoplasms with international incidence rates in the pediatric population averaging less than two cases per million children. Such low incidence rates have limited the detection of TP53 mutations to relatively small cohorts from independent clinical centres, with these studies reporting TP53 mutation frequencies ranging from 4% to 20%. A meta-analysis of all primary or metastatic ESFTs involving TP53 point mutations confirmed through direct sequencing reveals that TP53 mutations are observed in approximately 10% of cases (Table 1).

Table 1

TP53 Mutations in Ewings Sarcoma confirmed by DNA sequencing.

Sarcoma type	Study	Method	Exons sequenced	Mutation frequency	TP53 mutation
Ewing Sarcoma	Kovar et al., 1993 [5]	PCR-SSCP Sequencing	4–8	2/37	C277Y, R273C
	Komuro et al., 1993 [6]	PCR-SSCPSequencing	5–9	2/14	152(FS), G154V
	Hamelin et al., 1994 [7]	PCR-DGGE Sequencing	5–8	2/12	R175H, R248W
	Patiño-García and Sierrasesúmaga, 1997 [8]	PCR-DDGE Sequencing	5–8	1/5	R273H
	Radig et al., 1998 [9]	IHC PCR-SSCP Sequencing	4–8	1/24	Not Specified
	Tsuchiya et al., 2000 [10]	PCR-SCCP Sequencing	5–9	1/24	G154V
	López-Guerrero et al., 2001 [11]	Sequencing	5–8	3/19	C135F, A138D, P151R
	Park et al., 2001 [13]	PCR-SSCP Sequencing	4–9	3/35	K132M, C135S, Q287V
	Huang et al., 2005 [12]	IHC p53 GeneChip Sequencing	—	8/60	W146(STOP), M160L N239A, G244I, R248Q R273F, A276S, R342P
	Schaefer et al., 2008 [14]	Sequencing	5–8	2/17	C141Y, R248W
Total TP53 mutations in Ewing Sarcoma					25/247 (10.1%)

DDGE: Denaturing Gradient Gel Electrophoresis; IHC: Immunohistochemistry; PCR-SSCP: Polymerase Chain Reaction Single-Strand Conformational Polymorphism; FS: Frameshift; Mutations in bold indicate p53 “hotspot mutations.”

A similar analysis of the available literature suggests that the majority of other sarcoma types are also associated with low frequencies of TP53 mutations, ranging from approximately 6% in well-differentiated/dedifferentiated liposarcomas to 23% in osteosarcomas (Table 2). Although malignant fibrous histiocytomas are listed, they are a discredited entity and will soon be removed from the World Health Organisation (WHO) sarcoma atlas [17]. Collectively, Ewing Sarcoma was associated with the lowest frequencies of TP53 mutation across all sarcoma types.

Table 2

TP53 Mutations in sarcomas other than ESFTs confirmed by DNA sequencing.

Sarcoma type	Study	Method	Exons sequenced	Mutation frequency	TP53 mutation
Liposarcoma
Well-differentiated/de-differentiated liposarcoma (WD/DDLPS)	Pilotti et al., 1997 [18]	IHCPCR-SSCPSequencing	5–9	4/13	H179Y, R213(STOP) R282W, Gg>Gc (SS)
	Dei Tos et al., 1997 [19]	IHCPCR-SSCPSequencing	5–8	1/14	S127F
	Schneider-Stock et al., 1998 [20]	PCR-SSCPSequencing	4–8	0/8	—
	Schneider-Stock et al., 1999 [21]	IHCPCR-SSCPSequencing	5–8	0/13	—
	Barretina et al., 2010 [22]	Sequencing, mass spectrometry-based genotyping	—	0/50	—
				Total TP53 mutations	5/98 (5.1%)
Myxoid/Round cell liposarcoma	Pilotti et al., 1997 [18]	IHCPCR-SSCPSequencing	5–9	1/6	Del nts 1506-1507 (STOP)
	Schneider-Stock et al., 1998 [20]	PCR-SSCPSequencing	4–8	1/12	P128S
	Schneider-Stock et al., 1999 [21]	IHCPCR-SSCPSequencing	5–8	3/19	H214L, P250T, G245S
	Oda et al., 2005 [23]	IHCPCR-SSCPSequencing	5–9	5/77	Q167(STOP), H214Y V225A, C238Y, C242Y
	Barretina et al., 2010 [22]	Sequencing, mass spectrometry-based genotyping	—	0/21	—
				Total TP53 Mutations	10/135 (7.4%)
Pleomorphic liposarcoma	Schneider-Stock et al., 1998 [20]	PCR-SSCPSequencing	4–8	3/6	R248Q, E271(STOP) R273C
	Schneider-Stock et al., 1999 [21]	IHCPCR-SSCPSequencing	5–8	2/9	R248Q, R273C
	Barretina et al., 2010 [22]	Mass spectrometry-based genotyping	—	4/24	C135F, T155IC>TT (SS), C>CT (SS)
				Total TP53 mutations	9/39 (23.1%)
Undefined liposarcomas	Toguchida et al., 1992 [24]	PCR-SSCPSequencing	2–11	1/4	AGgt>AGtt (SS)
	Leach et al., 1993 [38]	IHCSequencing	5–8	3/13	Q144(STOP), N239SGGT>GAT (SS)
	Latres et al., 1994 [25]	PCR-SSCPSequencing	2–9	5/25	H168R, H193Y, M246V R248W, 344(STOP)
	Castresana et al., 1995 [26]	PCR-SSCPSequencing	5–8	1/4	V216A
	Nawa et al., 1999 [27]	PCR-SSCPSequencing	5–8	1/9	T253A
	Das et al., 2007 [28]	IHC Sequencing	2–11	1/3	377(FS)
				Total TP53 mutations	12/58 (20.7%)
Total TP53 mutations in liposarcomas					36/330 (10.9%)
Rhabdomyosarcoma	Felix et al., 1992 [29]	PCR-SSCPSequencing	4–8	1/6	R213P
	Toguchida et al., 1992 [24]	PCR-SSCPSequencing	2–11	0/4	—
	Latres et al., 1994 [25]	PCR-SSCPSequencing	2–9	0/2	—
	Castresana et al., 1995 [26]	PCR-SSCPSequencing	5–8	1/1	V218L
	Kusafuka et al., 1997 [30]	PCR-SSCPSequencing	5–8	1/10	R273H
	Nawa et al., 1999 [27]	PCR-SSCPSequencing	5–8	0/2	—
	Taylor et al., 2000 [31]	PCR-SSCPSequencing	5–9	1/20	Del nt 1004-1017
	Takahashi et al., 2004 [32]	PCR-SSCPSequencing	5–9	9/45	E204G, R209T, P223RM243T, G245C, N247D, R249G, C291Q, P295H
	Das et al., 2007 [28]	IHC Sequencing	2–11	1/4	D393N
Total TP53 mutations in rhabdomyosarcomas					14/94 (14.9%)
Synovial Sarcoma	Toguchida et al., 1992 [24]	PCR-SSCPSequencing	2–11	0/5	—
	Latres et al., 1994 [25]	PCR-SSCPSequencing	2–9	0/8	—
	Schneider-Stock et al., 1997 [33]	IHCPCR-SSCPSequencing	5–8	0/2	—
	Dei Tos et al., 1999 [34]	PCR-SSCPSequencing	—	4/20	Not Specified
	Nawa et al., 1999 [27]	PCR-SSCPSequencing	5–8	1/7	L194F
	Schneider-Stock et al., 1999 [35]	IHCPCR-SSCPSequencing	5–8	2/19	P128L, R248W
	Oda et al., 2000 [36]	IHCPCR-SSCPSequencing	5–9	9/49	C141Y, A159T,V173MI195F, R196Q, G199RR213(STOP), N235DC238Y
	Das et al., 2007 [28]	IHC Sequencing	2–11	5/7	9(STOP), A63P, S96C, P250T, P250T
	Barretina et al., 2010 [22]	Sequencing, mass spectrometry-based genotyping	—	0/23	—
Total TP53 mutations in Synovial Sarcomas					21/140 (15.0%)
Malignant Fibrous Histiocytoma	Toguchida et al., 1992 [24]	PCR-SSCPSequencing	2–11	2/13	R196(STOP), R273H
	Andreassen et al., 1993 [37]	CDGESequencing	5,7,8	3/12	V143M, Y163C, G244D
	Leach et al., 1993 [38]	IHCSequencing	5–8	1/11	R158H
	Latres et al., 1994[25]	PCR-SSCPSequencing	2–9	0/9	—
	Castresana et al., 1995 [26]	PCR-SSCPSequencing	5–8	3/12	Not Specified
	Schneider-Stock et al., 1997 [33]	IHCPCR-SSCPSequencing	5–8	2/15	Y220C, C277(STOP)
	Nawa et al., 1999[27]	PCR-SSCPSequencing	5–8	5/15	Y126F, R175H R213(STOP), S241T R248Q
	Das et al., 2007 [28]	IHC Sequencing	2–11	2/11	P77Q, 213(FS)
Total TP53 mutations in malignant fibrous histiocytoma					18/98 (18.4%)
Leiomyosarcoma	Andreassen et al., 1993 [37]	CDGESequencing	5,7,8	2/6	K132M, R248W
	Latres et al., 1994[25]	PCR-SSCPSequencing	2–9	5/13	Y163C, Y163C, H214R, G266E, ATgg>ATag (SS Intron 5)
	Patterson et al., 1994 [39]	PCR-SSCPSequencing	4–9	6/29	P151H, P152L, R158HV216M, C238F, V272M
	Castresana et al., 1995 [26]	PCR-SSCPSequencing	5–8	1/1	Not Specified
	Miller et al., 1996 [40]	PCR-SSCPSequencing	2–11	1/8	Q165(STOP)
	Hall et al., 1997 [41]	IHCPCR-SSCPSequencing	5–8	3/21	K163E, T211IA nt Del codon 246
	Schneider-Stock et al., 1997 [33]	IHCPCR-SSCPSequencing	5–8	0/7	—
	Nawa et al., 1999[27]	PCR-SSCPSequencing	5–8	0/3	—
	Zhai et al., 1999 [42]	IHC Sequencing	5–8	9/21	V173M, Y205C, S215R, R248Q, R249W, R273H, A276D, E285D, S303I
	Miyajima et al., 2001 [43]	IHCPCR-SSCPSequencing	5–9	8/13	A161T, D184N, T220CT220C, C238S, C238C R273H, G279V
	Das et al., 2007 [28]	IHC Sequencing	2–11	0/5	—
	Barretina et al., 2010 [22]	Sequencing, mass spectrometry-based genotyping	—	0/27	—
Total TP53 mutations in leiomyosarcomas					35/154 (22.7%)
Osteosarcoma	Toguchida et al., 1992 [24]	PCR-SSCPSequencing	2–11	14/76	46 (STOP), 112 (STOP), R175H H193Q, E221(STOP) 227(STOP), S241YG244V, P250L, D259V R273H, D281H, D281NaaTG > ggTCG (SS)
	Andreassen et al., 1993 [37]	CDGESequencing	5, 7, 8	2/11	D281E, E286K
	Castresana et al., 1995 [26]	PCR-SSCPSequencing	5–8	2/7	169(STOP), D281Y
	Miller et al., 1996 [40]	PCR-SSCPSequencing	2–11	13/42	H179Y, E224D, 239(FS), G245D, R248W, R248, 248(FS), R273H R273H, A276P R282Q, R282Q R282H
	Patiño-García and Sierrasesúmaga, 1997 [8]	PCR-DDGE Sequencing	5–8	6/37	R175H, R196(STOP) P250F, N268S, R273H R2735
	Gokgoz et al., 2001 [44]	PCR-SCCPSequencing	4–10	60/272	agTCC > aaTCC (SS) L43(STOP), L43(STOP), P47L, 73(FS), 73(FS), 83(FS), 83(FS) In-Frame Ins (GGT) Codon 107/108ACGgt/ACGtt(SS), agTAC/aaTAC (SS), V172D, R175H ATGgt/ATGat(SS), R181P, R196(STOP) V197G, Del codon 202–206, E204(STOP), Y205CR213(STOP), Y220CY220C, E221(STOP) 31bp Del (FS) Intron 6 to Exon 7229(FS), M237I, M237I, M237I, C238GC238G, C238G241(FS), C242Y, C242Y, G245S, G245S R248W, R248Q, R248Q, R248Q, R248Q, P250L, T256S15 bp In-frame Del (codon 265) R273H, R273H, R273H, R280H, D281H, D281H D281N, D281NE285K, Del codon 296–303298(FS) GAGgt/GAGct (SS) R337C, R342(STOP) E343(STOP)
	Overholtzer et al., 2003 [45]	PCR-SCCPPCR-LDR	5–8	12/32	V173G, V173M, R175H, Del codon 175, Y220C, E224D, V272M, R273H, R273C, D281H FS (Exon 6), FS (Del 17nt) Exon 5
Total TP53 mutations in osteosarcomas					109/477 (22.9%)

CDGE: Constant Denaturant Gel Electrophoresis; DDGE: Denaturing Gradient Gel Electrophoresis; IHC: Immunohistochemistry; PCR-SSCP: Polymerase Chain Reaction Single-Strand Conformational Polymorphism; PCR-LDR: Polymerase Chain Reaction Ligase-Detection Reaction; FS: Frame shift; Del: Deletion; Ins: Insertion; SS: Splice Site. Mutations in bold indicate p53 “hotspot mutations”. Recurrences with TP53 mutations have been omitted in studies that reported both the primary tumour and recurrence with the same mutation.

The International Agency for Research on Cancer (IARC) has recently released recommendations for the detection of TP53 mutations, and advise direct sequencing of exons 4 to 10 of the TP53 gene [46]. Studies listed in Tables 1 and 2 rarely fulfilled these recommendations. Furthermore, the reported frequencies of TP53 mutations may be marginally underestimated as the majority of these studies sequenced TP53 in patients where an initial screen detected overexpressed p53 protein by immunohistochemistry, rather than performing an unbiased sequencing of all cases. Although mutant p53 is typically stabilized in cancer cells, and overexpression of p53 protein is predictive of TP53 mutation [47], this indirect approach cannot detect heterozygous truncating mutations of TP53 [46]. Nevertheless, frequencies of TP53 mutation reported by the studies in Tables 1 and 2 are consistent with publicly available sequencing data from the IARC TP53 database in which TP53 point mutations were observed in 373 out of 2145 (17.4%) tumours from bone or soft tissue origins [46].

The integrity of the p53 pathway in Ewing Sarcomas is further supported by studies suggesting that gross chromosomal alterations involving the TP53 locus on chromosome 17p are relatively infrequent in ESFT samples [48]. This is in contrast to other bone malignancies, such as osteosarcomas, where chromosome alterations of the TP53 gene are frequently observed [24, 44, 49, 50].

3. Genetic Alterations in Regulators of the p53 Pathway

Typically, cancers that retain wild-type p53 have been shown to indirectly suppress the p53 regulatory and signaling pathways. One of the most common oncogenic defects observed involves amplification or overexpression of MDM2 [51]. The stability and activity of p53 are constitutively regulated by MDM2 using an auto-regulatory feedback loop which in normal cells prevents inappropriate activation of p53 [52]. MDM2 is an E3 ubiquitin ligase that antagonizes the tumour suppressor function of p53 by silencing the ability of p53 to transactivate target genes or promoting its degradation or nuclear exportation [53, 54]. However, MDM2 amplification is a rare event in Ewing Sarcomas and is only observed in approximately 2% of ESFT cases (Table 3). In contrast, virtually all well-differentiated and dedifferentiated liposarcomas contain complex marker chromosomes with multiple copies of the MDM2 locus [55].

Table 3

MDM2 amplification or CDKN2A deletion in Ewing Sarcomas.

Study	MDM2 amplification	Study	CDKN2A deletion
Kovar et al., 1993 [5]	0/17	Kovar et al., 1997 [5]	7/27#
Ladanyi et al., 1995 [58]	3/30	Wei et al., 2000 [59]	7/39#
Tsuchiya et al., 2000 [10]	0/24	López-Guerrero et al., 2001 [11]	4/19#
Park et al., 2001 [13]	0/35	Brownhill et al., 2007 [60]	6/42*
López-Guerrero et al., 2001 [11]	0/19	López-Guerrero et al., 2010 [48]	34/169*
Total MDM Amplifications	3/125 (2.4%)	Total CDKN2A Deletions	58/296 (19.6%)

#Homozygous deletion of CDKN2A

*Includes both homozygous and hemizygous deletions of CDKN2A.

Cancers can also attenuate p53 function through deletion of CDKN2A, the gene encoding p14ARF [48]. The p14ARF tumour suppressor is a positive regulator of p53 in response to specific stimuli such as oncogenic stress. The stability of p53 is enhanced by p14ARF through its ability to sequester MDM2, thus releasing p53 to activate downstream pathways of growth suppression or apoptosis [56, 57]. A summary of the available literature related to CDKN2A chromosomal alterations in ESFT cases revealed that either homozygous or hemizygous deletion of the CDKN2A locus are also relatively infrequent events, occurring in less than 20% of all cases (Table 3). CDKN2A alterations and TP53 mutations are mutually exclusive events in the majority of ESFT cases, suggesting that either genetic insult is sufficient to inactivate the p53 pathway in these cancers [11, 48].

MDM4 is another key negative regulator of the p53 pathway [61-63]. This oncoprotein is closely related to MDM2 with significant homology between their DNA binding domains; however, MDM4 has a more specific role in the negative regulation of p53 transcriptional activity. Amplification of the MDM4 gene has also been reported in several tumour types, with MDM4 amplification observed in 65% of retinoblastomas [64]. MDM2 and MDM4 amplification are rarely observed within the same tumour, suggesting that either event is sufficient to inactivate the p53 pathway. MDM4 amplification is a possible mechanism for functional inactivation of the p53 pathway in Ewing Sarcoma given the infrequent occurrences of MDM2 amplification, CDKN2A deletion, or TP53 mutation in these cancers. Unfortunately, previous cytogenetic studies are restricted to investigating allelic imbalance of the TP53, MDM2, and CDKN2A loci in ESFT patient material, hence the frequency of MDM4 amplifications in Ewing Sarcoma is currently unknown.

The small proportion of ESFT cases with either MDM2 amplification or p14ARF deletion cannot collectively account for the ability of Ewing Sarcomas to develop in a cellular context with wild-type p53. Further investigations of Ewing Sarcoma are warranted to conclusively determine if the mechanisms that attenuate the p53 response during sarcomagenesis occur at the genetic or posttranslational level. The presence of an EWS-ETS translocation event is a universal feature of Ewing Sarcoma and represents another possible genetic alteration responsible for the regulation of p53 in ESFT. Recent insights into the functional characterization of the resulting oncogenic gene product suggest a potential role of this ubiquitous translocation event in silencing p53 activity in ESFTs (see later discussion).

4. The p53 Signaling Pathways Are Functionally Intact in Ewing Sarcoma

Abrogation of the p53 pathway through TP53 mutation is typically associated with enhanced tumour invasive and metastatic capabilities, and poorer patient survival rates [46, 65]. Ewing Sarcoma is an aggressive malignancy with the lowest patient survival rates of all primary musculoskeletal tumours, traits rarely possessed by cancers that retain wild-type p53. Despite observations that TP53 alterations and MDM2 amplifications are infrequent events in primary Ewing Sarcomas, it has been speculated that the downstream signaling pathways of p53 may be inactive in these sarcomas. In order to test the functional intactness of these p53 signaling pathways in Ewing Sarcoma, Heinrich Kovar and colleagues investigated the response of several ESFT cell lines with varying p53 status to ectopic p53 expression [66]. A prolonged apoptotic or growth arrest phenotype was observed upon ectopic expression of wild-type p53 in the cell lines. The sensitivity of Ewing Sarcoma cell lines to X-irradiation was also dependent on the expression of an endogenous wild-type p53. These findings confirm the intactness of the p53 signaling pathways in Ewing Sarcoma.

The frequent normal functioning of the p53 signaling pathway in ESFTs is also demonstrated by the observation that almost all Ewing Sarcoma cell lines have acquired either TP53 mutations or CDKN2A deletions, suggesting selective pressure for these genetic alterations to permit in vitro growth [67]. Similar observations have been made in the clinic, with p53 mutation and CDKN2A deletion defining a lethal subset of ESFTs associated with poor response to chemotherapy [12]. In conclusion, the in vitro and in vivo evidence suggests the p53 signaling pathways are intact in a significant proportion of Ewing Sarcomas.

5. The Prognostic Significance of TP53 Alterations

ESFT patients with point mutation of TP53 are associated with a poor prognosis [12, 68, 69]. Logically, one would expect a genetic event that confers a growth advantage to be highly represented across a tumour type through selective pressure. This is not the case in Ewing Sarcoma, as TP53 mutations rarely occur, yet define a high-risk population of patients.

Huang and colleagues have provided the most compelling evidence thus far for TP53 mutation as an independent prognostic marker using a combined immunohistochemistry, Genechip and sequencing approach to detect TP53 mutations in 60 ESFTs [12]. TP53 mutations were identified in 8 of these 60 cases (13.3%), and all eight patients expressing mutant p53 ESFTs died within 21 months of diagnosis with a mean survival of 11 months, as compared to a mean survival of 99 months for patients with wild-type p53 ESFTs. Multivariate analysis identified TP53 mutation as the strongest independent prognostic factor [12]. This is the largest prognostic study to date that involves DNA sequencing of TP53 in ESFT patient material.

The prognostic involvement of p53 mutation in Ewing Sarcoma was recently challenged by findings from a retrospective study involving 308 ESFT cases collected from 1971 to 2007 [48]. Although overexpression of p53 protein was detected in 25% of these cases, this study restricted the classification of “p53 mutation” to these cases which did not express p21Waf1/Cip1, the strongest canonical p53 target. Subsequently, 15% of ESFT cases where deemed to express mutant p53 upon application of these criteria. This study showed that “mutant p53” expression was more frequent in disseminated disease than in primary localized tumors, indicating a role in the progression and metastasis of Ewing Sarcoma. However, there was no association between “mutant p53” expression and patient survival. The conclusions from this study are limited by the absence of actual TP53 sequencing of ESFT samples to confirm the presence of TP53 mutations. These observations need to be attested by further investigation into the prognostic value of TP53 mutations using unbiased direct sequencing approaches.

Gross chromosome rearrangements involving TP53 have been recently reported to influence the prognosis of Ewing Sarcoma [48, 70]. Alteration of 17p (the chromosomal arm containing the TP53 locus) was observed in 16.7% of ESFT samples and was associated with significantly poorer survival rates [70]. Lopez-Guerrero and colleagues recently showed that alteration the TP53 locus alone was a prognostic marker for poor patient outcome [48]. Gross alteration of TP53 gene was detected in 32 of 191 (17%) ESFT cases. Interestingly, the strongest prognostic information from these studies was observed upon loss of heterozygosity (LOH) of 16q, which occurred in 20.8% of cases and was the most significant indicator of poor outcome [70]. A recent study that used a combined comparative genomic hybridization (CGH) and expression microarray analysis identified the ANKRD11 locus at 16q24.3 as one of the most frequently deleted and down-regulated genes in Ewing Sarcoma [71]. It is noteworthy that ANKRD11 was recently reported as a p53 coactivator [72], suggesting that the loss of p53 activators may contribute towards the ability of Ewing Sarcomas to develop and progress in the presence of a wild-type p53.

Studies to date are consistent with the presence of TP53 point mutations in defining a high-risk population of ESFT patients. However, more detailed studies are warranted to conclusively evaluate the prognostic potential of TP53 point mutation. Collectively, these findings indicate that the prognostic potential of mutant p53 will be fully realized through the application of definitive approaches to detect TP53 mutations.

6. Oncogenic EWS-ETS Translocations

ESFTs are cytogenetically diagnosed through specific genetic rearrangement involving the EWS gene (official symbol EWSR1) and a member of the ETS transcription factor gene family. This chimeric fusion protein is present in over 90% of ESFTs [73, 74] and is widely considered to be causative of this malignancy. The EWS-FLI1 translocation, t(11;22)(q24;12), is a chromosomal aberration specific to ESFTs and accounts for 85% of translocation events in Ewing Sarcoma. This reciprocal translocation generates fusion of the 5′ segment of EWS on chromosome 22 with the 3′ segment of FLI-1 on chromosome 11. Antisense DNA studies have confirmed that continuous EWS-FLI1 expression is required for the in vitro proliferation and in vivo tumorigenic capacity of Ewing Sarcoma cells [75-78]. The second most common EWS translocation described involves an in-frame fusion of the EWS and ERG genes as a result of the t(21;22)(q22;q12) translocation, accounting for 5% of translocations in Ewing Sarcoma [74].

7. Suppression of p53 Activity by EWS-FLI1

Attenuation of the p53 tumour surveillance mechanisms during the development and progression of Ewing Sarcomas may be explained through the ability of the EWS-FLI1 oncoprotein to silence p53 activity. Two independent studies have shown that silencing of EWS-FLI1 expression in Ewing Sarcoma cell lines increases p53 activity [79, 80], suggesting that the EWS-FLI1 fusion protein plays a role in the constitutive silencing of p53 tumour suppressor activity. It appears that EWS-FLI1 can achieve this through either an indirect mechanism, involving the Notch signaling pathway [79], or through the formation of a protein complex involving EWS-FLI1 and p53 [80]. This study suggests that EWS-FLI1 attenuates p53 activity through physically sequestration facilitated by the EWS region of the fusion protein [80]. However, it is unclear whether interaction between p53 and EWS-FLI1 occurs directly or is mediated through other oncogenic binding partners. It is of great interest that the amino region of EWS enables the recruitment of p53 to EWS-FLI1, as this p53-binding region is present in almost all Ewing Sarcoma gene translocation events and numerous other translocation-based cancers (Table 4). Such observations suggest that these malignancies share a common mechanism involving EWS that may potentially involve the functional inactivation of p53.

Table 4

Involvement of the EWS gene in translocation-based malignancies.

Tumour type	Fusion gene	Translocation	Reference
Ewing Sarcoma	EWS-FLI1	t(11;22)(q24;q12)	[73]
	EWS-ERG	t(21;22)(q22;q12)	[81]
	EWS-ETV1	t(7;22)(p22;q12)	[82]
	EWS-ETV4	t(17;22)(q12;q12)	[83]
	EWS-FEV	t(2;22)(q33;q12)	[84]
Acute Leukemia	EWS-CIZ1	t(12;22)(p13;q12)	[85]
Angiomatoid Fibrous	EWS-ATF1	t(12;22)(q13;q12)	[86]
Histiocytoma	EWS-CREB1	t(2;22)(q33;q12)	[87]
Clear-cell Sarcoma	EWS-ATF1	t(12;22)(q13;q12)	[88]
	EWS-CREB1	t(2;22)(q33;q12)	[89]
Desmoplastic Small Round Cell Tumour	EWS-WT1	t(11;22)(p13;q12)	[90]
Extraskeletal Myxoid	EWS-CHN1	t(9;22)(q22–31;q11-12)	[91]
Chondrosarcoma	EWS-NR4A3	t(9;22)(q22;q12)	[92]
Myxoid Liposarcoma	EWS-DDIT3	t(12;22)(q13;q12)	[93]

The ability of EWS-FLI1 to suppress p53 activity in Ewing Sarcoma is reminiscent of the functional role of the oncogenic translocation product in synovial sarcomas. The presence of the SS18-SSX fusion protein as a result of the t(X:18)(p11.2;q11.2) translocation is a universal feature of synovial sarcoma [94]. These sarcomas abrogate p53 protein levels through its enhanced proteasomal degradation facilitated by SS18-SSX [95]. Reminiscent of Ewing Sarcoma, TP53 mutations are rare events in synovial sarcomas (Table 1), suggesting that these cancers rely on the ability of the SS18-SSX fusion protein to abrogate the p53 response, facilitating oncogenic transformation in the presence of a functional, wild-type p53.

8. EWS-FLI1 Expression Stimulates the p53 Pathway in Normal Cells, Fibroblasts, or Nonmesenchymal Cells

The development of an animal model to investigate the oncogenic properties of EWS-FLI1 has been limited by the toxic effects associated with the expression of this potent fusion protein in primary cells [96]. Due to the absence of an adequate transgenic animal model for Ewing Sarcoma, experimental approaches have been restricted to forced expression of EWS-FLI1 in various cell lines. Introduction of EWS-FLI1 into primary human fibroblasts resulted in a growth arrest through stimulation of the p53 pathway [96]. Subsequent specific inhibition of p53 activity in these fibroblasts rescued the growth arrest phenotype, allowing EWS-FLI1 to promote anchorage-independent growth of these fibroblasts. Similar effects have been observed in mouse embryonic fibroblasts (MEFs) in which expression of EWS-FLI1 induced the p53-dependent growth arrest or apoptosis [97]. This apoptotic or growth arrest response was considered to be p53-dependent, as MEFs null for p53, p19ARF, or p16 were unaffected in response to EWS-FLI1 expression. Normal cells retaliate this aberrant oncogene expression by mounting a p53-based defense mechanism resulting in cellular apoptosis or senescence [98, 99]. Consideration of the responses elicited by EWS-FLI1 in normal cells therefore suggests this fusion protein in functioning as a potent oncogene and is reminiscent of the responses associated with elevated levels of the MYC or RAS oncogenes.

The cellular response to ectopically expressed EWS-FLI1 varies in primary cells of different origin. Ewing Sarcomas are derived from mesenchymal progenitor cells (MPCs) [100], and forced expression of EWS-FLI1 in MPCs was shown to be stably maintained without growth arrest or apoptosis whilst inducing an gene expression profile similar to that of a Ewing Sarcoma, all in the presence of a functional, wild-type p53 [101-103]. Such findings raise questions surrounding the ability of primary MPCs to tolerate forced EWS-FLI1 expression without engaging a p53-dependent response to the oncogenic stress. Nevertheless, the mechanisms used by MPCs to abide the expression of EWS-FLI1 in the presence of wild-type p53, whilst such expression in other primary cell lines triggers a p53 response, remain largely unknown.

Attempts have been made to investigate the function of EWS-FLI1 in a transgenic mouse model. While the conditional expression of EWS-FLI1 in mouse MPCs did not induce the formation of Ewing Sarcomas, this expression of EWS-FLI1 in mice was able to influence sarcoma development in the absence of p53 [104]. Conditional TP53 deletion in mouse MPCs led to the development osteosarcomas with a median tumour onset time of 50 weeks from birth. However, when EWS-FLI1 was conditionally expressed in these p53-null MPCs, an accelerated tumour growth from a median time of 50 to 21 weeks was observed, with the histological phenotype of these malignancies shifting towards a more poorly differentiated sarcoma [104]. These data provide in vivo evidence to further support the cross-talk between p53 and EWS-FLI1 which is essential in primary MPCs for the development of Ewing Sarcomas.

9. IGF1R and p53 Signaling Pathways

The activity of insulin-like growth factor 1 receptor (IGF1R) is essential for tumour development and progression through the signaling of antiapoptotic and prosurvival pathways [105-107]. IGF1R is also often overexpressed at the cell surface of malignant cells and thus has emerged as an attractive therapeutic target in cancer. A role of the IGF signaling pathway in the development of Ewing Sarcoma is evident through the finding that silencing of the EWS-FLI1 oncoprotein showed upregulation and activation of IGFBP genes [108]. Ewing Sarcoma cell lines are highly sensitive to IGF1R inhibitors, especially in combination with conventional chemotherapy [16, 109]. The IGF1R antagonist, AMG 479, has shown promising results in the treatment of ESFTs in a phase I clinical trial, indicating that Ewing Sarcomas may be particularly sensitive to intervention of the IGF1R signaling pathway [110, 111]. However, these studies raise vital questions as to why striking responses are seen in some, but not all, patients treated with these agents.

IGF1R and p53 drive distinctly opposing biological outcomes, with a significant level of molecular cross-talk occurring between these two signaling pathways. Initial studies suggested that p53 retorts the antiapoptotic signaling of IGF1R through repression of IGF1R expression [112]. Further antagonism of IGF1R activity by p53 was demonstrated though the identification of insulin-like growth factor binding protein 3 (IGF-BP3) as a novel p53-regulated target gene [113]. Induction of IGF-BP3 gene expression by p53 enhanced secretion of an active form of IGF-BP3 capable of inhibiting IGF1R mitogenic signalling. Thus, the IGF1R signalling pathway is functionally antagonised by wild-type p53. Recent evidence has also shown that IGF1R is degraded by MDM2 [114, 115]. Sequestration of MDM2 in the nucleus by high levels of mutant p53 may be a possible explanation for the high levels of IGF1R observed in some cancers [115].

Interestingly, pharmacological inhibition of IGF1R signalling reduces MDM2 translational synthesis, which in turn stabilizes p53 [116]. IGF1R signalling therefore regulates the p53 pathway. Hence, the overexpression of IGF1R and frequent retention of a functional, wild-type p53 present an opportunity of combined use of specific IGF1R inhibitors with activators of the p53 pathway.

10. Pharmacological p53 Activation As a Systemic Therapy for Ewing Sarcoma

Ewing Sarcomas provide a unique tumor type in which the majority of cases retain the functionally intact p53 pathways that are kept in check by either EWS-FLI1 or through another unknown mechanism. At present, there is no evidence of permanent suppression of the p53 pathway by specific mutation of critical components. Therefore, the most likely scenario involves abrogation of p53 function via a reversible, posttranslational mechanism. This provides unique therapeutic opportunities through intervention with small molecules that directly stabilize and activate endogenous intracellular p53. This concept was first demonstrated using Nutlin-3a, a small molecule antagonist of MDM2, which has shown antitumour activity in vitro and in vivo through activation of the p53 pathway in tumour cells that retain wild-type p53 [117]. Nutlin-3a antagonizes the p53-MDM2 interaction by blocking the p53-binding pocket of MDM2 and as a consequence there is a rapid stabilization and accumulation of p53 protein levels [117]. Promising results from several preclinical studies have clearly demonstrated the therapeutic potential of Nutlin-3a in a variety of tumour types expressing wild-type p53, including liposarcoma [118], rhabdomyosarcoma [119], osteosarcoma [117], synovial sarcoma [120], neuroblastoma [121], retinoblastoma [64], and leukemia [122-124].

We have recently investigated the potential of a p53-targeted therapeutic approach for the treatment of Ewing Sarcoma using Nutlin-3a. Interestingly, exposure of Ewing Sarcoma cell lines to Nutlin-3a resulted in a robust apoptotic phenotype [125]. Nutlin-3a induced apoptosis required the presence of a wild-type p53 and did not influence the growth of ESFT cell lines expressing mutant p53. These findings provide confirmation of the functionality of downstream p53 pathways in ESFTs retaining wild-type p53 and suggest that p53 activators will provide a novel molecular-based therapeutic for the majority of ESFTs.

Due to its aggressive nature and early systematic spread, treatment of ESFT is highly challenging. Current treatment protocols for ESFT patients involve multiagent chemotherapy [15]. Prior to the introduction of combinational chemotherapy, 5-year survival for patients diagnosed with ESFT was less than 10% [126]. Since the introduction of intensive VACD-IE (vincristine, actinomycin D, cyclophosphamide, doxorubicin, etoposide, and ifosfamide) chemotherapy regimens, the current 5-year survival rates for patients with localised disease are ranging from 60 to 70% [15, 127]. Nevertheless, Ewing Sarcoma still has the lowest survival rates of any of the musculoskeletal tumours, with minimal improvements to patient outcomes observed over the last decade. Hence, there is an urgent need to develop targeted therapeutic approaches to augment the action of these cytotoxic agents. The integral role of p53 in the DNA damage response pathways stimulated by these genotoxic agents suggests that p53 activators such as Nutlin-3a may provide a novel approach to augment intensive chemotherapies. Indeed, Nutlin-3a demonstrated synergistic activity with numerous chemotherapies from the VACD-IE protocol in Ewing Sarcoma [125] and in rhabdomyosarcoma [119]. Such encouraging observations will require further evaluation by the conduct of clinical trials using p53 activators as systemic therapies in combination with the current chemotherapy regimens for the treatment of wild-type p53 sarcomas.

The identification of additional drugs that work on the same principle as Nutlin-3a is an area of active research within the p53 community. The majority of reported p53 activators (Table 5) have been shown to rapidly stabilize and activate p53 protein levels through inhibition of the MDM2-p53 interaction. One such example is the recently identified MI-219, a highly selective MDM2 antagonist [128]. MI-219 interacts with the p53 binding pocket of MDM2 with a higher affinity and selectivity than Nutlin-3a and hence attains a more potent stimulation of the p53 pathway. MI-219 was observed to achieve p53-dependent antitumor activity without causing visible signs of toxicity or gross abnormalities in mice [128]. In addition, MI-219 exhibits highly desirable pharmacokinetic properties and is currently in early-phase clinical trials. Small molecule p53 activators can also function through direct interaction with p53, as demonstrated with RITA (Reactivation of p53 and induction of tumour cell apoptosis) [129]. Preclinical studies show that RITA can induce a nongenotoxic activation of p53 through inhibition of the MDM2-p53 interaction via direct interaction with N-terminal domain of p53.

Table 5

Chemical structures and proposed mechanisms of small molecule p53 activators.

MDM2 antagonists have been demonstrated to elicit their most potent effects in cell lines where MDM2 is amplified or overexpressed [118]. Since this genetic event is not observed in all tumour types, small molecules that activate p53 through alternative pathways have been developed. Lain and colleagues identified the Tenovins, a class of p53 activators that enhance the acetylation of p53 [130]. The mechanism of action of Tenovin-1 and the water soluble analog Tenovin-6 involve the direct inhibition of SIRT1 and SIRT2, two members of the sirtuin family of class III histone deacetylases responsible for the deacetylation of p53 [131-133]. It is widely accepted that acetylation is an indispensible modification of p53 that occurs during specific activation of the p53 pathway [134, 135]. Interestingly, Tenovin-6 was shown to repress the growth of cancer cells using in vitro and in vivo models through hyperacetylation of p53 [130]. These studies imply that pharmacological inhibition of sitrins is an effective approach for p53 activation.

Although specific p53 activation in tumours is an attractive therapeutic approach, these recently developed small molecules are under investigation in preclinical or early-phase clinical trials. In an attempt to accelerate the implementation of p53 activators in the clinic, a study led by Choong and colleagues screened a library of clinically approved drugs and successfully identified actinomycin D as a compound which mimics the action of Nutlin-3a when administered at specific dosages [136]. Surprisingly, low doses of actinomycin D induced specific activation of the p53 pathway with cellular responses remarkably similar to that of Nutlin-3a. These concentrations of actinomycin D were also demonstrated to augment the cytotoxic actions of chemotherapeutic drugs in cancer cells with wild-type p53 [136]. As actinomycin D is an FDA-approved drug, these findings catalyze the immediate application of a p53-based targeted therapeutic approach in the clinic.

High doses of actinomycin D are associated with intercalation into the DNA and subsequent double-strand DNA breaks [137]. It is presently used in the clinic as a chemotherapeutic agent and is a component of the highly successful combination treatment for Wilm's tumour. The inclusion of actinomycin D in current multiagent chemotherapy regimes for Ewing Sarcoma is variable amongst different clinical centers and is often dependent on the age of the patient. Since the pharmacokinetic data available for actinomycin D is limited [138, 139], comparison between the required in vivo dosages of this drug with the low concentrations used in vitro which elicit specific activation of p53 remains a formidable challenge. Interestingly, Ewing Sarcoma cells are highly sensitive to actinomycin D in vitro, with potent antitumour activity observed within the ranges described as “low dose” specifically in Ewing Sarcoma cell line that retains wild type p53 [125]. Further studies are warranted to evaluate the potential of incorporation of low dose actinomycin D with the current standard of care for the treatment of patients with wild-type p53 ESFTs.

11. Conclusion

Ewing Sarcomas share the common genetic features including the universal presence of the EWS-ETS translocation and frequent retention of the wild-type p53 and its associated functional downstream pathways. Targeted exploitation of the p53 pathway holds great promise to enhance the activity of current ESFT treatment regimes and improve the currently poor survival rates associated with Ewing Sarcoma. Recent identification of the first clinically approved drug, actinomycin D, as a p53 activator has facilitated the translation of these targeted therapies into current ESFT treatment regimens. Low dose actinomycin D hold an exciting potential as a directed molecular-based approach to specifically activate wild-type p53 in ESFTs, and the organization of clinical trials currently in progress to attest the potential of this approach. In addition, complementation of these studies with direct TP53 sequencing of ESFT material would identify either patients with wild-type p53 tumours most likely to benefit from p53-based therapies or the less frequent “high risk” population of ESFTs containing point mutations in the TP53 gene.