Targeting p53 by small molecules in hematological malignancies

Nutlin induced cytotoxic and apoptotic response in both ALL [32‐34] and AML [35‐49] cells including the cell lines and/or patient samples with little effect on normal CD34+ hematopoietic progenitor cells [35]. Nutlin-mediated killing of ALL cells harboring wild type p53 and over-expressing MDM2 is clinically very much significant since all patients with leukemic cells over-expressing MDM2 are usually resistant to conventional therapy and have a poor prognosis [32]. Nutlin-induced apoptosis in AML cells can be mediated by both p53-transcription-dependent and -independent pathways [35] (Figure 1A). Moreover, nutlin displayed synergistic responses in AML cells with several anti-leukemic agents including a Bcl2 antagonist (ABT-737) [36], MEK inhibitors (PD98059) [37] and (AZD6244) [38], recombinant TRAIL [39, 40]; FI-700, an FLT3 inhibitor [41], a vitamin D metabolite (1-25D) [42], HDAC inhibitor (valporic acid) [43], PI3K/mTOR inhibitor (PI-103) [44], perifosine (an Akt inhibitor) [49], and sorafenib, a second generation protein kinase inhibitor which is in phase I clinical trial [45].

Studies have shown that selective p53 activation with nutlin variably induced apoptosis in both low- and high-risk subtypes of B-CLL [50‐60] and CML [61, 62] patient cells. Notably, nutlin induced cytotoxicity toward B-CLL cells at concentrations that were less toxic toward normal B lymphocytes, peripheral blood mononuclear cells (PBMCs), and bone marrow (BM) hematopoietic progenitor cells [54]. In addition to conventional p53 transcriptional pathway, nutlin also induced apoptosis in CLL cells by mitochondrial pathway (Figure 1A) [58, 60].

We and others have demonstrated potent anti-myeloma activity of nutlin by molecular and functional analysis of the p53 pathway in MM cell lines, primary MM patient samples and in the cells of the bone marrow microenvironment [63‐65]. We explored the molecular mechanisms for nutlin-induced apoptosis in MM cells and provided the evidence for association of both p53-transcription-dependent and -independent pathways (Figure 1A) [65]. Our study supports the concept that the transcriptional and mitochondrial functions of p53 are equally important for nutlin-triggered apoptosis, perhaps depending on cancer cell types and their local micro-environments [65, 66].

Among different lymphoma models, nutlin has been shown effective in inducing wild type p53-dependent apoptosis of Hodgkin’s lymphoma (HL) [67, 68], mantle cell lymphoma (MCL) [69‐72], ALK-positive anaplastic large cell lymphoma (ALCL) [73], B-cell lymphoma (BCL) [74, 75], Burkitt’s and follicular lymphoma [76‐78], and adult T cell leukemia [79]. Interestingly, when combined with geldanamycin (an HSP90 inhibitor) nutlin exerted its apoptotic activity in both p53 wild type and mutant HL cells since geldanamycin-induced apoptosis in HL cells was p53-independent [68]. MDM2 inhibition by nutlin successfully induced intrinsic mitochondrial apoptotic activation through increased expression of Noxa in refractory MCL cells, which had limited sensitivity to bortezomib alone. The Nutlin/bortezomib combination enhanced Noxa protein expression in mutant p53 cells but not in wild type p53 MCL cells [71]. Similar to our observations in MM [65], nutlin-induced apoptosis in ALCL cells involved both p53-mediated transcriptional and non-transcriptional mechanisms [71, 73]. Recently, by both in vitro and in vivo evidence Drakos et al. demonstrated that nutlin induced cell cycle arrest and apoptosis in DLBCL cells with functional p53, t(14;18)(q32;q21) translocation, and Bcl2 over-expression [75]. Importantly, combined treatment with nutlin and doxorubicin synergistically inhibited the growth of ALCL or DLBCL cells harboring either wild type or mutant p53 [73, 75]. These studies also demonstrated that nutlin induced increased expression of p73 in MCL, ALCL, or BCL cells harboring mutant p53 [72, 73, 75]. Activation of p53 by nutlin resulted in both cellular senescence and apoptosis in ATL-related cell lines harboring wild type p53 suggesting that cellular senescence might be an important event in p53-dependent cell death in ATL cells [79].

Among hematological malignancies, anti-tumor activity of RITA was first described in a panel of CLL and AML patient samples [8]. This study described a constitutive activation of the p53 pathway leading to cell cycle arrest and apoptosis by RITA in CLL and AML cells harboring wild type p53 [8]. However, RITA acted synergistically with fludarabine in CLL cells irrespective of p53 status and with PRIMA-1 in AML cells with or without p53 deletion [8].

Anti-tumor activity of RITA in MM cells was first described by our group in 2010 [83]. Our in vitro studies demonstrated that RITA displayed potent anti-myeloma activities in MM cells harboring wild type p53 without killing normal cells [83]. The in vitro observation was further confirmed in xenograft mouse model of MM where we have demonstrated significant inhibition of tumor growth and prolongation of survival in mice bearing MM tumors [84, 85]. RITA was initially thought to bind with amino terminal domain of p53, inducing a conformational change of the protein and increasing its half life and its accumulation in tumor cells. However, the results of a recent nuclear magnetic resonance (NMR) study indicated that RITA might affect p53 function by other mechanisms, not involving binding to its N-terminal, such as interaction with other binding proteins and cofactors [86]. In keeping with this theory, most recently we provided the evidence that RITA targeted c-Jun N-terminal Kinase (JNK) for the induction of apoptosis in MM cells suggesting that RITA might function as a multi-target molecule [87] (Figure 1A). Further studies are needed to identify the specific binding targets for RITA.

Interestingly, study by Jones et al. provided the evidence that continuous exposure of MCL and MM models to two different MDM2 inhibitors MI-63 and nutlin resulted in p53 point mutations as a mechanism of acquired drug resistance, and that RITA might overcome this resistance by restoring p53 function [81]. This study, therefore, suggests simultaneous restoration of p53 function and MDM2 inhibition as a rational strategy for clinical translation. In support of this, we showed that RITA in combination with nutlin displayed synergistic cytotoxic response in MM cells [83]. The combination of RITA with MI-63 resulted in synergistic response in both MCL and MM cell lines resistant to MI-63 or nutlin [81]. In addition, our studies showed that RITA exerted synergistic response in combination with current chemotherapeutic agents such as doxorubicin or dexamethasone or with the JNK activator 2-Cyano-3,12-dioxooleana-1,9-dien-28 oic Acid (CDDO) [87].

PRIMA-1 was initially tested on 60 human tumor cell lines including lymphoma tumor cell lines [26, 90]. A few years later, Nahi et al. reported the effect of PRIMA-1 in leukemic cells from CLL and AML patients with or without p53 deletion [88, 92]. There were no obvious differences in cytotoxic response of PRIMA-1 between hemizygous p53 deleted and non-deleted CLL samples [92]. However, PRIMA-1 was more cytotoxic to AML cells with hemizygous p53 deletion/mutation [88]. Several studies including ours showed that normal hematopoietic cells were relatively resistant to PRIMA-1 in the concentrations used to kill tumor cells [92‐94]. PRIMA-1 has been shown to display synergistic or additive response in combination with fludarabine in CLL [92] and AML [93]. The methylated analog of PRIMA-1, PRIMA-1^Met, has even greater potency [27], leading to its development as a candidate therapeutic drug under the code name APR246 which is in phase I/II clinical trial [91].

The therapeutic concept is that PRIMA-1^Met may selectively rescue mutant p53 and induce apoptosis in cancer cells, leaving wild-type p53 in normal cells mostly unaffected [26]. However, so far there is little information on how PRIMA-1^Met affects p53 in cancer cells with no mutation. At first sight, PRIMA-1^Met activates wild type p53 appears to contradict claims that the drug is specific for mutant p53 [26, 27]. However, there are good biochemical reasons to propose that the drug operates as an all-around rescuer of inactive p53, independent of p53 mutation status. In some cancer cells, p53 protein activity may be disrupted by other mechanisms with functional consequences equivalent to mutation. How such a functionally disrupted p53 may react to PRIMA-1^Met is unknown. In the meantime, the recent studies including our preliminary results added a new dimension to the potential of PRIMA-1^Met as a therapeutic drug by showing that it induced apoptosis in cells bearing wild type p53 or even in the absence of p53 (Figure 1B) [92‐94]. Thus, the therapeutic usage of PRIMA-1^Met may be extended for treatment of the vast majority of tumors with a broader spectrum.