Lead Optimization of 2-Phenylindolylglyoxylyldipeptide Murine Double

Apr 6, 2016 - In glioblastoma multiforme (GBM), translocator protein (TSPO) and murine double minute (MDM)2/p53 complex represent two druggable target...
0 downloads 6 Views 2MB Size
Subscriber access provided by TULANE UNIVERSITY

Article

Lead Optimization of 2-Phenylindolylglyoxylyldipeptide Murine Double Minute (MDM)2/Translocator Protein (TSPO) Dual Inhibitors for the Treatment of Gliomas Simona Daniele, Valeria La Pietra, Elisabetta Barresi, Salvatore Di Maro, Eleonora Da Pozzo, Marco Robello, Concettina La Motta, Sandro Cosconati, Sabrina Taliani, Luciana Marinelli, Ettore Novellino, Claudia Martini, and Federico Da Settimo J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01767 • Publication Date (Web): 06 Apr 2016 Downloaded from http://pubs.acs.org on April 8, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Medicinal Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Lead Optimization of 2Phenylindolylglyoxylyldipeptide Murine Double Minute (MDM)2/Translocator Protein (TSPO) Dual Inhibitors for the Treatment of Gliomas

Simona Daniele,‡§Valeria La Pietra,†§Elisabetta Barresi, ‡ Salvatore Di Maro,¶Eleonora Da Pozzo ‡

Marco Robello,‡Concettina La Motta,‡Sandro Cosconati,¶Sabrina Taliani,*,‡Luciana Marinelli,*,†Ettore Novellino, † Claudia Martini,‡Federico Da Settimo.‡



Department of Pharmacy, University of Naples Federico II, Naples, Italy.



Department of Pharmacy, University of Pisa, Pisa, Italy



DiSTABiF, Seconda Università di Napoli, Caserta, Italy.

§

These Authors contributed equally to this work.

KEYWORDS: MDM-2, p53, rational drug design, molecular docking, glioblastoma

ABSTRACT. In Glioblastoma Multiforme (GBM), Translocator Protein (TSPO) and Murine Double Minute (MDM2)/p53 represent two druggable targets. We recently reported the first dual 1 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 45

binder 3 possessing a higher anticancer effect in GBM cells than the standards PK11195 1 or Nutlin-3 2 singularly applied. Herein, through a Structure-Activity-Relationship study, we developed derivatives 4-10 with improved potencies toward both TSPO and MDM2. As a result, compound 9: (i) reactivated the p53 functionality; (ii) inhibited viability of two human GBM cells; (iii) impaired the proliferation of glioma cancer stem cells (CSCs), more resistant to chemotherapeutics, and responsible of GBM recurrence; (iv) sensitized GBM cells and CSCs to the activity of Temozolomide; (v) directed its effects preferentially toward tumour cells with respect to healthy ones. Thus, 9 may represent a promising cytotoxic agent which is worthy to be further developed for a therapeutic approach against GBM, where the downstream p53 signaling is intact and TSPO is over-expressed.

2 ACS Paragon Plus Environment

Page 3 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Introduction. Gliomas represent the most frequent primary malignant brain tumors in adults. They can occur anywhere in the central nervous system (CNS), but mostly are intracranial and arise in the glial tissue.1 Glioblastoma multiforme (GBM), which represents alone the 45% of all gliomas, has been defined by The World Health Organization as a grade IV cancer characterized as malignant, mitotically active, and predisposed to necrosis. Among all gliomas, GBM has the poorest overall prognosis, with only ~5% of patients surviving 5 years past diagnosis.2 Tumor surgical resection, radiotherapy, and chemotherapy strategies currently used to treat GBM have slowly evolved, and did not lead to significant increases in patient survival. The current standard of care for primary GBM is temozolomide (TMZ), a brain‐penetrant alkylating agent that methylates purines (A or G) in DNA and induces apoptosis.3 However, with TMZ use, risks arise from drug‐dependent DNA damage in healthy cells, which becomes heavier considering the possible inefficacy on GBM cells themselves. Along these lines, antiangiogenic therapy for solid tumors, including GBM, has been in the spotlight for many years, and culminated in the development of bevacizumab (Avastin®), a recombinant humanized monoclonal antibody against VEGF-A. Disappointingly, recent data show that bevacizumab, in combination with the standard treatment, did not significantly improve overall patient survival compared to standard treatment alone,4 while patients experienced a certain impact in terms of angiogenic side effects. It seems that Avastin is more a means to contain GBM growth, rather than to eliminate the tumor.5

In addition, GBM has an unfavorable prognosis mainly due to its high propensity for tumor recurrence, the causes of which are complex, and comprise the high proliferation ability of the tumor cells and their resistance to therapies, particularly as concern the cancer stem cells (CSCs). Recent literature data have demonstrated that this subpopulation of cells contribute to GBM genesis, as well as to the its extremely proliferative nature and to tumor recurrence. On this basis, it is not surprising that the number of CSCs positively correlates with tumor aggressiveness and resistance.68

3 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 45

Today, it is well known that one of the major causes for treatment failure is the acquired capability of cancer cells to escape apoptosis. In this respect, pharmacological induction of Mitochondrial Outer Membrane Permeabilization (MOMP) through apoptosis inducers9-11 has emerged as a novel and promising therapeutic approach in several cancers, including GBM.12 A number of proteins all directly or indirectly regulating the MOMP can be targeted including: (i) p53 tumour suppressor protein, negatively regulated by murine double minute (MDM)2 and MDM4, which can induce MOMP by direct interactions with multidomain proteins from the Bcl-2 family; (ii) peculiar components/modulators of the mitochondrial permeability transition pore (MPTP), such as voltage-dependent anion channel (VDAC)13 or the 18 kDa Translocator Protein (TSPO).14,15 Along these lines, TSPO ligands, such as 7-chloro-5-(4-chlorophenyl)-1-methyl-3H-1,4benzodiazepin-2-one

(Ro5-4864),16

1-(2-chlorophenyl)-N-methyl-N-(1-methylpropyl)-1-

isoquinoline carboxamide 1 (PK11195),17 and diazepam, have shown in vitro and in vivo antitumor properties, alone or in combination with etoposide or ifosfamide.10,18 Of note, 1 (Chart 1) has demonstrated to decrease or eradicate the antiapoptotic effect of Bcl-2-family proteins both in cellular cultures and in vivo, suggesting TSPO as an useful target to avoid Bcl-2-mediated chemoresistance.19 In this respect, we have confirmed that selective TSPO ligands trigger cellular apoptosis in rat and human glioma cells.20-22 Very recently, being confident that multi-target single molecules directed toward TSPO and MDM2/p53 would enhance the antitumor efficacy, and making use of indole which is known to be a “privileged structure” for drug discovery,23 we rationally designed and synthesized the first dual binder (compound 3, Chart 1) molecule;24 subsequently, we chemically modified 3 to develop an irreversible analogue able to elicit a long-lasting apoptosis of GBM cells.25 Compound 3, a 2phenylindolylglyoxylyldipeptide, binding TSPO (Ki 438 nM) and disrupting MDM2/p53 interaction (IC50 11.65 nM), caused MPTP opening, and transmembrane mitochondrial potential (∆ψm) dissipation without any steroidogenic effect, which in turn induced cell-cycle arrest and apoptosis of 4 ACS Paragon Plus Environment

Page 5 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

human GBM cells and, thus an anti-proliferative activity.24 Interestingly, compound 3 caused both early and late apoptosis of GBM cells, in a different way from what observed with the standard MDM2 inhibitor 4-{[4,5-bis(4-chlorophenyl)-2-(2-isopropoxy-4-methoxyphenyl)-4,5-dihydro-1Himidazol-1-yl]carbonyl}-piperazin-2-one 2 (Nutlin-3),26 Chart 1, and induced an higher ∆ψm collapse with respect to that produced by 1 or 2 alone.24 Along these lines, the percentage of GBM cell death was significantly higher with respect to 1 and 2 singularly applied, suggesting that the simultaneous targeting of MDM2 and TSPO could be useful in blocking tumor cell proliferation.

Chart 1. Structures of 1 (PK11195), 2 (Nutlin-3), known 3

24

and newly synthesized (4-10) 2-

phenylindolyl-glyoxylyldipeptides.

Herein, through a Structure-Activity-Relationship (SAR) study, we designed and synthesized a novel small library of 2-phenylindolylglyoxylyldipeptides 4-10 (Chart 1), featuring different dipeptide moieties on the glyoxyl bridge. All the new derivatives 4-10 were tested for their ability to bind TSPO and dissociate the MDM2/p53 complex. The best performing compound 9 was then 5 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 45

assayed for its ability to reactivate the p53 functionality and to inhibit proliferation/viability of different human GBM cells, as well as of CSCs, a subset of cells within the tumor intrinsically more resistant to chemotherapeutic agents, and responsible of GBM recurrence.27,28 In addition, as the coadministration of different drugs that produce common biological effects through the modulation of different individual targets can represent an advantageous strategy in the treatment of cancer, the effects elicited by the co-treatment of GBM cells with compound 9 and TMZ was studied to investigate a potential synergy of action. Finally, cell viability assays were performed on nontumoural human Mesenchymal Stem Cells (MSCs), to preliminarily assess the safety of compound 9. Results and Discussion Design and Synthesis. In our previous work,24 as indole is considered a “privileged structure” for drug discovery, and 2-phenylindolyglyoxylamides (PIGAs) are prone to be easily modified, an unsubstituted PIGA was docked in the MDM2 pocket giving rise to a three-dimensional complex. Subsequently, crystallographic data about p53 helix and MDM2 inhibitors, and the recent knowledge that MDM2 N-terminus region can establish van der Waals and hydrogen bond interactions with small ligands,29 both suggested how to decorate the PIGA scaffold to target MDM2 pocket. Thus, analogously to p53 transactivation domain helix, we decorated the glyoxylamide group with a Leu residue, and elongated our scaffold with the purpose to reach the region nearby the MDM2 N-terminus, by adding a Phe residue. The design strategy resulted in compound 3 (Chart 1), possessing an IC50 value of 11.65 nM.24 To corroborate the design hypothesis, docking calculation of 3 in the MDM2 binding site was attempted, and resulted in two main binding poses, where the terminal Phe residue was found within the MDM2 N-terminus region forming hydrophobic contacts with I19, L54, M50 and Y100, while the Leu side chain of 3 occupied the Leu26 pocket, interacting with the Y100, or leaved the Leu26 subpocket unoccupied,

6 ACS Paragon Plus Environment

Page 7 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

getting close to F55 located on the opposite site with respect to the Leu26 pocket (see Figure 1a, hereafter named as “Phe55 pocket”).24 Noteworthy, our three-dimensional complex showed that in the MDM2 N-terminus region, as well as in the Leu 26 or the Phe55 pocket, either aromatic or aliphatic residues were present. Thus, with the aim of improving the MDM2 IC50, we thought to convert the terminal phenyl of the lead 3 into aliphatic side chains of different length and/or shape, and the Leu side chain into a phenyl or other aliphatic chains. Moreover, to also improve the TSPO Ki of 3, our experimental (SARs) and theoretical data concerning the binding of PIGA-containing compounds with TSPO were taken into account for the choice of the possible side chain decoration.30 Indeed, our data clearly suggested that, although an N,N-disubstitution at the amide nitrogen was required for an optimal activity towards TSPO, N-monosubstituted glyoxylamides featuring highly steric demanding and flexible aliphatic or aromatic groups at the side chain are able to lead to an optimal occupancy of the TSPO L3/L4 lipophilic pockets of the receptor binding site, endowing ligands with nanomolar affinity.31-35 As a result, a number of combinations (compounds 4-10) were designed and synthesized. The experimental procedure for the synthesis of target compounds 4-10 is outlined in Scheme 1. The commercially available 2-phenylindole 11 was acylated with oxalyl chloride, in anhydrous ethyl ether, at 0 °C, to obtain the corresponding 2-phenylindolylglyoxyl chloride 12,which was allowed to react with the appropriate dipeptides (L-Valine-L-Isoleucine 13, L-Isoleucine-L-Valine 14, L-Isoleucine- L-Isoleucine 15, L-Leucine-L-Valine 16, L-Valine-L-Leucine 17, LPhenylalanine-L-Leucine 18, L-Phenylalanine-L-Isoleucine 19) in their methyl ester form, in the triethylamine presence, in dry toluene, at room temperature for 20-24 h (TLC analysis). At the end of the reaction, the suspension was filtered, and the collected precipitate was washed with a 5% NaHCO3 aqueous solution to yield products 4-10. Scheme 1 7 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 45

Dipeptides 13-19 were simply obtained in two steps as reported in Scheme 2. Reaction of the appropriate N-Boc protected amino acid with the proper amino acid methyl ester hydrochloride, in the presence of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) and hydroxybenzotriazole (HOBt) in DMF for 12 hours, yielded derivatives 20-26. Successive deprotection obtained with aqueous trifluoroacetic acid (TFA) yielded the desired dipeptides 13-19. Scheme 2

8 ACS Paragon Plus Environment

Page 9 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Dissociation of MDM2/p53 interaction and reactivation of p53 pathway in U87MG cells. The ability of the new compounds to bind MDM2 and disrupt the MDM2/p53 complex, was assessed by an immune-enzymatic assay on native human MDM2/p53 complex. All the tested compounds efficaciously dissociated the MDM2/p53 complex, with nanomolar potencies (IC50 values, Table 1), and with a potency similar or higher with respect to that of the reference MDM2 inhibitor, 2. The dose-response curves for the most potent compounds 4, 6, 8, 9 and 10 are shown in Figure 2. Interestingly, these derivatives showed comparable (4, 6, and 8) and higher (9 and 10) activity with respect to the lead 3 (IC50 11.65 nM), with compound 9 being the best performing one, yielding an IC50 value of 4.3 nM. For this reason, 9 was used in further experiments aimed at confirming the p53 reactivation upon cell treatment with a MDM2 inhibitor, by investigating the p53 protein accumulation and induction of its target genes. U87MG cells were used as representative GBM cells. This human cell line resulted an appropriate model to study MDM2-p53 and TSPO pathways, because: i) it maintains a wild type p53; ii) it is deficient for the tumor suppressor phosphatase and tensin homologue (PTEN), leading to an enhancement of MDM2 levels in the nucleus, thus inhibiting p53 functionality;36 iii) it expresses high levels of TSPO.35 9 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 45

As depicted in Figures 3A and 3B, challenging U87MG cells with 1 µM compound 9 for 8 h led to a significant increase in the levels of p53 (242 ± 25 % with respect to control cells). Noteworthy, the outcome produced by 9 resulted higher, if compared with that elicited by the parent compound 3 and the reference 2. In fact, the lead compound 3 (1 µM) lead to a maximal effect of 173 ± 10 %, while 10 µM 2 did not significantly affect p53 protein levels, at least the treatment time, in accordance with data previously reported.24,37 Accordingly, cell treatment with 1 µM 9 resulted in an increase in the mRNA levels of p53 target genes: MDM2, PUMA and p21 (Figure 4). Specifically, a 3.9-, 4.8- and 5.2-fold of induction of MDM2, PUMA and p21 mRNA, respectively, was observed. These results demonstrated that stabilization of p53 in cells incubated with 9 increased mRNA levels of MDM2, PUMA and p21 in a manner consistent with a re-activation of the p53 pathway. As a comparison, challenging U87MG cells with the lead compound 3 resulted in a 3.3-, 4.1- and 3.9-fold induction of MDM2, PUMA and p21 mRNA, respectively.24 Overall, these data suggest that 9 is more effective in the reactivation of p53 pathway with respect to both compound 3 and the reference standard 2. TSPO activity. The compounds exhibiting the highest ability to dissociate the MDM2/p53 complex (i.e., 4, 6, 8, 9 and 10) were challenged for their capability to bind TSPO. To this aim, radioligand binding assays employing the radioligand [3H]-1, which is TSPO-selective, were performed. All the tested molecules were able to displace the specific [3H]-1 binding, in a concentration-dependent manner (Figure 5), with Ki ranged in the submicromolar/nanomolar range (Table 2). The highest affinity was displayed by compound 9, that showed a Ki value about 5-fold lower than that of the lead 3 (9 Ki 87.2 nM vs 3 Ki 438 nM24). Of note, within this class, compound 9 displayed the highest ability to dissociate the p53/MDM2 complex (Table 1), and the highest affinity toward TSPO (Table 2), and both are superior than those of lead 3.24 9 was thus selected for further biological studies.

10 ACS Paragon Plus Environment

Page 11 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Molecular Modeling. To explain at molecular level the binding of 9 to MDM2 protein, docking studies were performed by means of Glide5.5 software .38 As shown in Figure 1, the pose found for 9 in its interaction with MDM2 is very similar to that already described for the lead compound 3. In particular, the phenyl substituent bound to the indole ring inserts in the Trp23 pocket establishing hydrophobic interaction with L57, I61, F86, F91, I99 and I103 side-chains, while the indole core occupies the MDM2 Phe19 subpocket interacting with I61, M62, V75 and Y67. At the same time, the glyoxylamide-NH H-bonds with the L54 carbonyl group, while the 9 side chain phenyl forms a π-π interaction with F55. On the other side of the ligand, the Leu side chain forms hydrophobic interaction with I19, Y100, L54 and M50, while the methyl ester moiety was found close with Q24. Differently from docking results of 3, pointing out that the ligand Leu side chain can occupy either Leu26 or Phe55 pockets, docking of 9 showed a clear binding pose, in which the Phe side chain of 9 interacts with F55. This finding would be in line with the lower IC50 of 9 and 10 with respect to 3 (12.1 nM for 3 vs 4.3 nM and 9.8 nM for 9 and 10, respectively) and all other compounds, which do not possess an aromatic R1 substituent (compounds 4-8). Finally, the proposed binding modes of 3 and 9 would explain the higher activity of 6 (IC50 11.7 nM), with respect to the one of 4, 5, 7, and 8 (IC50 15.2 nM, 77.7 nM, 154.6 nM, and 24.8 nM, respectively). Indeed, 6, by possessing two isoleucine residues, is able to fill up both the Leu26 or Phe55 pockets more efficiently than ligands featuring a valine residue at one of the two residues attached to the PIGA moiety (4, 5, 7 and 8).

Cell apoptosis and cell cycle. The effects of 9 on cellular apoptosis and cell cycle were analyzed. Incubation of U87MG cells with 9 (1 µM) for 24 h significantly induced both early and late apoptosis (phosphatidylserine externalization, in the absence or in the presence of 7-aminoactinomysin binding to DNA, respectively, Figures 6A and 6B). The total percentage of apoptotic U87MG cells was of 73.7 ± 7.0 (Figures 6A and 6B), whereas 1 µM compound 3 induced a percentage of cell death of 49.0 ± 3.5. Of note, the reference standard 2 did not significantly induce 11 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 45

apoptosis at this treatment time.37 The reported data are in line with the increase of PUMA mRNA level (Figure 4).

Moreover, cell cycle analysis demonstrated an increase in the G2/M fraction after 24h treatment with 9 (Figures 7A and 7B), similarly to previously reported data for compound 3.24 These results demonstrated that both the dual-target derivatives are able to trigger GBM cell apoptosis and to arrest cell-cycle progression in the G2/M-phase.

Antiproliferative Activity. The effects of 9 on U87MG cell growth/survival were examined by incubating cells with different compound concentrations. The results showed that the new compound 9 triggered a concentration-dependent inhibition of cell viability (Figure 8A), with pharmacological inhibition (IC50 value) of 1.2 ± 0.1 µM. As a comparison, compound 3 and the standard 2 showed IC50 values of 2.6 ± 0.4 µM and 6.5 ± 0.4 µM, respectively. In contrast, the standard TSPO ligand 1 significantly inhibited cell viability starting from 10 µM, reaching a maximal effect at 100 µM (inhibition at 100 µM, 45.0 ± 4.3 % vs control), in agreement with data reported in literature.39

Similar results were obtained in wild-type p53 U343MG cells (Figure 8B), where 3, 9 and the standard 2 showed IC50 values of 2.1 ± 0.3 µM, 1.6 ± 0.3 µM and 12.6 ± 1.0 µM, respectively. The GBM cell line T98G, expressing a p53 mutated, was used as a negative control.40 Following an analogous protocol, compound 9 induced a modest inhibition of T98G cell growth/survival, demonstrating about 60% of viable cells at 10 µM (Figure 8C). Treatment of T98G cells with 3 or the standard 2 gave similar results (Figure 8C), thus confirming that ligand-mediated antiproliferative effects required a wild-type p53. The standard 1 induced a maximal percentage of 20.3 ± 2.5 % versus control (Figure 8C).

12 ACS Paragon Plus Environment

Page 13 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

In addition, preliminary viability experiments were performed to evaluate the anti-proliferative potential of compound 9 on glioma CSCs, a GBM subpopolation that has been demonstrated to be less sensible to anticancer agents, and involved in GBM recurrence.8,27,28 As depicted in Figure 9A, after 24 h of incubation, compound 9 significantly decreased CSC proliferation at the highest 100 µM concentration; an enhanced antiproliferative effect was observed after prolonged treatment (seven days), with a 25% and 92% depletion of the CSC population at 1 µM and 100 µM 9, respectively. In parallel experiments, the lead compound 3 and the standard 2 induced a slight but significant inhibition of CSC viability, showing a 65% and 45% of CSC depletion at 100 µM after seven days of treatment, respectively (Figure 9A). The standard 1 didn’t reduce CSC growth, suggesting that targeting only TSPO is not sufficient to block CSC proliferation.

Co-administration of drugs which produce a common final biological effect by modulation of different individual targets may be viewed as an useful tool in the treatment of GBM. In this view, the effect elicited by the co-treatment of GBM cells with the dual-target compound 9 and the alkylating agent TMZ was studied to investigate a potential synergy in the antiproliferative activity. In these experiments, a minimum of 72 h cell incubation was chosen, in accordance with previous reports on TMZ.41 As depicted in Figure 9B, TMZ alone (50 µM) exhibited a reduction in U87MG cell viability of 55% after 72 h. When the alkylating agent was combined with 100 nM 9, an almost complete depletion of U87MG cells was observed, thus suggesting an additive/synergic effect on the reduction of U87MG cell viability (Figure 9B). Similar results were observed in U87MGderived CSCs (Figure 9C), thus demonstrating that the new compound 9 is able to sensitize GBM cells and CSCs to TMZ. Importantly, co-administered drugs are usually employed in lower concentrations than when singularly applied, thus allowing to obtain the same final effect with reduced side effects. Globally, these data highlight that 9, being able to target U87MG and U343MG cells, and the CSC subpopulation, can be regarded as a promising lead compound for the development of novel innovative and efficacious anti-proliferative strategy in GBM. Finally, as the 13 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 45

selectivity of anticancer drugs against tumor cells over the nonmalignant ones is a key issue to be taken into account, the safety of 9 was preliminary tested by evaluating its effect on the viability of non-tumoural human Mesenchymal Stem Cells (MSCs). To this aim, MSCs or U87MG were challenged with different concentrations (100 nM-10 µM) of compound 9 for 48h, and MTS assay was used to monitor cell viability. As shown in Figure 10, a slight, but significant, reduction of MSC viability was observed at 10 µM. Noteworthy, the effects of 9 on MSC viability were not rigorously related to the compound concentration, and, most importantly, the percentages of reduction of MSC viability were considerably lower if compared to those obtained in GBM cells. Similar results were obtained with the lead compound 3 (Figure 10), suggesting that antiproliferative effects elicited by these derivatives were directed preferentially toward tumour cells.

Serum stability and in silico physico-chemical characterization of 9. Metabolic stability of compound 9 was assessed in human serum as described in Experimental Section. As shown in Figure S1 (SI), 9 remains unaltered up to 3 hours. This means that human serum proteases do not cleave the amide bond of the dipeptide unit and serum esterases do not convert the methyl ester terminal into a carboxylic group at least up to 3 hours, which is the maximum time of our test. As regards the physical-chemical properties of 9, a qualitative prediction of ADMET-related properties has been performed by the Qikprop module of Maestro package.42 The analysis substantially showed that over 24 molecular descriptors,43 shown in Table S1 (SI), 5 violations with respect to ideal values ranges were present for 9. It has been demonstrated that for drug-like compounds a maximum of 5 violations are tolerated. Specifically, as shown in Table S1 (SI) the five violations regards PISA, QPlogPC, QPlogPo/w, and QPlogS, which are parameters that affect the, diffusional, passive blood-brain barrier (BBB) uptake. However, it has to be pointed out that the in silico prediction totally neglects uptake mechanisms such as carrier-mediated uptake, receptor-mediated transport, and active efflux that can greatly affect the BBB crossing of a drug. Furthermore, if the 14 ACS Paragon Plus Environment

Page 15 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

drug alone is not permeable itself, different approaches are nowadays being exploited to help pharmaceuticals to overcome the BBB, e.g. nanotechnologies,44 or the “Trojan horse” technology, where the drug is coupled to a molecule recognized by a receptor-mediated transport system.45

Conclusions. Starting from our recent finding of compound 324 as dual-targeting molecule (TSPO and MDM2/p53) endowed with enhanced antitumor efficacy in GBM cells with respect to the reference standards 1 or 2 alone, we rationally designed a small library of 2-phenylindolylglyoxylyldipeptides (4-10), featuring different dipeptide moieties on the glyoxyl bridge. Basically, our threedimensional complex of MDM2/3 showed that in the MDM2 N-terminus region, in the Leu26 as well as Phe55 pockets, either aromatic or aliphatic residues were present. Thus, we thought to convert the terminal phenyl of the lead 3 into aliphatic side chains of different length and/or shape, and the Leu side chain into a phenyl or other aliphatic chains (compounds 4-10). Subsequent biological single-target in vitro screening identified compound 9 as the best performing in terms of TSPO binding affinity and MDM2/p53 complex dissociation ability. As a result, 9 efficiently reactivated the p53 function and inhibited cell growth of GBM, at least in vitro, triggering cell cycle arrest and subsequent apoptosis. Of note, compound 9 resulted ineffective in a GBM cell line expressing mutant p53, whereas it was able to affect the viability of glioma cancer stem cells (CSCs), cells within the tumour that are less sensible to anticancer agents, and responsible of GBM recurrence.27,28 Finally, cell viability assays performed on non-tumoural human Mesenchymal Stem Cells (MSCs) showed that the anti-proliferative effect elicited by 9 was preferentially directed toward tumour cells. All these findings confirmed that dual targeting MDM2/p53 and TSPO is a valuable anticancer strategy and that 9 may represent a promising cytotoxic agent which is worthy to be further developed as a therapeutic against GBM, where the downstream p53 signalling is undamaged and TSPO is over-expressed.

15 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 45

Table 1. Effect of 2-phenylindol-3-ylglyoxylyldipeptides 3,21 and 4-10 on the dissociation of human p53/MDM2 complex.

a

compd

R1

R2

IC50 (nM)a

3b

-

-

12.1 ± 0.7

4

-CH(CH3)2

-CH(CH3)CH2CH3

15.2 ± 1.9

5

-CH(CH3)CH2CH3

-CH(CH3)2

77.7± 10.5

6

-CH(CH3)CH2CH3

-CH(CH3)CH2CH3

11.7 ± 1.2

7

-CH2CH(CH3)2

-CH(CH3)2

154.6 ± 0.8

8

-CH(CH3)2

-CH2CH(CH3)2

24.8 ± 2.7

9

-CH2C6H5

-CH2CH(CH3)2

4.3 ± 0.6

10

-CH2C6H5

-CH(CH3)CH2CH3

9.8 ± 0.9

2b

-

-

104.5 ± 8.5

Concentration (nM) leading to half-maximal inhibition of p53/MDM2 complex. Data

represent the mean values (± SEM) of three independent determinations.

16 ACS Paragon Plus Environment

Page 17 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Table 2. Binding Affinity of 2-phenylindol-3-ylglyoxylyldipeptides 3,24 and 4, 6, 8, 9 and 10 to TSPO. Displacement of specific [3H]-1 binding in mitochondrial membranes obtained from rat kidney.

a

compd

R1

R2

Ki (nM)a

3b

-

-

430 ± 39

4

-CH(CH3)2

-CH(CH3)CH2CH3

783 ± 60

6

-CH(CH3)CH2CH3

-CH(CH3)CH2CH3

1004 ± 85

8

-CH(CH3)2

-CH2CH(CH3)2

365 ± 30

9

-CH2C6H5

-CH2CH(CH3)2

87.2 ± 6.8

10

-CH2C6H5

-CH(CH3)CH2CH3

504 ± 30

1

-

-

3.6 ± 0.5

Data are expressed as means ± SEM derived from a curve-fitting procedure (GraphPad

Prism 5).

17 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 45

Figure 1. Docking poses of compound 3 (a) and 9 (b) in the MDM2 binding cleft. The ligand is shown as coral sticks, the protein surface as transparent green, and the interacting residues as light green sticks. MDM2 binding pockets are defined in red dots and labeled in accordance with the p53 interacting side chains

18 ACS Paragon Plus Environment

Page 19 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

Figure 2. Dissociation of human MDM2/p53 complex in vitro. MDM2/p53 complexes obtained from U87MG cell lysates, were treated with DMSO (control) or the indicated compounds. MDM2/p53 complex was quantified as reported in the Methods section. The data were expressed as a percentage with respect to control cells (set to 100, mean ± SEM, N = 3). A sigmoidal doseresponse analysis (GraphPad Prism 5 software) was used to obtain curves and IC50 values.

19 ACS Paragon Plus Environment

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 45

Figure 3. p53 protein accumulation in GBM cells. The U87MG cells were incubated with DMSO (Control), 1 µM 3, 1 µM 9 or 10 µM standard 2 for 8 h. At the end of treatment, cells were collected and Western blot analysis was performed on cell lysates, as reported in the Methods section. For each condition, a representative Western blot is shown (panel A). β-actin was the loading control. The quantitative analysis of immunoreactive bands is presented in panel B. The data were reported as the percentage of Optical Density (OD) versus control set to 100% (mean ± SEM, N = 3). **P