Phorbol Diesters and 12-Deoxy-16-hydroxyphorbol 13,16-Diesters Induce TGFα Release and Adult Mouse Neurogenesis

Phorbol Diesters and 12-Deoxy-16-hydroxyphorbol 13,16-Diesters Induce TGFαRelease and Adult Mouse Neurogenesis Abdellah Ezzanad, Ricardo Gómez-Oliva, Felipe Escobar-Montano, Mónica Díez-Salguero, Noelia Geribaldi-Doldan, Samuel Dominguez-Garcia, JoséManuel Botubol-Ares, Carolina de los Reyes, Rosa Durán-Patrón, Pedro Nunez-Abades, Antonio J. Macías-Sánchez, Carmen Castro, and Rosario Hernández-Galán* Cite This: J. Med. Chem. 2021, 64, 6070−6084 Read Online ACCESS Metrics & More Article Recommendations * sıSupporting Information ABSTRACT: A small library of phorbol 12,13-diesters bearing low lipophilicity ester chains was prepared as potential neurogenic agents in the adult brain. They were also used in a targeted UHPLC−HRMS screening of the latex of Euphorbia resinifera. Two new 12-deoxy-16-hydroxyphorbol 13,16-diesters were isolated, and their structures were deduced using two-dimensional NMR spectroscopy and NOE experiments. The ability of natural and synthetic compounds to stimulate transforming growth factor alpha (TFGα) release, to increase neural progenitor cell proliferation, and to stimulate neurogenesis was evaluated. All compoundsthatfacilitatedTGFαrelease promoted neural progenitor cell proliferation. The presence of two acyloxy moieties on the tigliane skeleton led to higher levels of activity, which decreased when a free hydroxyl group was at C-12. Remarkably, the compound bearing isobutyryloxy groups was the most potent on the TGFαassay and at inducing neural progenitor cell proliferation in vitro, also leading to enhanced neurogenesis in vivo when administered intranasally to mice. ■INTRODUCTION Brain damage caused by neurological and vascular disorders or by traumatic injuries is tightly associated with an irreversible neuronal loss, which may lead to cognitive impairment, motor dysfunction, and even alterations of the personality. 1−3 Despite the physiological capacity of the adult brain to generate new neurons from neural stem cells (NSC) in the subventricular zone (SVZ) 4 and the dentate gyrus (DG) of the hippocampus, 5 neuronal replacement in damaged brain regions rarely occurs, complicating functional recovery. An association of alterations in neurogenesis with cognitive impairment has been described in mouse models of different metabolic and neurological disorders, 6,7 and it is generally accepted that increased neurogenesis leads to improve cognitive performance. 8 Thus, this field demands the discovery of new drugs that regenerate damaged brain regions, facilitating functional recovery. Searching for molecules that potentiate the capacity of NSC to generate new neurons in the adult brain is a crucial step in this pathway. Our group has previously shown that several phorbol derivatives such as 12-deoxyphorbol 13-acetate (Prostratin), phorbol 12-myristate-13-acetate (PMA), 12-deoxyphorbol 13- phenylacetate (DPP), 12-deoxyphorbol 13-isobutyrate (DPB), and 12-deoxyphorbol 13-angelate (DPA) (Supporting Information, Figure S1) are capable of inducing adult neural progenitor cell (NPC) proliferation via PKC activation and of stimulating neurogenesis in the adult brain neurogenic niche of the SVZ and DG. Particularly, it was found that the 12- deoxyphorbol isolated from Euphorbia resinifera, DPB, showed the highest capacity to stimulate neurogenesis in vivo. 9 Furthermore, the long-term intranasal administration of DPB exerted not only a potent effect in stimulating neurogenesis in both the SVZ and DG, but the treatment of mice with this compound also improved their capacity to perform memory and learning tasks. 10 The comparison of the activity of all 12- deoxyphorbols tested with DPB 9 pointed at the role of the C- 13 acyl chains as a key element in defining the potency of these compounds at activating PKC and promoting neurogenesis in vivo. Additionally, unlike PMA, other 12-deoxyphorbols 13- acyl esters previously described 9 did not show tumorigenic Received: January 27, 2021 Published: May 4, 2021 Articlepubs.acs.org/jmc © 2021 The Authors. Published by American Chemical Society 6070 https://doi.org/10.1021/acs.jmedchem.1c00156 J. Med. Chem. 2021, 64, 6070−6084 Downloaded via UNIV DE SEVILLA on March 24, 2022 at 10:09:57 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. activity, suggesting 12-deoxyphorbols 13-acyl esters as good candidates to produce potent pharmacological drugs that promote neurogenesis and facilitate functional repair of damaged brain tissue. Regarding the mechanism of action of these phorbol derivatives, it has been reported that conventional PKC activation facilitates the secretion of neurotrophic factors that induce NPC proliferation such as the transforming growth factor alpha (TGFα). This neurotrophic factor stimulates the proliferation of undifferentiated NPC by binding and activating the epidermal growth factor receptor (EGFR). PKC-mediated phosphorylation of membrane-bound pro-TGFαfacilitates the ADAM17-mediated shedding of the soluble TGFαligand. 11−13 Furthermore, it has been recently described that DPB elicits TGFαrelease mimicking the effect of TGFαin cultures of NPC. 10 In this paper, we describe the preparation of selected phorbol 12,13-diesters (2−9) from commercial Croton oil, their use for identification of 12-deoxy-16-hydroxyphorbol 13,16-diesters through a screening-assisted ultrahigh performance liquid chromatography (UHPLC)−high-resolution mass spectroscopy (HRMS) from the latex of E. resinifera and also the isolation and characterization of two new 12-deoxy-16- hydroxyphorbol 13,16-diesters (10−11). Furthermore, we report the evaluation of these compounds as inducers of TGFαrelease, promoters of NPC expansion in the absence of epidermal growth factor, and as neurogenesis-stimulating small molecules. We also show structure−activity studies that can help the development of this class of compounds as effective neurogenic agents. ■RESULTS AND DISCUSSION Among the 12-deoxyphorbol esters tested previously, 9,10 those with a phenylacetate or isobutyrate chain on C-13 (DPP and DPB, respectively) presented good activity at the lower range of tested concentrations. Therefore, it is reasonable to conceive that the nature of the biological activity of this class of compounds depends on the number and type of ester chains on the phorbol skeleton. Consequently, it is logical to consider that compounds with two ester functional groups, like PMA, but bearing less lipophilic chains, like those found in 12- deoxyphorbol esters such as the ones mentioned above, would present a better activity profile. A review of the literature shows that not only phorbol esters at C-12 and C-13 have been described but also derivatives with no substituents at C-12, but with ester groups at C-13 and C- 16 (12-deoxy-16-hydroxyphorbol 13,16-diesters). 14,15 Both groups of compounds would be related to 12-deoxyphorbols like those described above (Figure S1) but with an extra acyloxy chain either at C-12 or C-16. Phorbol diesters bearing different chains at C-12 and C-13 could conceivably be prepared starting from a commercially available mixture of phorbols coming from Croton oil, an oil obtained from the seeds of Croton tiglium. 16 On the other hand, as no synthetic procedures have been described for the preparation of 12-deoxy-16-hydroxyphorbol 13,16-diesters, they would have to be obtained from natural sources. For instance, the presence of 12-deoxy-16-hydroxyphorbol derivatives in the latex of E. resinifera 17−19 is known. With this in mind, we have prepared a small library of phorbols bearing low-lipophilicity ester chains through chemical transformations on C-20-protected phorbol obtained from Croton oil. Based on our previous results 9 and with the aim to evaluate the effect of the presence of an extra acyloxy moiety on the tigliane skeleton, we decided to prepare 12,13- diester-bearing acetate, phenylacetate, isobutyrate, and methylbutenoate chains (like prostratin, DPP, DPB, and DPA, respectively). To analyze the effect of an extra free hydroxyl group, and as DPB has resulted to be the most potent compound when tested as a neural precursor promoter, 9 phorbol 13-isobutyrate was also prepared. On the other hand, to study the impact of the location of the chain on the tigliane skeleton on the capability of induction of the NPC proliferation, mixed 12,13-diesters carrying a combination of these chains were prepared, and both groups of compounds were used to identify possible 12-deoxy-16-hydroxyphorbol Scheme 1. Synthesis of Monoacyl and Diacylphorbols a a Reagents and conditions: (a) Ac2O/C5H5N, rt, 7 h for 2a (99% yield); isobutyric acid, DMAP, EDCI, dichloromethane for 3a,5a, and 9a (99, 99, and 82%, respectively); phenylacetic acid, DMPA, EDCI, dichloromethane for 4a,6a, and 7a (82, 67, and 73% yield, respectively); tiglic acid, DMAP, EDCI, dichloromethane for 8a (47% yield). (b) HClO4, rt 30 min for 2,3,4,5, and 6(99, 87, 90, 84, and 86%, respectively); HClO4,rt15 min for 7,8, and 9(75, 79, and 52% yield, respectively). Journal of Medicinal Chemistry pubs.acs.org/jmc Article https://doi.org/10.1021/acs.jmedchem.1c00156 J. Med. Chem. 2021, 64, 6070−6084 6071 13,16-diesters in the latex of E. resinifera through a targeted UHPLC−HRMS method. The exchange of chains between C- 12 and C-13 could tell us if the activity was associated with the location of each chain in a specific position or simply with their presence in that area of the molecule. Finally, a comparison of the effect of diesters in 12,13 with those of 13,16-diesters would inform us about the influence of the relative position of the ester group on the tigliane skeleton. Synthesis of Phorbol Diesters. A set of phorbol esters was prepared by selective transformations at C-12, C-13, and C-20 positions on the tigliane skeleton. Reactivity of hydroxyl groups in phorbol (1)toward esterification proceeds in the order C-20 > C-13 > C-12 due to steric effects. 20 Therefore, to selectively acylate the hydroxyl groups on C12 and/or C13 of phorbol, the primary allylic alcohol at C20 must be first protected. Croton oil was used as a source of phorbol (1), and a modified protocol reported by Bertolini et al. 20 was used to isolate phorbol 20-trityl ether (1a)(Supporting Information, Scheme S1). Once the hydroxyl group at C-20 was protected, we proceeded to functionalize C-12 and C-13 following a modification of the previously reported strategy. 20 A hydrolysis reaction of the trityl ether allows the recovery of the hydroxyl group at C-20. A careful chromatographic purification of precursor 1a must be carried out in order to avoid the co-occurrence of the epimer at C-4 ((4S)-phorbol derivatives). Epimerization at C-4 can take place through a retro-aldol-mediated rearrangement in a basic medium. 21,22 Due to the conformational change associated with the epimerization at C-4, (4S)-phorbol derivatives are not PKC activators. 20 20-Trityl phorbol esters containing identical chains at C-12 and C-13 (2a−4a) were obtained following the procedure shown in Scheme 1. While ester 2a was obtained by treatment of 1a with acetic anhydride in pyridine, esters 3a and 4a were prepared by treatment with either isobutyric or phenyl acetic acid, N-ethyl-N′-(3-dimethylaminopropyl) carbodiimide (EDCI) hydrochloride, and N,N-dimethylamino pyridine (DMAP) (Scheme 1). All attempts to obtain phorbol 12,13- diangelate were unsuccessful. As mentioned above, the C-13 hydroxyl group in phorbol derivatives is more reactive toward esterification than the one at C-12. 20 Therefore, C-13 monoacyl esters 5a and 6a could be prepared in a similar fashion to 3a and 4a, modifying slightly the reaction conditions (Scheme 1). In order to achieve selective acylation on C-13, temperature control and close monitoring of the reaction using thin layer chromatography (TLC) are required. Once selective conditions for the preparation of 13- acylphorbols were found, it was possible to prepare 12,13- diacylphorbols with differentchainsatC-12andC-13. Consequently, 20-tritylphorbol 12-phenylacetate-13-isobuty- rate (7a), 20-tritylphorbol 12-tigliate-13-isobutyrate (8a), and 20-tritylphorbol 12-isobutyrate-13-phenylacetate (9a) (Scheme 1) were prepared as they would incorporate at C- 12 ester groups already present in bioactive 12-deoxyphor- bols. 9 Treatment of 5a with phenylacetic acid or tiglic acid yielded 7a and 8a. Diester 9a was prepared, starting from 6a by treatment with isobutyric acid (Scheme 1). Finally, hydrolysis of trityl group at C-20 was achieved by treatment of corresponding diacylphorbol derivatives (2a−9a)with HClO4(0.01 M) at room temperature for 30 min to produce compounds 2−9. UHPLC−HRMS-Assisted Identification, Isolation, and Structural Elucidation of 12-Deoxyphorbol13,16-die- sters from E. resinifera.12-Deoxy-16-hydroxyphorbol 13,16- diesters and phorbol 12,13-diesters with the same ester moieties are isomeric and, therefore, are expected to show equivalent molecular ions in mass spectrometry. Consequently, using the chemical formula of compounds 2− 9as leads, extracted ion chromatograms (XIC) for pseudomolecular ions [M + Na]+, in electrospray ionization (ESI) positive mode and [M + HCOOH−H]−in ESI negative mode, were obtained by the UHPLC−HRMS analysis of the fractions obtained from the column chromatography of the latex of E. resinifera, as described in the Experimental Section. XIC for m/z575.2621 (calculated mass for C32H40O8Na, corresponding to [M + Na]+for both compounds 7and 9)in ESI positive mode and m/z597.2700 (calculated mass for C33H41O10, corresponding to [M + HCOOH−H]−for both compounds 7and 9)inESInegativemodeprovided chromatographic peaks with different retention times to those of compound 9, under equivalent elution conditions (Supporting Information, Figures S36 and S37). Therefore, the E. resinifera fraction where the above mentioned compounds were detected, isomeric to 7and 9, was subject to a careful isolation procedure, as detailed in the Experimental Section, yielding compounds 10 and 11 (Figure 1). A comparison of the HRMS of the isomeric compounds 9,10, and 11, in ESI positive mode, and with a data-independent acquisition (MSE), allowing the fragmentation of all ions within full m/zrange, 23 shows similarities and some differences on both sets of spectra. MSEof these compounds shows a relatively abundant fragmentation corresponding to [M-PhAcOH-(CH3)2CCO2H + Na]+(calculated m/z351.1572). Further fragmentations derived from this ion showed differences in compound 9, on the one hand, and both compounds 10 and 11, on the other, under equivalent MSE conditions (Figure 2 and Supporting Information, Schemes S2, S3, and S4). On the one hand, for compound 9, a further loss of water (one and two molecules) and CO leads to ions m/z 311.1647, m/z293.1541, and m/z265.1592 (calculated). These are considered diagnostic fragments, associated with different mass spectrometry strategies to identify phorbol esters in complex mixtures. 24,25 On the other, for compound 10, additional ions at m/z323.1623, m/z275.1436, and m/z 247.1487 (calculated) were observed, which were not apparent in the MSEof compound 9. As detailed in the proposed fragmentation map for compound 10 (Figure 2 and Supporting Information, Scheme S3), while the ion at m/z323.1623 could be explained by further loss of CO from the ion at m/z 351.1572, ions at m/z275.1436 and m/z247.1487 could be explained by further loss of water and CO, respectively, from m/z293.1541 ([M + H-2H2O-PhAcOH-(CH3)2CCO2H]+). Figure 1. Structures of 10−11. Journal of Medicinal Chemistry pubs.acs.org/jmc Article https://doi.org/10.1021/acs.jmedchem.1c00156 J. Med. Chem. 2021, 64, 6070−6084 6072 Interestingly, the latter fragmentation was the most abundant one observed in the ESI positive-ion mass spectra of compound 10 due to in-source fragmentations. Similar fragmentation patterns were observed for compound 11 (Figure 2 and Supporting Information, Scheme S4). Comparison of the 1H and 13C NMR spectra, as well as heteronuclear multiple−bond correlation (HMBC) coupling patterns of 10−11 with those of 7−9, indicates that the substituents were now located on C-13 and C-16 (Table 1). The 13C NMR spectra of compounds 10 and 11 showed the presence of only three methyl groups and three methylenes, indicating that the substituent on C-12 on phorbol 12,13- diesters must be now located on one of the methyls of the gemdimethyl group on C-15.1H NMR data for 10, and 11 showed the presence of phenylacetate (3.63, s, 7.31−7.23, m for 10 and 3.62, s, 7.32−7.19, m for 11) and either an isobutyrate (2.53, sp; 1.15, d; 1.14, d for 10) or a tigliate ester (6.84, q; 1.82, s; 1.81 d for 11), respectively. The main difference between the 1H NMR spectra of the natural compounds 10 and 11 and those of the synthetic compounds was the presence of two doublets (δH4.05 and 3.85 ppm in 10 and 4.14 and 3.84 ppm in 11), which correlated to a methylene group (δC 70.2 and 70.5 ppm, respectively) in the gHSQC experiments. These methylene groups were assigned to C-16 because of their correlation with the carbon signals assigned to C-17 (δC 11.7 ppm in 10 and 11.8 ppm in 11) in the gHMBC. The longrange couplings of the protons H2-16 with the carbonyl carbons at δC178.6 ppm in 10 and 169.5 ppm in 11 in the gHMBC spectra demonstrated that isobutyril and tiglyl groups are located at C-16. Consequently, phenylacetyl groups could only be located on the C-13 quaternary carbon. Furthermore, the presence of a methylene group at C-12 was confirmed by the correlations observed in the gHMBC experiments between themethylgroupsatC-11(δH0.84 and 0.89 ppm, respectively) and the C-12 carbons in each compound. Nuclear Overhauser effect spectroscopy (NOESY) correlations observed for H-16α, H-16β/H-14, and H3-17/H-8, H3-17/H- 11 in all compounds (Supporting Information) confirmed that H2-16 is α-oriented, as H-14, and H3-17 is β-oriented, as H-8 and H-11. Accordingly, compounds 10 and 11 were assigned as 12-deoxy-16-hydroxyphorbol 16-isobutyrate-13-phenyl acetate (DPPI) and 12-deoxy-16-hydroxyphorbol 13-phenyl- acetate-16-tigliate (DPPT), respectively. Both are new compounds and are described here for the first time. Effect of Phorbol Diesters and 12-Deoxy-16-hydrox- yphorbol 13,16-Diesters on TGFαRelease. A previous report shows that 12-deoxyphorbol DPB stimulated neurogenesis and improved memory in mice by facilitating TGFα release and promoting the proliferation of NSC. 10 In order to analyze whether the compounds isolated and synthesized had the ability to facilitate the release of TGFα, we built a fusion protein in which TGFαwas expressed flanked by green fluorescent protein (GFP) and mCherry in the C-terminal and N-terminal domains, respectively (Figure 3B). This construct was transfected into HEK293T cells and the ratio mCherry/ GFP in the cell membrane was quantified by using life imaging upon addition of the different compounds. By using this TGFα probe, cells expressing TGFαwill show a 1:1 ratio of mCherry to GFP. However, if the shedding of TGFαoccurs, cells will lose a percentage of mCherry fluorescence proportional to the amount of TGFαbeing released to the extracellular medium, whereas the GFP fluorescence will remain attached to the membrane. Thus, in the absence of shedding, the m-Cherry/ GFP ratio remains unaltered, whereas the shedding and release of soluble TGFαshould lower this ratio. As shown in Figure 3B,C, in the absence of compound (Ø; see Supporting Information, Movie 1) or in the presence of 0.5 μM phorbol (1; see Supporting Information, Movie 2), the mCherry/GFP ratio remains unaltered; however, the addition of compound 3 (see Supporting Information, Movie 3) and an identical Figure 2. Comparison of MSEspectra (Data-Independent Acquisition) 20 for (A) compound 9, (B) compound 10, and (C) compound 11 (data acquired in positive ionization with a ramp tramp collision energy of the high-energy function set at 60−120 eV). Journal of Medicinal Chemistry pubs.acs.org/jmc Article https://doi.org/10.1021/acs.jmedchem.1c00156 J. Med. Chem. 2021, 64, 6070−6084 6073 concentration of 4,7, and 9significantly reduced this ratio by 25%, whereas other compounds, such as 2(see Supporting Information, Movie 4) and 8, exerted a moderate effect on this ratio (15% reduction) and compound 5(see Supporting Information, Movie 5) barely reduced this ratio. These results indicated that, except for phorbol (1) and compound 5, the synthetic and naturally occurring compounds tested were able to facilitate TGFαrelease. Therefore, for these compounds to exert their effect, the hydroxyl groups at C-12 and C-13 or at C-13 and C-16 need to be esterified. Higher activities were observed when phenylacetate and isobutyrate esters were present, with the most active compound (3) bearing two isobutyrate chains. The analysis of TGFαrelease for compounds isolated from E. resinifera, where position C-12 was unsubstituted and positions C-13 and C-16 were acylated, was consistent with this hypothesis. As shown in Figure 3B,C, the effect of compounds 10 (see Supporting Information, Movie 6) and 11 on TGFαrelease was similar to that of compounds 2and 8. All these results agree with the reported effect of DPB, stimulating the release of TGFα, in a classical PKC-dependent manner. 10 Effect of Phorbol Diesters and 12-Deoxy-16-hydrox- yphorbol 13,16-diesters on NPC Proliferation In Vitro. In order to analyze whether the capacity of these compounds to release TGFαcorrelated with their proliferative activity on NPC, we tested the effect of phorbol (1) (as a negative control), compound 5(no activity on the TGFαassay), compound 2(mild effect), and compound 3(most potent on the TGFαassay). In addition, we chose compound 10 as the representative of 12-deoxy-16-hydroxyphorbol 13,16-diesters. We cultured NPCs isolated from the SVZ and cultured them under floating conditions in which the NPCs proliferate and form neurospheres. Cells were left in the absence and presence of increasing concentrations of the different compounds for 48 h. The size of the neurospheres formed was quantified afterward. We observed no effect in cultures treated with either phorbol (1) or compound 5(these two had shown no effect on TGFαrelease) (Figure 4A,B). However, all compounds that facilitated TGFαrelease, independently of their potency in the previous assay (compound 3showed a stronger effect compared with 2and 10), were able to promote NPC proliferation in vitro. This indicated that the amount of TGFαproduced by compounds 3,2, and 10 is higher than that required for these cells to proliferate. Interestingly, differences among these compounds are mainly based on the dose−response curve. Thus, compounds 3and 10 showed different dose−response curves with respect to compound 2(Figure 4A,B). The first two showed a bell-shaped dose−response curve, whereas the latter showed a linear dose response curve, within the concentration range explored. Maximum effect with 3and 10 was obtained with nanomolar concentrations, whereas no effect of compound 2was observed unless a concentration of 5 μM was used. These results indicated that all compounds with a capacity to facilitate TGFαrelease were able to promote the proliferation of NPC in vitro. However, differences were found in the doses required to exert its proliferative effect, with compounds 3and 10 being more potent than 2and also more potent than DPB. 10 The bell-shaped curve observed for compounds 3and 10 may be explained by their mechanism of action. As shown in Figure 3C, these compounds stimulate the release of the EGFR ligand TGFα. Previous reports have elucidated that the PKCαmediated release of TGFαinduced by DPB activates the receptor and its downstream signaling cascades, thus inducing proliferation 10 activating the receptor. However, another role for PKCαis mediated by its capacity to phosphorylate EGFR causing its feedback inhibition. 26 Thus, it is likely that higher doses of compounds 3and 10 result in the potent inhibition of EGFR and therefore in a reduction of proliferation and neurosphere size. Effect of Intranasal Administration of Compounds 2 and 3 on Adult Mouse Neurogenesis in the SVZ. In order to elucidate whether compounds with the high activity in vitro were able to stimulate neurogenesis in vivo in the adult mouse brain, compounds 2(moderate effect in TGFαand neurosphere assays) and 3(effect at lower concentrations in the neurosphere assay) were administered intranasally during 3 Table 1. Summary of NMR Data for DPPI (10) and DPPT (11) DPPI (10) DPPT (11) pos δH(J in Hz) a δC b δH(J in Hz) c δC d 1 7.53, m 160.8 7.53, m 160.8 2 135.0 134.6 3 210.5 210.5 4 74.7 74.7 5 2.50, d (19.0) 38.6 2.51, d (19.0) 38.6 2.42, d (19.0) 2.42, d (19.0) 6 142.4 142.4 7 5.56, bd (5.8) 129.7 5.58, bd (5.7) 129.6 8 3.08, t (5.8) 39.3 3.10, t (5.7) 39.3 9 77.6 77.7 10 3.16 d, (2.9) 57.2 3.16, d (2.7) 57.2 11 2.02, dt (11.0, 6.7) 37.4 2.03, m 37.4 12β1.57, dd (11.0, 14.8) 33.0 1.57, dd (11.0, 15.0) 33.1 12α2.13, dd (6.7, 14.8) 2.14, dd (7.0, 15.0) 13 65.1 65.2 14 1.12, d (5.8) 31.5 1.16, d (5.7) 31.5 15 28.3 28.3 16 4.05, d (11.4) 70.2 4.14, d (11.4) 70.5 3.85, d (11.4) 3.84, d (11,4) 17 1.13, s 11.7 1.14, s 11.8 18 0.84, d (6.7) 18.9 0.89, d (6,6) 18.9 19 1.73, dd (2.9, 1.4) 10.2 1.74, d (2.7, 0.7) 10.2 20 3.93, d (12,9) 68.2 3.94, d (12,9) 68.3 3.87, d (12,9) 3.84, d (12,9) 1′178.6 169.5 2′2.53, sp (7.0) 35.2 129.7 3′1.15, d (7.0) 19.4 1.82, s 12.2 4′1.14, d (7.0) 31.6 6.84, q (6.8) 138.8 5′1.81, d superimposed with H3-4″ 14.4 1″175.23 175.1 2″3.63, s 42.3 3.62, s 42.4 3″134.9 134.9 4″7.26−7.23, m 130.4 7.24−7.19, m 129.6 5″7.31−7.29, m 129.7 7.32−7.28, m 130.4 6″7.26−7.23, m 128.3 7.24−7.19, m 128.2 a 1H NMR (600 MHz, CD3OD). b 13C NMR (151 MHz, CD3OD). c 1H NMR (500 MHz, CD3OD). d 13C NMR (126 MHz, CD3OD). Journal of Medicinal Chemistry pubs.acs.org/jmc Article https://doi.org/10.1021/acs.jmedchem.1c00156 J. Med. Chem. 2021, 64, 6070−6084 6074 consecutive days. During the 3 days, the mice received injections of the thymidine analogue BrdU to label cells that proliferated during the duration of the treatment. Thereafter, the mice were sacrificed, their brains were sectioned, and immunohistochemistry was performed in the sections containing the SVZ to detect BrdU and the marker of newly generated neuronal progenitors doublecortin (DCX). As shown in Figure 5, and agreeing with the reported effects of DPB, 10 both compounds increased the number of proliferating cells (BrdU +) and the number of proliferative neuronal progenitors (BrdU +/DCX+). However, only compound 3increased the total DCX cells (DCX burden). Interestingly, the effect of compound 3on neurogenesis was observed when the concentration of the solution administered daily (12 μL/ day) was 10 nM, whereas no effect was observed when the concentration was 100 nM (10×higher) (Supporting Information, Figure S38) or 5 μM (data not shown). Thus, for compound 3, a daily dose of 12 μL of 10 nM solution had to be used in order to observe an in vivo effect. These findings are not surprising, considering the dose−response effect observed in vitro in which this compound showed a bellshaped dose−response curve with the maximum activity at 50 nM and a decline when higher doses of the compound were used. Although the actual concentrations used in vitro and in vivo should not be compared because the concentration of this compound in the SVZ or DG after the 3-day treatment period is not known, it is reasonable to suggest that the in vitro dose response curve and the in vivo effect on neurogenesis agree. The reason for the lack of activity at high doses of compound 3 can be explained by its mechanism of action. As explained above, the reported effect of DPB on adult hippocampal neurogenesis was mediated by its capacity to activate PKCα 10 inducing TGFαrelease and EGFR activation. Thus, it is reasonable to suggest that the mechanism of action of compound 3is similar and results in the activation of EGFR as well. In light of the results shown herein, it is evident that low concentrations of 3are enough to release the amount of TGFαrequired to saturate the EFGR receptor and initiated the EGFR-associated signaling pathways that promote proliferation. In this context, the lack of effect at high doses might be explained if elevated doses of the compound overactivate PKCαand directly phosphorylate EGFR, mediating the feedback inhibition of the receptor. 26 Physicochemical Properties and SAR Analysis. To investigate the influence of an additional ester group, we compared the previous results obtained with 12-deoxyphor- bols 9,10 with those described herein. Thus, in comparison to prostratin, the introduction of an additional acetyloxy moiety leads to an increased proliferative effect in compound 2 (phorbol 12,13-diacetate). A similar effect was observed when Figure 3. Phorbol diesters and 12-deoxy-16-hydroxyphorbol 13,16-diesters can induce mCherry-TGFα-GFP ectodomain shedding in vitro. (A) Images of mCherry-TGFα-GFP expressing, serum-starved HEK293T cells stimulated with diluent, with 1,2,3,5,or10 during 180 min. mCherry/ GFP ratio images at the indicated time points are shown in the intensity-modulated display mode. The color ranging goes from red to blue to represent mCherry/GFP ratios. The upper and lower limits of the ratio range are shown on the right. Scale bar = 20 μm. (B) Scheme of mCherry- TGFα-GFP construct and its action mechanism. (C) Quantitative analysis of the microscopic images obtained from the time-lapse assays of HEK293T cells expressing mCherry-TGFα-GFP protein fusion. mCherry/GFP ratios were normalized to the average mCherry/GFP ratio measured before stimulation with phorbol diesters and 12-deoxy-16-hydroxyphorbol 13,16-diesters. The mean normalized mCherry/GFP ratios are shown, n= 40. All compounds were tested at 0.5 μM. Ø: none, 1: phorbol (P), 2: phorbol 12,13-diacetate, 3: phorbol 12,13-diisobutyrate, 4: phorbol 12,13-diphenylacetate, 5: phorbol 13-isobutyrate, 7: phorbol 13-isobutyrate-12-phenylacetate, 8: phorbol 13-isobutyrate-12-tigliate, 9: phorbol 12-isobutyrate-13-phenylacetate, 10: DPPI, and 11: DPPT. Journal of Medicinal Chemistry pubs.acs.org/jmc Article https://doi.org/10.1021/acs.jmedchem.1c00156 J. Med. Chem. 2021, 64, 6070−6084 6075 comparing DPB with phorbol 12,13-diisobutyrate (3) and DPP with 12-deoxy-16-hydroxyphorbol 16-isobutyrate-13- phenyl acetate (10). Two additional factors directly influence the effect of the compound described in this work on NPC expansion: the nature and location of side chains on the tigliane skeleton. Comparing compounds with the same substituent located in different positions (9vs 10 or 7vs 9), it is observed that changing a substituent from C-12 to C-16 produces a significant reduction in the proliferative effect; however, no difference is observed when substituents are exchanged between C-12 and C-13. Remarkably, while the modification of the chain at C-12 in compounds bearing a phenylacetyl group at C-13 does not produce changes in the activity (compare compounds 4and 9,Figure 3C), significant differences were observed when the group located at C-13 is an isobutyrate (compare compounds 3,7, and 8,Figure 3C). Table S1 (Supporting Information) summarizes physicochemical parameters of the tested compounds, including molecular lipophilicity (log P), bulkiness (MW), H-bond features (HBA, HBD, and topological polar surface area (TPSA)), and molar refractivity. All of them possess a greater number of H-bond acceptors (HBA = 6−8) than H-bond donors (HBD = 3−5); only those with HBD > 3 (compounds 1and 5) proved to show no activity. On the other hand, the previously reported and active 12-deoxyphorbol esters 9,10 possess HBD = 3 (Supporting Information, Table S2), which is consistent with the analysis shown above, suggesting that the number of HBD seem to be essential for activity. Figure 4. Effect of phorbol diesters and 12-deoxy-16-hydroxyphorbol 13,16-diesters on NPC proliferation in vitro. (A) Representative phasecontrast microscopy images of neurospheres cultured for 72 h in the presence of 1,2,3,5, and 10. Scale bar = 100 μm. (B) Graph shows the effect of the different treatments at different concentrations on the neurosphere area expressed as the percentage of control (only bFGF treatment). Data shown are the mean ±SEM of nine independent measurements. Statistical analysis: ***p< 0.0001 in two-tailed unpaired Student’st-test comparing bFGF with bFGF + treatment. 1: phorbol, 2: phorbol 12,13-diacetate, 3: phorbol 12,13-diisobutyrate, 5: phorbol 13-isobutyrate, and 10: DPPI. Journal of Medicinal Chemistry pubs.acs.org/jmc Article https://doi.org/10.1021/acs.jmedchem.1c00156 J. Med. Chem. 2021, 64, 6070−6084 6076 Docking Analysis of Phorbol Derivatives with PKCδC1B (PDB: 1PTR) and PKCβC1B (PDB: 3PFQ). PKC activation is delicately controlled by several factors, involving its conformational change and membrane translocation. Phorbol esters show a higher affinity for the C-1 domain than the natural substrate (diacylglycerol, DAG), which, in part, explains their ability to promote PKC activation. Phorbol esters bind to the hydrophilic cleft of C-1 through hydrogen bonding interactions, which may result in a hydrophobic capping of the C-1 domain, depending on the nature of the substituents at C-12 and C-13 in the phorbol ester. 27 The phorbol ester/PKC complex can now eventually translocate to the membrane, which has been described as part of the activation process. A right balance in the lipophilicity of the substituents of the phorbol esters would translate into a balanced activation of PKC. 28 The crystal structure (PDB ID: 1PTR) 27 of PKCδC1B complexed with phorbol 13-acetate describes the binding pocket as a V-shape, surrounded by two long β-sheets. The PKCδC1B binds to phorbol 13-acetate through three residues: Gly-253, Thr-242, and Leu-251. Such residues target phorbol 13-acetate in specific atoms, where they made three hydrogen bonds with the hydroxyl group in position C-20 and another two with both oxygens in positions C-3 and C-4. 29 Transformation of any groups located at these positions would affect the ligand-binding affinity. On the other hand, phorbol esters’ ability to bind PKC isoforms depends on the structural characteristics of the C-1 domains in each one. 29 Previous results have shown that the capacity to promote proliferation of NPC is related to their ability to activate classical PKC. 10,30 Therefore, to further understand the structure−activity relationship, selected phorbol esters (1,3,and7), 12- deoxyphorbol esters (DPB and DPA), and 12-deoxy-16- hydroxyphorbol esters (10 and 11) were subjected to in silico molecular docking with C-1 domain models of a classical PKC. Previously, due to the availability of the crystal structure of the PKCδC1B/phorbol 13-acetate complex, we investigated possible interactions of the acyl groups at C-13, C-12/C-13, and C-13/C-16 with residues on the walls of the receptor groove PKCδ. Then, we extended our analysis such that we were able to identify key determinants of the interaction of phorbol esters with the C1 domain of PKCβ, involved in the promotion of NPC proliferation. 10,30 The subdomain C1 of PKCβwas determined from the full-length proteins PKCβII (PDB ID: 3PFQ). 31 The activator-binding residues of PKCβ were identified by comparison with the activator-binding residues of PKCδ-C1B (sequence homology was 62%, Figure 6). Despite differences in some residues, alignment of the crystal structures of δC1B (blue) and βC1B (gray) has shown a remarkable similarity in the activation of binding residues, including GLY-253 δC1B/GLY-124 βC1B, LEU-251 δC1B/ LEU-122 βC1B, and THR-242 δC1B/THR-113 βC1B). The differences between δC1B and βC1B were in the following residues Met-239 δC1B →Ser-110 βC1B, Trp-252 δC1B → Tyr-123 βC1B, Val-255 δC1B →Ile-126 βC1B, and Lys-256 δC1B →His-127 βC1B. Identical hydrogen bond interactions to those previously described for phorbol 13-acetate 29 were observed in all compounds except in phorbol (1), which showed an extra interaction between the hydroxyl group in C-13 and Ser110 of PKC β(equivalent to Met239 of PKCδ)(Supporting Figure 5. Effect of the intranasal administration of phorbol diesters on BrdU+ and DCX + cells in the SVZ in vivo. Intranasal administration of 2(5 μM), 3(10 nM), or only the vehicle was performed during 3 days to healthy 2 month-old adult mice. All mice were intraperitoneally injected with BrdU (100 mg/kg) on the last day of treatment as described in methods. (A) Representative confocal microscopy images of the SVZ of adult mice treated with 2,3, or only the vehicle. Slides were processed for the immunohistochemical detection of the proliferation marker BrdU and DCX. Scale bar = 50 μm. (B) Graph shows the number of proliferating cells marked with BrdU per mm, 3 the percentage of DCX burden, and the number of BrdU + cells per mm3that co-expressed DCX marker in the SVZ. Data shown are the mean ±SEM; n= 6 animals per group. Statistical analysis: *p< 0.05 by Student’st-test comparing with the control group. 2: phorbol 12,13-diacetate and 3: phorbol 12,13-diisobutyrate. Journal of Medicinal Chemistry pubs.acs.org/jmc Article https://doi.org/10.1021/acs.jmedchem.1c00156 J. Med. Chem. 2021, 64, 6070−6084 6077 Information, Figures S39 and S40). No differences were observed in the way in which the C-12, C-13-diester and C-13, C-16-diester were located inside the pocket, confirming that hydrogen bond interactions between oxygenated functions on A and B rings (C3O, C4OH, and C20OH and Gly-124, Thr- 113, and Leu-122 of PKCβ, equivalent to Gly-253, Thr-242, and Leu-251 of PKCδ) are the factors that determine how these compounds bind the C-1 domain of PKC. Experimentally, neurosphere assays have shown that 12- deoxyphorbols (DPP and DPB) 9 are weaker promoters of NPC proliferation than DPPI (10) and phorbol ester 3(Figure 4), confirming that the presence of an extra ester function (either at C-12 or at C-16) favors an activity increase. On the other hand, TGFαassays, which we have shown to correlate with the promotion of NPC proliferation, show that phorbol 12,13-diesters 3and 7are more potent than 12-deoxy-16- hydroxyphorbol 13,16-diesters 10 and 11, the difference between them being the disposition inside (10 and 11)or outside (3,7) of the C-1-domain receptor gap of the extra ester function, either at C-12 or at C-16, respectively (Supporting Information, Figure S40). This indicates that the additional side chains present on these compounds are key factors for activity, which would be consistent with an increase of the interaction of the phorbol ester/PKC complex with the membrane and a better modulation of the PKC activity. ■CONCLUSIONS In summary, we describe the preparation of a small library of phorbol 12,13-diesters, which were also used in a targeted UHPLC−HRMS screening of the latex of E. resinifera. This approach led to the identification and isolation of two new 12- deoxy-16-hydroxyphorbol 13,16-diesters. The ability of these compounds to induce TFGα, to increase NPC proliferation, and to stimulate neurogenesis was evaluated. All compounds that facilitated TGFαrelease were able to promote NPC proliferation when evaluated in vitro. The presence of an extra acyloxy moiety at the C or D rings of the tigliane skeleton leads to an increased level of activity compared with equivalent 12-deoxyphorbol 13-esters, which decreases if a free hydroxyl group is at C-12. While the number of hydrogen bond donors appears to be a determinant factor for activity, this is modulated by additional side-chain interactions. Remarkably, compound 3, bearing isobutyryloxy groups at C-12 and C-13, was the most potent compound on the TGFα assay and in the in vitro NPC proliferation assay, also leading to neurogenesis in vivo when administered intranasally to mice. Considering the effect of similar compounds like DPB at improving episodic memory in mice, 10 these results point out at phorbol diesters, like those described here, as relatively simple-to-manufacture pharmacological agents in the development of treatments for disorders associated with a reduction in neurogenesis and memory impairment such as neurodegenerative diseases. Figure 6. (A) Structural and sequence comparison of the zinc finger of PKCδC1B (blue) and PKCβC1B (gray). In silico interaction with phorbol 13-acetate is included for comparison with the crystal structure (PDB ID: 1PTR) of PKCδC1B complexed with phorbol 13-acetate. (B) Sequence alignment of PKCδC1B and PKCβII-C1B (amino acids (AA) in red are the AA around the ligand-binding site, and the AA marked in yellow are the AA that form hydrogen bonds with phorbol esters). Journal of Medicinal Chemistry pubs.acs.org/jmc Article https://doi.org/10.1021/acs.jmedchem.1c00156 J. Med. Chem. 2021, 64, 6070−6084 6078