Synthesis and Cytotoxicity Evaluation of Spirocyclic Bromotyrosine Clavatadine C Analogs

Marine-originated spirocyclic bromotyrosines are considered as promising scaffolds for new anticancer drugs. In a continuation of our research to develop potent and more selective anticancer compounds, we synthesized a library of 32 spirocyclic clavatadine analogs by replacing the agmatine, i.e., 4-(aminobutyl)guanidine, side chain with different substituents. These compounds were tested for cytotoxicity against skin cancer using the human melanoma cell line (A-375) and normal human skin fibroblast cell line (Hs27). The highest cytotoxicity against the A-375 cell line was observed for dichloro compound 18 (CC50 0.4 ± 0.3 µM, selectivity index (SI) 2). The variation of selectivity ranged from SI 0.4 to reach 2.4 for the pyridin-2-yl derivative 29 and hydrazide analog of 2-picoline 37. The structure–activity relationships of the compounds in respect to cytotoxicity and selectivity toward cancer cell lines are discussed.


Introduction
Natural products have a long history of use in the treatment of various diseases, including cancer [1]. Marine organisms are a rich source of novel compounds with medicinally relevant properties. Many marine-derived bioactive terpenes, alkaloids, macrolides, and other compounds isolated from aquatic fungi, cyanobacteria, algae, sponges, and tunicates have been found to exhibit anticancer activities [2,3]. To date, 15 drugs with marine origins have been approved by the U.S. Food and Drug Administration and/or European Medicines Agency. Nine of these drugs are registered for the treatment of different cancer types [4,5]. At present, 33 marine-based compounds are in clinical trials, out of which 29 are being evaluated for cancer therapy. Bromotyrosine alkaloids have acquired special importance in medicinal chemistry since the vast majority of these secondary metabolites possess potential anticancer [6], antimicrobial, antiviral, and antifungal activities [7,8]. Quinn and co-workers identified two new spirocyclic bromotyrosine compounds, clavatadine C 1 and clavatadine D 2 ( Figure 1). Both were isolated as trifluoroacetic acid (TFA) salts from the marine sponge Suberea clavata and their anticoagulative properties were described. [9]. Furthermore, moderate activity against MCF7, MDA-MB-231 (breast), and A549 (lung) cancer cell lines have been observed for clavatadine C 1-TFA [10]. Marine bromotyrosines with an isoxazoline moiety attached to the spiro center have exhibited anticancer properties [11,12]. The spirocyclic bromotyrosine is structurally more rigid and better occupies the chemical space than the open-chain bromotyrosine, making it an interesting scaffold for medicinal chemistry [13]. We have previously reported a set of simplified open-chain bromotyrosine better occupies the chemical space than the open-chain bromotyrosine, making it an interesting scaffold for medicinal chemistry [13]. We have previously reported a set of simplified open-chain bromotyrosine analogs of purpurealidin I 3 ( Figure 1) with potential antiproliferative activity [14]. As purpurealidin analogs are E-isomers having free rotation around the C-C σ bond, introduction of conformational restriction in the form of a spiro ring fusion offers a good strategy to improve selectivity toward the target cell of interest [15]. Cytotoxicity toward normal cells is a major challenge with anticancer compounds. Therefore, different approaches are required to develop a target-specific anticancer treatment. The structural simplification of natural products is one of the well-known strategies to improve pharmacokinetic profiles and to reduce side effects [16]. Initially, anticancer activity of the first simplified spiroisoxazolines 4 and 5 ( Figure 2) was reported in Ehrlich ascites tumor cells in mice [17,18].
In order to understand the structure-activity relationships (SARs) of spirocyclic bromotyrosines as cytotoxic agents toward cancer cells, we synthesized a library of simplified spirocyclic clavatadine analogs 11-42 (Table 1). The agmatine, i.e., 4-(aminobutyl)guanidine, side chain was replaced with different amino and hydrazide substituents. These analogs were tested against a melanoma cell line (A-375) and normal human skin fibroblast cell line (Hs27) for cytotoxicity. The clavatadine scaffold was selected for the library synthesis to limit the free C-C rotation of purpurealidin analogs, to have a stereochemically simpler spiro core than the one in other spirocyclic bromotyrosines, and to build on the proven anticancer activity of simplified clavatadine analogs, such as compound 4.
The compounds were synthesized according to the route presented in Scheme 1. Synthetic procedures and analytical data of the compounds are given in the Supporting Information.

Chemistry
The synthesis of the spirocyclic bromotyrosine scaffold started with esterification of L-tyrosine 6 using tert-butyl acetate in the presence of perchloric acid to give L-tyrosine tert-butyl ester 7. The ester 7 was oxidized with sodium tungstate and H2O2 to give the oxime 8 [10]. This resulting oxime was subjected to oxidative spirocyclization via treatment with [bis(trifluoro-acetoxy)iodo]benzene (PIFA) in 2,2,2-trifluoroethanol (TFE) in the case of non-halogenated compounds or N-bromosuccinimide (NBS) in N,N-dimethylformamide (DMF) in the case of brominated compounds to provide spirocyclic esters 9a/9b. The tert-butyl esters 9a/9b were deprotected with trifluoroacetic acid in dichloromethane (DCM) to give the spirocyclic carboxyl core 10a/10b. This spirocyclic core was coupled with various amines or hydrazides in the presence of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC·HCl) and 1-hydroxybenzotriazole hydrate (HOBt⋅H2O) to give the target spirocyclic bromotyrosine analogs 1-TFA and 11-42 (Table 1) with yields ranging 10-91%. We observed that the yields were typically higher in the case of dihydro carboxyl core 10a compared to the dibromo core 10b, and heterocyclic and aromatic amines compared to aliphatic amines and hydrazines. Cytotoxicity toward normal cells is a major challenge with anticancer compounds. Therefore, different approaches are required to develop a target-specific anticancer treatment. The structural simplification of natural products is one of the well-known strategies to improve pharmacokinetic profiles and to reduce side effects [16]. Initially, anticancer activity of the first simplified spiroisoxazolines 4 and 5 ( Figure 2) was reported in Ehrlich ascites tumor cells in mice [17,18]. Simplified spiroisoxazoline analogs 4 [17] and 5 [18].

Biological Activity
The cytotoxicities of the synthetic clavatadine C 1-TFA, dihydroclavatadine C 11, and compounds 12-42 against cancer cells were primarily evaluated in the human malignan melanoma A-375 cell line at the single concentration of 50 µM ( Table 2). The compound demonstrating over 80% cytotoxicity were selected for confirmatory dose-response exper iments in the same cell line and CC50 (cytotoxic concentration that caused death of 50% o cells) was calculated ( Table 2). The observed cytotoxicity (CC50) against the A-375 mela noma cell line for the compounds 1-TFA and 11-42 was in the range of 0.4-12.3 µM (Tabl 2). Furthermore, we aimed to evaluate the potential of the compounds to selectively per turb the growth of skin cancer cells. Therefore, the compounds were tested for cytotoxicity in normal human fibroblast cell line Hs27 (Table 2). Table 2. Cytotoxicity of compounds 1-TFA and 11-42 against human malignant melanoma cell line A-375 and normal human fibroblast Hs27 cells. Camptothecin, a compound with high selectivity to cancer cells, was used as a positive control. Purpurealidin I 3 was used as a reference compound [14]. CC50 = cytotoxic concentration that caused death of 50% of cells. ND = not determined. The primary screening was performed as a single experiment with three technical replicates per sample. The CC50 values are arithmetic means from 2-3 independent experiments performed in triplicate. The values in parentheses are standard deviations.  Simplified spiroisoxazoline analogs 4 [17] and 5 [18].
In order to understand the structure-activity relationships (SARs) of spirocyclic bromotyrosines as cytotoxic agents toward cancer cells, we synthesized a library of simplified spirocyclic clavatadine analogs 11-42 (Table 1). The agmatine, i.e., 4-(aminobutyl)guanidine, side chain was replaced with different amino and hydrazide substituents. These analogs were tested against a melanoma cell line (A-375) and normal human skin fibroblast cell line (Hs27) for cytotoxicity. The clavatadine scaffold was selected for the library synthesis to limit the free C-C rotation of purpurealidin analogs, to have a stereochemically simpler spiro core than the one in other spirocyclic bromotyrosines, and to build on the proven anticancer activity of simplified clavatadine analogs, such as compound 4.
The compounds were synthesized according to the route presented in Scheme 1. Synthetic procedures and analytical data of the compounds are given in the Supporting Information.  spacer compared to the analogs 14-17 having aliphatic substituents. In the case of the dihydro pyridin-2-yl analogs 32 and 37, the change of spacers to hydrazide 37 led to a modest improvement in selectivity, but in case of the corresponding bromo analogs 33 and 38 the SI was lower in hydrazide 38. The introduction of other heterocycles 40-42 gave similar lower SIs as observed for the aliphatic analogs 14-16. In comparison with our earlier reported open-chain bromotyrosine analogs [14], we found that most spirocyclic bromotyrosine analogs have lower CC50 to both cell lines tested.

Chemistry
The synthesis of the spirocyclic bromotyrosine scaffold started with esterification of L-tyrosine 6 using tert-butyl acetate in the presence of perchloric acid to give L-tyrosine tert-butyl ester 7. The ester 7 was oxidized with sodium tungstate and H 2 O 2 to give the oxime 8 [10]. This resulting oxime was subjected to oxidative spirocyclization via treatment with [bis(trifluoro-acetoxy)iodo]benzene (PIFA) in 2,2,2-trifluoroethanol (TFE) in the case of non-halogenated compounds or N-bromosuccinimide (NBS) in N,N-dimethylformamide (DMF) in the case of brominated compounds to provide spirocyclic esters 9a/9b. The tert-butyl esters 9a/9b were deprotected with trifluoroacetic acid in dichloromethane (DCM) to give the spirocyclic carboxyl core 10a/10b. This spirocyclic core was coupled with various amines or hydrazides in the presence of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC·HCl) and 1-hydroxybenzotriazole hydrate (HOBt·H 2 O) to give the target spirocyclic bromotyrosine analogs 1-TFA and 11-42 (Table 1) with yields ranging 10-91%. We observed that the yields were typically higher in the case of dihydro carboxyl core 10a compared to the dibromo core 10b, and heterocyclic and aromatic amines compared to aliphatic amines and hydrazines.

Biological Activity
The cytotoxicities of the synthetic clavatadine C 1-TFA, dihydroclavatadine C 11, and compounds 12-42 against cancer cells were primarily evaluated in the human malignant melanoma A-375 cell line at the single concentration of 50 µM ( Table 2). The compounds demonstrating over 80% cytotoxicity were selected for confirmatory dose-response experiments in the same cell line and CC 50 (cytotoxic concentration that caused death of 50% of cells) was calculated ( Table 2). The observed cytotoxicity (CC 50 ) against the A-375  (Table 2). Furthermore, we aimed to evaluate the potential of the compounds to selectively perturb the growth of skin cancer cells. Therefore, the compounds were tested for cytotoxicity in normal human fibroblast cell line Hs27 (Table 2). Table 2. Cytotoxicity of compounds 1-TFA and 11-42 against human malignant melanoma cell line A-375 and normal human fibroblast Hs27 cells. Camptothecin, a compound with high selectivity to cancer cells, was used as a positive control. Purpurealidin I 3 was used as a reference compound [14]. CC 50 = cytotoxic concentration that caused death of 50% of cells. ND = not determined. The primary screening was performed as a single experiment with three technical replicates per sample. The CC 50 values are arithmetic means from 2-3 independent experiments performed in triplicate. The values in parentheses are standard deviations.

Compound
100 The degree of selectivity toward cancer cells can be expressed by the selectivity index (SI). The SI was calculated as a ratio of CC 50 values between Hs27 fibroblasts and A-375 melanoma cells. High SI values show selectivity toward cancer cells, while values <2 show low selectivity [19] ( Table 2). The highest but still moderate selectivity to cancer cells (SI 2.4, Table 2) was observed for the pyridin-2-yl compound 29 and hydrazide analog of 2-picoline 37. To further elucidate mechanisms of cytotoxicity mediated by these compounds, we tested their ability to induce apoptosis. Apoptosis, or programmed cell death, is a mechanism utilized in the body for elimination of unwanted or damaged cells during development and aging. In cancer cells, apoptosis is typically inhibited and most selective anticancer agents act via induction of this pathway [20]. To elucidate potential effects of spirocyclic bromotyrosines on apoptosis, we tested their ability to induce the activity of caspases 3/7, a key protease involved in the apoptotic pathway. The results show that spirocyclic bromotyrosines induced the caspase pathway about twofold after 24 h, whereas no induction was observed after 48 h ( Figure 3A). Treatment with a positive control camptothecin resulted in about a 20-fold caspase induction after 24 h of treatment and a nearly fivefold induction after 48 h of treatment. The data obtained correlated with microscopic observations. Cell rounding up and shrinkage, typical for apoptosis, was observed for both camptothecin-treated cells and the cells treated with spirocyclic bromotyrosines. However, for camptothecin the effect was more profound ( Figure 3B-E).

Discussion
in Table 2. Clavatadine C 1-TFA, dihydroclavatadine C 11 and aliphatic morpholinoacetyl carbohydrazide 17 showed less than 25% primary cytotoxicity (at 50 µM in A-375 cells) while the rest of the compounds had over 75% primary cytotoxicity. This result was somewhat unexpected when compared to the known activity of clavatadine C 1-TFA

Discussion
To understand the preliminary cytotoxicity of clavatadine C 1 analogs comprised of dibromo and dihydro spirocyclic cores, we synthesized clavatadine C 1-TFA (overall yield 38%) along with various amides 11-42 having aliphatic, aromatic, and heterocyclic substitution with or without a carbon spacer. We then evaluated their effect on cytotoxicity of the melanoma cell line A-375 and the normal skin fibroblast cell line Hs27. Though the clavatadine C 1-TFA and its dihydro analog 11 did not show any primary cytotoxicity, its analogs showed cytotoxicity, as shown in Table 2. Clavatadine C 1-TFA, dihydroclavatadine C 11 and aliphatic morpholinoacetyl carbohydrazide 17 showed less than 25% primary cytotoxicity (at 50 µM in A-375 cells) while the rest of the compounds had over 75% primary cytotoxicity. This result was somewhat unexpected when compared to the known activity of clavatadine C 1-TFA against breast and colon cancer cell lines [10]. Binnewerg et al. recently found spiro-structured isofistularin-3 to display cell line dependent effects, and the three melanoma cell lines tested (SKMel-147, Mel-Juso, and Malme-3M) showed no significant decrease (and even increased in the case of SKMel-147) in cell viability, but on the contrary isofistularin-3 reduced cell viability of breast cancer cell line MCF-7 [21]. The highest cytotoxicity against the A-375 cell line was observed in dihydro 2,4-dichloro compound 18 (CC 50 0.4 ± 0.3 µM, with moderate SI 2). The selectivity indices of the most simplified dimethyl amides 13 and 14 were 1.8 and 1.9, respectively, and amides 14-16 containing aliphatic substituents showed low SI. While the introduction of aromatic substituent in 18-30 with or without a spacer exhibited relatively similar cytotoxicity, the 3-chloro-4-methoxyphenyl analog 21 showed a slight improvement in the selectivity index. As seen in our open chain library [14], the introduction of pyridinyl substituent 26-29 at the amide showed improvement in SI value to 2.4 in the case of pyridin-2-yl analog 29, whereas the introduction of the pharmacologically important trifluoromethyl group [22] to pyridine in 30 and 31 lowered the selectivity. We also introduced ethylene and methylene groups as spacers in 32-36 along with hydrazide spacers in 37-39 to pyridine to evaluate their effect on selectivity and cytotoxicity. Cytotoxicity and selectivity were increased in analogs 32-36 having a spacer compared to the analogs 14-17 having aliphatic substituents. In the case of the dihydro pyridin-2-yl analogs 32 and 37, the change of spacers to hydrazide 37 led to a modest improvement in selectivity, but in case of the corresponding bromo analogs 33 and 38 the SI was lower in hydrazide 38. The introduction of other heterocycles 40-42 gave similar lower SIs as observed for the aliphatic analogs 14-16. In comparison with our earlier reported open-chain bromotyrosine analogs [14], we found that most spirocyclic bromotyrosine analogs have lower CC 50 to both cell lines tested.
Overall, the SIs for this set of compounds stayed below 2.5, indicating relatively low selectivity toward A-375 cell line. The cytotoxic effect of the two most selective compounds was partially mediated by caspase-dependent apoptosis, although the low (twofold) level of apoptosis induction at CC 50 concentration suggests predominantly an unspecific cytotoxicity mechanism. Taking into account considerable dissimilarities in the biology of different cancer types [23], future work may include testing of the compounds in a panel of cancer cell lines representing various malignancies. In summary, biological data of synthesized spiro-structured clavatadine C 1 analogs demonstrate low selectivity toward skin cancer. However, structure-activity relationships indicate further structural optimization by modification of the side chain for potential development of these analogs into anticancer agents should be explored. and a positive control (camptothecin, Sigma-Aldrich, St. Louis, MO, USA) were prepared in DMSO and diluted into the assay medium (growth medium with 5% FBS) to the final concentration. Final DMSO concentration was 0.5% in all samples. The culture medium was removed from the plate and compounds added, 200 µL/well. After a 48-h incubation, the amount of ATP, which is directly proportional to the number of viable cells present in culture, was quantified using CellTiter-Glo ® Luminescent Cell Viability kit (Promega, Madison, WI, USA), according to manufacturer's instructions. Origin Graphing and Analysis, version 9.55 (OriginLab, Northampton, MA, USA), was used for determination of CC 50 values. The cancer cell selectivity index was calculated as a ratio of CC 50 values between Hs27 fibroblasts and A-375 melanoma cells. Standard deviation of selectivity indices was calculated using Equation (1)

Materials and Methods for Biological Testing
In the Equation (1), x and y are average CC 50 values in Hs27 and A-375 cells, respectively, and σ(x) and σ(y) are their standard deviations.
Apoptosis induction assay and imaging. A-375 cells were seeded to white 96-well plates and treated with compounds at 1 × CC 50 concentration, 100 µL/well, following the procedure described above. After 24 and 48 h, caspase-3/7 activity was measured using the ApoTox-Glo kit (Promega) following the manufacturer's instructions. The light microscopy images were taken using 4× phase contrast objective and Cytation5 automated imaging reader (Biotek).

General
All reactions were carried out using commercially available starting materials unless otherwise stated. The melting points were measured with a Stuart SMP40 automated melting point apparatus and were uncorrected. 1 H NMR and 13 C NMR spectra in CDCl 3 , d 6 -DMSO, d 6 -acetone, or CD 3 OD at ambient temperature were recorded on a Bruker Ascend 400 spectrometer. Chemical shifts (δ) are given in parts per million (ppm) relative to the NMR reference solvent signals (CDCl 3 : 7.26 ppm, 77.16 ppm; CD 3 OD: 3.31 ppm, 49.00 ppm; d 6 -DMSO: 2.50 ppm, 39.52 ppm; d 6 -acetone: 2.05 ppm, 29.84 ppm). Multiplicities are indicated by s (singlet), br s (broad singlet), d (doublet), dd (doublet of doublets), ddd (doublet of doublet of doublets), t (triplet), dt (doublet of triplets), q (quartet), p (pentet), and m (multiplet). The coupling constants J were quoted in hertz (Hz). LC-MS and HRMSspectra were recorded using a Waters Acquity UPLC ® -system (with Acquity UPLC ® BEH C18 column, 1.7 µm, 50 mm × 2.1 mm, Waters, Milford, MA, USA) with Waters Synapt G2 HDMS with the ESI (+), high resolution mode, and PDA. The mobile phase consisted of H 2 O (A) and acetonitrile (B), both containing 0.1% HCOOH. Microwave syntheses were performed in sealed tubes using a Biotage Initiator+ instrument equipped with an external IR sensor. The flash chromatography was performed with a Biotage Isolera One flash chromatography purification system with a 200-800 nm UV-VIS detector using SNAP KP-Sil 10 g, 25 g, or 50 g cartridges. The TLC plates were provided by Merck (Silica gel 60-F254) and visualization of the amine compounds was conducted using ninhydrin (a 0.2% w/v solution in a 3% solution of acetic acid in 1-butanol) staining.

tert-Butyl L-tyrosinate (7)
Mar. Drugs 2021, 19, x FOR PEER REVIEW All reactions were carried out using commercially available sta otherwise stated. The melting points were measured with a Stua melting point apparatus and were uncorrected. 1 H NMR and 13 C N d6-DMSO, d6-acetone, or CD3OD at ambient temperature were reco cend 400 spectrometer. Chemical shifts (δ) are given in parts per mil the NMR reference solvent signals (CDCl3: 7.26 ppm, 77.16 ppm; CD ppm; d6-DMSO: 2.50 ppm, 39.52 ppm; d6-acetone: 2.05 ppm, 29.84 pp indicated by s (singlet), br s (broad singlet), d (doublet), dd (doub (doublet of doublet of doublets), t (triplet), dt (doublet of triplets), q and m (multiplet). The coupling constants J were quoted in hertz (Hz spectra were recorded using a Waters Acquity UPLC ® -system (with C18 column, 1.7 µm, 50 mm × 2.1 mm, Waters, Milford, MA, USA) w HDMS with the ESI (+), high resolution mode, and PDA. The mob H2O (A) and acetonitrile (B), both containing 0.1% HCOOH. Micro performed in sealed tubes using a Biotage Initiator+ instrument equi IR sensor. The flash chromatography was performed with a Biota chromatography purification system with a 200-800 nm UV-VIS dete Sil 10 g, 25 g, or 50 g cartridges. The TLC plates were provided by F254) and visualization of the amine compounds was conducted us w/v solution in a 3% solution of acetic acid in 1-butanol) staining.

tert-Butyl L-tyrosinate (7)
To a stirred suspension of L-tyrosine 6 (25.0 g, 0.138 mol) in t mL) in an ice bath (0 °C), perchloric acid (15.7 mL, 0.276 mol, 2.0 equ wise. The reaction mixture was allowed to warm to room temperatu 17 h. The mixture was washed with H2O (300 mL) and a 1 M solutio mL). The aqueous phase was diluted with H2O (300 mL), followed by To a stirred suspension of L-tyrosine 6 (25.0 g, 0.138 mol) in tert-butyl acetate (100 mL) in an ice bath (0 • C), perchloric acid (15.7 mL, 0.276 mol, 2.0 equiv) was added dropwise. The reaction mixture was allowed to warm to room temperature and was stirred for 17 h. The mixture was washed with H 2 O (300 mL) and a 1 M solution of HCl in H 2 O (250 mL). The aqueous phase was diluted with H 2 O (300 mL), followed by the addition of solid K 2 CO 3 until the pH was 7. The resulting mixture was filtered, the filtrate was made alkaline (pH 9) by adding solid K 2 CO 3 , and then it was extracted with EtOAc (3 × 300 mL). The combined organic phases were washed with brine (300 mL), dried over anhydrous Na 2 SO 4 , filtered, and concentrated in vacuo to give an off-white solid; crude yield: 29 g (86%). The crude product was purified with automated flash chromatography (DCM/MeOH, gradient: 0→10%) to give the product 7 as a white solid (25 g, 76%). 1

tert-Butyl L-tyrosinate (7)
To a stirred suspension of L-tyrosine 6 (25.0 g, 0.138 mol) in t mL) in an ice bath (0 °C), perchloric acid (15.7 mL, 0.276 mol, 2.0 eq wise. The reaction mixture was allowed to warm to room temperatu 17 h. The mixture was washed with H2O (300 mL) and a 1 M solutio mL). The aqueous phase was diluted with H2O (300 mL), followed b K2CO3 until the pH was 7. The resulting mixture was filtered, the fi line (pH 9) by adding solid K2CO3, and then it was extracted with EtO combined organic phases were washed with brine (300 mL), d Na2SO4, filtered, and concentrated in vacuo to give an off-white so (86%). The crude product was purified with automated fla (DCM/MeOH, gradient: 0→10%) to give the product 7 as a white NMR (400 MHz, CDCl3)
General procedure for EDC-mediated coupling (B). Carboxylic acid 10a or 10b (0.3 mmol), amine (0.45 mmol, 1.5 equiv), HOBt hydrate (0.45 mmol, 1.5 equiv), and EDC·HCl (0.45 mmol, 1.5 equiv) were dissolved in anhydrous DCM (3 mL). The mixture was irradiated under microwave conditions at 60 • C for 2 h, after which it was diluted with DCM (10 mL). The solution was washed with a saturated solution of NH 4 Cl in H 2 O, water, and brine. The organic phase was dried over anhydrous Na 2 SO 4 , filtered, and concentrated in vacuo. The crude product was purified with automated flash column chromatography (n-heptane/EtOAc-EtOH 3:1 (12→100%) to give the pure product.
General procedure for deprotection of the Boc groups (C). To a solution of clavatadine bis-Boc-derivative (0.28 mmol) in DCM (2 mL), TFA (1 mL) at 0-5 • C was added dropwise. The reaction mixture was allowed to reach room temperature. The resulting mixture was stirred for 3 h at room temperature. The solvent was removed in vacuo to give the crude product, which was triturated in Et 2 O to give a solid trifluoroacetate salt.
General procedure for deprotection of the Boc groups (C). To tadine bis-Boc-derivative (0.28 mmol) in DCM (2 mL), TFA (1 mL) dropwise. The reaction mixture was allowed to reach room tempe mixture was stirred for 3 h at room temperature. The solvent was give the crude product, which was triturated in Et2O to give a solid