Design and Synthesis of Novel Antileishmanial Compounds

According to the WHO, infectious diseases, and in particular neglected tropical diseases in poor developing countries, still play a significant role in a vast number of deaths reported worldwide. Among them, leishmaniasis occurs as a complex and clinically diverse illness caused by protozoan Leishmania species which are transmitted through the bite of sandflies. They develop through a complex life cycle, from promastigotes in sandflies to amastigotes in humans. The severity of disease is determined by the type of infecting Leishmania species and also depends strongly on whether the parasite infection leads to a systemic involvement or not. Since the sensitivity towards diverse medicaments highly differs among the Leishmania species, it is advantageous to treat leishmaniasis with species-specific drugs. Towards this goal we report a synthetic methodology and characterization of novel small molecular agents active against both forms of L. major. This synthetic approach allows for rapid access to new active antileishmanial drug templates and their first derivatives in moderate to very good yields. Although the compounds reported here are bioactive, the detailed biological results are part of a more comprehensive study and will be reported separately by our collaborators.


Introduction
Infectious diseases are still a major concern worldwide [1,2]. In spite of the improved living conditions and advances in drug therapy leading to an increased protection from pathogen caused illnesses in industrialized countries, the ongoing research for new drugs is particularly important due to development of resistance to microorganisms [3], creation of new types of pathogenic agents, and the ease of transmission by globalization [2]. In the poor developing nations, which make up almost half of the world population, these kinds of diseases wreak havoc, eventually resulting in death [2]. The available medicaments against the so-called neglected tropical diseases such as leishmaniasis are often either harmful because of their side effects [4], not sufficiently effective, or expensive [5].
Leishmaniasis is caused by about 20 different protozoan Leishmania species [6] which are transmitted to humans through the bite of female phlebotomine sandflies [7,8]. During the dimorphic life cycle, the parasites develop in the guts of the sandfly into promastigotes [9]. Once the parasite is transmitted to humans, they grow up within the parasitophorous vacuole of human macrophages into amastigote forms [10], followed by host cell destruction to release Leishmania parasites for repeated infection of macrophages [11]. The severity of disease is in general determined by the grade of systemic involvement and in particular by the type of infecting Leishmania species [12,13]. The clinical picture shows diverse forms of leishmaniasis (visceral leishmaniasis, mucocutaneous leishmaniasis, and various cutaneous forms [6]). With no effective vaccines [14] and no prophylactic treatments currently available [15], with the development of resistance against the established drugs [16], and with irreversible and life-threatening side effects, leishmaniasis certainly paints a grim picture [6]. Most of the drugs such as stibogluconate, amphotericin B, pentamidine, and paromomycine need to be applied parenterally which complicates the administration, and miltefosine is the only orally applicable medicament so far ( Figure 1) [17].
Currently affecting about 350 million people with estimated 1.3 million new cases reported annually in 98 countries [18,19] the search for new orally bioavailable and costeffective medicaments with good toxicity profiles against leishmanial infections is extremely important [20]. Therefore, we have reported here the synthesis and characterization of novel small molecular compounds such as  (1) Pentamidine (2) Figure 1: The antileishmanial drugs miltefosine (1) and pentamidine (2) used as reference drugs in the bioactivity study against L. major. molecules 5, 11, 12, and 15 as potential lead structures for the treatment of leishmaniasis (L. major). The synthetic approach described here has the advantage of being amenable to rapid synthesis of novel antileishmanial drug templates and their derivatives. Although the new molecules reported in this work showed good biological activity against both forms of L. major in the range of the standard drugs miltefosine and pentamidine or even better, the detailed biological results are part of a more comprehensive study and will be reported separately by our collaborators [21].
Thus, our intention towards developing new antileishmanial drugs was to attach variable linkers to these heterocyclic structures which would enable us to link in subsequent synthetic steps other known antileishmanial active heterocyclic moieties. Since the first as linker part synthesized compound (rac)-5 already showed activity in the range of the standard drugs miltefosine (1) and pentamidine (2), we started to synthesize derivatives of 5 with varied linker according to chain length (4 and 10 carbon atoms), varied type of linkage (branched and unbranched), different terminal substituents in position A (hydrophobic, hydrophilic, and hydrogen bond forming substituents and aromatic moieties), and also cyclic structures. Eventually, also the structural part which was assumed as pharmacophore (substituent B) of these new templates was analogized ( Figure 3).

General Synthetic Procedures.
All synthesized novel antileishmanial compounds were based on the plain building block (rac)-5 (Scheme 1) and their general structure with various R 1 , R 2 , R 3 , and R 4 substituents which are shown in Figure 4.
All derivatives were synthesized from commercially available precursors purchased from Sigma Aldrich with different chain lengths according to standard chemical procedures,  followed by several conversion steps to obtain the derivatives needed for structure activity relationship studies. The general synthetic procedures are as follows.
The primary amine function of benzylamine (Scheme 1) was reacted with Boc 2 O in MeCN (dry) or 3,3-dimethylbutanoyl chloride in DCM (dry) at temperatures from 0 ∘ C up to 25 ∘ C to give the respective protected starting materials. Deprotonation with NaH in DMF (dry) at 0 ∘ C and subsequent reaction with (rac)-1,4-dibromopentane (Schemes 1 and 2), 1,4-dibromobutane, and 1,10-dibromodecane (Scheme 3) at 0 ∘ C to 25 ∘ C gave the bromine derivatives and their elimination products. The terminal bromine atom was used for the introduction of different nitrogen containing substituents such as azide, amine, and heterocycles by nucleophilic substitution reactions (Scheme 1).
In general, the bromine compounds were converted to azides using NaN 3 in DMF (dry) at room temperature with subsequent reduction to the corresponding amines by Staudinger reaction using a two-step synthetic protocol. The bromine and also the amine function were coupled with different heterocyclic moieties by nucleophilic substitution reactions.
In order to investigate which functional groups of the assumed pharmacophore were required for activity, analogs with diverse terminal functional groups in position B were synthesized. The original tert-butyloxycarbonylbenzylamino  group was replaced by a N-benzyl-3,3-dimethylbutanamide (Scheme 7) functionality as well as by a phthalimide group (Scheme 8).
Compounds with a terminal phthalimide functionality in position A were synthesized by the reaction of (rac)-1,4dibromopentane with potassium phthalimide in acetone at 60 ∘ C, followed by the introduction of the azide function using NaN 3 in DMF (dry) at room temperature (Scheme 5).

Results
(1) Synthesis of the Linker Moiety and the 4-Amino-7-chloroquinoline Substituted Compounds. For the synthesis of the linker molecule (Figures 3 and 4), we introduced the Boc protecting group to benzylamine according to a literature known protocol [32] using Boc 2 O in MeCN to give compound 6 in 98% yield, followed by the deprotonation with NaH and by a nucleophilic substitution reaction with 1,4dibromopentane in DMF (dry) at temperatures from 0 ∘ C up to 25 ∘ C (Scheme 1, step b). The major and the minor products of the reaction were (rac)-5 and the elimination product 7. The yield of the bromine compound (rac)-5 depended on the following factors: (a) the equivalents of NaH, (b) temperature, (c) reaction time, and (d) the total amount of substance in the reaction mixture. Compound (rac)-5 was obtained in varying yields from 56% to 77%, also obtaining elimination product 7. Interestingly, since the biological investigation showed that compound (rac)-5 was already antileishmanial active in the concentration range of the reference substances miltefosine (1) and pentamidine (2), we studied the structure activity relationships of those simple linkage molecules in more detail. The linker molecule was varied by the successful introduction of the azide functional group by using NaN 3 in DMF (dry) at room temperature to give compound (rac)-8 in 97% yield (Scheme 1, step c). (rac)-8 was further structurally modified in two ways: (a) by reacting with TFA in DCM at room temperature to remove the Boc protecting group to (rac)-9 (95%, Scheme 1, step d) and (b) by Staudinger reaction using PPh 3 in MeOH (dry) at room temperature yielding in amine (rac)-10 (96%, Scheme 1, step e). The free amine function of (rac)-10 was essential for the coupling to 4,7dichloroquinoline by a Buchwald-Hartwig amination reaction protocol using Pd 2 (dba) 3 as catalyst, the ligand ±-BINAP with basic KOtBu in 1,4-dioxane (dry) at 85 ∘ C, to give product (rac)-11 in 63% (Scheme 1, step f), which was Boc deprotected to substance (rac)-12 in 88% yield (Scheme 1, step g).
Derivatives with a tertiary aromatic amine function in position A and with two differently substituted terminal nitrogen atoms in position B were produced (Scheme 2). The Boc group was introduced to obtain 14 and then reacted with NaH followed by the benzyl bromide addition in DMF (dry) to give compounds 15 (23%) and 16 (23%) with the additionally benzylated aromatic amine function.
Since the bromine compound (rac)-5 showed antileishmanial activity, we checked if an in situ Boc deprotection and following nucleophilic substitution during the biotests might create (rac)-17 which could have shown activity instead of compound (rac)-5; thus the synthesis of cyclized derivatives followed (Scheme 3). By using TFA in DCM at room temperature (rac)-5 was converted to an intermediary benzylated amine that instantly cyclized to (rac)-17 in 86% yield. A similar but a quaternary amine compound (rac)-18 (54%) was obtained by using (rac)-1,4-dibromopentane and dibenzylamine in acetone at 60 ∘ C.
(2) Variation of the Linker and of Substituents in Position A. Since we assumed that the Boc group and the benzyl functionality attached to the terminal amine function in position B were the pharmacophore moiety, further derivatives with variable linker were synthesized preserving this moiety in position B. The reaction conditions that were previously successful in the synthesis of (rac)-5 from (rac)-1,4-dibromopentane did not work for the reaction of compound 6 with 1,2-dibromopropane, and hence a two-carbonatom side chain analog of (rac)-5 could not be synthesized for structure activity relationship studies. The influence of the branched side chain with the additional methyl group and the side chain length was moreover investigated by synthesizing analogs with an unbranched side chain and side chain length of four and 10 carbon atoms. Using the reaction conditions previously successful in the synthesis for (rac)-5, the unbranched starting materials 1,4-dibromobutane and 1,10-dibromodecane gave in general lower yields (Scheme 4). The bromine compounds 19 and 20 each were obtained in 27% and 44% yield, respectively, and the elimination products 21 and 22 were obtained in 9% and 5% yield, respectively (Scheme 4, step a). The conversion to the corresponding azides by NaN 3 in DMF (dry) at room temperature gave 23 (86%) and 24 (95%) (reaction monitored by NMR spectroscopy, Scheme 4, step b). Using a one-pot reaction procedure, compounds 19 and 20 were stirred separately, first with NaN 3 and PPh 3 in DMF at room temperature in an in situ Staudinger reaction and a final KOH addition step which hydrolyzes the DMF stable intermediary iminophosphorane compounds resulting in the formation of amines 25 (47%) and 26 (39%).
(4) Linkage to Various Heterocyclic Substituents. As a monocyclic, aromatic nitrogen containing heterocycle for structure activity relationship studies, imidazole was introduced to the bromine derivative (rac)-5 by using NaH in DMF (dry) to give compound (rac)-29 in 50% yield. Using equal reaction conditions and pyrrole as heterocyclic component, the desired product was not obtained. The introduction of the phthalimide heterocycle using potassium phthalimide in DMF at room temperature gave product (rac)-30 in good yields (81%, Scheme 6).
(5) Analogs of the Pharmacophore Moiety. Apart from varying the terminal substituents in position A, we also synthesized structural analogs of the assumed pharmacophore moiety in position B and replaced single atoms in order to investigate the significance of both the Boc and benzyl group in generating antileishmanial activity. Towards this, one oxygen atom of the Boc group was replaced by a carbon atom to examine the effect of the decreased free rotatability of the tert-butyl residue and the loss of the hydrogen bond acceptor ( Figure 5).
The key intermediate 31, the N-benzyl-3,3-dimethylbutanamide, was obtained in 94% yield using benzylamine and 3,3-dimethylbutanoylchloride (Scheme 7) and converted to compound (rac)-32 in very low yields of 5% using NaH, (rac)-1,4-dibromopentane in DMF (dry) at temperatures from 0 ∘ C up to 25 ∘ C where the elimination product 33 was obtained as the main product (22%). The introduction of the azide functionality using NaN 3 in DMF (dry) produced compound (rac)-34 (95%), followed by Staudinger reaction to obtain the amine (rac)-35 (89%), both in very good yields. The intended coupling to 4,7-dichloroquinoline under neat conditions at 120 ∘ C failed so far, and a Buchwald-Hartwig amination has not been tried yet but it is expected to be successful.
Compounds with a phthalimide group (36 to 39) instead of the assumed pharmacophore moiety in position B were designed as analogs with less steric hindrance and with the benzyl group involved in a fixed ring structure to decrease rotatability ( Figure 6).
A reaction between potassium phthalimide and (rac)-1,4dibromopentane [33] in acetone (dry) at 60 ∘ C produced not only the bromine derivate 36 (72%) but also the elimination product 37 and the disubstituted compound 38 as the minor products. Using NaN 3 in DMF (dry) gave product 39 in 90% yield. Because of the lack of biological activity, further derivatization of 39 was not pursued (Scheme 8).
In summary, the novel pharmacophore type present in the compounds 6 to 15 was investigated for its antileishmanial activity against promastigotes and amastigotes of Leishmania major. Since the building block (rac)-5, previously intended as linkage moiety, already showed activity against L. major in the concentration range of the reference substances miltefosine (1) and pentamidine (2), we examined the structure activity relationships of this novel substance class by the synthesis of various analogs.

Discussion.
Derivatives of the substance (rac)-5 with variation of both terminal substituents in positions A and B were synthesized to examine the necessary structural elements for antileishmanial bioactivity. The plain starting material compound 6 showed no activity at all against promastigotes of L. major hinting at the requirement of a terminally substituted (position A) alkyl side chain for activity. If the side chain was not substituted as in but-3enyl-(21), dec-3-enyl- (22), or pent-3-enyl-residues (7), the molecule showed very little to no activity (Table 1) (Table 2). If the methyl group in position 4 was lacking, compounds 19, 21, and 23 lost activity against promastigotes of L. major. The derivatives 22 and 24 without a terminal methyl group and with a longer side chain consisting of 10 carbon atoms showed no activity and 20 and 26 showed decreased activity ( Table 1).
The terminal substituents were varied using a 4-amino-7-chloroquinolinyl (11 and 14), a 6-methoxy-8-aminoquinolinyl (27 and 28), a phthalimide (30), and an imidazolyl substituent (29; Table 2). A larger, lipophilic, nitrogen containing substituent that could build hydrogen bonds and that could cause a trapping of the compound by protonation in acidic compartments of the Leishmania organism increased the activity (11). For substance 14 with a higher steric hindrance and higher lipophilic qualities caused by the benzyl group attached to the 4-amino nitrogen atom, the activity compared to compound 11 decreased. If the same side chain as of 11 was attached to the quinoline nitrogen atom, compound 40 [35] showed also decreased activity. Substituents such as the 6methoxy-8-aminoquinoline moiety showed no activity at all, and the terminal phthalimide residue 30 and an imidazolyl substituent as in compound 29 showed no activity. Similar to 11, compound 29 possessed the imidazolyl moiety which also could be enriched in acidic compartments as described for chloroquine [36] but showed no activity in comparison to the bicyclic derivative 11, hinting at the fact that a protontriggered enrichment in acidic compartments by protonation may not be the main reason for the activity of aminoquinoline substituted compounds. A slightly more spatially demanding substituent was introduced by a nonprotonable phthalimide group, but compound 30 was not active.
The bioactivities vary widely with the diverse terminal substituents and thus it can provide a hint about the nature of very specific interactions with cellular structures of the Leishmania species. The ability to build hydrogen bonds seems to be essential for the bioactivity values. The quinoline nitrogen atom of compound 11 could be accessible to form hydrogen bonds with H-bond donors (Figure 7). In contrast, the basic quinoline nitrogen atom of compound 28 has an opposite orientation to that in 11 and thus could be less capable of building hydrogen bonds. The free rotatability around the C,N-axes ( Figure 7) is probably limited by intramolecular hydrogen bonds which could fix the molecule in a particular conformation.
Due to the free rotatability around two C,N-axes of compounds 11 and 12, a higher flexibility for the spatial arrangement of the substituents in the target environment might be given compared to compounds 29 and 30. The nitrogen atoms of 29 and 30 as well as the quinoline nitrogen atoms of 11 and 28 are included as part of the ring structure. The nonincluded NH-function of 11, but rather not of 28, could contribute as hydrogen bond donor to a possible interaction with the unknown target.
In conclusion, we hypothesize that the interaction of the newly synthesized novel antileishmanial compounds with (rac)-28 (rac)-11 * * * * the unknown target on the protozoan might be influenced by the lipophilicity and size as well as the distance between the substituents in positions A and B, their flexible spatial orientation, and their ability to interact via hydrogen bonds.
To clarify the importance of the original tertbutyloxycarbonylbenzylamino substituent for the bioactivity, the four terminal functional groups which were used initially in position A (bromine, azide, amine substituent, and the terminal double bond) were retained, and the assumed pharmacophore moiety in position B was varied ( Figure 8).
The Boc-and benzyl-substituted amine function was replaced by a N-benzyl-3,3-dimethylbutanamide residue (31) which resulted in decrease in bioactivity for compound 32 relative to 5 and for 34 in comparison to 8 (Figures 8  and 9, Table 3). Compounds 31 and 6 showed no activity. These observations emphasize the need for hydrogen bond formation and of more freely rotatable bonds around the C,N-axis. The oxygen atom of the Boc group seems to be essential for the interaction (Figure 8). The introduction of a phthalimide function leads to a drastic decrease in activity for all compounds (36, 37, 38, and 39) ( Figure 10, Table 4).
We also investigated the activity of both the Boc and benzyl group as in the compounds 8/9, 11/12, 15/16, 27/28, and 40/41. The removal of the Boc group caused a significant loss in activity. Compound 8 lost all activity by removal of the Boc group (9) and substance 12 with a 4amino-7-chloroquinolinyl substituent showed a significant decrease in comparison to molecule 11. The 6-methoxy-8aminoquinoline derivatives 27 and 28 showed no activity against promastigotes of L. major. Similar to the initial observations, the removal of the Boc group from the 4amino-7-chloroquinolinium salt 40 leads to a decrease of activity 41 [35], and loss of the benzyl group leads to an activity loss of 15 to 16. This shows that it is extremely essential for the Boc and benzyl group to be attached to the amine function in order for these novel molecules to show antileishmanial activity against L. major.
The four most active compounds against promastigotes of L. major were also tested against amastigotes and showed even higher activity values. The bromine derivative 5 as a simple representative of the new substance class was active in the same concentration range as the reference substances miltefosine (1) and pentamidine (2), whereas azide 8 already showed a slightly higher activity against amastigotes. The most active representatives against promastigotes with a 4-amino-7-chloroquinoline substituent 11 and 12 showed higher activity against amastigotes than promastigotes with 8 International Journal of Medicinal Chemistry     Figure 9: General structure of analogs of the assumed pharmacophore moiety having a N-benzyl-3,3-dimethylbutanamide functionality.

11
showing slightly higher activity while 12 was significantly more active. The activity values of those compounds against promastigotes may be probably correlated with activities against amastigotes; amastigotes seem to react stronger in the case of the compounds 5, 8, 11, and 12. Investigation of bioactivity against L. donovani in the presence of these novel compounds showed no significant activities for all synthesized compounds, thus showing that these newly synthesized compounds might be highly selectively active against L. major.

Conclusion.
We have synthesized and characterized a novel class of compounds with highly selective activity  Figure 10: General structure of analogs of the assumed pharmacophore moiety with a phthalimide functionality.
against L. major. We also investigated the structure activity relationship against promastigotes of L. major and compared the activity values of promastigotes to the obtained activities of amastigotes. Terminal substituents in position A and the assumed pharmacophore moiety in position B were varied.
The most active substances were the 4-amino-7chloroquinolinyl substituted compound 11, followed by the N-benzylated derivative 15, prototype compound 5, and the Boc deprotected compound 12 (Scheme 9). The bromine substituted prototype substance 5 showed already good activity against promastigotes and amastigotes of L. major in the same concentration range as the reference substances miltefosine (1) and pentamidine (2).
The observed bioactivities and the short synthesis routes provide an excellent method for screening new drug targets Pentamidine (2) Heterocyclic substituents n Scheme 9: All active molecules of the study and reference drugs miltefosine (1) and pentamidine (2) in order from highest (11; 2.9 times more active than 1) to lowest (21; 2.8 times less active than 1) activity against promastigotes of L. major (against amastigotes of L. major: 12 < 11 < 8 < 5).
for activity against leishmaniasis in general and L. major in particular.  tert-Butyl 4-Aminopentyl(benzyl)carbamate (10). NaN 3 (26.8 mg, 0.41 mmol) and PPh 3 (102.6 mg, 0.39 mmol) were added to a solution of the bromine derivative 5 (42.2 mg, 0.12 mmol) in DMF (dry, 1 mL) at 25 ∘ C under nitrogen atmosphere. The reaction mixture was stirred for 2.5 hours until the conversion of 5 to the corresponding amine and iminophosphorane (thin-layer chromatography on silica gel and eluting with petroleum ether/EtOAc 5 : 1, followed by deactivation of the silica gel by gaseous ammonia using DCM/MeOH 10 : 1). For the hydrolysis of the iminophosphorane, KOH (39.3 mg, 0.70 mmol) was added and stirred at 25 ∘ C for 16 hours. Water was added and the aqueous phase was exhaustively extracted using dichloromethane. The combined organic extracts were dried (MgSO 4 ), filtered, and concentrated. DMF residues were removed using high  tert-Butyl 4-(7 -Chloroquinolin-4 -ylamino)butylcarbamate (14). Boc 2 O (199.5 mg, 0.91 mmol, and 210 L) was added to a solution of 13 [37] (203.0 mg, 0.81 mmol) in MeCN (dry, 6 mL) at 0 ∘ C under nitrogen atmosphere, and the reaction mixture was stirred for 1 hour at 25 ∘ C and concentrated. The residue was suspended in water and alkalized with aqueous NaOH solution and the mixture was exhaustively extracted with dichloromethane. The combined organic extracts were dried (MgSO 4 ), filtered, and concentrated. Purification tert-Butyl Benzyl (4-(benzyl(7 -chloroquinolin-4 -yl)amino)butyl)carbamate (15). NaH (51.4 mg of a 55% oily dispersion, 1.18 mmol) was added to a solution of the Boc protected 14 (134.8 mg, 0.39 mmol) in DMF (dry, 2 mL) under nitrogen atmosphere at 0 ∘ C and the reaction mixture was stirred for 30 min. Benzyl bromide (50.4 L, 72.5 mg, and 0.42 mmol) was added at 0 ∘ C, and the mixture was allowed to warm up to 25 ∘ C and was stirred for further 6 hours. DMF was removed using high vacuum and the residue was purified with flash column chromatography on silica gel (petroleum ether/EtOAc 2 : 1) to give compounds 15 (47.2 mg, 23%) and 16 (9.1 mg, 23%) both as highly viscous colorless oil.   13 (30). Potassium phthalimide (180.0 mg, 1.03 mmol) was added to a solution of the bromine derivative 5 (168.5 mg, 0.47 mmol) in DMF (dry, 5 mL) and the reaction mixture was stirred at 25 ∘ C for 72 hours. DMF was removed azeotropically with toluene under reduced pressure, and the residue was suspended in aqueous NaHCO 3 solution and exhaustively extracted with dichloromethane. Purification with flash column chromatography on silica gel (petroleum ether/EtOAc tert-Butyl 4-(1H-Imidazole-1-yl)pentyl(benzyl)carbamate (29). NaH (60.7 mg of a 55% oily dispersion, 1.39 mmol) was added to a solution of imidazole (47.3 mg, 0.69 mmol) in DMF (dry, 4 mL) at 0 ∘ C under nitrogen atmosphere and the reaction mixture was stirred for 30 min. The bromine compound 5 (312.2 mg, 0.88 mmol) was added at 0 ∘ C, and the reaction mixture was allowed to warm up to 25 ∘ C and was stirred for further 7 hours. The excess of NaH was carefully hydrolyzed using a small amount of water and the solvent mixture was removed azeotropically with toluene under reduced pressure. Purification with flash column chromatography on silica gel (EtOAc 100%) gave compound 29 (118.9 mg, 50%) as viscous  , 1 H, 4-H, B) (Me, B), 24.39 (C-3, A), 24.87 (C-3, B), 28.55 (tBu-Me, A), 28.62 (tBu-Me, B), 34.81 (C-2, A), 35.07 (C-2, B) (18). Cs 2 CO 3 (859.6 mg, 2.64 mmol) was added to a solution of (rac)-1,4-dibromopentane (500 L, 843.5 mg, and 3.67 mmol) and dibenzylamine (300 L, 307.8 mg, and 1.56 mmol) in acetone (10 mL) and the reaction mixture was stirred for 7 hours at 70 ∘ C. The mixture was concentrated, the residue was suspended in aqueous NaHCO 3 solution, and the aqueous layer was exhaustively extracted with dichloromethane. The combined organic extracts were dried (MgSO 4 ), filtered, and concentrated. Purification with flash column chromatography on silica gel (petroleum ether/EtOAc 50 : 1) gave compound 18 (289.   tert-Butyl Benzyl(4-bromobutyl)carbamate (19). NaH (3.048 g of a 55% oily dispersion, 69.8 mmol) was added to a solution of Boc protected benzylamine (4.789 g, 21.6 mmol) in DMF (dry, 50 mL) at 0 ∘ C under nitrogen atmosphere and the reaction mixture was stirred for 30 min at 0 ∘ C. 1,4-Dibromobutane (8.3 mL, 15.006 g, and 69.5 mmol) was added at 0 ∘ C and the mixture was stirred for further 3 hours. The excess of NaH was carefully hydrolysed with water and the mixture was exhaustively extracted with dichloromethane. The combined organic extracts were dried (MgSO 4 ) and filtered and the solvent mixture was removed azeotropically with toluene under reduced pressure. Purification with flash column chromatography on silica gel (petroleum ether/EtOAc 10 : 1) gave product 19 (2.019 g, 27%) and the elimination product 21 (501.4 mg, 9%) both as colorless oils.