Optimized syntheses of Fmoc azido amino acids for the preparation of azidopeptides

The rise of CuI‐catalyzed click chemistry has initiated an increased demand for azido and alkyne derivatives of amino acid as precursors for the synthesis of clicked peptides. However, the use of azido and alkyne amino acids in peptide chemistry is complicated by their high cost. For this reason, we investigated the possibility of the in‐house preparation of a set of five Fmoc azido amino acids: β‐azido l‐alanine and d‐alanine, γ‐azido l‐homoalanine, δ‐azido l‐ornithine and ω‐azido l‐lysine. We investigated several reaction pathways described in the literature, suggested several improvements and proposed several alternative routes for the synthesis of these compounds in high purity. Here, we demonstrate that multigram quantities of these Fmoc azido amino acids can be prepared within a week or two and at user‐friendly costs. We also incorporated these azido amino acids into several model tripeptides, and we observed the formation of a new elimination product of the azido moiety upon conditions of prolonged couplings with 2‐(1H‐benzotriazol‐1‐yl)‐1,1,3,3‐tetramethyluronium hexafluorophosphate/DIPEA. We hope that our detailed synthetic protocols will inspire some peptide chemists to prepare these Fmoc azido acids in their laboratories and will assist them in avoiding the too extensive costs of azidopeptide syntheses. Experimental procedures and/or analytical data for compounds 3–5, 20, 25, 26, 30 and 43–47 are provided in the supporting information. © 2017 The Authors Journal of Peptide Science published by European Peptide Society and John Wiley & Sons Ltd.

In general, there are two main approaches for the synthesis of derivatives of Fmoc-L-azidohomoalanine. Mostly, the protected Fmoc-L-glutamine is converted under Hofmann rearrangement conditions [41,[52][53][54]] to 2-Fmoc-4-aminobutanoic acid, followed by the azido transfer reaction [12,41,46]. The other approach involves protected L-aspartic acid, which is partially reduced via a mixed anhydride. The resulting alcohol is mesylated, and the corresponding mesyl derivative is replaced by the azido function [55].
The strategy of orthogonal functional group protection [56][57][58][59] of L-ornithine and L-lysine is usually used for the synthesis of their azido derivatives.
Here, we compared several of these reaction pathways and also investigated a few new routes for the preparation of the abovelisted Fmoc-protected azido amino acids in a multigram scale, including their incorporation into short model peptides. The advantages and drawbacks of the approaches are discussed.

Results and Discussion
Synthesis of β-Azido L-Alanine and D-Alanine Firstly, L-serine 1 was chemoselectively protected by the benzyloxycarbonyl (Z) group, and Z-Ser 3 was then esterified with tert-butyl bromide in the presence of a benzyltriethylammonium chloride phase catalyst and excess of potassium carbonate [60,61] (Scheme 1). These two transformations led to the intermediate 4.
Next, starting from D-serine 2 and through intermediates 9 and 12, we employed the synthetic pathway A and the following reactions for the preparation of Fmoc-β-azido-D-Ala 15.
Johansson and Pedersen [65] and others [35,66] claimed that dehydroalanine 10 is a perfect Michael acceptor, which undergoes racemization. To verify this statement, we carried out a simple experiment; the isolated dehydroalanine 10 was heated in DMSO with an excess of NaN 3 at 70°C overnight. However, no traces of the expected product 17 were found.
Next, we verified the optical purity of the compound 14 by NMR spectroscopy. For this, the dipeptide 18 was prepared by the reaction of 14 with L-Val-OBn·TsOH. The incorporation of valine added a new stereo center to the molecule. However, the presence of the sterically bulky Fmoc group resulted in the observation of geometrical isomers of the carbamate in a ratio of 60 : 40. Therefore, we had to prepare a fully deprotected molecule. Compound 14 was thus coupled with ester 20, which was prepared by esterification of L-valine 19 using tert-butyl acetate [67] as a source of (CH 3 ) 3 C + cation. Acidic hydrolysis gave free acid, which was attached to 2chlorotrityl chloride resin. This allowed the removal of the Fmoc protecting group by conveniently washing off poorly separable dibenzofulvene. A usual work-up and separation furnished dipeptide 22, which was manifested by only one set of signals in both 1 H and 13 C NMR spectra. The same protocol with the chlorotrityl resin was used for the preparation of free amino acid 16. The optical purity of water-soluble acid 16 was checked by the method of Inamoto et al. [68]. No splitting of signals on α-carbons and α-protons was observed after the addition of sodium [(S)-1,2diaminopropane-N,N,N 0 ,N 0 -tetraacetato]-samarate(III) [69] to a pHadjusted solution of 16 in a 2 : 1 molar ratio. In conclusion, NMR analyses unequivocally proved the high optical purity of compound 14.
We have not performed the same procedure with the optical isomer 15, which was prepared by the same reactions, but from the pure Dserine. However, the same high optical purity can be expected.
We also investigated two alternative reaction routes (Scheme 2) for the synthesis of Fmoc-β-azido-L-Ala 14 as outlined in Scheme 2. Firstly, L-asparagine 23 was protected by the Fmoc group; then carboxamide functionality of the resulting intermediate 25 was eliminated under Hoffmann rearrangement conditions. The last step of this reaction pathway B was a diazotransfer reaction, which allowed the conversion of the amino group to the corresponding azido acid 14. In parallel, we demonstrated that 14 can be obtained with a similar synthetic strategy, but using Boc protection and starting from L-asparagine 23 over intermediates 31 and 32 (reaction pathway C).
Compound 14, which was prepared using three different synthetic pathways (A, B or C), provided the same mass and NMR Scheme 2. Reagents, conditions and yields: (a) Na 2 CO 3 , Fmoc-OSu, dioxane and water, 0°C for 1 h, then RT overnight (87% for 25, 93% for 26); (b) PhI(OAc) 2 , CH 3 CN, ethyl acetate and water at RT overnight (75% for 27, 56% for 28); (c) TfN 3 , NaHCO 3 , CuSO 4 ·5H 2 O, water and methanol at RT overnight (80% for 14, 92% for 29); (d) Na 2 CO 3 , Boc 2 O, dioxane and water 0°C 1 h then RT overnight (73%); (e) PhI(OAc) 2 , CH 3 CN, ethyl acetate and water at RT overnight (75%); (f) TfN 3 , TEA, CuSO 4 ·5H 2 O, water and methanol at RT overnight; (g) TFA, DCM, 2 h at RT; (h) NaHCO 3 , Fmoc-OSu, dioxane and water, 0°C for 1 h, then RT overnight (37% over three steps). spectra and other physicochemical characteristics. Therefore, the number of synthetic steps, cumulative yields and costs are decisive for the choice of the optimal strategy. Pathway A includes six steps and gave 29% yield, pathway B was performed in three steps and with 52% yield and pathway C required five steps and gave 20% yield. Clearly, from this aspect, the preferred synthetic route is pathway B and the less convenient is pathway C. We also calculated the approximate costs of synthetic pathways A and B for the preparation of 14, using precursor and solvents purchased at standard prices from Fluka. Synthetic pathway A yielded 1 g of 14 for about €37 and pathway B for about €43. Pathway A can be completed within a week; pathway B is faster. Taking everything together, the method of choice for the preparation of 14 is pathway B, despite the fact that its diazotransfer reaction requires the use of an excess of rather costly trifluoromethanesulfonic anhydride (triflic anhydride).

Synthesis of β-Azido L-Homoalanine
Next, for the preparation of L-homoazidoalanine, we chose the straightforward pathway B (Scheme 2), starting from L-glutamine over intermediates 26 and 28. The product 29 was obtained in an excellent yield of 92%.
Synthesis of δ-Azido L-Ornithine and ω-Azido L-Lysine For the synthesis of azido acids 41 and 42, derivatives of L-ornithine and L-lysine, we applied the strategy of orthogonal functional group protection [56][57][58][59] (Scheme 3). Synthesis started from commercially available L-ornithine·HCl 33 or L-lysine·HCl 34, and the reaction with copper acetate monohydrate under basic conditions afforded [Orn (Boc)] 2 Cu 35 or [Lys(Boc)] 2 Cu 36, respectively, which were isolated by a perfect filtering off of their insoluble copper complexes. Metal was quantitatively removed using 8-quinolinol to furnish selectively Boc-protected intermediates 37 and 38 in forms of zwitterions. The alpha amino group was acylated with Fmoc-OSu, and the resulting diamino acids 39 and 40 were treated with TFA to liberate δ-free amine and ω-free amine, respectively. The final step is represented by a diazotransfer reaction, which leads to the required Fmocazido-L-norvaline 41 or Fmoc-azido-L-lysine 42.

Synthesis of Model Azido Tripeptides
Finally, we synthesized a series of model tripeptides 43-47, using the standard manual Fmoc solid-phase synthesis protocol [70] (Scheme 4). When the methodology was precisely followed, all required peptides were obtained in good yields and with a high Scheme 3. Reagents, conditions and yields: (a) CuAc 2 ·H 2 O, water, then Boc 2 O in acetone overnight (90% for 35, 92% for 36); (b) 8-hydroxyquinoline, water, 4 h (90% for 37, 88% for 38); (c) Fmoc-OSu, NaHCO 3 , water and dioxane, 0°C for 0.5 h, then RT overnight (93% for 39, 95% for 40); (d) TFA, DCM, 2 h at RT (e) TfN 3 , NaHCO 3 , CuSO 4 ·5H 2 O, water and methanol at RT overnight (92% for 41 over two steps, 89% for 42 over two steps). chemical purity. However, surprisingly, during the synthesis of 44 and only in the case of prolonged condensations (5 and 18 h) of Fmoc-azidoalanine 14 with the resin-bound Phe-Phe-NH 2 , we observed the massive appearance of a new compound representing a major product of the synthesis ( Figure 1). This product was isolated, and its chemical structure assigned using spectral methods and attributed to the compound 48. It appears that only alpha-keto functionality is present in 48 instead of methylazido and acetylamino moieties. We have not found any information in the available literature about this type of elimination of azido species. Some studies (e.g., [71]) reported on obtaining carbonyl compounds from α,β-disubstituded compounds. However, here, we can only speculate that the shorter side chain of 14 (i.e., the proximity of the azido moiety and the primary α-amino group) and/or possibly also the activating agent 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU) in the presence of DIPEA (7087-68-5) and/or longer couplings or treatment with 95% TFA/5% water may play a role in this side reaction. Interestingly, in addition, it seems that this elimination is also sequence specific, as it occurred only in the case of the coupling of 14 to Phe-Phe dipeptide and not to other sequences. Hence, at this stage, we rather do not suggest any plausible reaction mechanism for this process. The susceptibility of 14 to the elimination of its azido moiety during the synthesis of a simple tripeptide suggests that the use of azido amino acids in the peptide synthesis is not without risks and that some precautions should be taken (e.g., shorter reaction times and alternative reagents)

Conclusions
In conclusion, we investigated several synthetic protocols for the preparation of L-Fmoc-β-azidoalanine and D-Fmoc-β-azidoalanine (14 and 15, respectively). We found that pathway B starting from asparagine is the most straightforward one and it can also be used for the preparation of γ-azido L-homoalanine when starting from glutamine. NMR analysis confirmed the high optical purity of 14 prepared with these protocols. We also synthesized L-Fmoc-γazidohomoalanine 29, L-Fmoc-δ-azidoornitine 41 and L-Fmoc-εazidolysine 42. Several synthetic steps previously described in literature were improved and optimized, and several new reactions investigated. All synthetic procedures were described in detail, and the complete physicochemical characterization of all intermediates and final compounds was provided. We found that multigram quantities of these Fmoc-protected azido amino acids can be prepared within a week, at average costs of about €40 per gram of final compounds (excluding work and energies). This makes them incomparably cheaper than standard commercial counterparts. We also observed a new type of elimination which occurred during prolonged couplings upon the solid-phase synthesis of Ac-β-azido-Ala-Phe-Phe-NH 2 .

Experimental Part
Reagents and solvents (Sigma-Aldrich-Fluka, St. Louis, MO, USA) used in this study were of analytical grade. TLC analyses were performed on silica-gel-coated aluminum plates (Fluka). The compounds were visualized by exposure to UV light at 254 nm and by ninhydrin spraying (dark blue color) of Boc-protected amines or amines. Flash chromatography purifications were carried out on silica gel (40-63 μm, Fluka). Preparative RP-HPLC chromatography was carried out on a C18 Luna column (Phenomenex, Torrance, CA, USA, 250 × 21.2 mm, 10 μm) at a flow rate 9 ml/min (solvent A: 0.1% TFA; solvent B: 80% CH 3 CN, 0.1% TFA). Eluted compounds were detected at 218 and 254 nm and lyophilized from water. Melting points were determined on a Boetius block and are uncorrected. 1 H and 13 C NMR spectra were measured on a Bruker AVANCE-600 spectrometer (Billerica, MA, USA; 1 H at 600.13 MHz, 13 C at 150.9 MHz) in CDCl 3 , DMSO-d 6 , CD 3 OD or D 2 O solution at 300 K. The 2D-H,H-COSY, 2D-H,C-HSQC and 2D-H,C-HMBC spectra were recorded and used for the structural assignment of proton and carbon signals. IR spectra were recorded on Bruker IFS 55 Equinox apparatus. HRMS spectra were obtained on an FTMS mass spectrometer LTQ-orbitrap XL (Thermo Fisher, Bremen, Germany) in electrospray ionization mode.
Experimental procedures and analytical data for compounds 3-5 are provided in the supporting information.
tert-Butyl 2-(S)-(9-fluorenylmethyloxycarbonylamino)-3bromopropanoate 6 Ester 5 (14.6 g; 38.1 mmol) and CBr 4 (15.2 g; 45.7 mmol) were dissolved in 100 ml of DCM (CAS 75-09-2). The flask with the reaction mixture was immersed in an ice cooling bath, and PPh 3 (12 g; 45.7 mmol) in 100 ml DCM was added dropwise under stirring. Stirring at 0°C was continued for 1 h and then at room temperature (RT) overnight. DCM was removed under reduced pressure, and a brown residue was purified by flash chromatography on silica gel, using a linear gradient of ethyl acetate in petroleum ether. The product was a colorless oil, which was triturated in petroleum ether at À20°C.

tert-Butyl 2-(S)-(tert-butoxycarbonylamino)-3hydroxypropanoate 8
Z-L-Ser-OtBu 4 (12.8 g; 43.3 mmol) was put into a glass pressure bottle and dissolved in 300 ml of methanol, and then 500 mg of 10% Pd/C was added. The mixture was vigorously stirred and allowed to react under the atmosphere of hydrogen (15 psi) at RT overnight. TLC analysis revealed (toluene-ethyl acetate 50 : 50) that the starting compound had completely disappeared. The catalyst was filtered off through celite, and the filter cake was washed in 300 ml of methanol. The filtrate was evaporated in vacuo to give 7.2 g of light brown residue, which was immediately dissolved in a 100 ml saturated solution of NaHCO 3 . The flask was placed into the ice bath, and di-tert-butyl dicarbonate (Boc 2 O) (9.5 g; 43.3 mmol) in 100 ml dioxane was added dropwise under stirring. After the addition of all Boc 2 O, stirring was continued for 1 h at 0°C and then overnight at RT. Thereafter, 500 ml of water was poured in, and the reaction mixture was transferred to the separatory funnel and extracted four times with 150 ml of ethyl acetate. Combined organic layers were washed consecutively twice with 100 ml of water and twice with 100 ml of brine and dried over anhydrous Na 2 SO 4 . Evaporation of the filtrate under reduced pressure furnished a colorless oil, which was triturated in petroleum ether at À20°C: colorless solid, yield 9.7 g (86% over two steps). The spectra and physicochemical characteristics of the product are the same as those for 3 prepared by a direct alkylation of Boc-L-Ser.
Alternatively, L-serine 1 (10.5 g; 0.1 mol; [α] 20 D = +13, c = 5, 5 M HCl) was placed into a 1 l, round-bottom flask equipped with a magnetic spin bar. The compound 1 was dissolved in the solution of sodium hydrogen carbonate (16.8 g; 0.2 mol) in 150 ml of water. The flask was immersed into the ice cooling bath, and 1.1 eq. of Boc 2 O (24 g; 0.11 mol) in 100 ml of dioxane was added dropwise under vigorous stirring for 30 min. When the addition of Boc anhydride was completed, the reaction mixture was allowed to react for 1 h at 0°C and then overnight at RT. A saturated aqueous solution of citric acid was added carefully until acidic pH ≈ 3 was reached. The aqueous-organic solution was saturated with sodium chloride, followed by four extractions with ethyl acetate (each with 200 ml). Combined organic phases were washed four times with 100 ml of brine and dried over Na 2 SO 4 . The filtration of the drying agent, followed by evaporation of the filtrate under reduced pressure, furnished 25 g of the crude [(S)-2-(tert-butoxycarbonylamino)]-3hydroxypropanoic acid as a colorless oil (R f = 0.46 ethyl acetate-acetone-methanol-water 6 : 1: 1 : 0.5), which was used in the next step without additional purification.
Compound 8 was prepared in the same manner as 4 by the reaction of 25 g of Boc-L-Ser, 200 ml of tert-butyl bromide, potassium carbonate (69.1 g; 0.5 mol) and benzyl triethylammonium chloride (CAS 56-37-1, 11.4 g; 0.05 mol) in 450 ml of dimethylacetamide. The required product was a clear oil, which solidified upon standing in a refrigerator at 5°C. An analytical sample was prepared by trituration in petroleum ether at À20°C. tert-Butyl 2-(R)-(tert-butoxycarbonylamino)-3hydroxypropanoate 9 With the method described for 8, the title enantiomer 9 was prepared from D-serine (10. The combined organic layers were washed twice with 100 ml of water and twice with 100 ml brine and dried over Na 2 SO 4 . The drying agent was filtered off, and the filtrate was evaporated under reduced pressure to afford 20 g of yellow residue, which was purified by flash chromatography on silica gel, using a linear gradient of diethyl ether in petroleum ether. Two major compounds were isolated, the product of elimination 10 and the required azide 11. Yield 10 2.7 g (16% over two steps). Yield 11 9.7 g (47% over two steps).
Alternatively, 13 (9.7 g; 29.9 mmol) was dissolved in 50 ml of DMSO with NaN 3 (3.9 g; 59.8 mmol) and heated at 70°C overnight. After cooling, 100 ml of water was added, and the solution was extracted four times with 50 ml of diethyl ether. The combined organic layers were washed twice with 50 ml of water and twice with 50 ml of brine and dried over Na 2 SO 4 . The drying agent was filtered off, and the filtrate was evaporated under reduced pressure to afford 8.6 g of dark yellow residue, which was purified by flash chromatography on silica gel, using a linear gradient of diethyl ether in petroleum ether. Yield 10 2.9 g (40%), yield 11 3.5 g (41%).

tert-Butyl 2-(S)-(tert-butoxycarbonylamino)-3bromopropanoate 13
Ester 8 (11.5 g; 44 mmol) and CBr 4 (17.5 g; 52.8 mmol) were dissolved in 100 ml of DCM. The flask with the reaction mixture was immersed in the ice cooling bath, and PPh 3 (13.8 g; 52.8 mmol) in 50 ml of DCM was added dropwise under stirring. Stirring was continued for 1 h at 0°C and then at RT overnight. DCM was removed under reduced pressure, and the brown residue was purified by flash chromatography on silica gel, using a linear gradient of ethyl acetate in petroleum ether. The product was a colorless oil, which was triturated in petroleum ether at À20°C. 2-(S)-(9-Fluorenylmethyloxycarbonylamino)-3-azidopropanoic acid (Fmoc-β-azido-L-Ala) 14 Compound 11 (9.6 g; 33.5 mmol) was treated with the cleavage cocktail, consisting of 18.8 ml of DCM, 18.8 ml of TFA and 2.4 ml of water. The reaction started with severe liberation of CO 2 and isobutene and continued at RT overnight under stirring. Volatile materials were removed on the rotary evaporator. The yellow residue was dissolved in a solution of NaHCO 3 (11.3 g; 134 mmol) in 50 ml of water. The reaction mixture was cooled in the ice bath, and Fmoc-OSu (11.3 g; 33.5 mmol) in 50 ml dioxane was added dropwise under vigorous stirring. The reaction mixture was allowed to react for 1 h at 0°C and then overnight at RT. The flask was again ice cooled, and concentrated hydrochloric acid was carefully added until acidic pH ≈ 1 was reached. The reaction mixture was extracted thrice with 100 ml of ethyl acetate. Thereafter, the combined organic layers were successively washed once with 100 ml of water and twice with 100 ml of brine, followed by drying on Na 2 SO 4 . Evaporation of the filtrate gave a brown oil, which was purified by flash chromatography on silica gel, using a linear gradient of 1% CH 3 COOH/ethyl acetate in toluene. Evaporation of the product afforded a yellow semisolid, which was triturated in toluene at À20°C. Alternatively, compound 14 was prepared from 27. Triflic anhydride (13.3 g; 47.2 mmol) was added dropwise under ice cooling and vigorous stirring to the two-phase system of NaN 3 (15.3 g; 236 mmol) in 60 ml of water and 70 ml of DCM. The ice bath was removed and stirring continued for 2 h. The aqueous layer was separated and extracted twice with 50 ml of DCM. Thereafter, the combined organic phases were washed with 5% NaHCO 3 . The resulting solution of triflic azide (TfN 3 ) in DCM was immediately added dropwise to the suspension of 27 (7.7 g; 23.6 mmol), NaHCO 3 (19.8 g; 236 mmol) and CuSO 4 ·5H 2 O (60 mg; 23.6 μmol) in 50 ml of water and 150 ml of methanol, and the mixture was stirred at RT overnight. Volatile material was evaporated, and the remaining slurry was carefully acidified with concentrated HCl until pH 1-2 was reached. The reaction mixture was extracted four times with 100 ml of ethyl acetate. The combined organic layers were washed twice with 100 ml of water and twice with 100 ml of brine and dried over Na 2 SO 4 . Filtering off the drying agent and evaporation of the filtrate gave 8 g of brown residue, which was purified by flash chromatography on silica gel, using a linear gradient of 1% CH 3 COOH/ethyl acetate in toluene. The yellow oil was triturated in toluene at À20°C to give the pure product. Yield 6.6 g (80%). Physicochemical characteristics were consistent with the abovelisted ones.
Alternatively, compound 14 was prepared from 32. Compound 32 (8.4 g; 36.5 mmol) was treated with a mixture of 18.8 ml of DCM, 18.8 ml of TFA and 2.4 ml of water. After 2 h of stirring, volatile materials were evaporated, and the yellow oil was dissolved in 50 ml of water with NaHCO 3 (9.2 g; 109.5 mmol). The flask was immersed in an ice cooling bath, and Fmoc-OSu (12.3 g; 36.5 mol) in 100 ml of dioxane was added dropwise during a period 15 min under vigorous stirring. When the addition of Fmoc-OSu was complete, the slurry was allowed to react for 1 h at 0°C and then overnight at RT. The reaction mixture was again cooled in an ice bath, and concentrated HCl was added dropwise until pH~0-1 was reached. Thereafter, 150 ml of water was added, and the reaction mixture was extracted thrice with 100 ml of ethyl acetate. The combined organic layers were washed once with 100 ml of water and twice with 100 ml of brine and dried over Na 2 SO 4 . The filtrate was evaporated, and the resulting brown oil was subjected to flash chromatography on silica gel, using a linear gradient of 1% CH 3 COOH/ethyl acetate in toluene. The yellow oil was triturated in toluene at À20°C to afford the pure product. Yield 8.2 g (64%). Physicochemical characteristics were consistent with the abovelisted ones.

2-(S)-Amino-3-azidopropanoic Acid Trifluoroacetic Acid Salt 16
One gram of 2-Cl-Trt-chloride resin (Merck Novabiochem, Darmstadt, Germany, capacity 1.5 mmol/g, 100-200 mesh) was placed in a 20 ml syringe with a frit and preswollen in 10 ml DMF for half an hour. The solvent was removed, and 14 (0.575 g; 1.5 mmol) in 4 ml of DMF and DIPEA (783 μl; 4.5 mmol) in 2 ml of DMF were added. The syringe was agitated by shaking for 1.5 h, followed by washing (3× 10 ml of DMF). The reaction was terminated by two subsequent additions of a mixture of 5.1 ml of DCM, 0.6 ml of CH 3 OH and 0.3 ml of DIPEA, each for 5 min. The resin was washed thrice with 10 ml of DCM and thrice with 10 ml of DMF. The Fmoc group was cleaved with 20% (v/v) piperidine in DMF (10 ml for 5 and 30 min). The resin was washed thrice with 10 ml of DMF and thrice with 10 ml of DCM. Finally, the product was cleaved from the resin by three subsequent treatments with a mixture of 2 ml of AcOH, 2 ml of trifluoroethanol, CAS 75-89-8, and 6 ml of DCM, each for 15 min. Filtrates were evaporated to dryness, and the crude material was subjected to RP-HPLC. The following gradient was used: Experimental procedures and analytical data for compound 20 are provided in the supporting information.

Fmoc-β-azido-Ala-Val-OtBu 21
Fmoc-β-azido-Ala 14 (0.652 g; 2 mmol), L-Val-OtBu·TsOH 20 (0.419 g; 2 mmol), bromotripyrrolidinophosphonium hexafluorophosphate (1.17 g; 2.5 mmol) and DIPEA (1 g; 8 mmol) in 10 ml of DMF were stirred overnight at RT. The solvent was evaporated under reduced pressure, and the residue was purified by flash chromatography on silica gel, using a linear gradient of ethyl acetate in toluene. An analytical sample was gained by crystallization from a mixture of ethyl acetate-petroleum ether. Yield 700 mg (70%). Compound 21 (600 mg; 1.2 mmol) was treated with a mixture consisting of 2 ml of DCM, 2.5 ml of TFA and 40 μl of water. After 2 h, TLC analysis revealed completely deprotected tert-butyl moiety, and the volatile material was evaporated under reduced pressure. 2-Cl-Trt-chloride resin (0.8 g, Merck Novabiochem, capacity 1.5 mmol/g, 100-200 mesh) was placed into a 20 ml syringe with a frit and preswollen in 10 ml of DMF for half an hour. DMF was removed, and the crude acid Fmoc-β-azido-Ala-Val in 4 ml DMF and DIPEA (627 μl; 3.6 mmol) in 2 ml DMF were added. The syringe was agitated by shaking for 1.5 h, followed by washing thrice with 10 ml of DMF. In the next step, the Fmoc group was cleaved with 20% (v/v) piperidine in DMF (10 ml for 5 and 30 min). The resin was washed thrice with 10 ml of DMF and thrice with 10 ml of DCM. Finally, the product was cleaved from resin by three subsequent treatments with 10 ml of AcOH, 15 min each. Filtrates were combined and evaporated to dryness. The residue was sonicated for 10 min in 15 ml of diethyl ether. The white precipitate was decanted and dissolved in 2 ml of hot acetonitrile (60°C). After cooling, the crystals were filtered off and washed with 5 ml of diethyl ether. Yield 181 mg (67%). White solid, m.p. 187-189°C.

2-(S)-(tert-Butoxycarbonylamino)-3-aminopropionic Acid 31
Intermediate 30 (18.9 g; 81.4 mmol) was suspended in a mixture of acetonitrile (90 ml), 90 ml ethyl acetate (90 ml) and water (45 ml), and PIDA (31.4 g; 97.7 mmol) was added in five portions during 15 min. Then, 10 min after the addition of all amount of PIDA, the slurry turned clear followed by a rapid precipitation of the crude product. The cake of the filtrate was washed out with 200 ml of chilled ethyl acetate, and no additional purification was needed.

2-(S)-(tert-Butoxycarbonylamino)-3-azidopropionic Acid 32
With the previously described azido transfer reaction employed for 11, the reaction of 31 (12.7 g; 62.2 mmol), TEA (18.9 g; 186.6 mmol) and CuSO 4 ·5H 2 O (155 mg; 0.622 mmol) in a mixture of 100 ml of methanol and 50 ml of water with TfN 3 in dichloromethane gave 32. TfN 3 was prepared by the reaction of NaN 3 (40.4 g; 622 mmol) and triflic anhydride (35.1 g; 124.4 mmol) in 100 ml of water and 100 ml of DCM. A bright yellow oil (9 g) was obtained after the isolation by flash chromatography on silica gel and was used as a crude product in the following step.
Copper(II) Complex of N δ -tert-Butoxycarbonyl-L-ornithine 35  Compound 37 (5.9 g; 25.4 mmol) was placed in a 1 l round-bottom flask, equipped with a magnetic spin bar and dissolved in a solution of NaHCO 3 (4.3 g; 50.8 mmol) in 100 ml of water. The flask was immersed in an ice cooling bath and Fmoc-OSu (8.6 g; 25.4 mmol) in 100 ml of dioxane was added dropwise under vigorous stirring during 30 min. When the addition of Fmoc-OSu was complete, the reaction mixture was allowed to react for 1 h at 0°C and then overnight at RT. Thereafter, 200 ml of water was added and followed by the dropwise addition of concentrated citric acid until pH~2-3 was reached. The reaction mixture was extracted four times with 150 ml of ethyl acetate. The combined organic layers were washed twice with 150 ml of brine and twice with 150 ml of water and dried over Na 2 SO 4 . The filtrate was evaporated and the resulting brown oil was subjected to flash chromatography on silica gel, using a linear gradient ethyl acetate in ethyl acetate-acetonemethanol-water 6 : 1 : 1 : 0.5. The yellow oil was triturated in a mixture of ethyl acetate-petroleum ether at À20°C to afford the pure product. (CAS 6485-79-6) (95/2.5/2.5). The cleavage step was repeated under the same conditions, the resin was washed with 10 ml of glacial acetic acid and all cleavage solutions and acetic acid were combined and evaporated under reduced pressure. The brown residue was then sonicated for 10 min with 10 ml diethyl ether in an ice cooling bath. The slurry was centrifuged for 10 min at 10 000 g, diethyl ether was decanted and crude amorphous white tripeptides were dried in vacuo. The purity of the prepared tripeptides was checked by RP-HPLC; analytical samples were isolated using the following gradient: t = 0 min (20% B), t = 30 min (100% B).