Total Synthesis of Hemerocallisamine I Paved by Gram-Scale Synthesis of (2S,4S)-4-Hydroxyglutamic Acid Lactone

Total synthesis of the 2-formylpyrrole alkaloid hemerocallisamine I is presented, both in racemic and enantiopure form. Our synthetic strategy involves (2S,4S)-4-hydroxyglutamic acid lactone as the key intermediate. Starting from an achiral substrate, the target stereogenic centers were introduced by means of crystallization-induced diastereomer transformation (CIDT) in a highly stereoselective fashion. A Maillard-type condensation was crucial to constructing the desired pyrrolic scaffold.


Introduction
Daylilies (genus Hemerocallis) are beautiful flowering plants, now counting 16 species and over 98,000 cultivars, widely domesticated in much of the Northern Hemisphere [1,2]. Mentioned in poems compiled by Confucius more than 2500 years ago, in addition to ornamental function, they have a long history of culinary [3] and medicinal use in Eastern Asia. The flowers, buds, leaves, and roots of daylilies are edible and have been reported to be utilized, e.g., in the treatment of sleep disorders, depression, inflammation, jaundice, schistosomiasis, and chronic rheumatism [4]. In an attempt to elucidate the structures of the active ingredients, the chemical constitution of daylilies has been investigated in a number of studies. As of 31 December 2020, a total of 266 secondary metabolites have been identified in Hemerocallis plants, primarily focusing on species H. citrina, H. fulva, and H. minor [4,5].
In a search to identify sedative amino acid derivatives from H. fulva flower buds, in 2014, Matsuda et al. isolated hemerocallisamine I (1) as a novel 4-hydroxyglutamine metabolite [6] (Figure 1). The initially assigned (2R,4R)-1 configuration, on the basis of the Flack parameter, corresponded to the unnatural D-glutamine and was puzzling. Three years later, the first total synthesis of hemerocallisamine I (1) was reported by Brimble et al. and resulted in a revision of the previously proposed absolute configuration from (2R,4R)-1 to (2S,4S)-1 [7]. Until now, no other total synthesis of hemerocallisamine I (1) has been communicated.
Herein we describe an alternative synthetic access to hemerocallisamine I (1), star from the achiral substrate 6 (Scheme 1B). Given the importance of the 4-hydroxygluta motif in the chemistry of daylilies, we opted for facile stereoselective access to lacton as a key intermediate.
Herein we describe an alternative synthetic access to hemerocallisamine I (1), starting from the achiral substrate 6 (Scheme 1B). Given the importance of the 4-hydroxyglutamic motif in the chemistry of daylilies, we opted for facile stereoselective access to lactone 7 as a key intermediate.
Herein we describe an alternative synthetic access to hemerocallisamine I (1), starting from the achiral substrate 6 (Scheme 1B). Given the importance of the 4-hydroxyglutamic motif in the chemistry of daylilies, we opted for facile stereoselective access to lactone 7 as a key intermediate.

Synthesis of (±)-hemerocallisamine I and Optimization of the Maillard Reaction Step
Considering that the synthesis of racemic hemerocallisamine I remains undescribed and that this material might be of value, e.g., for biological activity investigation, our synthetic approach was first evaluated in its achiral variant.
The studies commenced with a routine construction of amino acid 9 by means of aza-Michael addition (Scheme 2). Shortly after mixing an excess of aq. NH 3 with aroylacrylic substrate 6 in MeOH, adduct 9 started to precipitate out of the reaction mixture and was filtered off in 77% yield. A consequent reduction step delivered a diastereomeric mixture of the corresponding γ-aryl-γ-hydroxy-α-amino acids, in a ratio ca 2:1. However, our one-pot CIDT protocol exploited configurational lability of the benzylic γ-hydroxy group in dilute aqueous HCl [28,29] and provided crystalline cis-lactone salt rac-10·HCl and then free amine rac-10 in excellent diastereomeric purity (>99:1 dr). After its N-Boc protection, the adjacent electron-rich anisole ring was oxidatively cleaved employing a catalytic amount of in situ prepared RuO 4 , with NaIO 4 as a reoxidant [30,31]. Our modified procedure furnished the 4-hydroxyglutamic acid lactone rac-7 in 71% yield, over two steps. When approaching the critical Maillard reaction step, we were inspired by the studies of Brimble et al. [7] and chose to introduce the p-methoxybenzyl (PMB) protected amide group. The coupling with PMBNH 2 was mediated by ethyl chloroformate and a subsequent Boc-removal provided amine rac-11 as a synthetic alternative to amine 3 (Scheme 1A). Pleasingly, each of the steps in the 6-11 sequence delivered crystalline compounds and are convenient to purify, isolate, and store. hydroxylations of L-glutamic substrates [21][22][23], and cycloadditions [24,25]. Since the number of synthetic methods available for the preparation of enantiopure 4-hydroxyglutamates is rather limited [26], we wished to develop a new, scalable, and configurationally flexible protocol based on crystallization-induced diastereomer transformations (CIDT) [27].

Synthesis of (±)-hemerocallisamine I and Optimization of the Maillard Reaction Step
Considering that the synthesis of racemic hemerocallisamine I remains undescribed and that this material might be of value, e.g., for biological activity investigation, our synthetic approach was first evaluated in its achiral variant.
The studies commenced with a routine construction of amino acid 9 by means of aza-Michael addition (Scheme 2). Shortly after mixing an excess of aq. NH3 with aroylacrylic substrate 6 in MeOH, adduct 9 started to precipitate out of the reaction mixture and was filtered off in 77% yield. A consequent reduction step delivered a diastereomeric mixture of the corresponding γ-aryl-γ-hydroxy-α-amino acids, in a ratio ca 2:1. However, our onepot CIDT protocol exploited configurational lability of the benzylic γ-hydroxy group in dilute aqueous HCl [28,29] and provided crystalline cis-lactone salt rac-10•HCl and then free amine rac-10 in excellent diastereomeric purity (>99:1 dr). After its N-Boc protection, the adjacent electron-rich anisole ring was oxidatively cleaved employing a catalytic amount of in situ prepared RuO4, with NaIO4 as a reoxidant [30,31]. Our modified procedure furnished the 4-hydroxyglutamic acid lactone rac-7 in 71% yield, over two steps. When approaching the critical Maillard reaction step, we were inspired by the studies of Brimble et al. [7] and chose to introduce the p-methoxybenzyl (PMB) protected amide group. The coupling with PMBNH2 was mediated by ethyl chloroformate and a subsequent Boc-removal provided amine rac-11 as a synthetic alternative to amine 3 (Scheme 1A). Pleasingly, each of the steps in the 6-11 sequence delivered crystalline compounds and are convenient to purify, isolate, and store. The Maillard condensations targeting the 2-formylpyrrole framework have often proved challenging [9,10]. The reported Maillard protocol towards hemerocallisamine I harnessed an open glutamine chain 3 (Scheme 1A). We hypothesized that the application of lactone 11 instead might reduce the complexity of the amine building block and provide better yields. Seeing that dihydropyranone 4 was frequently reported as a suitable sugar surrogate in Maillard reactions [7,[32][33][34][35][36][37], it was picked as a reaction partner for our The Maillard condensations targeting the 2-formylpyrrole framework have often proved challenging [9,10]. The reported Maillard protocol towards hemerocallisamine I harnessed an open glutamine chain 3 (Scheme 1A). We hypothesized that the application of lactone 11 instead might reduce the complexity of the amine building block and provide better yields. Seeing that dihydropyranone 4 was frequently reported as a suitable sugar surrogate in Maillard reactions [7,[32][33][34][35][36][37], it was picked as a reaction partner for our optimization studies (Table 1). In the published procedures, amines were usually used in 1.5-to 4-fold excess for the sake of better yields [32,33,35,36]. In our case, considering that amine 11 is not an effortless building block, it was reasonable to screen reaction conditions with equimolar amounts of rac-11 and 4. Somewhat confusingly, the currently recorded op-timized protocols vary in the suggested reaction conditions from case to case. Pyridine [7], undistilled dioxane [32], THF/H 2 O [33][34][35], and MeCN [36] were used as solvents and Et 3 N [32], AcOH [35], and pyridinium p-toluenesulfonate [7,36] as additives, at reaction temperatures ranging from rt to 60 • C. Dihydropyranone 4 was occasionally observed to remain partially unreacted and was recovered in 8-35% yields [33,34]. Taking into account all these data, we performed a broader investigation (selected examples are in Table 1, for details, see the Supplementary Materials), screening diverse solvents and additives (entries 2-12). As found, simple heating of rac-11 and 4 in dry toluene at 70 • C gave the best outcome, and the target 2-formylpyrrole rac-8 was isolated in 43% yield (entry 1). Interestingly, 2,5-diketopiperazine rac-12 was repeatedly detected as a side product in the reaction mixtures, in yields of up to 35% with respect to amine rac-11 [38]. These findings provide an important argument when explaining why Maillard condensations so often fail to deliver better yields and should be reflected in future synthetic strategies. optimization studies (Table 1). In the published procedures, amines were usually used in 1.5-to 4-fold excess for the sake of better yields [32,33,35,36]. In our case, considering that amine 11 is not an effortless building block, it was reasonable to screen reaction conditions with equimolar amounts of rac-11 and 4. Somewhat confusingly, the currently recorded optimized protocols vary in the suggested reaction conditions from case to case. Pyridine [7], undistilled dioxane [32], THF/H2O [33][34][35], and MeCN [36] were used as solvents and Et3N [32], AcOH [35], and pyridinium p-toluenesulfonate [7,36] as additives, at reaction temperatures ranging from rt to 60 °C. Dihydropyranone 4 was occasionally observed to remain partially unreacted and was recovered in 8-35% yields [33,34]. Taking into account all these data, we performed a broader investigation (selected examples are in Table 1, for details, see the Supplementary Materials), screening diverse solvents and additives (entries 2-12). As found, simple heating of rac-11 and 4 in dry toluene at 70 °C gave the best outcome, and the target 2-formylpyrrole rac-8 was isolated in 43% yield (entry 1). Interestingly, 2,5-diketopiperazine rac-12 was repeatedly detected as a side product in the reaction mixtures, in yields of up to 35% with respect to amine rac-11 [38]. These findings provide an important argument when explaining why Maillard condensations so often fail to deliver better yields and should be reflected in future synthetic strategies. CH2Cl2, rt 33 10 +0.1 equiv rac-11, CH2Cl2, rt 37 11 AcOH (cat.), wet CH2Cl2, rt Having 2-formylpyrrole rac-8 in hand, the synthesis of racemic hemerocallisamine I (1) could proceed to the concluding steps (Scheme 3). Under conditions developed by Okada et al. [39] and later tailored by Brimble et al. [7], rac-8 was simultaneously desilylated and converted into a mixture of methyl ethers rac-13 and rac-14 in a ratio of 85:15. The crude product was directly used in a subsequent lactone opening and oxidative cleavage of the PMB group, thus providing (±)-hemerocallisamine I, (2S*,4S*)-1 (Scheme 3). Having 2-formylpyrrole rac-8 in hand, the synthesis of racemic hemerocallisamine I (1) could proceed to the concluding steps (Scheme 3). Under conditions developed by Okada et al. [39] and later tailored by Brimble et al. [7], rac-8 was simultaneously desilylated and converted into a mixture of methyl ethers rac-13 and rac-14 in a ratio of 85:15. The crude product was directly used in a subsequent lactone opening and oxidative cleavage of the PMB group, thus providing (±)-hemerocallisamine I, (2S*,4S*)-1 (Scheme 3).

Synthesis of (-)-hemerocallisamine I
With lessons learned from successfully completing the racemic sequence, we moved towards synthesizing the natural (-)-hemerocallisamine I. The initial steps were devoted to a stereoselective assembly of 4-hydroxyglutamic acid lactone (S,S)-7 (Scheme 4). Crystallization-induced diastereomer transformations (CIDT) built around aza-Michael additions have been proven to provide convenient access to diversely functionalized α-amino acids in a highly stereoselective fashion [27]. (S)-/(R)-Phenylethan-1-amine has frequently been reported as a chiral auxiliary. Herein we chose to slightly alter the original procedure [40] and utilize a suitably cleavable (S)-1-(4-methoxyphenyl)ethan-1-amine ((S)-MPEA) as a chiral amine carrier. CIDT with 1.1 equiv of (S)-MPEA was monitored by means of HPLC, proceeded over 5 days, and crystalline amino acid (S,S)-15 was collected in 96:4 dr. The ensuing TFA-mediated acidolytic N-debenzylation in the presence of a silane scavenger [41] smoothly provided crystalline salt of (S)-9. At this point, the synthetic sequence previously elaborated on the racemic substrate 9 (Scheme 2) came into play, and after the reaction steps c.-e., the 4-hydroxyglutamic acid lactone (S,S)-7 was

Synthesis of (-)-hemerocallisamine I
With lessons learned from successfully completing the racemic sequence, we moved towards synthesizing the natural (-)-hemerocallisamine I. The initial steps were devoted to a stereoselective assembly of 4-hydroxyglutamic acid lactone (S,S)-7 (Scheme 4).
Continuing with (S,S)-7 and following the verified procedures (Table 1, Scheme 3), (-)-hemerocallisamine I (S,S)-(1) was obtained, albeit in lower yields (Scheme 5). As confirmed both for the Maillard product (S,S)-8 and the target (S,S)-1 itself, the stereochemical integrity of these species remained uncompromised (chiral HPLC traces of (S,S)-8 and (S,S)-1 are available in the Supplementary Materials). The relative configuration of (S,S)-1 was confirmed by means of X-ray analysis # (#: Crystallographic data for (

General Experimental Details
Unless otherwise noted, all chemicals were purchased from commercial sources and used without further purifications. Column chromatography was carried out using silica 60 Å, Davisil, purchased from Fisher Chemicals. Reactions were monitored by thin-layer chromatography (TLC) using Macherey-Nagel's pre-coated TLC sheets POLYGRAM SIL G/UV254, which were visualized under UV light (254 nm) or by staining with aqueous basic potassium permanganate or cerium molybdate solutions, as appropriate. HPLC analyses were performed on a Varian system using a Macherey-Nagel EC 250/4 Nucleodur Phenyl-Hexyl 5 μm, CHIRAL ART, Amylose-SA, 250 mm × 4.6 mm, 5 μm and Astec CHIROBIOTIC ® T, 250 mm × 4.6 mm, 5 μm column. All 1 H and 13

General Experimental Details
Unless otherwise noted, all chemicals were purchased from commercial sources and used without further purifications. Column chromatography was carried out using silica 60 Å, Davisil, purchased from Fisher Chemicals. Reactions were monitored by thin-layer chromatography (TLC) using Macherey-Nagel's pre-coated TLC sheets POLYGRAM SIL G/UV254, which were visualized under UV light (254 nm) or by staining with aqueous basic potassium permanganate or cerium molybdate solutions, as appropriate. HPLC analyses were performed on a Varian system using a Macherey-Nagel EC 250/4 Nucleodur Phenyl-Hexyl 5 µm, CHIRAL ART, Amylose-SA, 250 mm × 4.6 mm, 5 µm and Astec CHIROBIOTIC ® T, 250 mm × 4.6 mm, 5 µm column. All 1 H and 13 C NMR (Supplementary Materials) spectra were recorded using Bruker Avance NEO 400 MHz and/or Varian 400 MR spectrometers. Chemical shifts (δ) are given in parts per million (ppm). The 1 H NMR chemical shift scale is referenced to the TMS internal standard (δ = 0 ppm) or solvent residual peak (δ = 2.50 ppm for DMSO-d 6 and δ = 7.26 ppm for CDCl 3 ). The 13 C NMR chemical shift scale is referenced to the solvent residual peak (δ = 39.52 ppm for DMSO-d 6 and δ = 77.16 ppm for CDCl 3 ). Coupling constants (J) are given in hertz (Hz). The multiplicity of 1 H NMR signals is reported as follows: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, bs = broad singlet, "t" for dd with two identical or Molecules 2023, 28, 2177 7 of 15 similar coupling constants, "dt" or "td" for ddd with two identical or similar coupling constants, and "q" for ddd with three identical or similar coupling constants. Optical rotations were recorded using the JASCO P-2000 polarimeter, with [α] D values measured at 589 nm and the concentration (c) given in g/100 mL. High-resolution mass spectra were measured using a Thermo Scientific mass spectrometer with an Orbitrap analyzer and HESI ionization.

(3S*,5S*)-3-Amino-5-(4-methoxyphenyl)dihydrofuran-2(3H)-one (rac-10)
Acid 9 (12.3 g, 55.1 mmol) was suspended in a mixture of MeOH (200 mL) and H 2 O (40 mL). NaBH 4 (7.30 g, 0.193 mol, 3.5 equiv) was added portionwise over 30 min at rt. The reaction progress was monitored by HPLC. After 1 h, the reaction mixture was concentrated in vacuo, providing the crude hydroxy acid. The product was suspended in H 2 O (120 mL), and conc. HCl (36%, 60 mL) was added. The reaction mixture was stirred at rt for 4 h. After completion of the reaction, the insoluble white precipitate was collected by filtration and washed with 1 M HCl (15 mL) and Et 2 O (2 × 30 mL). The white solid was then suspended in H 2 O (80 mL), and 10% aqueous K 2 CO 3 solution (100 mL) was added. The resulting mixture was extracted with CH 2 Cl 2 (3 × 150 mL). The combined organic layers were dried over anhydrous Na 2 SO 4 and concentrated in vacuo, yielding amino lactone rac-10 Step c: Lactone rac-10 (8.4 g, 40.5 mmol) was dissolved in THF (405 mL) and Et 3 N (12.4 mL, 89.2 mmol, 2.2 equiv) was added. Boc 2 O (9.73 g, 44.6 mmol, 1.1 equiv) was dissolved in a small amount of THF (15 mL) and the resulting solution was added to the first one. The reaction mixture was stirred at rt and monitored by TLC. After 4 h, the reaction mixture was cooled to 0 • C and the pH was adjusted to 2-3 with 2 M HCl, followed by extraction with EtOAc (3 × 140 mL). The combined organic layers were washed with brine (200 mL), dried over anhydrous Na 2 SO 4, and concentrated in vacuo. The crude product was crystallized from a mixture of EtOAc and heptane, yielding the corresponding N-Boc lactone (11.7 g, 38.1 mmol, 94%) as a white solid. Rf 0.69 (EtOAc); mp 175.6-176. Step d: The N-Boc lactone (4.0 g, 13.0 mmol) was dissolved in a mixture of CH 3 CN (65 mL) and EtOAc (65 mL). In parallel, NaIO 4 (54.1 g, 252.7 mmol, 19.4 equiv) was dissolved in H 2 O (184 mL) and both solutions were combined. Consequently, RuCl 3 (162 mg, 0.782 mmol, 0.06 equiv) was added and the reaction mixture was stirred at 10 • C in a water bath. After 1 h, the bath was removed, and the reaction mixture was stirred for an additional 2 h at rt. Then the resulting thick suspension was decanted, and the white residue was washed with EtOAc (4 × 100 mL). Et 2 O (50 mL) was added to the combined solutions and the resulting mixture was stirred for 30 min at rt. Afterwards, the resulting black suspension was filtered through a pad of Celite, and the pad was washed with EtOAc (3 × 50 mL). The combined filtrate was washed with 20% aqueous NaCl (3 × 100 mL) and dried over anhydrous Na 2 SO 4 . The filtrate volume was reduced in vacuo to ca 10 mL, and Et 2 O (30 mL) was added. The resulting suspension was placed in a freezer for 30 min. The white precipitate was collected by filtration, washed with a small amount of Et 2 O, and dried in vacuo, yielding the desired acid rac-7 (2.4 g, 9.79 mmol, 75%) as a white solid. Step e: Acid rac-7 (1.43 g, 5.83 mmol) was dissolved in dry THF (145 mL) at 0 • C and the solution was treated with Et 3 N (1.48 g, 2.0 mL, 14.6 mmol, 2.5 equiv). Ethyl chloroformate (1.30 g, 1.1 mL, 11.7 mmol, 2.0 equiv) was added dropwise at 0 • C, under argon. After 15 min, para-methoxybenzyl amine (2.20 g, 2.1 mL, 15.7 mmol, 2.7 equiv) was added and the reaction mixture was stirred at 0 • C for 2 h, under argon. After completion, the reaction mixture was diluted with EtOAc (100 mL) and washed with 1M HCl (50 mL). The organic layer was dried over anhydrous Na 2 SO 4 and concentrated under reduced pressure, yielding a pale-yellow solid. The crude product was treated with MeOH (5 mL), resulting in a white suspension that was cooled to 0 • C. The white precipitate was collected by filtration, washed with ice-cold MeOH, and dried in vacuo, yielding the corresponding amide (1. Step f: The amide (580 mg, 1.59 mmol) was dissolved in CH 2 Cl 2 (6.4 mL), the solution was cooled to 0 • C, and trifluoroacetic acid (1.92 g, 1.3 mL, 16.8 mmol, 10.6 equiv) was added. The reaction mixture was stirred at rt for 3 h while being monitored by TLC. The mixture was concentrated in vacuo, Et 2 O (10-15 mL) was added to the residue, and the mixture was kept in the ultrasonic bath for 30 min to form a white suspension. The white solid was filtered off, washed with a small amount of Et 2 O (5 mL), and dried in vacuo, providing rac-11·TFA (574 mg, 1.52 mmol, 95%) as an off-white solid. Step g: Salt rac-11·TFA (570 mg, 1.51 mmol) was suspended in CH 2 Cl 2 (15 mL), and the resulting suspension was washed with 10% aqueous NaHCO 3 (7 mL). The aqueous layer was extracted with CH 2 Cl 2 (2 × 10 mL). The combined organic layers were dried 3.2.7. (±)-Hemerocallisamine I, (2S*,4S*)-1 Step b: The mixture of rac-13 and rac-14 (85:15) (100 mg, 0.256 mmol for mole-fractionweighted M = 391.21 g.mol −1 ) was dissolved in dry MeOH (2.9 mL). MeONa (25w% solution in MeOH) (4.2 mg, 21 µL, 0.078 mmol, 0.3 equiv) was added to the solution at 0 • C, under an argon atmosphere. The reaction mixture was stirred at 0 • C and was monitored by TLC. After 30 min, a saturated aqueous NH 4 Cl solution (2 mL) was added at 0 • C and the resulting mixture was extracted with EtOAc (3 × 10 mL). The collected organic layers were washed with brine (10 mL), dried over anhydrous Na 2 SO 4 , and concentrated in vacuo, yielding rac-14 as a thick pale-yellow oil (108 mg, quantitative).
Step c: The crude methyl ester rac-14 (100 mg, 0.239 mmol) was dissolved in a mixture of CH 2 Cl 2 (2.3 mL) and a phosphate buffer (NaH 2 PO 4 , Na 2 HPO 4 ; pH 7, c 1M; 0.481 mL). Subsequently, 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (163 mg, 0.717 mmol, 3.0 equiv) was added to the solution. The resulting mixture was stirred at rt while the reaction progress was monitored by HPLC and TLC. After 24 h, the reaction mixture was diluted with CH 2 Cl 2 (5 mL), dried with Na 2 SO 4 , and filtered through a Celite pad. The pad was washed first with CH 2 Cl 2 (10 mL), then with dioxane (3 × 20 mL). The combined filtrates were concentrated in vacuo and purified using flash column chromatography (CH 2 Cl 2 :MeOH, 13:1). The desired product was isolated after trituration from MeOH-Et 2  The salt (S)-9·TFA (13.0 g, 38.5 mmol) was suspended in a mixture of MeOH (128 mL) and H 2 O (28 mL). NaBH 4 (5.1 g, 0.135 mol, 3.5 equiv) was added portionwise over 30 min to the suspension at rt. The reaction was monitored by HPLC. After 1 h, the reaction mixture was concentrated in vacuo, providing the crude hydroxy acid. The crude product was suspended in H 2 O (85 mL), and conc. HCl (36%, 45 mL) was added. The reaction mixture was stirred at room temperature for 4 h. After completion of the reaction, the insoluble white precipitate was collected by filtration and washed with 1M HCl (15 mL) and Et 2 O (2 × 30 mL). The white solid was then suspended in H 2 O (80 mL), and a 10% aqueous K 2 CO 3 solution (100 mL) was added. The resulting mixture was extracted with CH 2 Cl 2 (3 × 100 mL). The combined organic layers were dried over anhydrous Na 2 SO 4 and concentrated in vacuo, providing amino lactone (S,S)-10 (6. Step d: The lactone (S,S)-10 (5.5 g, 26.5 mmol) was dissolved in THF (265 mL), and Et 3 N (8.1 mL, 5.91 g, 58.4 mmol, 2.2 equiv) was added. Boc 2 O (6.37 g, 29.2 mmol, 1.1 equiv) was dissolved in a small amount of THF (10 mL) and the resulting solution was added to the first one. The reaction mixture was stirred at room temperature and was monitored by TLC. After 4 h, the reaction mixture was cooled to 0 • C and the pH was adjusted to 2-3 with 2M HCl, followed by extraction with EtOAc (3 × 90 mL). The combined organic layers were washed with brine (130 mL), dried over anhydrous Na 2 SO 4 , , and concentrated in vacuo, yielding crude N-Boc lactone as a pale-yellow solid. The crude product was crystallized from EtOAc-heptane, providing the N-Boc lactone (6.6 g, 21.5 mmol, 81%) as white crystals. Step e: The N-Boc lactone (4.0 g, 13.0 mmol) was dissolved in a mixture of CH 3 CN (65 mL) and EtOAc (65 mL). In parallel, NaIO 4 (54.1 g, 252.7 mmol, 19.4 equiv) was dissolved in H 2 O (184 mL) and both solutions were combined. Consequently, RuCl 3 (162 mg, 0.782 mmol, 0.06 equiv) was added, and the reaction mixture was stirred in a water bath (10 • C). After 1 h, the bath was removed, and the reaction mixture was stirred for an additional 2 h at rt. The thick suspension was then decanted, and the white residue was washed with EtOAc (4 × 100 mL). Et 2 O (70 mL) was added to the combined solutions, and the resulting mixture was stirred for 30 min at rt. Afterward, the resulting black suspension was filtered through the pad of Celite, and the pad was washed with EtOAc (3 × 50 mL). The combined filtrate was washed with a 20% aqueous NaCl solution (3 × 100 mL), dried over anhydrous Na 2 SO 4, and concentrated in vacuo to provide (S,S)-7 (2. Step a: Acid (S,S)-7 (1.23 g, 5.02 mmol) was dissolved in dry THF (125 mL) at 0 • C and the solution was treated with Et 3 N (1.27 g, 1.7 mL, 12.5 mmol, 2.5 equiv). Ethyl chloroformate (1.12 g, 1.0 mL, 10.0 mmol, 2.0 equiv) was added dropwise at 0 • C, under argon. After 15 min, para-methoxybenzyl amine (1.90 g, 1.8 mL, 13.5 mmol, 2.7 equiv) was added and the reaction mixture was stirred at 0 • C for 2 h. After completion, the reaction mixture was diluted with EtOAc (85 mL) and washed with 1M HCl (1 × 40 mL). The aqueous phase was extracted with EtOAc (2 × 50 mL). The combined organic layers were dried over anhydrous Na 2 SO 4 and concentrated under reduced pressure. The crude product was purified by flash column chromatography (EtOAc:Hex, 2:3 → 1:1→ 4:1), yielding the corresponding amide (930 mg, 2.55 mmol, 51%) as a pale-yellow solid. Step b: The amide (650 mg, 1.78 mmol) was dissolved in CH 2 Cl 2 (7.2 mL), the solution was cooled to 0 • C, and trifluoroacetic acid (2.15 g, 1.4 mL, 18.9 mmol, 10.6 equiv) was added. The reaction mixture was stirred at rt while being monitored by TLC. After 3 h, the mixture was concentrated in vacuo, Et 2 O (40 mL) was added to the residue, and the mixture was kept in the ultrasonic bath for 30 min to form a white suspension. Step c: Salt (S,S)-11·TFA (643 mg, 1.70 mmol) was suspended in CH 2 Cl 2 (20 mL), and the resulting suspension was washed with 10% aqueous NaHCO 3 (5 mL). The aqueous layer was extracted with CH 2 Cl 2 (2 × 10 mL). The combined organic layers were dried over anhydrous Na 2 SO 4 , and concentrated in vacuo, yielding (S,S)-11 (430 mg, 1.63 mmol, 96%) as an orange solid. 2-Formylpyrrole (S,S)-8 (145 mg, 0.298 mmol) was dissolved in CH 2 Cl 2 (4.5 mL), and a solution of para-toluenesulfonic acid monohydrate (113 mg, 0.596 mmol, 2.0 equiv) in MeOH (1.5 mL) was added. The resulting mixture was stirred for 4 h at rt while being monitored by TLC. After completion, the reaction mixture was neutralized with a saturated aqueous NaHCO 3 solution (3 mL), and the aqueous layer was extracted with CH 2 Cl 2 (2 × 10 mL). The combined organic solutions were washed with brine (15 mL), dried over anhydrous Na 2 SO 4 , and concentrated in vacuo. The resulting crude product was purified by flash column chromatography (EtOAc:petroleum ether, 3:1 → 4:1), yielding a pale-yellow oil (107 mg, 0.274 mmol for mole-fraction-weighted M = 391.21 g.mol −1 , 92%). The isolated product contained 85% of (S,S)-13 and 15% of (S,S)-14, as confirmed by HPLC and 1 H NMR. This mixture was used in the next step. Step f: The mixture of (S,S)-13 and (S,S)-14 (85:15) (96 mg, 0.245 mmol for molefraction-weighted M = 391.21 g.mol −1 ) was dissolved in dry MeOH (2.8 mL). MeONa (25w% solution in MeOH) (5.4 mg, 27 µL, 0.099 mmol, 0.4 equiv) was added to the solution at 0 • C, under an argon atmosphere. The reaction mixture was stirred at 0 • C. After 30 min, a saturated aqueous NH 4 Cl solution (3 mL) was added at 0 • C, and the resulting mixture was extracted with EtOAc (3 × 10 mL). The combined organic layers were washed with brine (10 mL), dried over anhydrous Na 2 SO 4 , and concentrated in vacuo, yielding crude (S,S)-14 as a thick pale-yellow oil (96 mg, 0.229 mmol, 93%).

Conclusions
The synthesis of the pyrrolic alkaloid (-)-hemerocallisamine I, featuring crystallizationinduced diastereomer transformation (CIDT) and the Maillard reaction as the key synthetic strategies, was achieved in 12 steps and a 1.6% overall yield. The sequence involved the preparation of (2S,4S)-4-hydroxyglutamic acid lactone in gram quantities, from an achiral substrate. In parallel, the first synthesis of (±)-hemerocallisamine I was described in 11 steps and a 5.9% overall yield. A detailed inspection of the Maillard reaction conditions revealed diketopiperazine 12 as the dominant side product, arising from a cannibalistic reaction of the amine 11. This transformation might be responsible for the often-reported depletion of the starting amino-acid-derived amines in the Maillard-type condensations.