Discovery of hybrid Glypromate conjugates with neuroprotective activity against paraquat-induced toxicity

Abstract

Neurodegenerative disorders comprise a series of heterogeneous conditions that affect millions of people worldwide, representing a significant health burden in both developed and developing countries. Without disease-modifying treatments currently available, the development of effective neurotherapeutics is a health priority. In this work, a new series of peptide-conjugates of the Glypromate neuropeptide is reported to determine the interplay of annular constriction and neuroprotective activity. To this end, (1R,3S,4S)-2-azanorbornane-3-carboxylic acid was used as an l-proline and l-pipecolic acid surrogate in addition to functionalization of the glutamate residue with relevant active pharmaceutical ingredients (APIs), namely amantadine, memantine, and (R)-1-aminoindane. Using non-differentiated SH-SY5Y cells, conjugates 14a and 15a, functionalized with amantadine, significantly reduced protein aggregation, with 15a outperforming both Glypromate (2-fold enhancement, p < 0.05) and an equimolar mixture of Glypromate and amantadine (p < 0.0001). On the other hand, in SH-SY5Y differentiated cells, conjugate 18c functionalized with (R)-1-aminoindane counteracted the toxicity elicited by paraquat (p < 0.0001), while Glypromate was found to exacerbate the neurotoxicity. Altogether, this work adds new insights into Glypromate research by demonstrating that chemical conjugation and annular constriction are effective strategies to tune neuroprotective responses against different neurotoxic stimuli, paving the way for the development of new neurotherapeutics.

A series of bicyclic-based Glypromate conjugates with reduction of protein aggregation elicited by Aβ_25–35 and neuroprotective activity against paraquat-induced toxicity is reported, paving the way for the discovery of novel neurotherapeutics.

Introduction

Neurodegenerative diseases of the central nervous system (CNS) comprise a series of multifactorial and complex disorders characterized by the progressive degeneration of neuronal cells, resulting in the impairment of the structural and functional features of the nervous system.¹ Alzheimer's disease (AD) is the most common neurodegenerative disease and the leading cause of progressive dementia, affecting approximately 30 million people around the world.^2,3 Next comes Parkinson's disease (PD), with nearly 10 million people affected worldwide, ranking first place as the most common movement disorder.^4–6 Less prevalent neurodegenerative disorders, such as Huntington's disease and amyotrophic lateral sclerosis, among others, also have a great impact on society, leading to serious and life-threatening conditions.⁷

So far, no curative or restorative treatments are available for these neurological conditions. Currently, pharmacological therapies are focused on symptom alleviation and prevention of disease-related complications.⁷ However, as the neurodegenerative conditions progress, symptom management usually becomes less and less effective.⁷

With the global rise in life expectancy, the prevalence of neurodegenerative diseases is projected to substantially increase in the following decades and cause a severe social and economic burden.^2,5 For this reason, the discovery of more effective treatments to tackle neurodegenerative diseases is an imperative medical need of both developed and developing countries.

Glypromate (glycyl-l-prolyl-l-glutamic acid, also known as GPE, Fig. 1) is an endogenous neuropeptide obtained by the cleavage of the N-terminal fragment of insulin-like growth factor 1 (IGF-1), which in turn has potent neurotrophic and antiapoptotic activity.⁸ Furthermore, Glypromate is known to be metabolized by carboxypeptidases to afford glutamate along with cyclo-glycyl-l-proline (cGP), a nootropic compound (Fig. 1).⁸

Fig. 1. Chemical structures of Glypromate neuropeptide and its main metabolites.

Glypromate is well described in the literature as an active neuroprotective compound in several in vitro and in vivo models of neurodegenerative diseases.⁸ Experiments performed with male Wistar rats provided evidence of the vast repertoire of biological properties exhibited by this small neuropeptide. Glypromate has been shown to reduce quinolinic acid-induced injury (model of Huntington's disease),⁹ hypoxic-ischemic injury,¹⁰ and ischemic-reperfusion injury,¹⁰ decrease 6-hydroxydopamine (6-OHDA) neurotoxin-induced injury (model of PD),¹¹ and partially block somatostatin depletion induced by Aβ (model of AD).⁸

Despite its biological repertoire, the clinical application of Glypromate is hindered by its peptide nature, which translates in poor metabolic stability and low oral bioavailability.⁸ Over the past decades, intense research has been conducted with the aim of developing peptidomimetics of this neuropeptide with improved biological activity and pharmacological profiles.⁸

Among the main structural modifications reported in the literature, the use of non-proteinogenic cyclic amino acids as l-proline surrogates is extensively explored as a chemical strategy to increase the lipophilicity and metabolic stability of Glypromate analogs.^8,12 A paradigmatic example in this regard is the proton-to-methyl substitution at the α-position of the prolyl moiety, which not only increases conformational constriction but also improves metabolic resistance towards proteases when compared to Glypromate neuropeptide.^8,13 This simple yet effective substitution led to the discovery of Trofinetide (Daybue™) by Neuren Pharmaceuticals, which has found application as the first and only treatment approved for Rett syndrome by the Food and Drug Administration in 2023 (with commercial rights given to Acadia Pharmaceuticals).^14–16 Moreover, Trofinetide has also been explored for the treatment of patients with Fragile X syndrome, showing clinical improvements during phase 2 of clinical trials.^14–16

Recently, the chemical conjugation of Glypromate at both N- and C-terminal positions of the peptide motif with bioactive compounds has gained momentum as an effective strategy to develop new potential neurotherapeutics with tunable neuroprotective activity and improved pharmacological properties.^17–25

In this work, a novel series of bicyclic-based Glypromate analogs were synthetized by chemical conjugation with relevant APIs and their neuroprotective activity was assessed using cell-based models of AD and PD to explore the interplay between annular constriction and neuroprotective activity.

Results and discussion

Design and chemistry

Recently, our research group explored the conjugation of both Glypromate neuropeptide and a Glypromate analog incorporating l-pipecolic acid as an l-proline surrogate by functionalization of either α- or γ-carboxylic acid of the glutamate residue with an API as a strategy to leverage the neuroprotective activity and overall lipophilicity (Fig. 2).²⁵

Fig. 2. Glypromate conjugates 1–12(a–c) previously reported.^24,25.

Chemical conjugation of Glypromate with relevant APIs used in neurodegenerative disorders, such as amantadine (a, PD), memantine (b, AD), and (R)-1-aminoindane (c, neuroprotective), has proven to be effective with several conjugates outperforming the parent neuropeptide, namely 2c, 3b, and 5c (Fig. 2).²⁵

Among this series of conjugates, the results showed dependency on the API used and site of functionalization (α- or γ-carboxylic acid of the glutamic acid residue). Structure–activity relationship studies hint that an increased annular ring constriction at the central position of the peptide motif seems to be detrimental to improving cytotoxicity profiles, with l-pipecolyl-based conjugates, 7–12(a–c), exhibiting high cytotoxicity in comparison with l-prolyl counterparts, 1–6(a–c) (Fig. 2).²⁵

To test this hypothesis and enlighten the interplay between annular constriction at the central residue and both cytotoxicity and neuroprotective activity, a new series of bicyclic-based Glypromate conjugates were rationally designed by merging the structures of l-prolyl and l-pipecolyl-based conjugates as highlighted in Scheme 1.

Scheme 1. Rational design of bicyclic-based Glypromate conjugates 13–18(a–c) by incorporation of (1R,3S,4S)-I as an l-prolyl and l-pipecolyl hybrid surrogate.

(1R,3S,4S)-2-Azanorbornane-3-carboxylic acid, (1R,3S,4S)-I, has long been recognized as a valuable non-proteinogenic amino acid merging the pyrrolidine and piperidine cyclic structures of l-proline and l-pipecolic acid, respectively, into a bicyclic hybrid construct.^26,27 In view of this, heterobicyclic scaffold (1R,3S,4S)-I was selected as an adequate surrogate to investigate the influence of enhanced annular constriction and lipophilicity at the central residue of Glypromate neuropeptide. Importantly, (1R,3S,4S)-I is a structural isomer of 7-azanorbornane-1-carboxylic acid, a bicyclic scaffold that has been explored by García-López's group for the assembly of a Glypromate peptidomimetic exhibiting 60-fold enhancement in the binding affinity for glutamate receptors in comparison to Glypromate.²⁸ However, no correlation was found between binding affinity for glutamate receptors and neuroprotective effects on cultured hippocampal neurons exposed to N-methyl-d-aspartate (100 μM).²⁸

Bicyclic scaffold (1R,3S,4S)-I has been successfully applied as an l-proline surrogate in the development of novel neurotherapeutic drugs, including positive allosteric modulators of D₂ receptors²⁹ and new ligands for the peptidylprolyl isomerase FKBP12.³⁰ Moreover, it has also been used as an l-proline and l-pipecolic acid surrogate for the development of inhibitors of hepatitis C NS3-NS4A serine protease.²⁶

The use of (1R,3S,4S)-I for the assembly of bicyclic-based Glypromate conjugates with APIs (a–c) is expected to contribute to relevant structure–activity relationship insights while improving the overall lipophilicity and confer additional enthalpic binding contributions to target biomolecules.^12,31,32

Organic synthesis

The enantiopure bicyclic scaffold (1R,3S,4S)-I was obtained following a diastereoselective aza-Diels-Alder protocol previously described by our research group,³³ using (−)-8-phenylmenthol (8PM) as a chiral auxiliary (Scheme 2A).

Scheme 2. Preparation of (A) (1R,3S,4S)-I and (B) functionalized glutamates III–IV(a–c).^24,33.

Briefly, 8PM is acylated with acryloyl chloride to provide the corresponding ester, which undergoes catalytic di-hydroxylation with OsO₄ (Upjohn dihydroxylation) and oxidative cleavage using NaIO₄ to provide 8-phenylmenthyl glyoxylate. Then, the glyoxylate is reacted with (R)-1-phenylethan-1-amine to provide the corresponding imine intermediate which undergoes acid-catalyzed aza-Diels–Alder reaction with cyclopentadiene to afford II as a sole diastereoisomer (Scheme 2A).^33,34 After the cleavage of the N-phenylethyl group by a transprotection reaction and saturation of the olefin under reductive conditions, the chiral auxiliary 8PM is recovered by saponification followed by acid-treatment, affording enantiopure bicyclic amino acid (1R,3S,4S)-I (Scheme 2A).³³

Next, functionalized glutamates with APIs at either α- or γ-positions, namely III(a–c) and IV(a–c), respectively, were prepared using a synthesis protocol previously described from adequate commercially available glutamates, Boc-Glu(OMe)-OH and Boc-Glu(OH)-OMe (Scheme 2B).²⁴ In brief, these Boc-protected glutamates are coupled with APIs a–c using 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium tetrafluoroborate (TBTU) in the presence of triethylamine (Et₃N) obtaining the corresponding amides followed by cleavage of the carbamate group by acidolysis with trifluoroacetic acid (TFA) to provide the deprotected functionalized glutamates III(a–c) and IV(a–c), depicted in Scheme 2B.

With (1R,3S,4S)-I and functionalized glutamates III(a–c) and IV(a–c) in hand, conjugates 13(a–c) and 16(a-c) were synthesized using an efficient one-pot peptide synthesis protocol in solution-phase previously described by our research group (Scheme 3).^24,25,35

Scheme 3. Assembly of bicyclic-based Glypromate conjugates 13–18(a–c). Conditions and reagents: i) Et₃N, Boc-Gly-OSu, TBTU, III(a–c),^24,25 anhydrous CH₂Cl₂; ii) Et₃N, Boc-Gly-OSu, TBTU, IV(a–c),^24,25 anhydrous CH₂Cl₂; iii) TFA, anhydrous CH₂Cl₂; iv) formaldehyde solution 37% in water, NaBH(OAc)₃, 1,2-dichloroethane.

Following this protocol, fully deprotected amino acid (1R,3S,4S)-I reacts with N-(tert-butyloxycarbonyl)glycine succinimidyl ester (Boc-Gly-OSu) in the presence of Et₃N to afford the corresponding amide-carboxylate intermediate (not isolated). Then, this intermediate is activated in situ with TBTU to generate the corresponding activated ester. Finally, upon the addition of the appropriate functionalized glutamates III(a–c) or IV(a–c),^24,25 the activated ester intermediates undergo aminolysis to afford 13(a–c) and 16(a–c), respectively, in high global yields (79–91%, from (1R,3S,4S)-I).

Next, cleavage of the carbamate group from 13(a–c) and 16(a–c) was performed by acidolysis with TFA, affording the corresponding conjugates 14(a–c) and 17(a–c) as ammonium trifluoroacetate salts in very good to excellent yields (93–100%). The final step of the synthesis route consisted of N,N-dimethylation by reductive amination using formaldehyde solution 37% in water and sodium triacetoxyborohydride, NaBH(OAc)₃, as the reducing agent. Under these conditions, 15(a–c) and 18(a–c) were obtained in moderate to very good yields (58–96%).

All the conjugates were characterized by proton (¹H), proton-decoupled carbon-13 (¹³C{¹H}), and distortionless enhancement by polarization transfer-135 (DEPT-135) nuclear magnetic resonance (NMR), as well as high-resolution mass spectrometry (HRMS). The obtained spectroscopic and spectrometric data confirmed the structures of the target conjugates.

Cytotoxicity assays in non-differentiated SH-SY5Y cells

An initial toxicological screening was performed to evaluate the cytotoxicity of conjugates 13–18(a–c) at 100 μM using non-differentiated SH-SY5Y cells.²⁵ The conjugates were dissolved in dimethyl sulfoxide (DMSO) 0.5% in phosphate-buffered saline (PBS) and the results obtained are depicted in Fig. 3.

Fig. 3. Cytotoxicity of conjugates 13–18(a–c) evaluated through the MTT reduction assay in non-differentiated SH-SY5Y neuronal cells. The vehicle represents the condition of DMSO 0.5% in PBS. Data are expressed as a percentage of control and are presented as mean ± standard deviation. The results were obtained from at least three independent experiments, each performed in triplicate. Statistical analyses were performed using the analysis of variance (ANOVA) test followed by uncorrected Fisher's LSD (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 vs. vehicle).

Most of the bicyclic-based Glypromate conjugates exhibited significant cytotoxicity, except for conjugates 14c (96.04 ± 5.30%) and 15a (97.93 ± 8.67%). Curiously, all the conjugates with memantine (13b–18b) exhibited the highest cytotoxic effects with values of MTT reduction ranging from 32.77 ± 3.99% (p < 0.0001) to 83.18 ± 2.71% (p < 0.0001), when compared to the control. This trend is in agreement with the results previously obtained in the series of conjugates 1–12(a–c), in which memantine-based derivatives yielded highly cytotoxic effects in this model.²⁵

After defining a cut-off of 85% for cell viability, conjugates 13a (90.79 ± 5.20%, p < 0.01), 13c (86.49 ± 4.51%, p < 0.0001), 14a (93.91 ± 9.23%, p < 0.05), 16a (89.94 ± 7.78%, p < 0.01), 17a (89.15 ± 10.69%, p < 0.001), 17c (92.66 ± 7.35%, p < 0.05) and 18c (90.95 ± 10.55%, p < 0.01) were selected, together with conjugates 14c and 15a, to undergo protein aggregation studies in non-differentiated SH-SY5Y cells.

Protein aggregation

The selected bicyclic-based Glypromate conjugates (cell viability > 85%) were evaluated to assess their ability to reduce the protein aggregation caused by Aβ_25–35. The Aβ peptide is the major contributing component of neuritic plaques in AD.³⁶ Aβ_25–35 is an active fragment of the full-length Aβ_1–42 peptide that displays similar biophysical and biochemical properties.³⁶ To this purpose, non-differentiated SH-SY5Y cells were incubated with Aβ_25–35 (10 μM) and conjugates 13a, 13c, 14a, 14c, 15a, 16a, 17a, 17c, and 18c (100 μM), and then protein aggregates were quantified by the thioflavin T assay (Fig. 4), as previously described by our research group.³⁷

Fig. 4. Protein aggregation in non-differentiated SH-SY5Y cells exposed to Aβ_25–35 (10 μM) and co-incubated with the selected conjugates (100 μM). In this assay, the results obtained after thioflavin T staining represent the percentage of fluorescence compared to the positive control (Aβ_25–35, 10 μM) and are presented as mean ± standard deviation. The results were obtained from at least three independent experiments, each performed in triplicate. Statistical analyses were performed using the analysis of variance (ANOVA) test followed by uncorrected Fisher's LSD (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 vs. Aβ_25–35; ^#p < 0.05 vs. Glypromate; ^&&&&p < 0.0001 vs. Glypromate + a).

In this assay, among the conjugates tested, 14a (79.51 ± 15.86%, p < 0.001) and 15a (75.69 ± 13.54, p < 0.0001), both functionalized with amantadine (a), were able to significantly reduce the formation of protein aggregates by 20 and 24%, respectively (Fig. 4). The conditions of Glypromate, amantadine (a), and Glypromate + a (equimolar mixture) were also included for comparison (Fig. 4). Although both Glypromate (87.78 ± 7.13%, p < 0.05) and a (82.96 ± 11.35%, p < 0.01) per se were able to reduce the protein aggregation caused by Aβ_25–35 by 12 and 17%, respectively, the condition Glypromate + a (equimolar mixture) was devoid of a significant reduction of protein aggregation. In this series, 15a outperformed the parent neuropeptide (2-fold improvement, p < 0.05), by effectively counteracting protein aggregation elicited by Aβ_25–35.

While in this series of bicyclic-based Glypromate conjugates 15a was the best-performing compound, its proline-based counterpart (5a, Fig. 2) was unable to counteract protein aggregation induced by Aβ_25–35.²⁵ This result demonstrates that the capacity of these conjugates to counteract protein aggregation induced by Aβ_25–35 can be regulated by annular constriction at the central position of the peptide motif. Moreover, while proline-based conjugates functionalized with (R)-1-aminoindane (c) led to the discovery of two highly active conjugates (2c and 5c, Fig. 2) against protein aggregation, in this series, (R)-1-aminoindane-based conjugates 13c, 14c, 17c, and 18c were found to be inactive or to even significantly exacerbate the Aβ_25–35-induced protein aggregation (Fig. 4).

Altogether, these results highlight that both neurotoxicity and neuroprotective effects do not depend exclusively on the API used but the interplay between the API and the heterocyclic scaffold present at the central residue of the peptide motif.

Cytotoxicity assays in differentiated SH-SY5Y cells

SH-SY5Y neuronal cells were used as a cell-based model of PD. To achieve a dopaminergic phenotype, SH-SY5Y cells underwent a 6-day differentiation protocol with retinoic acid (RA) and 12-O-tetradecanoylphorbol-13-acetate (TPA), following a protocol described by our research group.^38–40

An initial screening was performed to evaluate the cytotoxicity of conjugates 13–18(a–c) at 100 μM concentration (Fig. 5).

Fig. 5. Cytotoxicity of conjugates 13–18(a–c) evaluated through the MTT reduction assay in differentiated SH-SY5Y neuronal cells. Cells were incubated for 48 h with 13–18(a–c). Data are expressed as a percentage of control (PBS) and are presented as mean ± standard deviation. The results were obtained from 3–5 independent experiments, each performed in quadruplicate. Statistical analyses were performed using the analysis of variance (ANOVA) test followed by uncorrected Fisher's LSD (*p < 0.05, ***p < 0.001, and ****p < 0.0001 vs. control).

In differentiated neuronal cells (Fig. 5), the majority of the bicyclic-based Glypromate conjugates exhibited a significant cytotoxicity when compared with the control, except for conjugates 13b (98.10 ± 4.50%), 14a (102.46 ± 3.19%), 14c (98.48 ± 4.06%), and 15c (100.56 ± 3.21%). As shown, the cytotoxicity observed in this model was less pronounced when compared with the results obtained for non-differentiated SH-SY5Y cells. This can be explained by changes in metabolism and signaling regulation, making the RA-differentiated cells more resistant to toxins in comparison with non-differentiated SH-SY5Y cells.⁴¹ However, in differentiated cells, conjugate 18b (9.07 ± 2.77%, p < 0.0001) was found to be highly cytotoxic.

After defining a cut-off of 90% for cytotoxicity, conjugates 16c (94.07 ± 6.80%, p < 0.001), 17a (96.82 ± 6.68%, p < 0.05), and 18c (96.47 ± 6.06%, p < 0.05) were also selected (in addition to conjugates 13b, 14a, 14c, and 15c) to undergo neuroprotection studies in differentiated SH-SY5Y cells exposed to 6-OHDA and paraquat (PQ), two neurotoxins commonly used in models of PD.

Neuroprotection against 6-OHDA

6-OHDA is a catecholamine derivative, which permeates catecholaminergic neurons through dopamine or noradrenaline transporters and accumulates inside the cells triggering the production of reactive oxygen species and quinones, therefore leading to oxidative stress and cell death.^40,42 This neurotoxin is widely used in in vitro and in vivo models of PD.⁴³

To this end, besides the MTT reduction assay for the determination of the neurotoxicity induced by 6-OHDA, the NR uptake assay was also envisioned as a complementary classical toxicity assay. Contrary to the MTT reduction assay, which measures the reduction of a tetrazolium component (MTT) into an insoluble formazan product by dehydrogenases (mainly mitochondrial),³⁸ the NR uptake assay is based on the uptake of an eurhodin dye (NR) that is accumulated in the lysosomes of cells via active transport.⁴⁴ The obtained results in the MTT reduction and the NR uptake assays are depicted in Fig. 6 and 7, respectively.

Fig. 6. MTT reduction assay performed in differentiated SH-SY5Y cells exposed to the selected conjugates (100 μM, 48 h) and co-incubated with 6-OHDA (125 μM). Data are expressed as a percentage of control (PBS) and are presented as mean ± standard deviation. The results were obtained from three independent experiments in quadruplicate. Statistical analyses were performed using the analysis of variance (ANOVA) test followed by uncorrected Fisher's LSD (****p < 0.0001 vs. control; ^####p < 0.0001 vs. 6-OHDA).

Fig. 7. Neutral red uptake assay performed in differentiated SH-SY5Y cells exposed to the selected conjugates (100 μM, 48 h) and co-incubated with 6-OHDA (125 μM). Data are expressed as a percentage of control (PBS) and are presented as mean ± standard deviation. The results were obtained from six independent experiments in quadruplicate. Statistical analyses were performed using the analysis of variance (ANOVA) test followed by uncorrected Fisher's LSD (****p < 0.0001 vs. control; ^####p < 0.0001 vs. 6-OHDA).

In both assays, none of the synthesized conjugates were able to significantly reduce the neurotoxicity elicited by 6-OHDA (Fig. 6 and 7). As expected, in the MTT reduction assay, Glypromate (60.12 ± 8.59%, p < 0.0001) was able to significantly counteract the 6-OHDA-induced cytotoxicity by 12% (Fig. 6).²⁵ However, this effect was not observed in the NR uptake assay (Fig. 7). In this case, in contrast to the tested conjugates, Glypromate (25.32 ± 5.76%, p < 0.0001) further exacerbated the neurotoxicity triggered by 6-OHDA (61.31 ± 12.93%, p < 0.0001). This is the first report on the effect of Glypromate neuropeptide on 6-OHDA-induced cytotoxicity by the NR uptake assay using differentiated SH-SY5Y cells.

Neuroprotection evaluation against PQ

The same conjugates were selected to study the neuroprotective activity against PQ (300 μM) in differentiated SH-SY5Y cells including Glypromate neuropeptide (unprecedented). PQ is a common toxic herbicide associated with an increased risk of developing PD.⁴⁵ It is known that this toxin produces subcellular changes associated with PD, including increased production of reactive oxygen species, aggregation of α-synuclein, and selective nigral injury.⁴⁶ Noteworthily, the chemical structure of PQ resembles that of 1-methyl-4-phenylpyridinium, a metabolically active product from 1-methyl-4-phenyl-1,2,3,6-tetrahydropyrimidine, a potent neurotoxin known to produce severe irreversible brain damage, particularly in the substantia nigra, leading to symptoms similar to those of PD.^45,47

The results of the MTT reduction and NR uptake assays are depicted in Fig. 8 and 9, respectively.

Fig. 8. MTT reduction assay performed in differentiated SH-SY5Y cells exposed to the selected conjugates (100 μM, 48 h) and co-incubated with PQ (300 μM). Data are expressed as a percentage of control (PBS) and are presented as mean ± standard deviation. The results were obtained from four to six independent experiments in quadruplicate. Statistical analyses were performed using the analysis of variance (ANOVA) test followed by uncorrected Fisher's LSD (**p < 0.01 and ****p < 0.0001 vs. control; ^###p < 0.001 and ^####p < 0.0001 vs. PQ).

Fig. 9. NR uptake assay in differentiated SH-SY5Y cells exposed to the selected conjugates (100 μM, 48 h) and co-incubated with PQ (300 μM). Data are expressed as a percentage of control (PBS) and are presented as mean ± standard deviation. The results were obtained from four independent experiments in quadruplicate. Statistical analyses were performed using the analysis of variance (ANOVA) test followed by uncorrected Fisher's LSD (*p < 0.05 and ****p < 0.0001 vs. control; ^#p < 0.05, ^###p < 0.001, and ^####p < 0.0001 vs. PQ).

In the MTT reduction assay (Fig. 8), conjugate 13b (38.81 ± 4.18%, p < 0.0001) significantly exacerbated the toxicity elicited by PQ (46.02 ± 3.61%, p < 0.0001), while conjugates 14a (50.66 ± 2.91%, p < 0.001), 14c (50.44 ± 5.31%, p < 0.001), and 18c (57.88 ± 6.16%, p < 0.0001) were able to counteract the cytotoxicity elicited by PQ by 5, 4, and 12%, respectively (Fig. 8). Among the tested compounds, 18c stands out as the best-performing conjugate by reducing the cytotoxicity elicited by PQ up to 12%, while Glypromate (25.32 ± 5.76%, p < 0.0001), c (23.65 ± 4.07%, p < 0.0001), and the equimolar mixture of Glypromate + c (24.12 ± 7.69%, p < 0.0001) were found to exacerbate the PQ-induced cytotoxicity (Fig. 8).

In the NR uptake assay (Fig. 9), conjugates 14c (40.76 ± 7.35%, p < 0.05), 17a (40.94 ± 8.90%, p < 0.05), and 18c (51.71 ± 4.88%, p < 0.0001) were able to counteract the cytotoxicity elicited by PQ (36.78 ± 9.57%, p < 0.0001). As observed in the MTT reduction assay with PQ (Fig. 8), compound 18c was the best-performing conjugate in the NR uptake assay (Fig. 9), exhibiting neuroprotective activity up to 15% against the toxicity elicited by PQ (in comparison with PQ condition). On the other hand, while c (22.95 ± 11.36%, p < 0.0001) and an equimolar mixture of Glypromate + c (30.14 ± 12.63%, p < 0.001) exacerbated the cytotoxicity triggered by PQ, Glypromate (36.33 ± 9.65%) was devoid of neuroprotective activity. These experiments constitute the first report on the effect of both Glypromate neuropeptide and (R)-1-aminoindane (c) on PQ-induced cytotoxicity using differentiated SH-SY5Y cells.

Interestingly, the neuroprotective activity observed for conjugate 18c against PQ-induced cytotoxicity in both MTT reduction and NR uptake assays demonstrates that the site of functionalization (γ-position) is a key structural determinant, since isomer 15c, functionalized at the α-position, was devoid of neuroprotective activity against this neurotoxin.

In silico prediction of oral bioavailability and blood-brain barrier (BBB) permeability

For oral-available drugs targeting the CNS, gastrointestinal absorption and BBB permeability have long been the limiting steps for the successful treatment of neurological pathologies. In this sense, in silico experiments were performed using cheminformatic resources to gain insights about the oral bioavailability and BBB permeability of the best-performing bicyclic-based conjugates in the protein aggregation experiments and/or in the neuroprotection against PQ-induced cytotoxicity (14a, 14c, 15a, 17a, and 18c), including Glypromate for comparison purposes (Table 1).^48–50

Estimated oral bioavailability and BBB permeability for bioactive bicyclic-based Glypromate conjugates and Glypromate using SwissADME,⁴⁸ ADMETsar 3.0,⁴⁹ and LightBBB⁵⁰ cheminformatic resources.

	Glypromate	14a	14c	15a	17a	18c
Log Po/w48	−1.98	1.36	1.05	1.99	1.47	1.72
TPSA (Å2)49	150.03	130.83	130.83	108.05	130.83	108.05
Bioavailability score48	0.11	0.55	0.55	0.55	0.55	0.55
HIA (%)49	9	69.1	57.9	86.7	71.8	86.2
Log BB50	−1.08	−0.72	−0.83	−0.78	−0.68	−0.78

First, the partition coefficient between octan-1-ol and water (log P_o/w) was predicted using the consensus estimation of log P_o/w,⁴⁸ the arithmetic mean of log P_o/w determined by different methods (including XLOGP3, WLOGP, MLOGP, SILICOS-IT, and iLOGP), which provides more accurate results.⁴⁸ The results obtained from the consensus log P_o/w are listed in Table 1. The results show that while Glypromate is considered highly hydrophilic (log P_o/w = −1.98) and predicted to exhibit low permeability across biological membranes, all the selected bicyclic-based Glypromate conjugates exhibited enhanced lipophilicity (1.05 < log P_o/w < 1.99, Table 1).

Another relevant parameter is the topological polar surface area (TPSA, in Ų), which provides information about the polarity and dictates interaction with biological membranes. Molecules with higher TPSA values are less lipid-soluble and expected to be less extensively and more slowly absorbed and exhibit reduced distribution in comparison with molecules with lower TPSA values. Molecules with a TPSA ≥ 140 Ų tend to present a low ability to permeate cell membranes.⁵¹ For molecules intended to cross the BBB, a TPSA less than 90 Ų is usually more favourable. In this series, all the conjugates exhibit lower TPSA (108.05 ≤ TPSA ≤ 130.83 Ų, Table 1) than Glypromate (TPSA = 150.03 Ų, Table 1), which suggests improved absorption and distribution.

The TPSA is recognized as a good predictor (TPSA ≤ 140 Ų) of human intestinal absorption (HIA),⁵² deemed as an essential prerequisite for the apparent efficacy of oral drugs.⁴⁹ In this study, a HIA ≥ 30% indicates good absorption profiles. The results showed that while Glypromate is predicted to display limited HIA (HIA = 9%, Table 1), all the selected bioactive conjugates exhibited favourable HIA profiles (57.9 ≤ HIA ≤ 86.3%, Table 1). As a result, the selected bicyclic conjugates demonstrate a 5-fold improvement in bioavailability compared with Glypromate (score of 0.55 vs. 0.11, respectively, Table 1).

The distribution into the CNS can be calculated by the logarithmic ratio of the concentration of a drug in the brain to its concentration in the blood (log BB), where a log BB ≥ −1 is classified as “BBB permeable”, while a log BB < −1 is labeled as “BBB non-permeable”. The data obtained show that the bicyclic conjugates are predicted to display better log BB (−0.83 ≤ log BB ≤ −0.62, Table 1) than Glypromate (log BB = −1.08, Table 1), being classified as BBB permeable, while the parent neuropeptide has limited BBB permeability.

Altogether, the data shows that the selected bicyclic conjugates are likely to exhibit improved oral bioavailability and enhanced BBB permeability in comparison with the parent neuropeptide.

Conclusion

Herein, a novel series composed of 18 bicyclic-based Glypromate conjugates incorporating the bicyclic scaffold (1R,3S,4S)-I as an l-proline and l-pipecolic acid surrogate and functionalized at the glutamate residue with several APIs, namely conjugates 13–18(a–c), was design, synthesized, and biologically evaluated using cell-based models of AD and PD.

In this series of bicyclic-based Glypromate conjugates (Fig. 10), the use of scaffold (1R,3S,4S)-I was found to reduce the overall cytotoxicity of the conjugates in comparison with the l-pipecolic acid-based counterparts,²⁵ leading to the discovery of bioactive conjugates against protein aggregation induced by Aβ_25–35 (14a and 15a) and the toxicity elicited by PQ (18c).

Fig. 10. Overview of the main structure–activity relationships of bicyclic-based Glypromate conjugates.

In the case of 15a, this conjugate outperformed the parent neuropeptide exhibiting a 2-fold reduction of protein aggregation elicited by Aβ_25–35 (p < 0.05). Moreover, 15a exhibited a superior effect in comparison with an equimolar mixture of Glypromate and amantadine (p < 0.0001). On the other hand, in differentiated SH-SY5Y cells, bicyclic-Glypromate conjugate 18c (100 μM) functionalized with (R)-1-aminoindane exhibited neuroprotective effects between 12 and 15% (p < 0.0001) against the toxicity elicited by PQ in the MTT reduction and NR uptake assays, while the parent neuropeptide was inactive (MTT reduction assay) or even exacerbated the PQ-induced neurotoxicity (NR uptake assay).

This work adds new biological information about Glypromate neuropeptide by demonstrating, for the first time, exacerbation of toxicity elicited by either 6-OHDA (NR uptake assay) and PQ (MTT reduction assay) using differentiated SH-SY5Y cells and lack of neuroprotective activity against PQ using the NR uptake assay.

It is worth mentioning that the range of neuroprotective activity elicited by 18c (12–15%) using PQ as a neurotoxic insult was similar to that of Glypromate (12%) using 6-OHDA. These results are of utmost importance hinting that the biological activity of Glypromate can be biased through chemical constriction at the central residue and conjugation at the C-terminal position.

Overall, the results obtained in the neuroprotective assays highlight the importance of using different neurotoxins with dissimilar underlying neurotoxicity mechanisms,^53–56 and the use of complementary toxicity assays for the discovery of new Glypromate-based neuroprotective hits.

These results provide useful structure–activity insights (highlighted in Fig. 10) for the development of new Glypromate-based conjugates with improved activity, biased biological responses, and enhanced oral bioavailability and BBB permeability, paving the way for the development of new neurotherapeutics.

Experimental section

Materials and methods

Organic synthesis

General data

All chemicals were of reagent grade and were obtained from Fluorochem (Hadfield, United Kingdom), Bachem (Bubendorf, Switzerland), Alfa Aesar (Karlsruhe, Germany), Acros Organics (Geel, Belgium), Fisher Scientific (Loures, Portugal), Sigma-Aldrich (Algés, Portugal), and ABCR GmbH (Karlsruhe, Germany). All air-sensitive reactions were carried out under an argon atmosphere. Analytical TLC was carried out on pre-coated silica gel plates (Merck 60 F254, 0.25 mm) using UV light and an ethanolic solution of phosphomolybdic acid (followed by gentle heating) for visualization. Flash chromatography was performed on silica gel (Merck 60, 230–240 mesh). HPLC analyses were performed on a Merck-Hitachi Lachrom Elite instrument equipped with a diode array (DAD) from the Department of Chemistry and Biochemistry, Faculty of Sciences of the University of Porto (FCUP|DQB – Lab&Services). HPLC column: LiChroCART (250 mm × 4.0 mm; particle size, 5 μm). Elution conditions: 1–100% B in A (A = H₂O with 0.05% of TFA; B = acetonitrile) with a flow rate of 1 mL min⁻¹ for 30 min. The target conjugates were confirmed to have at least 95% purity.

Apparatus

¹H, ¹³C{¹H}, and DEPT-135-NMR spectra were recorded at Centro de Materiais da Universidade do Porto (CEMUP) with a Bruker Avance III 400 at 400.15 MHz and 100.62 MHz, respectively. The NMR spectra were calibrated using residual signals from deuterated solvents (CDCl₃: δ_H = 7.26, δ_C = 77.16, CD₃OD: δ_H = 3.31, δ_C = 49.00, DMSO-d₆: δ_H = 2.50, δ_C = 39.52)⁵⁷ and are reported in ppm. The nomenclature used for the assignment of protons and/or carbons for each α-amino acid residue in the peptide chains was made using a three-letter system in subscript for the canonical amino acid residues (Gly: glycine; Glu: l-glutamic acid) indicating the proton (or group of protons) and/or the carbons in the structures by starting the numeration at the carbonyl carbon of each α-amino acid residue, while for (1R,3S,4S)-I, the IUPAC nomenclature was used instead. Assignment of protons and/or the carbons relative to the API scaffolds is indicated in subscripts as follows: amantadine (a), memantine (b), and (R)-1-aminoindane (c). Mass spectra were recorded on an LTQ Orbitrap XL hybrid mass spectrometer (Thermo Fischer Scientific, Bremen, Germany) controlled using Xcalibur 2.1.0 and LTQ Tune Plus 2.5.5 (Centro de Materiais da Universidade do Porto, CEMUP). The capillary voltage of the electrospray ionization source (ESI) and the temperature were set to 3.1 kV and 275 °C, respectively, and the sheath gas was set at 6 (arbitrary units) according to the software settings. The capillary and the tube lens voltage were set to 35 and 110 V, respectively. Optical rotations were measured on a JASCO P-2000 thermostated polarimeter using a sodium lamp and are reported as follows: [α]^θ_D expressed in (°) (dm⁻¹) (g⁻¹), in which θ is the temperature in Celsius and c (g per 100 mL, solvent). Melting points were determined using a STUART Scientific, model SMP1, and are not corrected. Solvents were removed in a Büchi rotavapor.

General protocols

The synthesis of (1R,3S,4S)-I was conducted by asymmetric aza-Diels–Alder reaction, following an optimized protocol previously described by our research group employing (−)-8-phenylmenthol as the chiral auxiliary.³³ The synthesis of functionalized glutamates III–IV(a–c) was performed using a protocol previously described.²⁴ All the analytical and structural data are in agreement with the previously reported data.^24,33 The synthesis of conjugates 13–18(a–c) was performed using methodologies described by our research group,²⁴ as described below.

(A) One-pot synthesis for the assembly of N-Boc protected conjugates²⁴

A solution of (1R,3S,4S)-I (1 equiv.) in anhydrous CH₂Cl₂ (20 mL) was prepared in a round bottom flask followed by the addition of Et₃N (3 equiv.) and Boc-Gly-OSu (1.1 equiv.). The reaction was left overnight (12 h) with magnetic stirring, then TBTU (1.2 equiv.) was added, and the solution was stirred for 30 min followed by the addition of the appropriate functionalized glutamate (1.2 equiv.). The system was allowed to stir for an additional 2 h at room temperature. The solvent was then removed in vacuo and the crude residue was dissolved in EtOAc (50 mL), transferred into a separatory funnel, and washed with the addition of saturated NaHCO₃ solution (3 × 50 mL). The organic extract was dried over anhydrous Na₂SO₄, filtered, and the filtrate was concentrated in vacuo. The resulting residue was purified with a chromatographic column using an appropriate eluent, specified for each compound.

(B) Removal of the N-Boc group by acidolysis^24,25

A solution of N-Boc protected compound (1 equiv.) in anhydrous CH₂Cl₂ (30 mL) was prepared in a round-bottom flask followed by the addition of TFA (30 equiv.). The reaction was left stirring for 2 h at room temperature, then the volatiles were removed using a rotatory evaporator

(C) N,N-Dimethylation by reductive amination^24,25

A solution of the N-deprotected conjugate (1 equiv.) in 1,2-dichloroethane (30 mL) was prepared in a round-bottom flask followed by the addition of CH₂O (37% m/v, 3 equiv.), and NaBH(OAc)₃ (6 equiv.). The resulting suspension was stirred for 2 h. The solvent was removed using a rotatory evaporator and CH₂Cl₂ (25 mL) was added to induce the precipitation of reductive agent excess and inorganic salts followed by filtration. The filtrate was concentrated in vacuo and the resulting residue was purified with a chromatographic column using an appropriate eluent, specified for each compound.

Synthesis of 13a

Following the general protocol A, starting from (1R,3S,4S)-I (0.2308 g, 1.316 mmol), and after the typical workup, the crude oil was chromatographed using EtOAc as the eluent, affording 13a (0.6101 g, 1.0616 mmol) as a white solid. Yield: 81%. Melting point: 49–51 °C. [α]¹⁷_D: −80.90 ± 0.11 (c1.120, CHCl₃). ¹H-NMR (CDCl₃, 400 MHz) δ ppm (rotamers): 7.70 (d, J = 7.8 Hz, CONH), 6.35–5.96 [6.25 (s), 6.06 (s), 1H, CONH], 5.52–4.53 [5.39 (br s), 4.61 (br s), 1H, OCONH], 4.36–3.83 (m, 5H, H-1 + H-3 + H_Glu-2 + H_Gly-2), 3.69–3.64 [3.67 (s), 3.66 (s), 3H, COOCH₃], 2.93–2.81 [2.90 (br s), 2.84 (br s), 1H, H-4], 2.58–2.33 (m, 2H, H_Glu-3), [2.15–1.90 (m, 12H), 1.89–1.70 (m, 4H), 1.71–1.61 (m, 8H), H_Glu-4 + H-5 + H-6 + H-7 + H_a], 1.44 (s, 9H, Boc). ¹³C{¹H}-NMR/DEPT-135 (CDCl₃, 101 MHz) δ ppm (rotamers): [175.6 (C), 169.5 (C), 169.3 (C), 168.6 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], 155.9 (C, Boc), 79.9 (C, Boc), [66.9 (CH), 66.5 (CH), C-3], [57.8 (CH), 57.6 (CH), C-1], [53.4 (CH), 53.0 (CH), C_Glu-2], [52.3 (C), 52.2 (C), C_a], [52.1 (CH₃), 52.0 (CH₃), COOCH₃], 43.4 (CH₂, C_Gly-2), [41.6 (CH₂), 41.4 (CH₂), 3CH₂, C_a], [40.6 (CH), 40.5 (CH), C-4], [36.5 (CH₂), 36.4 (CH₂), C-7 + C_a], [31.2 (CH₂), 31.1 (CH₂), C_Glu-4], [30.4 (CH₂), 30.3 (CH₂), C-6], [29.6 (CH), 29.5 (CH), 3CH, C_a], 28.5 (3CH₃, Boc), [27.3 (CH₂), 26.7 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₃₀H₄₇N₄O₇]⁺ 575.3439, found 575.3416.

Synthesis of 13b

Following the general protocol (A), starting from (1R,3S,4S)-I (0.2197 g, 1.237 mmol), and after the typical workup, the crude oil was chromatographed using EtOAc as the eluent, affording 13b (0.5911 g, 0.9806 mmol) as a yellow oil. Yield: 79%. [α]²³_D: +14.60 ± 0.10 (c1.000, CHCl₃). ¹H-NMR (CDCl₃, 400 MHz) δ ppm (rotamers): 7.79–7.17 [7.71 (d, J = 7.9 Hz), 7.24 (d, J = 7.9 Hz), 1H, CONH], 6.44–6.09 [6.33 (s), 6.19 (s), 1H, CONH], 5.39 (br s, 1H, CONH), 4.33–4.24 (m, 1H, H_Glu-2), 4.23–4.13 [4.21 (br s), 4.15 (br s), 1H, H-3], 4.10–3.78 (m, 3H, H_Gly-2 + H-1), 3.70–3.62 [3.66 (s), 3.65 (s), 3H, COOCH₃], 2.92–2.76 [2.87 (d, J = 4.0 Hz), 2.81 (d, J = 3.8 Hz), 1H, H-4], 2.55–2.30 (m, 2H, H_Glu-4), [2.13–2.00 (m, 3H), 1.95–1.80 (m, 1H), 1.82–1.72 (m, 4H), 1.68–1.45 (m, 7H), 1.42 (s, 9H), 1.38–1.22 (m, 4H), 1.17–1.06 (m, 2H), H_Glu-3 + H-5 + H-6 + H-7 + Boc + H_b], 0.82 (s, 6H, 2CH₃, H_b). ¹³C{¹H}-NMR/DEPT-135 (CDCl₃, 101 MHz) δ ppm (rotamers): [174.5 (C), 169.9 (C), 169.8 (C), 168.3 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], 155.8 (C, Boc), 79.9 (C, Boc), [66.9 (CH), 66.5 (CH), C-3], 57.5 (CH, C-1), 53.8 (C, C_b), [53.3 (CH), 52.0 (CH), C_Glu-2], [52.1 (CH₃), 52.0 (CH₃), COOCH₃], [50.7 (CH₂), 47.5 (CH₂), 47.4 (CH₂), 43.3 (CH₂), 42.7 (CH₂), C_Gly-2 + C_b], [40.7 (CH), 40.4 (CH), C-4], [40.0 (CH₂), 36.6 (CH₂), 36.5 (CH₂), 32.4 (C, C_b), 31.1 (CH₂), C-7 + C_Glu-4 + C_b], 30.3 (CH₂, C-6), 30.2 (2CH₃, C_b), 30.1 (CH, C_b), 28.5 (3CH₃, Boc), [27.3 (CH₂), 27.3 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₃₂H₅₁N₄O₇]⁺ 603.3752, found 603.3740.

Synthesis of 13c

Following the general protocol (A), starting from (1R,3S,4S)-I (0.1989 g, 1.120 mmol), and after the typical workup, the crude oil was chromatographed using EtOAc as the eluent, affording 13c (0.5729 g, 1.0292 mmol) as a white solid. Yield: 91%. Melting point: 73–78 °C. [α]²⁵_D: −6.07 ± 0.12 (c1.060, CHCl₃). ¹H-NMR (CDCl₃, 400 MHz) δ ppm (rotamers): 7.50–7.10 (m, 5H, CONH + H_c), 6.94–6.61 (m, 1H, CONH), 5.53–5.36 (m, 1H, H_c), 4.75–4.61 [4.69 (br s), 4.62 (br s), OCONH], 4.59–4.50 (m, 1H, H_Glu-2), 4.03 (br s, 1H, H-3), 3.87 (br s, 1H, H-1), 3.78–3.67 [3.73 (s), 4H, COOCH₃ + H_Gly-2a], 3.46 (dd, J = 16.5, 3.0 Hz, 1H, H_Gly-2b], [3.02–2.89 (m, 1H), 2.87–2.77 (m, 2H), 2.62–2.49 (m, 1H), 2.42–2.31 (m, 2H), 2.30–2.12 (m, 2H), 2.11–1.88 (m, 2H), 1.86–1.54 (m, 4H), 1.39 (s, 9H, Boc), 1.24 (br s, 1H), H-4 + H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + H_c]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [172.4 (C), 171.8 (C), 169.7 (C), 168.0 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], 155.8 (C, Boc), [143.4 (C), 143.2 (C), C_c], [127.8 (CH), 126.9 (CH), 126.6 (CH), 124.8 (CH), 124.7 (CH), 124.5 (CH), 124.1 (CH), C_c], [79.8 (C), 79.6 (C), Boc], [66.3 (CH), 66.0 (CH), C-3], [57.3 (CH), 57.3 (CH), C-1], [54.7 (CH), 54.7 (CH), C_c], [52.6 (CH), 52.3 (CH), C_Glu-2], 51.5 (CH₃, COOCH₃), 43.1 (CH₂, C_Gly-2), 40.4 (CH, C-4), [36.6 (CH₂), 34.2 (CH₂), 34.0 (CH₂), 32.3 (CH₂), 32.0 (CH₂), 31.1 (CH₂), 30.3 (CH₂), C_Glu-4 + C-6 + C-7 + C_c], [28.4 (CH₃), 28.2 (CH₃), Boc], [28.1 (CH₂), 27.1 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₉H₄₁N₄O₇]⁺ 557.2970, found 557.2952.

Synthesis of 14a

Following the general protocol (B), starting from compound 13a (0.8294 g, 1.443 mmol) and after the typical work-up, 14a (0.7858 g, 1.335 mmol) was afforded as a white solid. Yield: 93%. Melting point: 141–143 °C. [α]¹⁸_D: −92.01 ± 0.10 (c1.320, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 4.35–4.20 (m, 2H, H-3 + H_Glu-2), 4.08–3.94 (m, 2H, H_Gly-2), 3.81–3.72 (m, 1H, H-1), 3.70–3.66 [3.68 (s), 3.67 (s), 3H, COOCH₃], 2.85–2.69 (m, 1H, H-4), 2.46–2.34 (m, 2H, H_Glu-3), [2.12–1.96 (m, 12H), 1.96–1.80 (m, 3H), 1.77–1.66 (m, 8H), H_Glu-4 + H-5 + H-6 + H-7 + H_a]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [175.2 (C), 175.1 (C), 172.2 (C), 171.9 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [67.1 (CH), 67.0 (CH), C-3], [59.0 (CH), 58.7 (CH), C-1], [54.6 (CH), 54.4 (CH), C_Glu-2], [53.1 (C), 53.1 (C), C_a], [52.3 (CH₃), 52.3 (CH₃), COOCH₃], [42.9 (CH), 42.8 (CH), C-4], [42.3 (CH₂), 42.2 (CH₂), 42.2 (CH₂), 42.2 (CH₂), 3CH₂, C_a], [41.7 (CH₂), 41.7 (CH₂), C-7], 37.4 (3CH₂, C_a), [36.6 (CH₂), 36.6 (CH₂), 31.7 (CH₂), 31.2 (CH₂), 31.0 (CH₂), C_Gly-2 + C-6 + C_Glu-4], 30.9 (3CH, C_a), [28.6 (CH₂), 28.5 (CH₂), 28.3 (CH₂), 28.3 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₅H₃₉N₄O₅]⁺ 475.2915, found 475.2903.

Synthesis of 14b

Following the general protocol (B), starting from compound 13b (1.2118 g, 2.0104 mmol) and after the typical work-up, 14b (1.1980 g, 1.9427 mmol) was afforded as a white solid. Yield: 97%. Melting point: 54–57 °C. [α]²⁴_D: +8.00 ± 0.20 (c0.9700, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 4.35–4.21 (m, 2H, H_Glu-2 + H-3), [4.11–3.97 (m, 2H), 3.84–3.76 (m, 1H), H_Gly-2 + H-1], 3.71–3.64 [3.68 (s), 3.67 (s), 3H, COOCH₃], 2.86–2.70 (m, 1H, H-4), 2.48–2.34 (m, 2H, H_Glu-4), [2.15–1.76 (m, 8H), 1.72–1.48 (m, 7H), 1.42–1.26 (m, 4H), 1.15 (br s, 2H), H_Glu-3 + H-5 + H-6 + H-7 + H_b], 0.85 (s, 6H, 2CH₃, H_b). ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [175.3 (C), 172.2 (C), 171.9 (C), 165.8 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [67.1 (CH), 67.0 (CH), C-3], [59.1 (CH), 58.7 (CH), C-1], 54.7 (C, C_b), [54.5 (CH), 54.1 (CH), C_Glu-2], 52.4 (CH₃, COOCH₃), [51.6 (CH₂), 48.1 (CH₂), 43.7 (CH₂), C_b], [42.9 (CH), 42.8 (CH), C-4], [41.6 (CH₂), 41.2 (CH₂), 40.7 (CH₂), 36.5 (CH₂), 33.2 (C, C_b), 31.6 (CH₂), C_Gly-2 + C-7 + C_Glu-4], 31.5 (CH, C_b), [31.3 (CH₂), 31.2 (CH₂), C-6], 30.2 (2CH₃, C_b), [28.5 (CH₂), 28.5 (CH₂), 28.3 (CH₂), 28.3 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₇H₄₃N₄O₅]⁺ 503.3228, found 503.3215.

Synthesis of 14c

Following the general protocol (B), starting from compound 13c (0.6476 g, 1.1630 mmol) and after the typical work-up, 14c (0.6610 g, 1.1585 mmol) was afforded as a colorless oil. Yield: 100%. [α]²⁴_D: +16.30 ± 0.21 (c1.010, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 7.28–7.13 (m, 4H, H_c), 5.43–5.29 (m, 1H, H_c), 4.66–4.23 [4.60 (br s), 4.28 (br s), 2H, H-3 + H_Glu-2], 4.19–3.54 [3.68 (s), 3.67 (s), 6H, H-1 + H_Gly-2 + COOCH₃], [3.09–2.93 (m, 1H), 2.92–2.70 (m, 2H), 2.56–2.32 (m, 3H), 2.27–1.73 (m, 6H), 1.70–1.37 (m, 3H), H-4 + H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + H_c]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [176.7 (C), 175.0 (C), 173.3 (C), 172.1 (C), 165.6 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [145.5 (C), 144.3 (C), C_c], [129.0 (CH), 128.9 (CH), 127.7 (CH), 127.6 (CH), 125.7 (CH), 125.7 (CH), 125.2 (CH), 125.1 (CH), C_c], [67.1 (CH), 67.0 (CH), C-3], [59.1 (CH), 58.7 (CH), C-1], 55.9 (CH, C_c), [54.5 (CH), 54.3 (CH₂), C_Glu-2], [53.9 (CH₃), 52.3 (CH₃), COOCH₃], [42.9 (CH), 42.8 (CH), C-4], [41.5 (CH₂), 41.4 (CH₂), C_Gly-2], [36.6 (CH₂), 34.2 (CH₂), 34.1 (CH₂), 31.7 (CH₂), 31.2 (CH₂), 31.2 (CH₂), 31.0 (CH₂), C_Glu-4 + C-6 + C-7 + C_c], [28.5 (CH₂), 28.4 (CH₂), 28.3 (CH₂), 28.3 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₄H₃₃N₄O₅]⁺ 457.2446, found 457.2425.

Synthesis of 15a

Following the general protocol (C), starting from compound 14a (0.2689 g, 0.4567 mmol), and after the typical workup, the crude oil was chromatographed using CH₂Cl₂/CH₃OH (4:1) as the eluent, affording 15a (0.2194 g, 0.4365 mmol) as a colorless oil. Yield: 96%. [α]¹⁸_D: +47.52 ± 0.20 (c1.030, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 4.64–4.23 [4.60 (br s), 4.39–4.23 (m), 2H, H-3 + H_Glu-2], 4.12–3.94 [4.09 (s), 3.98 (s), 1H, H-1], 3.81–3.36 [3.79 (s), 3.75 (s), 3.74 (s), 3.72 (s), 3.62–3.35 (m), 5H, COOCH₃ + H_Gly-2], 2.84–2.67 [2.80 (d, J = 3.7 Hz), 2.71 (d, J = 3.3 Hz), 1H, H-4], 2.64–2.53 [2.62 (s), 2.60 (s), 2.57 (s), 2.55 (s), 6H, N(CH₃)₂], [2.48–2.34 (m, 2H), 2.14–1.96 (m, 10H), 1.95–1.81 (m, 3H), 1.80–1.74 (1H), 1.73–1.64 (m, 7H), 1.63–1.50 (m, 1H), 1.49–1.26 (m, 1H), H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + Ca]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [175.2 (C), 175.0 (C), 172.2 (C), 167.7 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], 67.0 (CH, C-3), [61.0 (CH₂), 60.8 (CH₂), C_Gly-2], [59.2 (CH), 58.4 (CH), C-1], [54.5 (CH), 54.1 (CH), C_Glu-2], [53.1 (CH₃), 52.3 (C), 52.2 (CH₃), COOCH₃ + C_a], [45.5 (CH₃), 45.2 (CH₃), N(CH₃)₂], 42.9 (CH, C-4), [42.3 (CH₂), 42.2 (CH₂), 3CH₂, C_a], [37.4 (CH₂), 36.6 (CH₂), C-7 + C_a], [31.8 (CH₂), 31.3 (CH₂), 31.2 (CH₂), C_Glu-4 + C-6], 30.9 (3CH, C_a), [28.5 (CH₂), 28.4 (CH₂), 28.3 (CH₂), 28.1 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₇H₄₃N₄O₅]⁺ 503.3228, found 503.3209.

Synthesis of 15b

Following the general protocol (C), starting from compound 14b (0.2867 g, 0.4649 mmol). After the typical workup, the crude oil was chromatographed using CH₂Cl₂/CH₃OH (4 : 1) as the eluent, affording 15b (0.1591 g, 0.2998 mmol) as a white solid. Yield: 64%. Melting point: 64–67 °C. [α]²⁵_D: −14.74 ± 0.15 (c1.040, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 4.66–4.15 [4.64–4.56 (m), 4.40–4.17 (m), 2H, H-3 + H_Glu-2], [4.11–3.77 (m, 2H), 3.73–3.40 (m, 4H), H-1 + H_Gly-2 + COOCH₃], 2.86–2.53 [2.66 (s), 2.64 (s), 2.62 (s), 2.59 (s), 7H, N(CH₃)₂ + H-4], 2.49–2.34 (m, 2H, H_Glu-3), [2.15–1.75 (m, 8H), 1.72–1.45 (m, 7H), 1.43–1.28 (m, 4H), 1.16 (s, 2H), H_Glu-4 + H-5 + H-6 + H-7 + H_b], 0.85 (s, 6H, 2CH₃, H_b). ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [175.1 (C), 175.0 (C), 172.3 (C), 172.0 (C), 167.2 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [67.0 (CH), 66.5 (CH), C-3], [60.6 (CH₂), 60.6 (CH₂), C_Gly-2], [59.2 (CH), 58.5 (CH), C-1], 54.7 (C, C_b), [54.4 (CH), 54.2 (CH), C_Glu-2], [52.3 (CH₃), 52.2 (CH₃), COOCH₃], [51.7 (CH₂), 48.2 (CH₂), 48.1 (CH₂), C_b], [45.4 (CH₃), 45.2 (CH₃), N(CH₃)₂], 43.7 (CH₂, C_b), [42.9 (CH), 42.8 (CH), C-4], [40.8 (CH₂), 40.7 (CH₂), 36.0 (CH₂), 34.4 (CH₂), 34.3 (CH₂), C-7 + Glu-4], 33.3 (C, C_b), [31.8 (CH₂), 31.8 (CH₂), C-6], 31.6 (CH, C_b), [31.1 (CH₂), 31.0 (CH₂), 30.7 (2CH₃, C_b), 28.6 (CH₂), 28.4 (CH₂), 28.2 (CH₂), 28.1 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₉H₄₇N₄O₅]⁺ 531.3541, found 531.3527.

Synthesis of 15c

Following the general protocol (C), starting from compound 14c (0.6521 g, 1.143 mmol), and after the typical workup, the crude oil was chromatographed using CH₂Cl₂/CH₃OH (4 : 1) as the eluent, affording 15c (0.3971 g, 0.8194 mmol) as a white solid. Yield: 72%. Melting point: 38–40 °C. [α]²⁵_D: +14.10 ± 0.21 (c1.060, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 7.30–7.14 (m, 4H, H_c), 5.45–5.30 (m, 1H, H_c), 4.68–4.25 [4.62 (br s), 4.50 (dd, J = 9.4, 4.8 Hz), 4.45 (dd, J = 9.5, 4.7 Hz), 4.31 (br s), 2H, H-3 + H_Glu-2], 4.08–3.68 [3.76 (s), 3.73 (s), 6H, H-1 + H_Gly-2 + COOCH₃], [3.05–2.93 (m, 1H), 2.90–2.81 (m, 1H), 2.78–2.68 (m, 6H, N(CH₃)₂), 2.56–2.19 (m, 4H), 2.10–1.94 (m, 3H), 1.92–1.38 (m, 6H), H-4 + H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + H_c]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [174.6 (C), 173.5 (C), 172.1 (C), 165.4 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [144.5 (C), 144.4 (C), C_c], [128.9 (CH), 128.8 (CH), 127.7 (CH), 125.7 (CH), 125.7 (CH), 125.1 (CH), 125.0 (CH), C_c], [67.1 (CH), 66.6 (CH), C-3], [60.0 (CH₂), 59.8 (CH₂), C_Gly-2], [59.2 (CH), 58.6 (CH), C-1], 55.8 (CH, C_c), [53.7 (CH), 53.3 (CH), C_Glu-2], [53.0 (CH₃), 52.8 (CH₃), COOCH₃], [45.1 (CH₃), 44.9 (CH₃), N(CH₃)₂], 43.0 (CH, C-4), [36.4 (CH₂), 34.4 (CH₂), 34.3 (CH₂), 33.1 (CH₂), 31.8 (CH₂), 31.0 (CH₂), 30.5 (CH₂), C_Glu-4 + C-6 + C-7 + C_c], [28.6 (CH₂), 28.4 (CH₂), 28.0 (CH₂), 27.6 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₆H₃₇N₄O₅]⁺ 485.2759, found 485.2744.

Synthesis of 16a

Following the general protocol (A), starting from (1R,3S,4S)-I (0.2157 g, 1.214 mmol), and after the typical workup, the crude oil was chromatographed using EtOAc as the eluent, affording 16a (0.5479 g, 0.9533 mmol) as a white solid with a low melting point. Yield: 79%. [α]²³_D: −44.63 ± 0.15 (c1.120, CHCl₃). ¹H-NMR (CDCl₃, 400 MHz) δ ppm (rotamers): 7.80–7.33 [7.72 (br s), 7.39 (br s), 1H, CONH], 6.00–5.75 [5.96 (d, J = 6.6 Hz), 5.82 (br s), 1H, CONH], 5.46 (br s, 1H, OCONH), 4.67–3.59 [3.70 (s) + 3.68 (s), 8H, H-1 + H-3 + H_Glu-2 + H_Gly-2 + COOCH₃], 2.81 (br s, 1H, H-4), 2.66 (d, J = 7.7 Hz, 2H, H_Glu-3), 2.24–1.54 [2.01 (br s), 1.93 (br s), 1.62 (br s), 23H, H_Glu-4 + H-5 + H-6 + H-7 + H_a], 1.40 (s, 9H, Boc). ¹³C{¹H}-NMR/DEPT-135 (CDCl₃, 101 MHz) δ ppm (rotamers): [172.6 (C), 171.7 (C), 169.9 (C), 168.4 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], 156.0 (C, Boc), 79.9 (C, Boc), [66.5 (CH), 66.1 (CH), C-3], [57.6 (CH), 57.3 (CH), C-1], [52.5, 52.1 (C), 51.8, C_Glu-2 + C_a + COOCH₃], 43.2 (CH₂, C_Gly-2), 41.4 (3CH₂, C_a), 40.5 (CH, C-4), 36.5 (CH₂, C-7), 36.4 (3CH₂, C_a), [33.5 (CH₂), 33.3 (CH₂), C_Glu-4], 31.1 (CH₂, C-6), 29.5 (3CH, C_a), 28.4 (3CH₃, Boc), [28.3 (CH₂), 27.5 (CH₂), 27.0 (CH₂), 25.6 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₃₀H₄₇N₄O₇]⁺ 575.3439, found 575.3416.

Synthesis of 16b

Following the general protocol (A), starting from (1R,3S,4S)-I (0.2301 g, 1.295 mmol), and after the typical workup, the crude oil was chromatographed using EtOAc as the eluent, affording 16b (0.6826 g, 1.1324 mmol) as a low melting point white solid. Yield: 87%. [α]²³_D: −27.53 ± 0.20 (c1.021, CHCl₃). ¹H-NMR (CDCl₃, 400 MHz) δ ppm (rotamers): 7.60–7.30 [7.54 (d, J = 7.7 Hz), 7.37 (d, J = 7.8 Hz), 1H, CONH], 5.85 (br s, 1H, CONH), 5.38 (br s, 1H, CONH), 4.52–4.38 (m, 1H, H_Glu-2), 4.15 (br s, 1H, H-3), 4.07–3.73 [4.02 (dd, J = 17.2, 4.3 Hz), 3.95 (br s), 3.91 (d, J = 16.7, 4.7 Hz), 3H, H_Gly-2 + H-1], 3.70 (s, 3H, COOCH₃), 2.89–2.79 [2.85 (d, J = 3.0 Hz), 2.82 (d, J = 3.4 Hz), 1H, H-4], [2.29–1.95 (m, 6H), 1.91–1.59 (m, 10H), 1.43 (s, 9H), 1.37–1.22 (m, 4H), 1.18–1.06 (m, 2H), H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + Boc + H_b], 0.84–0.79 [0.82 (s), 0.82 (s), 6H, 2CH₃, H_b]. ¹³C{¹H}-NMR/DEPT-135 (CDCl₃, 101 MHz) δ ppm (rotamers): [172.4 (C), 171.3 (C), 169.8 (C), 168.3 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], 155.9 (C, Boc), 79.9 (C, Boc), [66.5 (CH), 66.2 (CH), C-3], [57.6 (CH), 57.2 (CH), C-1], [53.9 (C), 53.6 (C), C_b], [52.5 (CH), 52.3 (CH), C_Glu-2], 51.8 (CH₃, COOCH₃), [50.7 (CH₂), 47.5 (CH₂), 43.3 (CH₂), 42.8 (CH₂), C_b + C_Gly-2], 40.3 (CH, C-4), [40.1 (CH₂), 37.6 (CH₂), 33.3 (CH₂), 33.2 (CH₂), C-7 + C_Glu-4], [32.5 (C), 32.4 (C), C_b], 31.2 (CH₂, C-6), 30.2 (2CH₃, C_b), 28.5 (3CH₃, Boc), [27.0 (CH₂), 26.8 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₃₂H₅₁N₄O₇]⁺ 603.3752, found 603.3737.

Synthesis of 16c

Following the general protocol (A), starting from (1R,3S,4S)-I (0.1875 g, 1.056 mmol), and after the typical workup, the crude oil was chromatographed using EtOAc as the eluent, affording 16c (0.5103 g, 0.9167 mmol) as a low melting point white solid. Yield: 87%. [α]²⁵_D: +15.14 ± 0.21 (c1.061, CHCl₃). ¹H-NMR (CDCl₃, 400 MHz) δ ppm (rotamers): [7.91 (d, J = 8.1 Hz), 7.34 (d, J = 7.9 Hz), 1H, CONH], 7.20–7.09 (m, 5H, CONH + H_c), 5.50–5.38 (m, 1H, H_c), 4.92–4.34 [4.83 (br s), 4.64–4.32 (m), OCONH + H_Glu-2], 4.23–3.28 [4.18 (br s), 3.89 (br s), 3.67 (s), 3.59 (s), 3.37 (dd, J = 17.1, 3.3 Hz), 7H, H-1 + H-3 + COOCH₃ + H_Gly-2], [3.13–2.92 (m, 1H), 2.89–2.69 (m, 2H), 2.59–2.39 (m, 3H), 2.29–2.17 (m, 1H), 2.13–1.52 (m, 7H), 1.47–1.37 (m, 10H), H-4 + H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + H_c + Boc]. ¹³C{¹H}-NMR/DEPT-135 (CDCl₃, 101 MHz) δ ppm (rotamers): [176.0 (C), 174.5 (C), 170.6 (C), 170.0 (C), 169.2 (C), 168.5 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [155.7 (C), 155.6 (C), Boc], [144.0 (C), 143.5 (C), 143.3 (C), 143.1 (C), C_c], [128.0 (CH), 127.7 (CH), 126.6 (CH), 126.4 (CH), 124.8 (CH), 124.7 (CH), 124.2 (CH), 124.1 (CH), C_c], [79.9 (C), 79.7 (C), Boc], [67.0 (CH), 66.1 (CH), C-3], [57.6 (CH), 57.0 (CH), C-1], [54.6 (CH), 54.4 (CH), C_c], [53.5 (CH), 53.0 (CH), C_Glu-2], [52.1 (CH₃), 52.0 (CH₃), COOCH₃], [43.3 (CH₂), 42.9 (CH₂), C_Gly-2], 40.5 (CH, C-4), [36.8 (CH₂), 36.4 (CH₂), 33.9 (CH₂), 33.6 (CH₂), 31.0 (CH₂), 30.8 (CH₂), 30.6 (CH₂), 30.5 (CH₂), 30.3 (CH₂), 30.3 (CH₂), C_Glu-4 + C-6 + C-7 + C_c], 28.5 (3CH₃, Boc), [27.5 (CH₂), 27.3 (CH₂), 26.7 (CH₂), 25.7 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₉H₄₁N₄O₇]⁺ 557.2970, found 557.2952.

Synthesis of 17a

Following the general protocol (B), starting from compound 16a (1.1124 g, 2.149 mmol) and after the typical work-up, 17a (1.1052 g, 1.8776 mmol) was afforded as a low melting point white solid. Yield: 97%. [α]¹⁸_D: −92.01 ± 0.17 (c1.320, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 4.64–4.25 [4.61 (br s), 4.42 (dd, J = 8.8, 5.0 Hz), 4.36 (dd, J = 9.3, 4.8 Hz), 4.29 (br s), 2H, H-3 + H_Glu-2], 4.07–3.75 (m, 3H, H_Gly-2 + H-1), 3.74–3.70 [3.73 (s), 3.72 (s), 3H, COOCH₃], 2.92–2.73 [2.91–2.82 (m), 2.78 (d, J = 4.1 Hz), 1H, H-4], 2.28–2.19 (m, 2H, H_Glu-3), [2.18–2.08 (m, 1H), 2.07–1.98 (m, 10H), 1.96–1.77 (m, 3H), 1.76–1.68 (m, 7H), 1.61–1.41 (m, 2H), H_Glu-4 + H-5 + H-6 + H-7 + H_a]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [173.9 (C), 173.5 (C), 172.2 (C), 165.2 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [67.1 (CH), 66.6 (CH), C-3], [59.1 (CH), 58.7 (CH), C-1], [53.8 (CH), 53.4 (CH), C_Glu-2], [52.9 (CH₃), (52.9 (C), 52.8 (C), 52.8 (CH₃), C_a + COOCH₃], 43.1 (CH, C-4), 42.3 (3CH₂, C_a), [41.5 (CH₂), 41.3 (CH₂), C-7], 37.5 (3CH₂, C_a), [36.3 (CH₂), 34.7 (CH₂), 34.1 (CH₂), 31.7 (CH₂), C_Gly-2 + C-6 + C_Glu-4], 30.9 (3CH, C_a), [28.8 (CH₂), 28.4 (CH₂), 28.1 (CH₂), 28.0 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₅H₃₉N₄O₅]⁺ 475.2915, found 475.2902.

Synthesis of 17b

Following the general protocol (B), starting from compound 16b (1.3757 g, 2.2823 mmol) and after the typical work-up, 17b (1.3871 g, 2.2493 mmol) was afforded as a white solid. Yield: 99%. Melting point: 48–49 °C. [α]²⁴_D: −7.50 ± 0.20 (c1.070, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 4.66–4.24 [4.61 (br s), 4.40 (dd, J = 8.9, 5.0 Hz), 4.35 (dd, J = 9.2, 4.9 Hz), 4.29 (br s), 2H, H_Glu-2 + H-3], [4.12–3.93 (m, 2H), 3.85–3.75 (m, 1H), H_Gly-2 + H-1], 3.74–3.71 [3.73 (s), 3.72 (s), 3H, COOCH₃], 2.91–2.73 [2.86 (d, J = 4.3 Hz), 2.77 (d, J = 4.1 Hz), 1H, H-4], [2.29–2.18 (m, 2H), 2.16–1.75 (m, 8H), 1.73–1.52 (m, 6H), 1.50–1.28 (m, 5H), 1.15 (br s, 2H), H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + H_b], 0.85 (s, 6H, 2CH₃, H_b). ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [174.0 (C), 173.5 (C), 172.2 (C), 165.2 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [67.0 (CH), 66.6 (CH), C-3], [59.0 (CH), 58.7 (CH), C-1], [54.5 (C), 54.5 (C), C_b], [53.8 (CH), 53.4 (CH), C_Glu-2], [52.9 (CH₃), 52.8 (CH₃), COOCH₃], [51.7 (CH₂), 48.3 (CH₂), 44.7 (CH), 43.8 (CH₂), C_b + C-4], [41.5 (CH₂), 41.3 (CH₂), 40.7 (CH₂), 36.4 (CH₂), 33.2 (C, C_b), 34.7 (CH₂), 34.1 (CH₂), C_Gly-2 + C-7 + C_Glu-4], [31.7 (CH₂), 31.5 (CH), 30.5 (CH₃), 30.4 (CH₂), C-6 + C_b], [28.7 (CH₂), 28.4 (CH₂), 28.1 (CH₂), 27.9 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₇H₄₃N₄O₅]⁺ 503.3228, found 503.3215.

Synthesis of 17c

Following the general protocol (B), starting from compound 16c (0.5380 g, 0.9665 mmol) and after the typical work-up, 17c (0.5498 g, 0.9636 mmol) was afforded as a colorless oil. Yield: 100%. [α]²⁵_D: +16.51 ± 0.20 (c0.950, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 7.27–7.05 (m, 4H, H_c), 5.41–5.31 (m, 1H, H_c), 4.66–4.21 [4.56 (br s), 4.43 (dd, J = 9.0, 5.1 Hz), 4.37 (dd, J = 9.1, 4.9 Hz), 4.22 (br s), 2H, H-3 + H_Glu-2], 4.10–3.86 [4.02 (s), 3.95 (s), 2H, H-1 + H_Gly-2a], 3.82–3.69 [3.74 (s), 3.73 (s), 4H, H_Gly-2b + COOCH₃], [3.06–2.94 (m, 1H), 2.92–2.77 (m, 2H), 2.55–2.44 (m, 1H), 2.41–2.34 (m, 2H), 2.29–2.16 (m, 1H), 2.13–1.96 (m, 2H), 1.93–1.76 (m, 3H), 1.73–1.52 (m, 2H), 1.50–1.40 (m, 1H), H-4 + H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + H_c]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [174.6 (C), 173.5 (C), 172.2 (C), 165.2 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [144.5 (C), 144.4 (C), C_c], [128.9 (CH), 127.7 (CH), 125.7 (CH), 125.0 (CH), C_c], [67.1 (CH), 66.6 (CH), C-3], [59.0 (CH), 58.7 (CH), C-1], 55.9 (CH, C_c), [53.7 (CH₂), 53.5 (CH₂), C_Glu-2], [53.0 (CH₃), 52.8 (CH₃), COOCH₃], 43.0 (CH, C-4), [41.5 (CH₂), 41.4 (CH₂), C_Gly-2], [36.4 (CH₂), 34.7 (CH₂), 34.3 (CH₂), 33.2 (CH₂), 31.8 (CH₂), 31.0 (CH₂), 30.5 (CH₂), C_Glu-4 + C-6 + C-7 + C_c], [28.6 (CH₂), 28.3 (CH₂), 28.1 (CH₂), 27.8 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₄H₃₃N₄O₅]⁺ 457.2446, found 457.2423.

Synthesis of 18a

Following the general protocol (C), starting from compound 17a (0.2859 g, 0.4857 mmol), and after the typical workup, the crude oil was chromatographed using CH₂Cl₂/CH₃OH (4:1) as the eluent, affording 18a (0.2318 g, 0.4611 mmol) as a white solid. Yield: 95%. Melting point: 105–108 °C. [α]¹⁸_D: +47.57 ± 0.23 (c1.030, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 4.65–4.31 [4.60 (br s), 4.47–4.33 (m), 2H, H-3 + H_Glu-2], 4.14–3.93 [4.09 (br s), 3.97 (br s), 1H, H-1], 3.79–3.68 [3.74 (s), 3.72 (s), 3H, COOCH₃], 3.65–3.33 (m, 2H, H_Gly-2), 2.86–2.70 [2.81 (d, J = 3.4 Hz), 2.73 (d, J = 3.2 Hz), 1H, H-4], 2.66–2.40 [2.55 (s), 2.52 (s), 6H, N(CH₃)₂], [2.30–2.12 (m, 3H), 2.11–1.94 (m, 10H), 1.93–1.79 (m, 3H), 1.78–1.66 (7H), 1.63–1.50 (m, 1H), 1.46–1.28 (m, 1H), H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + Ca]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [173.9 (C), 173.6 (C), 172.3 (C), 167.9 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [67.2 (CH), 66.6 (CH), C-3], [61.4 (CH₂), 61.1 (CH₂), C_Gly-2], [59.3 (CH), 58.4 (CH), C-1], [53.7 (CH), 53.1 (CH), C_Glu-2], [52.9 (CH₃), 52.8 (C), 52.8 (CH₃), COOCH₃ + C_a], [45.6 (CH₃), 45.3 (CH₃), N(CH₃)₂], 42.9 (CH, C-4), [42.3 (CH₂), 43.3 (CH₂), 3CH₂, C_a], [37.5 (CH₂), 36.4 (CH₂), C-7 + C_a], [34.5 (CH₂), 34.0 (CH₂), 32.0 (CH₂), C_Glu-4 + C-6], 30.9 (3CH, C_a), [28.8 (CH₂), 28.5 (CH₂), 28.2 (CH₂), 27.9 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₇H₄₃N₄O₅]⁺ 503.3228, found 503.3209.

Synthesis of 18b

Following the general protocol (C), starting from compound 17b (0.2327 g, 0.3773 mmol), and after the typical workup, the crude oil was chromatographed using CH₂Cl₂/CH₃OH (4 : 1) as the eluent, affording 18b (0.1161 g, 0.2188 mmol) as a white solid. Yield: 58%. Melting point: 103–109 °C. [α]²⁵_D: −15.10 ± 0.20 (c1.120, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 4.67–4.29 [4.63 (br s), 4.47–4.32 (m), 2H, H-3 + H_Glu-2], 4.11–3.95 [4.06 (br s), 3.99 (br s), 1H, H-1], 3.85–3.55 [3.74 (s), 3.72 (s), 5H, H_Gly-2 + COOCH₃], 2.85–2.73 [2.83 (d, J = 3.7 Hz), 2.74 (d, J = 3.5 Hz), 1H, H-4], 2.72–2.59 [2.69 (s), 2.63 (s), 6H, N(CH₃)₂], [2.30–2.09 (m, 4H), 2.08–1.99 (m, 1H), 1.95–1.78 (m, 5H), 1.73–1.54 (m, 6H), 1.47–1.28 (m, 5H), 1.18–1.14 (m, 2H), H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + H_b], 0.91–0.80 [0.86 (s), 0.85 (s), 6H, 2CH₃, H_b]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [173.9 (C), 173.5 (C), 172.2 (C), 166.8 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [67.1 (CH), 66.6 (CH), C-3], [60.8 (CH₂), 60.4 (CH₂), C_Gly-2], [59.3 (CH), 58.5 (CH), C-1], 54.5 (C, C_b), [53.8 (CH), 53.2 (CH), C_Glu-2], [52.9 (CH₃), 52.8 (CH₃), COOCH₃], [51.7 (CH₂), 48.2 (CH₂), C_b], [45.4 (CH₃), 45.1 (CH₃), N(CH₃)₂], 43.8 (CH₂, C_b), 42.9 (CH, C-4), [40.8 (CH₂), 40.7 (CH₂), 36.0 (CH₂), 34.5 (CH₂), 34.0 (CH₂), C-7 + Glu-4], 33.2 (C, C_b), 31.9 (CH₂, C-6), 31.6 (CH, C_b), 30.7 (2CH₃, C_b), [28.8 (CH₂), 28.4 (CH₂), 28.1 (CH₂), 27.8 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₉H₄₇N₄O₅]⁺ 531.3541, found 531.3524.

Synthesis of 18c

Following the general protocol (C), starting from compound 17c (0.5583 g, 0.9785 mmol), and after the typical workup, the crude oil was chromatographed using CH₂Cl₂/CH₃OH (4 : 1) as the eluent, affording 18c (0.2998 g, 0.6186 mmol) as a yellow oil. Yield: 63%. [α]²³_D: −50.09 ± 0.20 (c1.050, CH₃OH). ¹H-NMR (CD₃OD, 400 MHz) δ ppm (rotamers): 7.34–7.06 (m, 4H, H_c), 5.44–5.32 (m, 1H, H_c), 4.64–4.31 [4.59 (br s), 4.47–4.33 (m), 2H, H-3 + H_Glu-2], 4.20–3.84 [4.17 (br s), 3.95 (br s), 3.91 (br s), 1H, H-1], 3.67 (s, 3H, COOCH₃), 3.47–3.19 (m, 2H, H_Gly-2), [3.09–2.77 (m, 2H), 2.74–2.62 (m, 1H), 2.53–2.42 (m, 3H), 2.41–2.28 (m, 6H), 2.26–2.09 (m, 1H), 2.07–1.90 (m, 3H), 1.87–1.74 (m, 2H), 1.72–1.37 (m, 3H), H-4 + H_Glu-3 + H_Glu-4 + H-5 + H-6 + H-7 + N(CH₃)₂ + H_c]. ¹³C{¹H}-NMR/DEPT-135 (CD₃OD, 101 MHz) δ ppm (rotamers): [175.1 (C), 173.2 (C), 172.5 (C), 169.5 (C), C_Glu-1 + C_Glu-5 + C_Gly-1 + CONH], [144.5 (C), 144.4 (C), C_c], [128.8 (CH), 128.8 (CH), 127.6 (CH), 125.6 (CH), 125.8 (CH), 128.6 (CH), 125.1 (CH), 124.9 (CH), C_c], [67.1 (CH), 66.9 (CH), C-3], [61.8 (CH₂), 61.3 (CH₂), C_Gly-2], [59.3 (CH), 59.1 (CH), 58.2 (CH), 58.1 (CH), C-1 + C_c], [54.2 (CH), 53.9 (CH), C_Glu-2], [52.3 (CH₃), 52.3 (CH₃), COOCH₃], [45.6 (CH₃), 45.6 (CH₃), 45.5 (CH₃), N(CH₃)₂], [42.8 (CH), 42.7 (CH), C-4], [36.7 (CH₂), 36.7 (CH₂), 34.4 (CH₂), 34.3 (CH₂), 34.0 (CH₂), 33.9 (CH₂), 31.9 (CH₂), 31.8 (CH₂), 31.3 (CH₂), 31.2 (CH₂), 31.2 (CH₂), 31.1 (CH₂), 31.0 (CH₂), C_Glu-4 + C-6 + C-7 + C_c], [28.6 (CH₂), 28.5 (CH₂), 28.2 (CH₂), 28.1 (CH₂), C-5 + C_Glu-3]. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for [C₂₆H₃₇N₄O₅]⁺ 485.2759, found 485.2744.

Biological assays

Materials

The SH-SY5Y cell line was obtained from Sigma-Aldrich (Taufkirchen, Germany). Dulbecco's modified Eagle's medium (DMEM) [DMEM (1×) + GlutaMAX], DMEM/F-12, and fetal bovine serum (FBS) were obtained from Gibco, Alfagene (Carcavelos, Portugal). Penicillin and streptomycin were obtained from Biotecnómica (Porto, Portugal). PBS (without calcium and magnesium) was obtained from Biochrom (Berlin, Germany). Trypsin/ethylenediaminetetraacetic acid (EDTA) solution, trypan blue solution 0.4% (w/v) RA, TPA, NR dye, PQ, Aβ_25–35, sodium chloride, potassium phosphate monobasic, potassium chloride, sodium phosphate dibasic, sodium bicarbonate, d-glucose and 2-(4-amidinophenyl)-6-indolecarbamidine dihydrochloride (DAPI) were obtained from Sigma-Aldrich (Taufkirchen, Germany). DMSO and Hanks' balanced salt solution (HBSS) were obtained from Merck (Darmstadt, Germany). MTT and thioflavin T were obtained from Alfa Aesar (Kandel, Germany). The cells were maintained in a Heraeus incubator (Hanau, Germany) at 37 °C with 5% CO₂ for maintenance and experiments.

Cytotoxicity in non-differentiated human neuronal SH-SY5Y cells

SH-SY5Y cells were maintained in a complete medium DMEM/F-12 containing 10% FBS and 1% penicillin/streptomycin and cells were kept at 37 °C with 5% CO₂ throughout all the procedures. For experiments, cells were seeded in multiwell plates as previously described by our research group,⁵⁸ and cells were exposed to the compounds of interest, 13–18(a–c). After 24 h, the MTT reduction assay was performed as a classical cytotoxicity assay.³⁹ All the compounds tested were solubilized in sterile DMSO (5% in PBS final concentration) and the results are presented as the percentage of the vehicle.

Quantification of protein aggregates

On black bottom-96-well plates, non-differentiated SH-SY5Y cells were seeded at a density of 3 × 10⁴ cells per well.⁵⁸ On the following day, they were incubated with the selected compounds in the presence of Aβ_25–35 at 10 μM.⁵⁸ After an incubation period of 24 h, cells were washed with HBSS and incubated for 30 min with thioflavin T at 5 μM (prepared in HBSS).⁵⁸ At this point, fluorescence was read at 450 (excitation)/482 nm (emission).⁵⁸ Results represent the fold decrease in fluorescence compared to the positive control of Aβ_25–35 at 10 μM alone.

Cytotoxicity in human neuronal SH-SY5Y differentiated cell culture

Before differentiation, SH-SY5Y cells were maintained in a complete DMEM medium supplemented with 10% (v/v) FBS and 1% (v/v) of antibiotics (100 units per mL penicillin and 100 μg mL⁻¹ streptomycin) and maintained at 37 °C in a 5% CO₂ incubator throughout all the procedures. For experiments, cells were seeded in multiwell plates at a density of 25 000 cells per cm² in a complete DMEM medium containing 10 μM RA, and maintained for 3 days. On day 3, cells were then exposed to 80 nM TPA in complete DMEM medium and kept for another 3 days to obtain a dopaminergic phenotype.^36,38 After the differentiation protocol, on day 6, cells were exposed to the compounds under study [13–18(a–c), Glypromate, and (R)-1-aminoindane] at 100 μM for cytotoxicity evaluation. In a new set of experiments, selected compounds were co-incubated with 6-OHDA at 125 μM, or PQ at 300 μM for 48 h in new complete DMEM to determine neuroprotection, using MTT reduction and NR uptake assays. 6-OHDA and PQ were used herein as toxins that elicit features resembling the PD phenotype.^38,39 All the drugs tested were solubilized in sterile PBS (6-OHDA was prepared immediately before incubation).

MTT reduction assay

The tetrazolium salt is reduced to the corresponding formazan by dehydrogenases, predominantly mitochondrial.^38,48 After the 24 h (for undifferentiated SH-SY5Y cells) or 48 h incubation (for differentiated SH-SY5Y cells), MTT was added and the protocol was followed as described.^39,58 The values are expressed as a percentage of control/vehicle incubated cells.

NR uptake assay

The NR dye is incorporated into viable cells and accumulates in lysosomes.⁴⁴ This assay is based on the uptake of NR, an eurhodin dye that is retained within fully functioning lysosomes via active transport.⁴⁴ After the 48 h incubation, the cellular medium was changed to 250 μL per well of NR solution (33 μg mL⁻¹ of NR in DMEM). Then, plates were incubated for 1.5 h (at 37 °C, protected from light). After this time, the NR solution was removed and 250 μL per well of warm HBSS was used for a washing step. Then, this solution was replaced by the solvent (50% ethanol/1% acetic acid) to extract the NR dye from viable cells, followed by 15 min agitation protected from light. The absorbance was measured at 540 and 690 nm.⁴⁴ The values are expressed as a percentage of control cells that were set to 100%, after subtracting the absorbance reference value of each well.

Statistical analysis

The results are expressed in mean ± standard deviation. An ordinary ANOVA statistical analysis was done followed by uncorrected Fisher LSD. The GraphPad Prism 8.3 (CA, USA) software was used to perform all statistical analyses.

Abbreviations

6-OHDA

6-Hydroxydopamine

8PM

(−)-8-Phenylmenthol

Aβ

Amyloid β

AD

Alzheimer's disease

API

Active pharmaceutical ingredient

BBB

Blood-brain barrier

CNS

Central nervous system

DAPI

2-(4-Amidinophenyl)-6-indolecarbamidine dihydrochloride

DMEM

Dulbecco's modified Eagle medium

DMSO

Dimethyl sulfoxide

DEPT-135

Distortionless enhancement by polarization transfer-135

EDTA

Ethylenediaminetetraacetic acid

ESI

Electrospray ionization

Et₃N

Triethylamine

EtOAc

Ethyl acetate

FBS

Fetal bovine serum

GPE

Glycyl-l-prolyl-l-glutamic acid

HBSS

Hanks' balanced salt solution

HIA

Human intestinal absorption

IGF-1

Insulin like growth factor 1

MTT

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

NMR

Nuclear magnetic resonance

NR

Neutral red

PBS

Phosphate-buffered saline

PD

Parkinson's disease

PQ

Paraquat

RA

Retinoic acid

TBTU

2-(1H-Benzotriazol-1-yl)-1,1,3,3-tetramethyluronium tetrafluoroborate

TFA

Trifluoroacetic acid

TPA

12-O-Tetradecanoylphorbol-13-acetate

TPSA

Topological polar surface area

Data availability

The data supporting this article have been included as part of the ESI.† This includes copies of NMR (¹H, ¹³C{¹H}, DEPT-135) spectra (PDF).

Author contributions

Sara C. Silva-Reis: conceptualization, formal analysis, investigation, methodology, project administration, writing – original draft. Vera M. Costa: conceptualization, formal analysis, funding acquisition, methodology, project administration, supervision, writing – review & editing. Daniela Correia da Silva: investigation. David M. Pereira: writing – review & editing. Xavier Cruz Correia: formal analysis, writing – review & editing. Xerardo García-Mera: formal analysis, project administration, supervision. José E. Rodríguez-Borges: formal analysis, funding acquisition, project administration, supervision. Ivo E. Sampaio-Dias: conceptualization, formal analysis, funding acquisition, methodology, project administration, supervision, writing – review & editing.

Conflicts of interest

There are no conflicts to declare.