The role of carotenoids from red mamey fruit ( Pouteria sapota ) against amyloid-β monomers in Alzheimer's disease: Computational analysis and ADMET prediction

Abstract

Background

Carotenoids, potent antioxidants in fruits and vegetables, have recently garnered attention for their potential therapeutic effects against neurodegenerative diseases. This study focuses on the interaction and anti-aggregation properties of conventional and unconventional carotenoids found in red mamey fruit, a nutraceutical fruit that is a rich source of these compounds.

Objective

To assess computational the interaction between of amyloid-β (Aβ) peptide with a set of carotenoids and three carotenoids previously explored in experimental assays as well as to assess ADMET prediction of carotenoids selected by computational analysis results.

Methods

We analyzed the interaction between these carotenoids and Aβ peptides using molecular docking, a key factor in Alzheimer's disease (AD) pathology. Selected carotenoids were compared with the reference compounds cryptocapsin (1), zeaxanthin (2), lutein (3), and cryptocapsin-5,6-epoxide (24), which previously demonstrated significant anti-aggregation activity against Aβ. Using SwissADME and ADMET Predictor^® software to determine the pharmaceutical analysis prediction.

Results

Computational analysis identified (5R,8R)-sapotexanthin-5,8-epoxide (19) and (5S,8S)-cryptocapsin-5,8-epoxide (26) as the most promising candidates due to their strong binding affinity and potential anti-aggregation properties against Aβ. The pharmaceutical analysis identified compounds (5S,8S)-cryptocapsin-5,8-epoxide (26) and (5R,8R)-cryptocapsin-5,8-epoxide (28) as the most promising compounds. Our findings suggest that specific modifications in the carotenoid structure, particularly modifications in the type of epoxidation and stereochemistry, can significantly influence the biological activity of carotenoids and biopharmaceutical performance.

Conclusions

These results provide valuable insights for future in vitro studies of most potential carotenoids (19 and 26) and the development of potential therapeutic agents for AD.

Keywords

ADMET prediction Alzheimer's disease AutoDock 4 AutoDockVina carotenoid molecular docking nutraceutic red mamey fruit Smina

Introduction

Alzheimer's disease (AD) is a complex and challenging disorder affecting older adults 60 years and older in the world.¹ The etiology and pathology of AD are still not clear and are multifactorial.^2–5 Currently, AD affects 55 million and by 2050 estimates show that 150 million people will be affected.⁶ AD is clinically challenging because of a lack of biomarkers for early diagnosis and a lack of pharmaceutical treatments to clinically manage the disease.⁶ The pathology of the disease includes changes in genomics stability, over-expression of amyloid-β peptide (Aβ), accumulation of misfolded proteins, metal accumulation, changes in cell signaling pathways, alterations in gene expressions, phosphorylation of tau, neurofibrillary tangles, and others.^7–9 The current focus in AD pathology includes Aβ peptide aggregation and the mechanism of Aβ induced cell dysfunction leading to neurodegeneration in AD.^8,9 Ongoing efforts are focused on identifying small molecules, both synthetic and natural, in modulating Aβ production through augmenting β-secretase pathway in AβPP metabolism and preventing Aβ aggregation to protect cellular dysfunction.¹⁰ To this end, screening of novel molecules through omics approach can lead to novel therapeutics for AD.¹¹ Molecules from natural sources, which have unique structural properties, have become ideal molecules to evaluate as future drugs for AD as these products often contain neuroprotective properties.¹² There are studies on the role of natural molecules as curcumin, carotenoids and others in the prevention of AD related pathology.^3,13–18 These natural molecules have issues with reference to solubility and blood-brain-barrier penetration, but have very high protective activity for cell survival, with less side effects than other molecules.¹⁸ In particular, carotenoids have been identified as natural products that can inhibit neurodegenerative pathways.^19–21 As they are lipophilic, most carotenoids can cross the blood‒brain barrier.²² Over 750 carotenoids have been found, but the anti-aggregation activity of only a few has been explored.^19,22 Carotenoids generally contain 40 carbons (C₄₀) and are composed of two terminal members (A and A’) connected by a conjugated long chain (B) (Figure 1). Carotenoids can be classified as carotenes and xanthophylls, which can be acyclic or cyclic.^23–25

Figure 1.

Schematic representation of a carotenoid.

While carotenes only contain carbon and hydrogen, xanthophylls consist of carbon, hydrogen, and oxygen; however, there are few exceptions, including apocarotenoids (containing oxygen and fewer than C₄₀) and acyclic carotene (without oxygen and C₄₀). Most carotenoids share a common beta-ionone ring on both sides (A and A’) of the molecule, as seen in β-carotene. Unconventional carotenoids also contain a keto-kappa group on one or both sites, such as cryptocapsin (1) or capsorubin (30). The conjugated long chain (B) can be either a trans-carotenoid isomer or a cis-carotenoid isomer, and the former is most predominant in nature (Figure 1).^26–30

In the present investigation, we focus on novel carotenoids from red mamey fruit, as the fruit contains unconventional carotenoids with ĸ-ring end groups, such as sapotexanthin (16), cryptocapsin (1), capsorubin (30) and other carotenoids (Figure 2).^16,20,31,32 Previously, our group reported that cryptocapsin (1) exhibited greater anti-amyloidal activity than zeaxanthin (2) and that the anti-amyloidal activity of 1 decreased when the compound contained an epoxide, such as cryptocapsin-5,6-epoxide (24).¹⁹ A computational analysis comparing cryptocapsin with lutein, lycopene, astaxanthin, fucoxanthin, and bixin revealed that cryptocapsin and lutein (3) exhibited the greatest anti-amyloid activity.²⁰ Additionally, we recently summarized the chemistry and biological benefits of carotenoids from red mamey fruit.¹⁶ Although the red mamey fruit contains more than thirty carotenoids (Figure 2), only two carotenoids (1, 24) have been experimentally evaluated for their anti-amyloidogenic activity against AD.¹⁶ Due to the isolation of carotenoids from red mamey fruit require high amount of fruit and the experimental process can take a long time, we can use computational analyses to explore which of the 30 carotenoids have more potential to inhibit Aβ aggregation.¹⁶ Selected carotenoids were compared with the reference compounds cryptocapsin (1), zeaxanthin (2), lutein (3), and cryptocapsin-5,6-epoxide (24), which previously demonstrated significant anti-aggregation activity against Aβ. Zeaxanthin (2) and lutein (3) are not present in red mamey fruit (Figure 2).

Figure 2.

Carotenoids from P. sapota (red mamey fruit) selected for computational analysis.

We conducted molecular docking and molecular dynamic (MD) experiments on the interaction of thirty carotenoids from mamey fruit with Aβ₄₀ and Aβ₄₂ monomers to predict potential molecules that block Aβ aggregation kinetics. Using computational tools, we aimed to identify the most likely carotenoid(s) that can prevent the formation of Aβ aggregation. The findings from this study may guide the selective extraction of natural sources or facilitate semisynthetic carotenoids for future preclinical trials.

Methods

The aim of our study was to investigate the anti-aggregation properties of a series of carotenoids derived from natural products against the Aβ peptide. The first step involved performing geometry optimization and charge calculations for these carotenoids using density functional dispersion-corrected density functional theory (DFT) calculations with the B3LYP-D3 functional and 6–31 + g(d) basis set. To account for solvent effects, we employed the SMD model.³³ All calculations were conducted using Gaussian 16 software.³⁴

To account for the flexibility of the Aβ peptide, which is an important factor in studying anti-aggregation properties, we employed an MD protocol based on methodology described by Orjuela et al.¹⁴ Three Aβ-peptide models were used: Aβ₄₀ (PDB code: 1AML),³⁵ Aβ₄₀ (PDB code: 1BA4),³⁶ and Aβ₄₂ (PDB code: 1Z0Q).³⁷ Through exploring the conformational space of these models, we employed a robust molecular docking protocol, utilizing multiple computational packages to mitigate biases in the scoring function and ensure reliable results.

Molecular dynamics protocol of Aβ peptide

To capture the diverse range of conformations exhibited by the Aβ peptide and simulate the interactions with carotenoids over time, we performed MD simulations using the following parameters. The simulations were conducted using the Amber 18 packages.³⁸ The Aβ-peptide models were described using the FF14SB force field,³⁹ while the solvent system consisted of a periodic box of water with a NaCl concentration of 0.1 nM, employing the TIP3P force field.⁴⁰

The simulation protocol began with a 2 ns MD simulation in a canonical (NVT) ensemble to gradually heat the system from 0 to 300 K. Subsequently, the system density was equilibrated through a 5 ns MD simulation in an isothermal-isobaric ensemble. Finally, a 100 ns production phase was performed using an NVT ensemble to explore the conformational space of the Aβ peptide.

To identify representative conformations from the production phase, we employed the CPPTRAJ package⁴¹ and applied the K-means algorithm⁴² with 10 centroids, utilizing RMSD as the distance metric. All obtained conformations were analyzed using PyMOL⁴³ for further characterization and visualization.

Molecular docking of Aβ peptide

Considering all representative conformations of models of Aβ_40/42 to molecular docking with carotenoids to calculate the carotenoid binding to every Aβ peptide model, we used a weighted scoring function for each carotenoid using the percentage of the individual representation with the following equation:

S_{w} = \sum X_{i} R_{i}

(1)

where S_w is the binding score weighted of model i, X_i is the binding score of representation i, and R_i is the percentage of representation i. To consider all score functions of the molecular docking software, we used an exponential consensus ranking (ECR) described by Palacio et al.⁴⁴ We used the following molecular docking programs: AutoDock 4 (ATD),⁴⁵ AutoDock Vina (Vina),⁴⁶ Smina (Vinardo scoring function),⁴⁷ QuickVina 2,⁴⁸ QuickVina W,⁴⁹ and PSOVina.⁵⁰ Files used in molecular docking programs were prepared in AutoDock Tools 4.⁴⁵ To describe the charges of the Aβ peptide, we calculated the Gasteiger charges. The charges of carotenoids were obtained from the DFT calculations. The binding site of carotenoids is not known. The search space of molecular docking covers the entire surface of the peptide to determine the best affinity position. The interactions were analyzed with Maestro software.⁵¹

The choice of IZOQ and 1AML for the analyses was based on their specific structural characteristics relevant to our study. Despite their low percentile ranks in the PDB, these structures possess crucial features for understanding the interactions under investigation. Additionally, 100 ns of molecular dynamics simulations were performed to relax the structures at physiological pH, improving several PDB percentiles. Although other structures exist, these models have been previously used in similar studies. Therefore, IZOQ and 1AML were considered the most appropriate to ensure the reliability and relevance of our findings.

We considered that these 100 ns simulations were sufficient to achieve convergence and obtain reliable results consistent with experimental observations. Additionally, 10 clusters were obtained for each model to perform the molecular dynamics, ensuring a thorough sampling of the conformational space. Therefore, we determined that 100 ns was adequate for the current study. This approach balances computational efficiency with the accuracy needed to capture the relevant dynamics and interactions. In previous studies, the same models of Aβ peptides were used with 100 ns of molecular dynamics to reproduce the experimental studies on the anti-aggregating properties of carotenoids.¹⁴

Prediction of physicochemical and ADMET properties

The SwissADME is widely used for predicting physicochemical properties and we used it to analyze 30 compounds. We also used an ADMET Predictor^® in order to determine these properties.^52,53 Lipinski's rules were used to analyze the compounds that do not respect the threshold values of these principles, including a molecular weight of less than 500 g/mol, no more than 5 donor bonds, no more than 10 acceptor bonds, and a partition coefficient (LogP) no more than 5 and routable bonds equal or minor of 5. If it complies, it will have adequate passive diffusion across cell membranes and effective absorption from the intestine to the blood. If more than two do not comply, poor absorption and permeability are expected. This rule is used for oral administration of drug.

In other way, Veber's rule states that a compound that has 10 or fewer rotatable bonds and a polar surface area no greater than 140 Å² is likely to exhibit good oral bioavailability and less than 60 Å² good permeability in brain membrane.

ADMET analysis

Software ADMET Predictor^® version 10.4 (Simulations Plus Inc., Lancaster, CA, USA) was used as in silico simulation tool. All compounds were subjected to an ADMET analysis for the purpose of gaining insight into the biopharmaceutics and pharmacokinetic properties which includes prediction of several properties such as the absorption, distribution, metabolism, excretion, and toxicity.^52–55 Simulations of nine selected carotenoids was performed with a human oral dose of 10 mg until 120 h. For each structure various pharmacokinetic parameters were assessed, namely, fraction absorbed (Fa%), bioavailability (F%), Human volume of distribution (Vd), and clearance (L/h) and ADMET parameters. These evaluations offer interesting details into the potential behavior and suitability of the compounds for further development.

Results

Aβ monomer molecular dynamics

To consider the high flexibility of the Aβ peptide, we started from a representative structure of the model reported in PDB (code: 1AML). This is a model of Aβ₄₀ for which the MD simulations showed a significant fluctuation in three regions. The first region, from Asp1 to Tyr10, describes a nonstructural region. The second region, from Gly25 to Gly29, relates to a random coil. Finally, the third/last region, from Val35 to Val40, describes another nonstructural region, as shown in Supplemental Figure 1. In some representative structures, the α-helices were unfolded (Supplemental Figure 2). In comparison, the Aβ₄₀ 1BA4 and 1Z0Q models exhibited the same behavior in the first and last regions as the 1AML model. However, the random coil was less mobile in these models than the 1AML model. This difference is clearly seen in the positions of the α-helices within the model representations in Supplemental Figure 2.

Molecular docking

The results of the molecular docking simulations are presented in Table 1. The carotenoids (5R,8R)-sapotexanthin-5,8-epoxide (19) and (5R,8R)-cryptocapsin-5,8-epoxide (28) exhibit the best ranking across all three models of the Aβ peptide. Despite the differences in the representations of the models for Aβ, due to the molecular dynamics, the results highlight the robustness of our methods.

Table 1.

Exponential consensus ranking (ECR) of carotenoids and Aβ peptide.

Model Aβ₁₋₄₀ 1AML
Carotenoid	ATD	VINA	Smina	Qvina2	QvinaW	PSOVINA	Rank
19	−7.18	−7.13	−9.25	−7.14	−7.10	−6.39	1.26
28	−7.30	−7.06	−9.37	−7.08	−7.10	−6.22	1.26
26	−6.95	−7.11	−9.20	−7.13	−7.07	−6.22	1.25
20	−7.15	−6.96	−8.59	−6.95	−6.86	−6.14	1.22
1	−7.40	−6.75	−8.83	−6.69	−6.66	−5.67	1.21
3	−7.39	−6.60	−8.69	−6.66	−6.64	−5.86	1.21
2	−6.97	−6.72	−8.85	−6.63	−6.61	−5.99	1.21
Model Aβ₁₋₄₀ 1BA4
Carotenoid	ATD	VINA	Smina	Qvina2	QvinaW	PSOVINA	Rank
19	−7.34	−7.30	−9.21	−7.39	−7.44	−6.66	1.28
28	−7.64	−7.30	−9.02	−7.32	−7.36	−6.54	1.28
4	−7.17	−7.13	−8.94	−7.12	−7.16	−6.46	1.25
26	−6.85	−7.18	−8.94	−7.13	−7.19	−6.24	1.24
3	−7.44	−6.89	−8.97	−6.87	−6.93	−6.04	1.24
2	−7.08	−6.91	−9.02	−6.96	−6.96	−5.99	1.23
1	−7.46	−6.85	−8.67	−6.90	−6.87	−5.99	1.23
Model Aβ₁₋₄₂ 1Z0Q
Carotenoid	ATD	VINA	Smina	Qvina2	QvinaW	PSOVINA	Rank
19	−7.18	−7.45	−9.52	−7.36	−7.35	−6.86	1.29
26	−7.16	−7.40	−9.44	−7.37	−7.35	−6.91	1.29
28	−7.23	−7.33	−9.25	−7.33	−7.28	−6.60	1.28
4	−6.74	−7.30	−9.28	−7.18	−7.21	−6.55	1.26
3	−6.84	−7.16	−9.03	−7.16	−7.21	−6.30	1.25
1	−7.03	−7.16	−9.07	−7.03	−7.08	−6.02	1.24
2	−6.39	−7.19	−9.19	−7.13	−7.17	−6.19	1.24

The molecular docking calculations performed for the Aβ₄₂ peptide model (1Z0Q) exhibit a strong correlation with the experimental data of the carotenoids cryptocapsin (1), zeaxanthin (2), and lutein (3). Furthermore, the experimental models contain monomers of the Aβ₄₂ peptide, allowing us to investigate the interactions between these models and the identified ranking.

In all models of the Aβ peptide, (5R,8R)-sapotexanthin-5,8-epoxide (19) showed alkyl–alkyl interactions with Val12, His6, and Phe19, as well as an alkyl–π interaction with Phe19. These interactions occur along the peptide and are produced by the hydrophobic amino acids of the α-helix. Furthermore, the formation of a hydrogen bond between His6 and Ser8 contributes to the overall stabilization of the amyloid peptide. These interactions could impede the aggregation of the amyloid beta peptide by stabilizing the alpha-helices and preventing amyloid monomers to form beta-sheets. A representation of these interactions is presented in Figure 3.

Figure 3.

(a) Sapotexanthin (16) and (b) (5R,8R) sapotexanthin epoxide (19) interactions with Aβ₄₂: line (H-bond), dash dots (π–σ interaction), dash line (alkyl–alkyl interaction). Distances are in angstroms.

When the interaction of carotenoid 19, specifically (5R,8R)-sapotexanthin epoxide] was compared with the base structure of carotenoid 16 (sapotexanthin), a reduction in hydrophobic interactions was observed. However, notably, the binding mode with the amyloid peptide remained unchanged, as illustrated in Figure 3.

Carotenoid 26, (5R,8R)-cryptocapsin-5,8-epoxide, interacts with the Aβ peptide similarly to carotenoid 21. However, in the case of carotenoid 26, its hydroxyl group forms hydrogen bonds with His6. This conformation reduces the hydrophobic interactions with the peptide's side chains, as depicted in Figure 4. Interestingly, the binding mode observed in cryptocapsin (1) remains consistent with carotenoid 26, highlighting the significance of His6 in forming hydrogen bonds with the carbonyl group. Although this specific interaction is absent, the hydrogen bonds with Ser8 persist, as well as the hydrophobic interactions with Val12, Lys16, and Phe19.

Figure 4.

(a) Cryptocapsin (21) and (b) (5S,8S) Cryptocapsin 5,8 epoxide (26) interactions with Aβ₄₂: line (H-bond), dash dots (π–σ interaction), dash line (alkyl-alkyl interaction). Distances are in angstroms.

Regarding carotenoid 4, β-carotene, we identified unique π–alkyl and alkyl–alkyl interactions distinct from the previously observed interactions, in which hydrophobic interactions were predominant. Figure 5 illustrates this distinction, highlighting the significant involvement of Tyr10, Phe4, and Ala21 in these interactions.

Figure 5.

β-carotene (4) interaction with Aβ₄₂: line (π–σ interaction) and dash dots (alkyl–alkyl interaction). Distances are in angstroms.

For carotenoid 3, lutein, we observed distinct interactions that differed from the previously examined cases. Notably, these interactions help stabilize the two α-helices. Specifically, hydrogen bonds are formed with Ile41 and Gln15, as depicted in Figure 6.

Figure 6.

Lutein interaction with Aβ₄₂: line (H-bond), dash line (alkyl-alkyl interaction). Distances are in angstroms.

Based on this computational analysis, we can observe three regions (a-c) important in cryptocapsin (Figure 7): region a is essential for increase or decrease the biological activity of this carotenoid against Aβ₄₂ by π–σ interaction_. Meanwhile, region b keeps the interaction with this peptide through H-bond interaction. The region c can be considered an auxophoric region because it is not involved in the interaction with the peptide but It's susceptible to modification that can improve the biological activity or the analysis of ADMET.

Figure 7.

Cryptocapsin area essential for biological activity by computational analysis.

Prediction of physicochemical and ADMET properties

We predicted the physicochemical properties of 30 compounds. Results are shown in the Supplemental Material. The results of the physicochemical properties prediction of nine selected compounds, according to stronger biological activity, are shown in Table 2. For all nine compounds the water solubility was between 1.31 × 10⁻⁴ to 1.599 × 10⁻⁶, brain/blood partition coefficient (Log BB) were more than 0.3, molecular weight more than 500 g/mol, donor bonds no more than 5, acceptor bonds no more than 10, a partition coefficient (LogP) were between 8.2 to 11.6 and routable bonds more than 5. Zeaxanthin (2), lutein (3) β-Carotene (4), (5R,8R)-Sapotexanthin-5,8-epoxide (19), (5S,8S)-Cryptocapsin-5,8-epoxide (26), (5R,8R)-Cryptocapsin-5,8-epoxide (28) have 10 or fewer rotatable bonds and a polar surface area no greater than 140 Å² and have less than 60 Å².

Table 2.

Prediction of the physicochemical properties of carotenoids analyzed using SwissADME^® and ADMET Predictor^®.

Compounds	Formule*	WS* (mg/mL)	Log BB*	MW*(g/mol)	Log p*	Log p**	HBA*	HBD*	TPSA (Å²)*	RB**
Cryptocapsin (1)	C₄₀H₅₆O₂	3.48 × 10⁻⁵	1.21	568.89	9.46	10.84	2	1	37.30	11
Zeaxanthin (2)	C₄₀H₅₆O₂	2.17 × 10⁻⁵	1.12	568.89	9.32	10.55	2	2	40.46	10
Lutein (3)	C₄₀H₅₆O₂	8.22 × 10⁻⁵	1.10	568.89	9.36	10.40	2	2	40.46	10
β-Carotene (4)	C₄₀H₅₆	1.59 × 10⁻⁶	1.55	536.89	11.62	12.61	0	0	0.00	10
Sapotexanthin (16)	C₄₀H₅₆O	1.12 × 10⁻⁵	1.36	552.89	10.43	11.86	1	0	17.07	11
(5R,8R)-Sapotexanthin-5,8-epoxide (19)	C₄₀H₅₆O₂	2.98 × 10⁻⁵	1.31	568.89	10.21	11.08	2	0	26.30	10
(5R,6S)-Capsanthin-5,6-epoxide (26)	C₄₀H₅₆O₄	1.31 × 10⁻⁴	0.94	600.89	8.24	9.02	4	2	70.06	11
(5S,8S)-Cryptocapsin-5,8-epoxide (28)	C₄₀H₅₆O₃	8.22 × 10⁻⁵	1.12	584.89	9.99	10.05	3	1	46.53	10

*ADMET Predictor^®, **SwissADME.

WS: Water solubility; Log BB: Brain blood partition coefficient; MW: Molecular weight; HBA: Hydrogen acceptor bonds; HBD: Hydrogen donor bonds; TPSA: Topological Polar Surface Area; RB: Rotatable bonds.

In the following table, the formula, water solubility, molecular weight, hydrogen acceptor bonds, hydrogen donor bonds and topological polar surface area report the same results in ADMET Predictor^® and SwissADME. On the other hand, the Log P obtained differs slightly in both software's and the brain blood partition coefficient was obtained only with ADMET Predictor^®.

Prediction of ADMET properties

We predicted physicochemical properties of 30 compounds. Results are shown in the Supplemental Material. Based on the Rank score, compounds cryptocapsin (1), zeaxanthin (2), lutein (3) β-Carotene (4), sapotexanthin (16), (5R,8R)-Sapotexanthin-5,8-epoxide (19), (5R,6S)-Capsanthin-5,6-epoxide (23), (5S,8S)-Cryptocapsin-5,8-epoxide (26) and (5R,8R)-Cryptocapsin-5,8-epoxide (28) were selected for ADMET analysis and results are shown in Tables 3 to 5. Absorbed fraction (Fa%) and Bioavailable fraction (Fb%) were low for all compounds, however epoxide derivatives of carotenoid increase twice these values. All compounds were P-glycoprotein inhibitors and P-glycoprotein substrates. All compounds have the potential to cross the blood–brain barrier. There is a chance that the compounds will interact with any of the specific cytochrome P450 enzymes because it was predicted that they would all be CYP3A4 substrate, CYP3A4 inhibitor, CYP2D6 substrate, CYP2D6 inhibitor. Additionally, compound 19 was a CYP1A2 inhibitor and CYP2C19 inhibitor. Cryptocapsin (1), sapotexanthin (16), (5R,8R)-Sapotexanthin-5,8-epoxide (19), (5R,6S)-Capsanthin-5,6-epoxide (23), (5S,8S)-Cryptocapsin-5,8-epoxide (26), (5R,8R)-Cryptocapsin-5,8-epoxide (28) are considered OCT2 inhibitors. Oral rat acute toxicity (LD 50) and Oral rat chronic toxicity were predicted, and results are shown in Table 5. According to the Ames test, cryptocapsin (1), zeaxanthin (2), lutein (3), β-Carotene (4) and sapotexanthin (16) are expected to not introduce Ames toxicity.

Table 3.

Prediction of carotenoid absorption and distribution processes analyzed using ADMET predictor^® software.

Compounds	Absorption		Distribution
Compounds	Fa%	Fb%	hum-fup%	Vd (L)	Perm. Jejunal(cm/s×10⁴)	Perm. Skin (cm/s×10⁷)	OATP1B1 inhibitor	P-gp inhibitor	P-gp substrate	BBB filter	hERG filter
Cryptocapsin (1)	2.63%	0.08%	5.30%	1881.46	4.42	868.85	Yes (98%)	Yes (88%)	Yes	High	No
Zeaxanthin (2)	1.73%	0.06%	5.86%	1820.26	6.14	355.04	Yes (98%)	Yes (79%)	Yes	High	No
Lutein (3)	1.80%	0.05%	5.89%	1838.14	5.99	380.37	Yes (98%)	Yes (79%)	Yes	High	No
β-Carotene (4)	2.19%	0.06%	4.77%	3194.79	12.00	1710.30	Yes (98%)	Yes (70%)	Yes	High	No
Sapotexanthin (16)	2.48%	0.13%	4.84%	2382.47	8.29	592.47	Yes (98%)	Yes (88%)	Yes	High	No
(5R,8R)-Sapotexanthin-5,8-epoxide (19)	3.29%	0.09%	4.80%	2258.33	7.63	537.70	Yes (98%)	Yes (88%)	Yes	High	No
(5R,6S)-Capsanthin-5,6-epoxide (23)	4.68%	0.20%	6.24%	1409.89	2.54	281.57	Yes (98%)	Yes (88%)	Yes (95%)	High	No
(5S,8S)-Cryptocapsin-5,8-epoxide (28)	4.00%	0.17%	5.44%	1721.38	5.48	419.22	Yes (98%)	Yes (88%)	Yes	High	No

Fa%: Absorbed fraction; Fb%: Bioavailable fraction; hum-fup%: Fraction Unbound (human); Vd: Volume of distribution (human); Perm. Jejunal: Effective human jejunal permeability; Perm. Skin: Skin permeability; OATP1B1 inhibitor: Organic Anion Transporting Polypeptide Type 1B1 inhibitor; P-gp inhibitor: P-glycoprotein inhibitor; P-gp substrate: P-glycoprotein substrate; BBB filter: Blood-Brain Barrier filter; hERG filter: human Ether-a-go-go-Related Gene filter.

Table 4.

Prediction of carotenoid metabolism and excretion processes analyzed using ADMET Predictor^® software.

Compounds	Metabolism							Excretion
Compounds	CYP3A4 subst.	CYP3A4 inhib.	CYP2D6 subst.	CYP2D6 inhib.	CYP1A2 inhib.	CYP2C9 inhib.	CYP2C19 inhib.	t_1/2(h)	Cl (L/h)	OCT2 subst.	OCT2 inhib.
Cryptocapsin (1)	Yes	Yes (59%)	Yes	Yes (55%)	No	No	Yes	18.26	71.40	No (61%)	Yes (50%)
Zeaxanthin (2)	Yes (84%)	Yes (55%)	Yes	Yes (55%)	No	No	Yes	16.78	75.21	No (51%)	No (66%)
Lutein (3)	Yes	Yes (59%)	Yes	Yes (55%)	No	No	Yes	17.15	74.30	No (50%)	No (60%)
β-Carotene (4)	Yes (82%)	Yes (39%)	Yes	Yes (51%)	No	No	Yes	29.91	74.04	No (52%)	No
Sapotexanthin (16)	Yes (84%)	Yes (55%)	Yes	Yes (43%)	No	No	Yes	24.74	66.75	No (61%)	Yes (41%)
(5R,8R)-Sapotexanthin-5,8-epoxide (19)	Yes (87%)	Yes (39%)	Yes	Yes (46%)	Yes	No	Yes	27.11	68.13	No (68%)	Yes (53%)
(5R,6S)-Capsanthin-5,6-epoxide (23)	Yes	Yes (51%)	Yes	Yes (47%)	No	No	No	13.93	70.15	No (66%)	Yes (53%)
(5S,8S)-Cryptocapsin-5,8-epoxide (28)	Yes (92%)	Yes (49%)	Yes	Yes (49%)	No	No	Yes	17.16	69.52	No (72%)	Yes (57%)

CYP3A4 subst.: Cytochrome P450 type 3A4 substrate; CYP3A4 inhib.: Cytochrome P450 type 3A4 inhibitor; CYP2D6 subst.: Cytochrome P450 type 2D6 substrate; CYP2D6 inhib.: Cytochrome P450 type 2D6 inhibitor; CYP1A2 inhib.: Cytochrome P450 type 1A2 inhibitor; CYP2C9 inhib.: Cytochrome P450 type 2C9 inhibitor; CYP2C19 inhib.: Cytochrome P450 type 2C19 inhibitor; t_1/2: Half-life; Cl: Clearance; OCT2 subst.: Organic Cation Transporter 2 substrate; OCT2 inhib.: Organic Cation Transporter 2 inhibitor.

Table 5.

Prediction of the toxicity of carotenoids analyzed using ADMET predictor^® software.

Compounds	Toxicity
Compounds	ORAT LD50(mg/kg)	ORCT (mg/kg/day)	Max RTD Below 3.16 mg/kg/day	BDG	Ames Toxicity	T. pyriformis Toxicity(mmol/L)	MT(mg/L)	D. magna Toxicity(mg/L)
Cryptocapsin (1)	1027.91	67.83	91%	No (59%)	Negative	3.43	0.019	0.66
Zeaxanthin (2)	409.28	310.63	80%	No (58%)	Negative	2.31	0.034	0.87
Lutein (3)	337.87	383.23	80%	No (58%)	Negative	2.49	0.063	0.31
β-Carotene (4)	1235.87	380.35	65%	No (63%)	Negative	3.27	0.004	0.39
Sapotexanthin (16)	3438.83	106.92	74%	No (66%)	Negative	3.85	0.015	0.31
(5R,8R)-Sapotexanthin-5,8-epoxide (23)	971.15	67.61	80%	No (66%)	Positive MUT 100 (40%)	3.76	0.006	0.19
(5R,6S)-Capsanthin-5,6-epoxide (26)	487.96	21.99	91%	No (62%)	Positive MUT 100 (59%)	2.54	0.056	2.04
(5S,8S)-Cryptocapsin-5,8-epoxide (28)	656.54	41.46	91%	No (60%)	Positive MUT 100 (38%)	3.10	0.012	0.29

ORAT LD50: Oral Rat Acute Toxicity LD50; ORCT: Oral Rat Chronic Toxicity; Max RTD: Maximum Tolerated Dose (human); BDG.: Biodegration; Ames Toxicity: DNA mutagenic toxicity assay; T. pyriformis Toxicity: Toxicity test with Tetrahymena pyriformis; MT: Minnow Toxicity; D. magna Toxicity: Toxicity test with Daphnia magna.

Prediction of pharmacokinetic profiles after simulation of oral administration of all compounds (Figure 8) presented high volume of distribution and plasma concentration decrease gradually with the half-life of the compound longer. Compounds (5R,6S)-Capsanthin-5,6-epoxide (23), (5S,8S)-Cryptocapsin-5,8-epoxide (26) and (5R,8R)-Cryptocapsin-5,8-epoxide (28) presented lower volume of distribution than other compounds with a shorter half-life.

Figure 8.

Simulated human plasma concentration versus time after oral administration of 10 mg of each carotenoid.

The selected test compounds Cryptocapsin (1), (5R,8R)-Sapotexanthin-5,8-epoxide (19), (5S,8S)-Cryptocapsin-5,8-epoxide (26) and (5R,8R)-Cryptocapsin-5,8-epoxide (28) were subjected to Swiss ADME web tool to generate its bioavailability graphical radar plot (Figure 9). The pink area represents the optimal range for each property (lipophilicity: XLOGP3 between − 0.7 and + 5.0, size: MW between 150 and 500 g/mol, polarity: TPSA between 20 and 130 Å², solubility: log S not higher than 6, saturation: fraction of carbons in the sp 3 hybridization not less than 0.25, and flexibility: no more than 9 rotatable bonds). All compounds could show difficult for orally bioavailable because compounds presented more flexibility and size than allowed. However, derivatives Cryptocapsin presented less flexibility than original compounds.

Figure 9.

Radar plot of carotenoids selected.

Discussion

In the current computational study, we sought to clarify the interaction between Aβ peptides and conventional and unconventional carotenoids, focusing on their potential anti-aggregation properties. Our previous research showed that cryptocapsin, followed by zeaxanthin and cryptocapsin-5,6-epoxide, exhibited significant biological activity against Aβ aggregation.¹⁹ Katayama et al. found that lutein was more biologically active than β-cryptoxanthin, β-carotene, α-carotene, and zeaxanthin.²² Furthermore, our previous computational analysis²⁰ suggests that lutein was superior to cryptocapsin (1). Based on these findings, we selected cryptocapsin (1), zeaxanthin (2), lutein (3), and cryptocapsin-5,6-epoxide (24) as reference controls to evaluate other carotenoids in red mamey fruit (Figure 2). We were looking for different chemical interactions between each carotenoid and the Aβ monomers, indicating that the selected carotenoids from red mamey fruit can block the self-assembly of Aβ aggregates or disaggregation of fibrils.

We observed a strong correlation between our computational results and the experimental data. Cryptocapsin (1) and zeaxanthin (2) demonstrated similar biological activity, while cryptocapsin-5,6-epoxide (24) was less active. This behavior was replicated in the computational analysis using the 1Z0Q model. Lutein (3) exhibited greater activity than zeaxanthin (2), corroborating our previous findings obtained using different software²⁰ (Figure 10).

Figure 10.

Exponential consensus ranking (ECR) of model 1Z0Q Aβ₄₂ numbered according to Figure 2. Systems 1, 2, and 3 are the experimental references.

The computational analysis revealed that the activity decreased in the following order: (5R,8R)-sapotexanthin-5,8-epoxide (19) and (5S,8S)-cryptocapsin-5,8-epoxide (26) exhibited the highest activity, followed by (5R,8R)-cryptocapsin-5,8-epoxide (28), β-carotene (4), (5S,8R)-sapotexanthin-5,8-epoxide (20), (5R,8S)-cryptocapsin-5,8-epoxide (27), lutein (3), (3R,5'S,6’R)-beta-cryptoxanthin-5,6-epoxide (10), sapotexanthin (16), (5R,8S)-sapotexanthin-5,8-epoxide (21), cryptocapsin (1), and zeaxanthin (2). This ranking is depicted in Figure 11.

Figure 11.

Comparison between base structures exponential consensus ranking (ECR).

Figure 12 shows a distinct variation in activity between the base carotenoid structures of conventional and unconventional carotenoids. Among the conventional carotenoids, β-carotene (4) is the most promising, closely followed by lutein (3). Among the unconventional carotenoids, notable activity was observed for sapotexanthin (16), which is known for its provitamin A properties,³¹ and cryptocapsin (1), the highly active anti-amyloid compound found in red mamey fruit.¹⁹ After these compounds, three conventional carotenoids, namely, zeaxanthin (2), β-cryptoxanthin (6), and violaxanthin (13), exhibited the next highest activity. Lastly, three unconventional carotenoids, namely, 3′-deoxycapsanthin (14), capsanthin (22), and capsorubin (30), exhibited comparatively lower in silico activity. Furthermore, the activity between carotenes and xanthophylls may be distinct. Among the evaluated carotenes, β-carotene (4) demonstrates the most promising activity, while xanthophylls, such as lutein (3), sapotexanthin (16), and cryptocapsin (1), exhibit comparable levels of activity, showcasing the distinct nature of carotenes and xanthophylls.

Figure 12.

Exponential consensus ranking (ECR) of most active carotenoid derivates compared to cryptocapsin (1) reference in Model 1z0q Aβ₄₂. (a) Sapotexanthin group; (b) cryptocapsin group.

Notably, compared to the other 5,6 epoxidations, the 5,8 epoxidations of sapotexanthin and cryptocapsin exhibited greater promise Figure 12. Our results also suggested that the type of epoxidation could influence the anti-aggregation potential against Aβ₄₂. Compared to the 5,6-epoxidation in sapotexanthin and cryptocapsin, 5,8-epoxidation appeared more promising, indicating the possibility of adjusting the activity via different epoxidation types. Moreover, stereochemistry was also influential; for instance, (5R,8R)-sapotexanthin-5,8-epoxide demonstrated greater activity than that of its (5S,8R) and (5S,8S) counterparts.

Regarding sapotexanthin-5,8-epoxide, variations in activity were observed among different forms as follows: (5R,8R) > (5S,8R) > (5S,8S). Similarly, cryptocapsin-5,8-epoxide exhibited distinct activity patterns as follows: (5S,8S) > (5R,8R) > (5R,8S). Notably, all forms of sapotexanthin-5,8-epoxide demonstrated higher activity than that of cryptocapsin, indicating the importance of conducting in vitro analyses to validate these in silico findings. Additionally, three variants of cryptocapsin-5,8-epoxide exhibited higher activity than cryptocapsin itself. Conversely, sapotexanthin-5,6-epoxide, cryptocapsin-5,6-epoxide, and β-carotene-5,6-epoxide displayed lower activity than cryptocapsin Figure 11, highlighting that the type of epoxide influences their respective activities. These molecular analyses will guide future in vitro investigations, which will focus on sapotexanthin-5,8-epoxide and cryptocapsin-5,8-epoxide.^{19,25,31,56–59} Computational analysis helps us prioritize carotenoids with higher potential therapeutic properties, thus saving considerable time and effort in the research process.

We also examined the possible chemical interactions between each carotenoid and the Aβ monomers. Our aim was to determine whether these selected carotenoids from red mamey fruit could block the self-assembly of Aβ aggregates. Our findings suggest that the biological activity and molecular docking of selected carotenoids are correlated with Aβ monomers. Computational analysis indicated that specific modifications to the terminal side of the carotenoid structure could significantly affect biological activity.

Based on that, the different in biological activity between Sapotexanthin (16) and (5R,8R) sapotexanthin epoxide (19) is due to compound 19 by epoxide increase the π–σ interaction between Phe19 with the methyl from compound 19. However, this interaction decreased with the sapotexanthin (16) as showed in Figure 3. A similar behavior occurs with Cryptocapsin (21) and (5S,8S) Cryptocapsin 5,8 epoxide (26) where the first has π–σ interaction between Phe19 with the methyl from compound 21 and H-bond interaction of Ser8 with the carbonyl of compound 21 as showed in Figure 4. However, the compound 26 reduce the π–σ interaction but increase the H-bond interaction with Ser8 and His6 as showed in Figure 4. The computational analysis showed that hydroxyl group in compounds 16, 19, 21 and 26 is not involved in the biological activity, for this reason, this hydroxyl group can be considered an auxophoric part of these compounds. Meanwhile, the two-hydroxyl group of Lutein (3) are involved in the biological activity by forming H-bond interaction where one hydroxyl group with Gln15 and another hydroxyl group with the carbonyl of Ile41 of Aβ₄₂. But also, the long chain of the carotenoid allows to block the Aβ₄₂ self-assemble.

Further studies are needed to experimentally validate these computational results. This study underscores the potential for harnessing natural compounds in the treatment of complex diseases. While synthesized compounds play a crucial role in modern pharmacology, nature also offers a vast array of complex molecules that may possess therapeutic properties. In this context, red mamey fruit and other rich sources of carotenoids could become vital resources in the development of new drugs to treat AD.

Physicochemical results indicate that these compounds do not adhere to the Lipinskís rules because more than two values are not complied, that means these compounds could have low absorption and permeability trough passive diffusion and suggest that mediated active transport is necessary. This is due to their low water solubility and their molecular weight of more than 500 g/mol. However, Veber's rule states that Cryptocapsin (1), sapotexanthin (16), (5R,8R)-Sapotexanthin-5,8-epoxide (19), (5R,6S)-Capsanthin-5,6-epoxide (23), (5S,8S)-Cryptocapsin-5,8-epoxide (26) and (5R,8R)-Cryptocapsin-5,8-epoxide (28) compounds exhibit high probability of good oral bioavailability and good permeability in brain membrane. ADMET properties suggest that epoxide derivatives of carotenoid have a two-fold increase in absorbed fraction (Fa%) and bioavailable fraction (Fb%). Nine compounds could obstruct the efflux transporters that allow medications to be pumped out of cells and could be able to detect and transport them. All compounds have the potential to cross the blood-brain barrier. According to the Ames test, cryptocapsin (1), zeaxanthin (2), and lutein (3) β-Carotene (4) and sapotexanthin (16) were expected to not introduce Ames toxicity, indicating that it is unlikely that they will result in cell mutations. Prediction of pharmacokinetic profiles after simulation of oral administration of nine selected compounds presented high volume of distribution that means there is too much drug in the tissues, i.e., little drug available to be eliminated and thus the decrease in plasma concentration will be gradual. Therefore, if the volume of distribution is high, the half-life of the compound will be longer. Compounds (5R,6S)-Capsanthin-5,6-epoxide (23), (5S,8S)-Cryptocapsin-5,8-epoxide (26) and (5R,8R)-Cryptocapsin-5,8-epoxide (28) displayed lower distribution volume, as compared to compounds with shorter half-lives. According to physicochemical and ADMET prediction analysis compounds (5S,8S)-Cryptocapsin-5,8-epoxide (26) and (5R,8R)-Cryptocapsin-5, 8-epoxide (28) presented better overall properties to continue to in vitro and in vivo studies; and it is important to evaluate other routes of administration as well as new delivery systems, such as vectors.

From biopharmaceutical point of view, the higher value of Log P indicates that more lipophilic is the compound and present high solubility in fat, oils, lipids and nonpolar solvents. Carotenoids showed high Log P that suggests compound could present limitations in passive diffusion and oral absorption. This suggest that mediated active transport is necessary for achieve absorption of the compounds by oral route. In the context of pharmacodynamics, the hydrophobic effect is the major driving force for the binding of drugs to their target receptors. On the other hand, hydrophobic drugs tend to be more toxic because they are generally retained for a longer time, have a wider distribution within the body (e.g., intracellular), are somewhat less selective in their binding to proteins, and finally are often extensively metabolized.

Inhibition of the three major cytochrome P450 (CYP) isoforms (CYP2C19, CYP2D6, and CYP3A4) could be certainly one major cause of pharmacokinetic-related drug–drug interactions and it were expected for our derivatives due to their physicochemical properties (large and lipophilic molecules). This point is important taking in account because drug-drug interactions (DDIs) and inhibition by one drug may lead to systemic toxicity due to elevated levels in the blood or subtherapeutic levels of another in the liver. Molecules were not CYP1A2 and CYP2C9 inhibitors, then there are selective inhibitions for CYP isoforms and administration of this molecules have to take care for specific drug and avoid together consumption. But also, is important recognize that process enzymatically converts lipid-soluble compounds to more water-soluble compounds to facilitate the excretion from the body.

For future in vivo studies it is important to evaluate other routes of administration as dermal route, since these derivatives showed good skin permeability, as well as new drug delivery systems in order to efficiency transport derivatives to achieve the desired action and avoid possible side effects. If oral route is preferred, then suitable drug formulations is required in order to assess the best oral absorption with the require deposition and activity.

Finally, this study provides an important advancement in our knowledge on the potential role of carotenoids in neurodegenerative disease. However, it is important to note that computational analysis is only the beginning. The promising candidates identified in this research must be validated through further in vitro and in vivo studies. If these subsequent studies confirm our findings, these compounds could become a valuable addition to the development of new treatments for AD.

Conclusions

According to computational and pharmaceutical analysis, compounds 19, 26, and 28 are the most promising carotenoids. Thirty carotenoids from red mamey fruit were evaluated using the following molecular docking programs: AutoDock 4 (ATD), AutoDock Vina (Vina), Smina (Vinardo Scoring function), QuickVina 2, QuickVina W, and PSOVina. A strong correlation between our computational results and the experimental data was observed. Cryptocapsin (1) and zeaxanthin (2) demonstrated similar biological activity, while cryptocapsin-5,6-epoxide (24) was less active in vitro and in silico. The computational analysis revealed the following order of decreasing activity: (5R,8R)-sapotexanthin-5,8-epoxide (19) and (5S,8S)-cryptocapsin-5,8-epoxide (26) exhibited the highest activity, followed by (5R,8R)-cryptocapsin-5,8-epoxide (28), β-carotene (4), (5S,8R)-sapotexanthin-5,8-epoxide (20), (5R,8S)-cryptocapsin-5,8-epoxide (27), lutein (3), (3R,5'S,6’R)-beta-cryptoxanthin-5,6-epoxide (10), sapotexanthin (16), (5R,8S)-sapotexanthin-5,8-epoxide (21), cryptocapsin (1), and zeaxanthin (2). The computational analysis helped us prioritize carotenoids with higher potential therapeutic properties, thus saving considerable time and effort in the progression of research. Pharmaceutical analysis, compound (5S,8S)-Cryptocapsin-5,8-epoxide (26) and (5R,8R)-Cryptocapsin-5,8-epoxide (28) are the most promising compounds evaluated. Future studies of these compounds will aim to determine the in vitro biological analysis and in vivo performance.

Supplemental Material

sj-docx-1-alz-10.1177_13872877241291172 - Supplemental material for The role of carotenoids from red mamey fruit (Pouteria sapota) against amyloid-β monomers in Alzheimer's disease: Computational analysis and ADMET prediction

Supplemental material, sj-docx-1-alz-10.1177_13872877241291172 for The role of carotenoids from red mamey fruit (Pouteria sapota) against amyloid-β monomers in Alzheimer's disease: Computational analysis and ADMET prediction by Adrián L Orjuela, Marisín Pecchio, Jessica Cruz-Mora, Lakshmi Sowmya Emani, Maria B Carreira, Oleg V Larionov, Muralidhar L Hegde, K S Jagannatha Rao, Jorge Alí-Torres and Johant Lakey-Beitia in Journal of Alzheimer's Disease

Footnotes

Acknowledgments

M.P, M.B.C, J. A-T, and J.L.-B would like to acknowledge SENACYT Project [FID23-003]. K.S.J.R., O.V.L., and J.L.-B. would like to acknowledge SENACYT Project [APY-GC2018-010]. M.B.C and J.L.-B would like to acknowledge the National System of Investigation (SNI) for supporting their research. K.S.J.R. and J.L.-B. would like to acknowledge Melo Brain Grant (Panama) and INDICASAT Internal Grant [JR04-2020]. O.V.L. would like to acknowledge Welch Foundation grant number AX-0047. M.L.H is supported by the National Institute of Neurological Disorders and Stroke (NINDS) and the National Institute of Aging (NIA) of the National Institutes of Health (NIH) under award number RF1NS112719, and Everett E. and Randee K. Bernal for their support via Centenial Endowed Directorship of DNA Repair. A.L.O and J.A-T. want to thank Vicerrectoría de Investigación DIEB-UNAL, and the Center of Excellence in Scientific Computing (CoE–SciCo) for continuous support. LSE thankful to KELF for financial support through fellowship.

ORCID iDs

Adrián L Orjuela

Marisín Pecchio

Jessica Cruz-Mora

Maria B Carreira

Oleg V Larionov

Muralidhar L Hegde

Jorge Alí-Torres

Johant Lakey-Beitia

Author contributions

Adrián L Orjuela (Data curation; Formal analysis; Investigation; Methodology; Software; Validation; Writing – original draft; Writing – review & editing); Marisín Pecchio (Conceptualization; Formal analysis; Methodology; Software; Supervision; Visualization; Writing – original draft; Writing – review & editing); Jessica Cruz-Mora (Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Writing – original draft); Lakshmi Sowmya Emani (Investigation; Visualization; Writing – original draft; Writing – review & editing); Maria B Carreira (Investigation; Visualization; Writing – original draft; Writing – review & editing); Oleg V Larionov (Supervision; Visualization; Writing – original draft; Writing – review & editing); Muralidhar L Hegde (Supervision; Visualization; Writing – original draft; Writing – review & editing); K S Jagannatha Rao (Conceptualization; Resources; Supervision; Writing – original draft; Writing – review & editing); Jorge Alí-Torres (Conceptualization; Formal analysis; Funding acquisition; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Writing – original draft; Writing – review & editing); Johant Lakey-Beitia (Conceptualization; Formal analysis; Funding acquisition; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Writing – original draft; Writing – review & editing).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by Secretaría Nacional de Ciencia, Tecnología e Innovación [SENACYT grant number [FID23-003], [APY-GC2018-010], Instituto de Investigaciones Científicas y Servicios de Alta Tecnología [INDICASAT AIP grant number [JR04-2020], the Welch Foundation grant number AX-0047, National Institute of Neurological Disorders and Stroke (NINDS), the National Institute of Aging (NIA) of the National Institutes of Health (NIH) under award number RF1NS112719, Everett E. and Randee K. Bernal for their support via Centenial Endowed Directorship of DNA Repair, Vicerrectoría de Investigación DIEB-UNAL, and the Center of Excellence in Scientific Computing (CoE–SciCo) for continuous support.

Declaration of conflicting interests

Jorge Ali-Torres is an Editorial Board Member of this journal but was not involved in the peer-review process of this article nor had access to any information regarding its peer-review.

The other authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

All data generated or analyzed during this study are included in this published article and its supplemental material.

Supplemental material

Supplemental material for this article is available online.

References

2023 Alzheimer’s disease facts and figures. Alzheimers Dement 2023; 19: 1598–1695.

Cenini

Voos

. Mitochondria as potential targets in Alzheimer disease therapy: an update. Front Pharmacol 2019; 10: 902.

Lakey-Beitia

Burillo

La Penna

, et al. Polyphenols as potential metal chelation compounds against Alzheimer’s disease. J Alzheimers Dis 2021; 82: S335–S357.

Lakey-Beitia

Berrocal

Rao

, et al. Polyphenols as therapeutic molecules in Alzheimer’s disease through modulating amyloid pathways. Mol Neurobiol 2015; 51: 466–479.

Abyadeh

Tofigh

Hosseinian

, et al. Key genes and biochemical networks in various brain regions affected in Alzheimer’s disease. Cells 2022; 11: 987.

Kerwin

Abdelnour

Caramelli

, et al. Alzheimer’s disease diagnosis and management: perspectives from around the world. Alzheimers Dement 2022; 1: e12334.

Deture

Dickson

. The neuropathological diagnosis of Alzheimer’s disease. Mol Neurodegener 2019; 14: 32.

Murphy

Levine

. Alzheimer’s disease and the amyloid-β peptide. J Alzheimers Dis 2010; 19: 311–323.

Hampel

Hardy

Blennow

, et al. The amyloid-β pathway in Alzheimer’s disease. Mol Psychiatry 2021; 26: 5481–5503.

10.

Decourt

Noorda

, et al. Review of advanced drug trials focusing on the reduction of brain beta-amyloid to prevent and treat dementia. J Exp Pharmacol 2022; 14: 331–352.

11.

Mohd Sairazi

Sirajudeen

KNS

. Natural products and their bioactive compounds: neuroprotective potentials against neurodegenerative diseases. Evid Based Complement Alternat Med 2020; 2020: 6565396.

12.

Kumar

Bora

. Molecular docking studies on inhibition of Stat3 dimerization by curcumin natural derivatives and its conjugates with amino acids. Bioinformation 2012; 8: 988–993.

13.

Puentes-Díaz

Chaparro

Morales-Morales

, et al. Role of metal cations of copper, iron, and aluminum and multifunctional ligands in Alzheimer’s disease: experimental and computational insights. ACS Omega 2023; 8: 4508–4526.

14.

Orjuela

Lakey-Beitia

Mojica-Flores

, et al. Computational evaluation of interaction between curcumin derivatives and amyloid-β monomers and fibrils: relevance to Alzheimer’s disease. J Alzheimers Dis 2021; 82: S321–S333.

15.

Lakey-Beitia

González

Doens

, et al. Assessment of novel curcumin derivatives as potent inhibitors of inflammation and amyloid-β aggregation in Alzheimer’s disease. J Alzheimers Dis 2017; 60: S59–S68.

16.

Lakey-Beitia

Vasquez

Mojica-Flores

, et al. Pouteria sapota (red mamey fruit): chemistry and biological activity of carotenoids. Comb Chem High Throughput Screen 2021; 25: 1134–1147.

17.

González

Mojica-Flores

Moreno-Labrador

, et al. Polyphenols with anti-inflammatory properties: synthesis and biological activity of novel curcumin derivatives. Int J Mol Sci 2023; 24: 3691.

18.

Sánchez-Martínez

Valdés

Gallego

, et al. Blood–brain barrier permeability study of potential neuroprotective compounds recovered from plants and agri-food by-products. Front Nutr 2022; 9: 924596.

19.

Lakey-Beitia

Doens

Kumar

, et al. Anti-amyloid aggregation activity of novel carotenoids: implications for Alzheimer’s drug discovery. Clin Interv Aging 2017; 12: 815–822.

20.

Lakey-Beitia

Kumar

Hegde

, et al. Carotenoids as novel therapeutic molecules against neurodegenerative disorders: chemistry and molecular docking analysis. Int J Mol Sci 2019; 20: 5553.

21.

Parker

Swanson

You

C-S

, et al. Bioavailability of carotenoids in human subjects. Proc Nutr Soc 1999; 58: 155–162.

22.

Katayama

Ogawa

Nakamura

. Apricot carotenoids possess potent anti-amyloidogenic activity in vitro. J Agric Food Chem 2011; 59: 12691–12696.

23.

Hornero-Méndez

Gómez-Ladrón De Guevara

Mínguez-Mosquera

. Carotenoid biosynthesis changes in five red pepper (Capsicum annuum L.) cultivars during ripening. Cultivar selection for breeding. J Agric Food Chem 2000; 48: 3857–3864.

24.

Maiani

Castón

MJP

Catasta

, et al. Carotenoids: actual knowledge on food sources, intakes, stability and bioavailability and their protective role in humans. Mol Nutr Food Res 2009; 53: 194–218.

25.

Murillo

Giuffrida

Menchaca

, et al. Native carotenoids composition of some tropical fruits. Food Chem 2013; 140: 825–836.

26.

Kopsell

. Accumulation and bioavailability of dietary carotenoids in vegetable crops. Trends Plant Sci 2006; 11: 499–507.

27.

Hirschberg

. Carotenoid biosynthesis in flowering plants. Curr Opin Plant Biol 2001; 4: 210–218.

28.

Chemler

Yan

Koffas

. Biosynthesis of isoprenoids, polyunsaturated fatty acids and flavonoids in Saccharomyces cerevisiae. Microb Cell Fact 2006; 5: 20.

29.

Werner

Böhm

. Bioaccessibility of carotenoids and vitamin e from pasta: evaluation of an in vitro digestion model. J Agric Food Chem 2011; 59: 1163–1170.

30.

Wolak

Hamias

Volvich

, et al. Carotenoids and flavonoids synergistically attenuate inflammation in endothelial cells induced by TNF-α. A J Hypertens 2014; 8: e77.

31.

Murillo

McLean

Britton

, et al. Sapotexanthin, an A-provitamin carotenoid from red mamey (Pouteria sapota). J Nat Prod 2011; 74: 283–285.

32.

Deli

Osz

Molnár

, et al. Reduction of capsorubin and cryptocapsin. Helv Chim Acta 2001; 84: 3810–3817.

33.

Marenich

Cramer

Truhlar

. Universal solvation model based on solute electron density and on a continuum model of the solvent defined by the bulk dielectric constant and atomic surface tensions. J Phys Chem B 2009; 113: 6378–6396.

34.

Frisch

Trucks

Schlegel

, et al. Gaussian 16 revision C.01.

35.

Sticht

Bayer

Willbold

, et al. Structure of amyloid A4(1-4O)-peptide of Alzheimer’s disease. Eur J Biochem 1995; 233: 293–298.

36.

Coles

Bicknell

Watson

, et al. Solution structure of amyloid beta-peptide(1-40) in a water-micelle environment. Biochemistry 1998; 37: 11064–11077.

37.

Tomaselli

Esposito

Vangone

, et al. The α-to-β conformational transition of Alzheimer’s aβ-(1-42) peptide in aqueous media is reversible: a step by step conformational analysis suggests the location of β conformation seeding. ChemBioChem 2006; 7: 257–267.

38.

Case

Ben-Shalom

Brozell

, et al. AMBER 2018. San Francisco: University of California, 2018.

39.

Maier

Martinez

Kasavajhala

, et al. ff14SB: improving the accuracy of protein side chain and backbone parameters from ff99SB. J Chem Theory Comput 2015; 11: 3696–3713.

40.

Joung

Cheatham

. Determination of alkali and halide monovalent ion parameters for use in explicitly solvated biomolecular simulations. J Phys Chem B 2008; 112: 9020–9041.

41.

Roe

Cheatham

. PTRAJ And CPPTRAJ: software for processing and analysis of molecular synamics trajectory data. J Chem Theory Comput 2013; 9: 3084–3095.

42.

Jin

Han

. K-medoids clustering. In: Sammut C, Webb GI (eds) Encyclopedia of machine learning and data mining. Boston: Springer, 2017, pp.564–565.

43.

Schrödinger

LLC.

The {PyMOL} molecular graphics system, Version∼1.8. November 2015.

44.

Palacio-Rodríguez

Lans

Cavasotto

, et al. Exponential consensus ranking improves the outcome in docking and receptor ensemble docking. Sci Rep 2019; 9: 5142.

45.

Morris

Huey

. Autodock4 and AutoDockTools4: automated docking with selective receptor flexibility. J Comput Chem 2009; 30: 2785–2791.

46.

Trott

Olson

. Autodock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization and multithreading. J Comput Chem 2010; 31: 455–461.

47.

Koes

Baumgartner

Camacho

. Lessons learned in empirical scoring with smina from the CSAR 2011 benchmarking exercise. J Chem Inf Model 2013; 53: 1893–1904.

48.

Alhossary

Handoko

, et al. Fast, accurate, and reliable molecular docking with QuickVina 2. Bioinformatics 2015; 31: 2214–2216.

49.

Hassan

Alhossary

, et al. Protein-ligand blind docking using QuickVina-W with inter-process spatio-temporal integration. Sci Rep 2017; 7: 15451.

50.

Tai

Jusoh

Siu

SWI

. Chaos-embedded particle swarm optimization approach for protein-ligand docking and virtual screening. J Cheminform 2018; 10: 62.

51.

Schrödinger release 2024-3: Maestro. New York: Schrödinger, LLC, 2024.

52.

Abchir

Khedraoui

Nour

, et al. Integrative approach for designing novel triazole derivatives as α-glucosidase inhibitors: QSAR, molecular docking, ADMET, and molecular dynamics investigations. Pharmaceuticals 2024; 17: 261.

53.

Deb

Reeves

Lafortune

. Simulation of physicochemical and pharmacokinetic properties of vitamin d3 and its natural derivatives. Pharmaceuticals 2020; 13: 160.

54.

Yamari

Abchir

Nour

, et al. Identification of new dihydrophenanthrene derivatives as promising anti-SARS-CoV-2 drugs through in silico investigations. Main Group Chem 2023; 22: 469–484.

55.

Duchowicz

Talevi

Bellera

, et al. Application of descriptors based on Lipinski’s rules in the QSPR study of aqueous solubilities. Bioorg Med Chem 2007; 15: 3711–3719.

56.

Murillo

Turcsi

Szabó

, et al. Carotenoid composition of the fruit of red mamey (Pouteria sapota). J Agric Food Chem 2016; 64: 7148–7155.

57.

Gulyás-Fekete

Agócs

Turcsi

, et al. Cryptocapsinepoxide-type carotenoids from red mamey, Pouteria sapota. J Nat Prod 2013; 76: 607–614.

58.

Turcsi

Murillo

Kurtán

, et al. Isolation of β-cryptoxanthin-epoxides, precursors of cryptocapsin and 3’-deoxycapsanthin, from red mamey (Pouteria sapota). J Agric Food Chem 2015; 63: 6059–6065.

59.

Murillo

Agócs

Nagy

, et al. Isolation and identification of sapotexanthin 5,6-epoxide and 5,8-epoxide from red mamey (Pouteria sapota). Chirality 2020; 32: 579–587.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.57 MB