From Multi-Species Screening to Targeted Investigation: Discovery of ACE Inhibitory Peptides in Gigantidas platifrons via Peptidomics, Virtual Screening, and Molecular Dynamics Simulations

Abstract

Deep-sea mollusks represent untapped resources for searching novel biologically active peptides effectual against many chronic diseases. Here we presented the identification of four novel angiotensin I-converting enzyme (ACE) inhibitory peptides from the deep-sea mollusk Gigantidas platifrons by using a combined approach of peptidomics and virtual screening. Fifteen protein hydrolysates from five deep-sea macroorganisms were prepared using three different proteases and were determined for their ACE inhibitory activities. Pepsin hydrolysate of G. platifrons protein (GPp) demonstrated the highest inhibition rate against ACE at 400 μg/mL. Then, targeted investigation was conducted on the GPp with peptidomic profiling; more than 3000 peptides were de novo identified, which were then subject to virtual screening using the docking software Smina. Subsequently, 29 peptides were selected and synthesized based on the affinity threshold and the interactions with ACE active sites. More than 58% peptides were biologically active, showing more than 50% inhibition to ACE at 400 μM. Four peptides, LAAHFAR, YAAPYR, NGAGPYGRP, and FTTFGK, exhibited low micromolar inhibition. The most potent peptide, LAAHFAR with an IC₅₀ of 6.01 ± 1.06 μM, was subject to molecular dynamics simulations for revealing atomistic interaction analysis. LAAHFAR forms comprehensively stable hydrogen bonds with the classic active site of ACE, and its N terminal arginine residue is anchored by additional hydrogen bonding to Cys370, Asp377, and Thr372. This study highlights deep-sea mollusks as an important source of novel ACE inhibitory peptides, contributing to the development of new therapeutic ingredients or functional food agents against hypertension.

1. Introduction

Hypertension is a major global health concern, classified as one of the most common chronic diseases worldwide. It significantly increases the risk of serious pathological conditions, including cardiovascular diseases, stroke, and arteriosclerosis, contributing to an estimated 9 million deaths annually [1,2]. The regulation of blood pressure is critically mediated by the renin–angiotensin–aldosterone system (RAAS) [3]. At the core of this system is the angiotensin I-converting enzyme (ACE), which acts as a key therapeutic target for antihypertensive agents [4]. ACE is an enzyme that raises blood pressure through two primary mechanisms: it converts the inactive decapeptide angiotensin I into the potent vasoconstrictor angiotensin II, and it simultaneously inactivates the vasodilator bradykinin. Current clinical treatment heavily relies on synthetic ACE inhibitors, such as captopril, benazepril, and enalapril. However, these synthetic drugs frequently induce undesirable side effects, including dry coughs, taste disturbance, and skin rashes, prompting an urgent need for safe and effective alternatives [5]. Bioactive peptides derived from food proteins are increasingly recognized as promising candidates for the nutritional and pharmacological management of hypertension, due to their natural origin, minimal toxicity, and potential for safe application [6].

The marine environment provides an abundant and diverse resource of novel bioactive compounds, encompassing peptides with antihypertensive properties [7,8]. The unique amino acid compositions of marine proteins make them attractive resources for discovering novel ACE inhibitory peptides. Indeed, numerous ACE inhibitory peptides have been successfully isolated and identified from various marine organisms, including mollusks like blue mussels Mytilus edulis and oysters [9,10]. Studies, for example, have shown that mussel-derived NADH dehydrogenase and AMP-activated protein kinase are protein precursors for releasing high amounts of ACE-binding peptides. Nevertheless, the investigation of ACE inhibitory peptides originating from deep-sea organisms remains significantly underexplored [11]. Deep-sea ecosystems, such as hydrothermal vents and cold seeps, present extreme conditions characterized by challenges like high hydrostatic pressure, low temperatures, and low-nutrient availability. Organisms adapting to these harsh environments, such as deep-sea mussels Gigantidas platifrons, exhibit distinct molecular mechanisms compared to shallow-water species [12], suggesting they may harbor structurally novel and functionally unique proteins and peptides. Supporting this potential, an exceptionally potent ACE inhibitory peptide (KLLWNGKM) was isolated from the deep-sea hydrothermal vent mussel G. Vrijenhoeki [13]. This high efficacy highlights the immense, largely untapped, potential of deep-sea invertebrates as a reservoir for discovering potent and novel ACE inhibitory peptides.

Hypertension’s clinical burden and the therapeutic relevance of ACE inhibitors establish the need to discover peptide scaffolds beyond those found in well-studied marine organisms. In this context, the present work examined five representative deep-sea macrofaunal species from the South China Sea—Archivesica marissinica (AM), G. haimaensis (GH), G. platifrons (GP), Lamellibrachia columna (LC), and Shinkaia crosnieri (SC). Protein extracts were hydrolyzed using trypsin, pepsin, and bromelain, among which the pepsin hydrolysate of G. platifrons showed the most pronounced ACE inhibitory activity. Subsequent peptidomic profiling mapped the origins and compositional features of the resulting peptide sequences. A virtual screening was conducted using all the de novo identified peptides as ligands of ACE, which led to identification of 29 potential peptide candidates for synthesis and experimental validation. Four peptides, LAAHFAR, YAAPYR, NGAGPYGRP, and FTTFGK, exhibited significant ACE inhibitory activities, with IC₅₀ values of 6.01 μM, 10.17 μM, 11.86 μM, and 41.15 μM, respectively. Integrating molecular docking and molecular dynamics simulations further clarified the structural determinants that underlie their inhibitory potency, offering mechanistic insight into how deep-sea-derived peptides engaged and suppressed ACE. This work highlighted deep-sea mussels as an important source of potent ACE inhibitory peptides and expanded the structural landscape available for antihypertensive peptide discovery.

2. Results and Discussion

2.1. Preparation of Protein Hydrolysate and Determinations of ACE Inhibitory Effects

Before hydrolysis, the amino acid compositions of the five deep-sea animal powders were analyzed to characterize their intrinsic protein profiles (

Table 1. Amino acid compositions of five deep-sea specimens.

	Amino Acid Type	G. haimaensis	G. platifrons	A. marissinica	S. crosnieri	L. columna
		[%]	[%]	[%]	[%]	[%]
1	Asp #	4.05	3.37	5.75	4.01	3.70
2	Thr *	1.94	1.67	3.01	1.78	1.80
3	Ser	2.13	1.95	2.96	1.76	1.81
4	Glu #	6.37	4.92	6.78	5.16	3.74
5	Gly #	7.68	5.83	6.32	3.38	2.18
6	Ala	2.32	1.92	3.791	2.29	1.58
7	Cys	0.04	0.15	0.39	0.05	0.61
8	Val *	1.78	1.63	2.48	1.87	1.62
9	Met *#	0.07	0.17	0.97	0.14	0.22
10	Ile *	1.62	1.51	2.43	1.64	0.95
11	Leu *#	2.58	2.32	3.58	2.51	1.57
12	Tyr #	1.31	1.17	1.64	1.06	2.47
13	Phe *#	1.43	1.33	2.25	1.73	1.440
14	His *	1.48	1.29	3.23	5.04	4.59
15	Lys *#	2.85	3.42	3.52	2.33	2.35
16	Arg #	3.21	3.14	2.85	2.79	2.67
17	Pro	2.27	1.65	1.61	1.75	1.08
	TAA a	43.14	37.47	53.57	39.29	34.38
	EAA b	13.75	13.35	21.48	17.04	14.54
	MAA c	29.57	25.67	33.67	23.11	20.34
	EAA/TAA × 100	31.87	35.645	40.10	43.37	42.28
	MAA/TAA × 100	68.53	68.52	62.85	58.81	59.14

Note: ^a, essential amino acids (*); ^b, total amino acids; ^c, medicinal amino acids (^#).

). Glycine was the most abundant amino acid in GH and GP, whereas glutamic acid dominated in AM and SC. In LC, histidine was the most enriched amino acid. Among the five deep-sea species, AM exhibited the highest total amino acid content (TAA, 53.57%) and contained the greatest proportion of essential amino acids (EAA, 21.48%), suggesting its comparatively high nutritional value. SC showed the highest ratio of essential to total amino acids (43.47%), indicating that its protein-derived peptides may have considerable bioactive potential. Additionally, three deep-sea mollusks possessed a relatively high ratio of medicinal amino acids (MAA), the MAA/TAA value ranging from 62.85% to 68.53%.

The five species were subsequently hydrolyzed using pepsin, trypsin, and bromelain to obtain peptide-rich hydrolysates. To assess their antihypertensive potential, all fifteen hydrolysates were evaluated for ACE inhibitory activity at 400 μg/mL (Figure 1). More than half exhibited inhibition rates above 30%, with the pepsin hydrolysate of GP (GPp) showing the strongest activity, reaching approximately 70% inhibition against ACE. A previous study predicted that about 23% of the in silico-generated peptides were potentially anti-hypertensive [14]. This study highlighted the antihypertensive potential from deep-sea biomass through experimental evaluation. Consequently, we conducted a systematic analysis of GPp to characterize its constituent peptides with ACE inhibitory activity.

2.2. Peptidomics Analysis and De Novo Peptide Identification

To characterize the peptide constituents of GPp with high confidence, nano-HPLC-MS/MS analysis was performed, enabling sensitive and accurate sequence identification. Peptides with de novo scores above 80 were retained for detailed analysis, and GPp was analyzed in triplicate to ensure reproducibility. Across all replicates, 3088 unique peptides were identified, and their overlap was visualized (Figure 2A). Annotation of these peptides against a mollusk-specific protein database revealed that most sequences originated from actin of the Placopecten magellanicus (UniProt ID: Q26065) (Figure 2B). The prominence of actin-derived fragments underscores the potential of molluscan proteins as precursors of bioactive peptides. Only a small number of precursor proteins were matched overall, and many of these belonged to shallow-water mollusks rather than true deep-sea organisms. This observation highlights a major gap in existing peptide and protein databases for deep-sea fauna, despite their clear potential as sources of structurally novel bioactive peptides. The discovery of ACE inhibitory peptides from GPp, together with their deep-sea origin, positions G. platifrons as a particularly promising species for mining therapeutic peptide leads and illustrates the broader need to expand proteomic resources for deep-sea invertebrates.

Figure 2C–I show the sequence feature for the peptides with four to ten amino acid residues, hydrophobic amino acids (Leu, Gly, Val, Pro, Phe), with positively or negatively charged amino acids (Asp, Glu, Arg) being particularly enriched in their sequences. Leucine was the most common residue at both N- and C-terminal positions, consistent with the known cleavage preference of pepsin. Empirical results from previous studies indicated that existence of hydrophobic amino acids and branched amino acids contributed to increased ACE inhibitory activity [15].

2.3. Virtual Screening to Identify Potential ACE Inhibitory Peptides

The overall virtual screening workflow is illustrated in Figure 3A, which is divided into molecular docking, interactions filtering, and experimental verification. The molecular docking was performed using the docking software Smina, and the crystal structure of human ACE (PDB ID: 1O86) was selected as the target. During docking, the protein was treated as rigid and all peptides as fully flexible. A total of 2871 peptides were successfully docked, and 362 peptides with affinities below −10.4 kcal/mol were selected for further analysis (Figure 3B). Candidate peptides were ranked according to binding affinity derived from their top-scoring poses. A clear length-dependent trend emerged: octapeptides, nonapeptides, and decapeptides consistently achieved lower (better) docking scores, whereas both shorter and longer peptides showed progressively weaker predicted affinities (Figure 3C). Within the top 362 scoring sequences, peptides of seven to eleven residues were markedly enriched, while short peptides of four to five residues were substantially reduced in number (Figure 3D). This pattern indicated that ACE exhibited a structural preference for binding peptides of moderate length, particularly those with eight to ten amino acid residues. All the peptides with docking scores below the threshold of −10.4 kcal/mol underwent visual inspection of binding interactions in Pymol software (Version 2.6), and 29 peptides were retained for further evaluation. These peptides formed polar interactions with key catalytic-site residues of ACE, including Glu162, Glu384, Glu376, Asp453, as well as with the catalytic Zn²⁺ ion, indicating their potential inhibitory activity.

All 29 peptides were then synthesized and experimentally evaluated through in vitro enzymatic determinations (Figure 4A). Of the peptides predicted to be active by molecular docking, seventeen (or 58.6%) exhibited more than 50% ACE inhibition at a concentration of 400 μM (

Table 2. Docking scores and ACE inhibitory activities of the 29 selected peptides from GPp.

No.	Peptide Sequences	Inhibitory Potency (%)	Docking Affinity (kcal/mol)
1	APADF	35.50 ± 0.92	−10.8
2	APAKLWAL	87.73 ± 3.73	−10.7
3	ASLKF	71.18 ± 2.63	−10.4
4	DPVF	42.02 ± 0.89	−10.7
5	DVVY	81.51 ± 17.29	−10.5
6	ELRYY	78.71 ± 10.67	−11.4
7	EPRF	46.20 ± 0.20	−10.8
8	EVTF	37.15 ± 1.47	−10.4
9	FADY	42.50 ± 0.49	−11.1
10	FGPRL	37.00 ± 0.84	−10.7
11	FPLHAVR	72.21 ± 3.30	−10.8
12	FTDF	15.46 ± 0.09	−11.1
13	FTTFGK	91.42 ± 2.82	−11.0
14	FVDY	20.64 ± 9.68	−11.2
15	GPRF	48.50 ± 2.64	−11.0
16	KDDLSHGYY	81.17 ± 2.75	−11.7
17	LAAHFAR	93.84 ± 2.25	−11.0
18	NGAGPYGRP	91.06 ± 2.83	−11.7
19	NGPRF	23.87 ± 0.62	−11.0
20	TPGGLLFF	74.05 ± 2.67	−11.2
21	TPTAREF	38.84 ± 1.29	−10.8
22	VAARY	87.74 ± 2.94	−10.9
23	VAPDF	19.22 ± 0.31	−11.0
24	VGLGGPDGRP	88.31 ± 3.11	−11.5
25	VLRF	56.85 ± 3.66	−10.5
26	VLRY	66.05 ± 3.08	−10.9
27	YAAPYR	95.82 ± 2.95	−11.8
28	YDAF	52.75 ± 2.53	−11.4
29	YPSVPPSF	53.89 ± 2.18	−11.9

). This high hit rate underscored the utility of structure-based virtual screening for identifying ACE inhibitory candidates from complex deep-sea hydrolysates. Peptides showing greater than 90% inhibition were further subjected to concentration–response assays, and IC₅₀ values were calculated using a logistic model. Four peptides—LAAHFAR, YAAPYR, NGAGPYGRP, and FTTFGK—displayed potent ACE inhibition, with IC₅₀ values of 6.01 ± 1.06 μM, 10.17 ± 1.03 μM, 11.86 ± 1.12 μM, and 41.15 ± 1.06 μM, respectively (Figure 4B); their MS/MS spectra are shown in Figures S1–S4. These sequences represented previously unreported inhibitory scaffolds, broadening the structural diversity of known ACE inhibitory peptides.

Marine shellfish have garnered significant attention in recent years as a prominent source of ACE inhibitory peptides [16,17]. To date, a variety of ACE inhibitory peptides have been successfully isolated from common shellfish species, including oysters [18], clams [19,20], mussels [21], and scallops [22]. Their IC₅₀ values generally fall within the millimolar (mM) to micromolar (μM) range, with certain short peptides demonstrating potent in vitro inhibitory activity, achieving IC₅₀ values as low as 10–100 μM [23,24,25]. For example, a purified peptide with sequence VVYPWTQRF was isolated from the protein hydrolysate of Crassostrea talienwhanensis Crosse, and its IC₅₀ was 66 μM [26]. In contrast to the shallow-water shellfish, bioactive peptides derived from shellfish inhabiting extreme marine environments, such as deep-sea cold seeps, remain largely unexplored. The cold-seep environment is characterized by high pressure, low temperature, low oxygen levels, and an unique chemosynthesis-based ecosystem. To adapt to these extreme conditions, the proteins in these shellfish may undergo special modifications or possess enhanced stability, potentially leading to novel or more potent ACE inhibitory peptides. Furthermore, cold-seep shellfish rely on symbiotic bacteria or specialized food sources, and their metabolic products may contain rare amino acid sequences, providing a diverse resource for the discovery of ACE inhibitory peptides. A previous study on the deep-sea hydrothermal vent mussel G. Vrijenhoeki led to the identification of several potent ACE inhibitory peptides such as KLLWNGKM, possessing an IC₅₀ of 0.007 μM [13]. In the present study, we further highlighted the great potential of deep-sea organisms for mining ACE inhibitory peptides. More than half of the synthetic peptides identified from the cold-seep G. platifrons were observed to exhibit significant ACE inhibitory effects, and four peptides possessed micromolar inhibition potency ranging from 41.15 μM to 6.01 μM.

In this study, we also observed a low matching rate of known protein precursors for the sequences identified from GPp, indicating highly unexplored peptide sequences. Interestingly, we manually searched the Uniprot database with the “Peptide search” tool to find the potential protein precursors of the peptides LAAHFAR and NGAGPYGRP (https://www.uniprot.org/peptide-search, at least seven amino acids long, available until 6 December 2025). No known proteins or peptides contain the peptide fragment NGAGPYGRP, while one reviewed protein of microbial origin, dihydroorotate dehydrogenase (quinone), was found containing the sequence LAAHFAR at positions 132–138. In this study, ACE inhibitory peptides with single-digit micromolar activity were first identified from cold-seep shellfish, such as LAAHFAR, highlighting the considerable potential of deep-sea fauna as a source of novel antihypertensive peptide leads.

2.4. Molecular Docking Visualization and Interaction Analysis

Docking visualization revealed distinct interaction patterns between the four peptides and the active pocket of ACE, highlighting structural determinants associated with their inhibitory activities. ACE has three classic active pockets, including the S1, S1’, and S2’. The S1 pocket consists of Ala354, Glu384, and Tyr523. The S2′ pocket has five residues, including Gln281, His353, Lys511, His513, and Tyr520, while the S1′ pocket contains one residue Glu162 [4,27]. Among the candidates, LAAHFAR exhibited the most extensive hydrogen-bonding network, forming interactions with Arg522, Ala356, Tyr523, Ala354, His513, His353, Lys511, Gln281, Cys370, Asp377, Glu162, and Thr372 (The bold font indicates the classic binding sites). Several of these residues—particularly Glu162 and Asp377—are recognized catalytic or substrate-recognition residues within the ACE active center (Figure 5A). Prior studies established Glu162, Glu384, Glu376, Asp377, Asp453, and the catalytic Zn²⁺ ion as core determinants of inhibitor binding [15]. The ability of LAAHFAR to engage multiple key residues suggests a highly stable anchoring mode, thereby enhancing its inhibitory effect. By comparison, FTTFGK interacted with a smaller subset of residues (Ala356, Tyr523, His353, His513, Lys511, Gln281, Glu162) and is coordinated with the Zn²⁺ ion (Figure 5B). NGAGPYGRP formed even fewer hydrogen bonds, mainly involving Ala356, Thr282, Glu376, Gly276, and Zn²⁺ (Figure 5C). YAAPYR, although capable of contacting several residues—including Ala356, Tyr523, His353, Asp415, Ala354, Gln281, Lys454, Cys370, Asp377, Cys352, Glu162, and Zn²⁺—still formed fewer interactions with catalytically essential acidic residues compared with LAAHFAR (Figure 5D).

Taken together, LAAHFAR displayed the highest number of hydrogen bonds among all peptides, including stable interactions with both Glu162 and Asp377, two residues central to ACE catalysis. This dense interaction network likely stabilizes the peptide, improving its capacity to interfere with substrate access to the Zn²⁺-containing catalytic motif. The docking data therefore align with the biochemical assays, supporting LAAHFAR as the peptide with the strongest ACE inhibitory activity due to its superior engagement with key residues in the ACE active site.

2.5. Molecular Dynamics Simulations

To understand the atomistic interactions, molecular dynamics (MD) simulations of 100 ns were performed using the Smina docking structure as the initial model for the deep-sea peptides. We first calculated and compared the total binding free energy. Then subsequent in-depth analysis was focused on the peptide LAAHFAR-ACE complex, given that this peptide showed remarkable inhibitory potency. Throughout the simulation, ACE maintained its structural integrity, with a backbone RMSD of 1.78 Å, indicative of a highly consistent dynamical profile. The peptide LAAHFAR underwent an initial conformational adjustment, showing pronounced deviations during the first 40 ns, with an average RMSD of 3.81 Å (Figure 6A). We extracted snapshots of the complex at 0 and 21.5 ns and found that the three C-terminal residues of the peptide L¹A²A³H⁴F⁵A⁶R⁷—His⁴, Phe⁵, and Arg⁷—underwent notable conformational changes. Specifically, the benzene ring plane of Phe⁵ rotated by more than 90 degrees from its initial position. In the latter half of the trajectory, the peptide RMSD converged to an average of 1.88 Å, and the overall mean across the full simulation was 2.65 Å. These observations indicate that although the peptide experienced early structural deflection, it ultimately adopted a stable pose within the ACE active pocket.

Structural robustness was further supported by a stable radius of gyration (Rg) of 2.47 nm and low RMSF values for residues located in the active site (Figure 6B,C), reflecting restricted local flexibility. The protein, composed predominantly of α-helical elements, showed no evident secondary-structure transitions during the simulation (Figure 6D,E). Hydrophobic and hydrophilic solvent-accessible surface areas (SASA) remained essentially constant, with mean values of 137.9 nm² and 129.4 nm² (Figure 6F), respectively, confirming that ACE preserved a native-like conformation suitable for subsequent binding and energetic analysis.

Binding free energy calculations were performed on snapshots extracted from the final 30 ns of the MD trajectory using the molecular mechanics/Poisson–Boltzmann surface area (MM/PBSA) approach. The computed overall binding free energy for the ACE-LAAHFAR complex was −38.23 kcal/mol, which was comparable with that of the peptide YAAPYR-ACE complex (Figure 7A). Energy decomposition revealed that van der Waals interactions dominated complex stabilization (−54.55 kcal/mol). Among the four peptides investigated, LAAHFAR contributed the largest polar interaction component to the binding free energy (45.97 kcal/mol). Per-residue decomposition highlighted Tyr523 and Glu411 of ACE as major contributors (Figure 7B), with energies of −3.61 kcal/mol and −2.45 kcal/mol, respectively.

The stability of the ACE-peptide complex was further supported by intermolecular hydrogen-bond analysis. Throughout the simulation, LAAHFAR maintained an average of 2.23 hydrogen bonds with the protein (Figure 7C). The most persistent hydrogen bonds were formed with Tyr523 (34.31%), His513 (25.14%), His383 (15.67%), and Gln281 (12.97%) (

Table 3. Analysis of hydrogen-bond interactions in the LAAHFAR-ACE complex, and the most prevalent hydrogen bonds are highlighted in bold.

No.	H-Bond Donor->Hydrogen…H-Bond Acceptor	Occupancy Frequency (%)	H-Bond Distance (Å)	H-Bond Angle (°)
1	GLN(281)@NE2(3918)->HE22(3920)…LIGAND@O5(9258)	6.54	3.05 ± 0.21	18.36 ± 6.96
2	GLN(281)@NE2(3918)->HE21(3919)…LIGAND@O6(9264)	12.97	2.99 ± 0.19	13.33 ± 6.96
3	HIS(353)@NE2(5060)->HE2(5061)…LIGAND@O3(9233)	2.20	2.96 ± 0.20	16.77 ± 7.39
4	HIS(353)@NE2(5060)->HE2(5061)…LIGAND@O4(9239)	3.59	2.99 ± 0.19	16.36 ± 7.62
5	HIS(353)@NE2(5060)->HE2(5061)…LIGAND@N5(9243)	2.11	3.06 ± 0.18	13.41 ± 7.31
6	ALA(354)@N(5066)->HN(5067)…LIGAND@O2(9227)	1.43	3.02 ± 0.18	15.77 ± 7.41
7	ALA(354)@N(5066)->HN(5067)…LIGAND@NE2(9255)	4.18	3.12 ± 0.16	13.21 ± 6.80
8	LYS(511)@NZ(7527)->HZ3(7530)…LIGAND@O1(9222)	1.09	2.81 ± 0.15	16.49 ± 7.05
9	HIS(513)@NE2(7564)->HE2(7565)…LIGAND@O1(9222)	17.27	2.98 ± 0.21	14.47 ± 7.16
10	HIS(513)@NE2(7564)->HE2(7565)…LIGAND@O2(9227)	25.14	2.96 ± 0.19	16.06 ± 7.33
11	HIS(513)@NE2(7564)->HE2(7565)…LIGAND@O3(9233)	1.30	3.05 ± 0.21	19.24 ± 7.38
12	TYR(520)@OH(7668)->HH(7669)…LIGAND@O5(9258)	10.67	2.74 ± 0.14	12.85 ± 6.64
13	TYR(523)@OH(7732)->HH(7733)…LIGAND@O2(9227)	2.20	2.82 ± 0.19	16.76 ± 7.23
14	TYR(523)@OH(7732)->HH(7733)…LIGAND@NE2(9255)	34.31	2.95 ± 0.15	12.21 ± 6.18
15	LIGAND@N5(9243)->H1(9244)…SER(516)@OG(7610)	1.83	3.22 ± 0.15	15.35 ± 7.58
16	LIGAND@ND1(9252)->HD1(9311)…GLU(411)@OE2(5980)	1.62	3.02 ± 0.18	15.17 ± 7.35
17	LIGAND@OXT(9270)->HXT(9271)…HIS(383)@ND1(5530)	3.71	3.33 ± 0.13	20.28 ± 6.84
18	LIGAND@OXT(9270)->HXT(9271)…HIS(383)@NE2(5535)	15.67	3.11 ± 0.17	16.74 ± 7.14
19	LIGAND@NH2(9277)->HH21(9279)…GLU(376)@OE2(5430)	1.26	3.10 ± 0.18	16.04 ± 6.89
20	LIGAND@NH2(9277)->HH22(9280)…GLU(376)@OE2(5430)	1.24	3.11 ± 0.19	19.29 ± 6.70
21	LIGAND@NH1(9278)->HH12(9282)…GLU(376)@OE1(5429)	3.30	3.11 ± 0.18	15.20 ± 7.25
22	LIGAND@NH1(9278)->HH12(9282)…GLU(376)@OE2(5430)	3.96	3.10 ± 0.18	15.09 ± 6.97
23	LIGAND@NH1(9278)->HH12(9282)…ASP(453)@OD1(6561)	2.79	3.11 ± 0.19	16.93 ± 7.26
24	LIGAND@NH1(9278)->HH11(9281)…ASP(453)@OD1(6561)	3.52	3.07 ± 0.19	16.37 ± 7.59
25	LIGAND@NH1(9278)->HH12(9282)…ASP(453)@OD2(6562)	4.51	3.10 ± 0.18	16.61 ± 6.94
26	LIGAND@NH1(9278)->HH11(9281)…ASP(453)@OD2(6562)	3.07	3.11 ± 0.19	17.28 ± 7.65

Footnote: Hydrogen-bond interactions are denoted in the DonorAtom->Hydrogen…AcceptorAtom format. The residue and atom names follow standard conventions, and numbers in parentheses correspond to global atom indices in the simulation system. LIGAND indicates the peptide LAAHFAR.

). These interactions were consistent with the binding residues identified in the preceding molecular docking study, reinforcing that the peptide remained anchored within critical regions of the ACE active site.

Dynamic cross-correlation matrix (DCCM) analysis revealed coordinated residue motions across ACE in the presence of LAAHFAR (Figure 7D), suggesting favorable global dynamical coupling in the complex. In particular, Glu162 displayed a clear positively correlated motion with Gln281 and residues within the Ala510–Gln530 region, suggesting concerted movement between spatially separated structural elements upon peptide binding. In contrast, Glu162 showed strong negative correlation with residues in the Val350–Val380 region, indicating opposing directional motions among these segments. Projection of the MD trajectory onto the first two principal components (PC1 and PC2) yielded a well-defined free-energy landscape (FEL) (Figure 7E,F). A deep, dominant low-energy basin was observed for the ACE–LAAHFAR system, corresponding to the stable peptide conformation attained in the later stages of the simulation. The minimum-energy structure differed from the docking pose by only 1.88 Å, indicating that the final MD conformation represented a refined but closely related bound state. The existence of a single deep energy well suggested that this configuration was energetically favored and was unlikely to undergo transitions across significant energy barriers.

3. Materials and Methods

3.1. Materials and Chemicals

Specimens of the deep-sea organisms Gigantidas platifrons (GP), Shinkaia crosnieri (SC), and Lamellibrachia columna (LC) were obtained in 2017–2018 from the cold-seep site “F” (22°06′ N, 119°17′ E) in the South China Sea. While the deep-sea species, Archivesica marissinica (AM) and Gigantidas haimaensis (GH) were collected from Haima cold seep (16°43′46″ N, 110°28′22″ E, depth 1389 m) during the HYDZ6–202205 cruise in 2022. All samples were immediately frozen in liquid nitrogen and transferred to a −80 °C refrigerator for storage. All samples were freeze-dried and ground into powder for use after being transported back to the laboratory.

Pineapple bromelain was obtained from Shanghai Yuanye Biotechnology Co., (Shanghai, China). Trypsin was purchased from the China National Pharmaceutical Group Co., Ltd. (Beijing, China), and pepsin was sourced from the Macklin Biochemical Co., Ltd. (Shanghai, China). Angiotensin-converting enzyme and the substrate hippuryl-his-leu-OH (HHL) were obtained from Sigma Aldrich Co., Ltd. (Shanghai, China).

3.2. Preparation of Deep-Sea Animal Protein Hydrolysates

The protein hydrolysates were prepared with three different proteases. Briefly, powders from five deep-sea animals (AM, GH, GP, LC, SC) were hydrolyzed using pepsin, trypsin, and bromelain. An amount of 10 g of deep-sea animal powders was added to 200 mL of deionized water and thoroughly blended. The pH was adjusted and the reaction temperatures set for each enzyme: pepsin, pH = 2, 37 °C; trypsin, pH = 8, 37 °C; bromelain, pH = 7, 50 °C. Finally, proteases were then added at a 1% (w/w) enzyme-to-substrate (E/S) ratio. After a hydrolysis duration of 4 h, the reaction mixtures were heated to 95 °C for 15 min to deactivate the protease. Subsequently, the mixtures were centrifuged at 8000 g for 10 min to collect the supernatant, which was then freeze-dried to obtain the protein hydrolysates. The protein hydrolysates were stored at −80 °C until use.

3.3. Amino Acid Composition Analysis

Seventeen hydrolyzed amino acids (HAAs) were measured using an amino acid analyzer (MembraPure GmbH, Hennigsdorf, Germany) with minor modifications [28]. Approximately 10 mg of sample powder was subjected to acid hydrolysis in 6 mol/L HCl at 110 °C for 24 h in sealed tubes. After hydrolysis, the solutions were filtered, rinsed with ultrapure water, evaporated at 100 °C, and stored in airtight, moisture-resistant containers. Before analysis, the dried residues were reconstituted in 10 mL of 0.02 mol/L HCl, vortexed thoroughly, and passed through syringe filters. Filtrates of at least 1.5 mL (excluding the initial drops) were collected for subsequent quantification.

The instrument was operated under the following conditions: cleaning flow rate 0.1; eluent flow rate/pressure 0.45/36–37 Pa; and derivatization reagent flow rate/pressure 0.25/8–9 Pa. Amino acid concentrations were calculated according to standard quantitative formulas:

$$Contents = {\displaystyle \frac{sample peak area}{standard peak area}} \times standard concentration \times {\displaystyle \frac{10 \mathrm{mL}}{sample weight}} \times 100\%$$

(1)

3.4. ACE Inhibitory Activity Assay

The ACE inhibitory activity was assessed following previously reported procedures, with minor adjustments [29]. Three reaction systems were prepared. (1) Reagent blank: 50 μL of borate-buffered saline (BBS; 50 mM sodium borate, 300 mM NaCl, pH 8.2). (2) Negative control (100% ACE activity): 20 μL of ACE solution (0.1 U/mL) mixed with 30 μL of BBS. (3) Test samples: 20 μL of ACE solution combined with 20 μL of peptide sample and 10 μL of BBS. All mixtures were pre-incubated at 37 °C for 30 min with gentle shaking. Reactions were initiated by adding 50 μL of the substrate solution (5 mM hippuryl-histidyl-leucine dissolved in BBS). After incubation at 37 °C for 1 h, reactions were quenched by adding 150 μL of 1 M HCl. The mixtures were then filtered through 0.22 μm nylon membranes, and 20 μL of the filtrate was analyzed by RP-HPLC (Shimadzu, Kyoto, Japan) using an Eclipse XDB-C18 column (4.6 × 150 mm, 5 μm; Agilent, Santa Clara, CA, USA). Chromatographic separation was achieved under isocratic elution with a mobile phase of acetonitrile/water (22:78, v/v) containing 0.1% trifluoroacetic acid, at a flow rate of 0.8 mL/min. The column temperature was kept at 30 °C, and hippuric acid (HA) was detected at 228 nm using a UV detector (Tecan Group Ltd., Männedorf, Switzerland). The ACE inhibition rate was calculated using the following equation:

$$\begin{array}{c}\mathrm{ACE} \mathrm{inhibitory} \mathrm{rate}\left(\mathrm{\%}\right)=\left({\displaystyle \frac{{\mathrm{A}}_{\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}-{\mathrm{A}}_{\mathrm{e}\mathrm{x}}}{{\mathrm{A}}_{\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}-{\mathrm{A}}_{\mathrm{b}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{k}}}}\right)\times 100\%\end{array}$$

(2)

where A_control is the HA peak area of the negative control, A_ex is the HA peak area of the test group, and A_blank is the HA peak area of the reagent blank. The 100% ACE inhibitory rate refers to the full catalytic capacity of ACE in the reaction mixtures containing only the enzyme and substrate, free of any inhibitors. For determining the IC₅₀ of the selected peptides, their final concentrations in the reaction mixtures were from 0.625 μM to 40 μM. The IC₅₀ value is calculated using the log(inhibitor) vs. response–variable slope model with GraphPad Prism 9.5 software.

3.5. Peptidomics Analysis

Protein hydrolysates from GP underwent reduction in 10 mM dithiothreitol at 56 °C for 1 h, followed by alkylation in 50 mM iodoacetamide at room temperature for 40 min in the absence of light. The treated samples were evaporated to dryness, dissolved in 0.1% formic acid, and approximately 1 μg of peptides was injected into an Ultimate 3000 chromatographic system interfaced with a Q Exactive^TM Hybrid Quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). Peptide separation was achieved on a ReproSil-Pur C18-AQ analytical column (150 μm × 15 cm, 1.9 μm, 100 Å, Dr. Maisch High Performance LC GmbH, Ammerbuch, Germany) operated at a flow rate of 580 nL/min with a multi-segment gradient of acetonitrile in 0.1% formic acid. The mass spectrometer was operated in data-dependent acquisition (DDA) mode. MS data were acquired at a resolving power of 70,000 (m/z 400) using higher-energy collisional dissociation. The resulting spectra were processed and de novo sequenced in PEAKS Studio (version 10.6, BSI, Canada), and the identified peptide sequences were matched against the Mollusca protein database compiled from UniProtKB (release of 9 October 2024) selected based on taxonomic relevance and annotation completeness. Carbamidomethylation of cysteine residues was set as a fixed modification, while methionine oxidation and protein N-terminal acetylation were considered as variable modifications. The mass tolerances for precursor and fragment ions were set according to Orbitrap high-resolution standards. Peptide-spectrum matches were filtered using a false discovery rate (FDR) threshold of 1% at both peptide and protein levels.

3.6. Molecular Docking and Virtual Screening

The crystal structure of human ACE (PDB ID: 1O86) served as the receptor model for the docking analysis. Three-dimensional conformations of the oligopeptides were constructed in PyMOL. The docking grid was centered at coordinates x = 40.55, y = 37.39, z = 43.47, with a defined box dimension of 20 × 20 × 20 Å. Docking simulations were performed with Smina 2020.12.10 (https://github.com/mwojcikowski/smina, available until 6 December 2025), a modified implementation of AutoDock Vina 1.12, to evaluate peptide–ACE interactions [30,31,32]. Predicted binding affinities were used to select candidate peptides for chemical synthesis and subsequent in vitro evaluation of ACE inhibitory activity. To verify the uniqueness of the peptides, all sequences were queried against publicly available repositories, including the BIOPEP-UWM database [33].

3.7. Peptide Synthesis

The peptides were produced via standard Fmoc-based solid-phase peptide synthesis by GenScript Biotech Corporation (Nanjing, China). Their chemical purity was assessed by high-performance liquid chromatography, and the corresponding molecular masses were confirmed using electrospray ionization mass spectrometry. All peptides possess free C- and N-terminal ends. The purity of the peptides was above 95%, measured by HPLC. The HPLC spectra and MS spectra of the synthesized peptides (Figures S5–S62) are attached in the Supplementary Materials.

3.8. Molecular Dynamics Simulations

Molecular dynamics simulations were carried out with GROMACS 2024.2 to characterize the interactions and binding behavior of the protein–ligand complexes [34]. Complex structures of the peptides were parameterized using the CHARMM36 force field for ACE and the CHARMM General Force Field (CGenFF) parameters for the ligand molecules [35,36]. The catalytic Zn²⁺ ion present in ACE was retained in its coordinated, protonated state throughout the simulations and was treated as an integral part of the protein group. Each system was placed in a dodecahedral simulation box filled with TIP3P water, ensuring at least 10 Å between the solute and the box edges, and electrically neutralized by adding the appropriate counterions. To ensure overall charge neutrality and approximate physiological ionic conditions, appropriate numbers of Na⁺ counterions were added to the solvent. Both the protein and peptide ligands were simulated with charged N- and C-termini, consistent with the standard CHARMM36 and CGenFF force-field conventions. Energy minimization proceeded until the maximum force fell below 1000 kJ/mol. The systems were subsequently equilibrated under NVT and NPT ensembles at 310 K and 1 atm. Production trajectories were generated for 100 ns, with coordinates recorded every 10 ps. Structural stability and dynamic properties were evaluated through the root mean square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration (Rg), solvent-accessible surface area (SASA), hydrogen-bond profiles, and principal component analysis. Binding free energies (ΔG) were computed using the MM/PBSA approach based on the final 30 ns of the trajectories [37]. Residue-wise energy decomposition was also performed to quantify the energetic contributions of amino acids within 3 Å of the ligand.

3.9. Statistical Analysis

All experiments were repeated at least three times. Data analysis and visualization were performed using GraphPad 8 (GraphPad Software, San Diego, CA, USA). Data were analyzed with one-way analysis of variance (one-way ANOVA). The results are shown as the mean ± SEM (n = 3), and differences were considered significant when p < 0.05 and marked with an asterisk (*).

4. Conclusions

In this study, we successfully identified four novel ACE inhibitory peptides from the deep-sea mollusk Gigantidas platifrons by integrating peptidomics with virtual screening. The most potent peptide, LAAHFAR, exhibited significant inhibitory activity with an IC₅₀ value of 6.01 μM. Molecular dynamics simulation provided initial insights into its potential binding mode. Our work validated the deep-sea mollusk as an important source for bioactive peptides and demonstrated the efficiency of the combined in silico and peptidome approach for peptide discovery. However, the study was limited to in vitro assays and computational analysis; the in vivo antihypertensive efficacy and physiological stability of these peptides remain to be evaluated. Furthermore, the proposed binding mode of the lead peptide relied on computational simulations without experimental validation such as mutagenesis or binding affinity assays. The screening was focused exclusively on ACE inhibition, leaving potential off-target effects or interactions with other regulatory pathways unexamined. Future work should focus on in vivo validation using hypertensive animal models, alongside investigations into the absorption, metabolism, and potential toxicity of the peptide to assess its therapeutic potential.