Chemophysical Characterization of Carotenoproteins From Crustacean Shell Waste for Pharmaceutical Applications Authors: Luisa Di Paola, Gabriella Caruso, Francesca Manganiello, Chiara Bianca Maria Platania, Adriana Albini, Douglas Noonan, Claudio Serangeli Published: 2026 ABSTRACT Commercial crustacean species represent some of the most economically valuable fishery products; however, their full potential remains underexploited due to the routine disposal of shellfish waste, which is rich in high‐value bioactive compounds. Among these, carotenoid pigments have garnered increasing interest for their antioxidant and anti‐inflammatory properties. In this study, we investigate the potential of recovering bioactive compounds from the shells of the European lobster (Homarus gammarus Linnaeus, 1758) and the Atlantic blue crab (Callinectes sapidus Rathbun, 1896), with a particular focus on carotenoprotein complexes with promising applications in the pharmaceutical and cosmetic industries. We adopted a computational approach integrating protein contact networks, molecular docking, and molecular dynamics simulations to characterize the structural and functional properties of crustacean carotenoproteins as nanocarriers for astaxanthin. Then, we explored in silico the antioxidant activity of astaxanthin, as claimed by many studies. Specifically, we applied the computational analyses to the interaction between astaxanthin and Kelch‐like ECH‐associated protein 1 (Keap1‐DC domain), a key regulator of the Nrf2 signaling pathway involved in oxidative stress. Results suggest that astaxanthin would compete with Nrf2 at the Keap1‐DC domain, then promote Nrf2 induction and activate the antioxidant cellular machinery. These findings are supported by preliminary experimental evidence and highlight the potential of astaxanthin extracted from crustacean shell waste as a bioactive agent, with possible applications in diseases associated with oxidative stress, including cancer. Further preclinical and clinical in vivo studies are needed to validate astaxanthin and the carotenoprotein–astaxanthin complex for efficacy and safety, with potential relevance for oncology. 1. Introduction The seafood processing industry generates 50%–80% of waste by weight [1], produced by fishing and farming activities. Seafood waste refers to both edible and inedible raw materials that are discarded across the food chain, with consequent losses deriving from practices adopted by producers, retailers, food service operators, and consumers [2]. Crustaceans—including shrimp, lobster, crab, and crayfish—are among the most economically important species in the global seafood industry, representing approximately 5%–10% of total seafood production volume [3, 4]; their high market prices further amplify their share in terms of overall market value. Crustacean processing (especially of crab and lobster) generates as much as 30%–85% solid waste, including shells and heads [2, 5, 6], depending on the species and on the processing method, handling procedures, and storage conditions [7, 8]. Crustacean shell waste alone generates an estimated 6–8 million tons annually, thus being one of the largest streams of by‐products arising from seafood processing [9]. The biorefinery approach allows for transforming seafood processing waste and by‐products into value‐added products, thereby addressing environmental concerns while complying with circular economy principles [10]. In this scenario, crustacean wastes offer a unique opportunity for the sustainable valorization of products with high economic potential, which would otherwise be lost. In this work, we focus on by‐products from two decapod crustacean species: the European lobster (Homarus gammarus Linnaeus, 1758) and the Atlantic blue crab (Callinectes sapidus Rathbun, 1896). H. gammarus is a commercially appreciated lobster species widely distributed in the European region. Its distribution in the benthic ecosystem is controlled bottom‐up by food availability and top‐down by predation [11]. Lobster processing generates a high amount of waste (heads, shells, eggs, and livers; over 50,000 tons annually [12]), covering 50%–70% of the total shellfish waste for all commercial species. The valorization of lobster processing waste is highly recommended thanks to the presence of several high‐value compounds such as proteins, chitin, lipids, minerals, and pigments with nutraceutical, cosmetic, and pharmaceutical applications. Likewise, the shell waste from C. sapidus is a resource that deserves to be exploited for its biotechnological potential. C. sapidus is an invasive species in the Mediterranean region, representing an emerging ecological and socioeconomic threat [13]. Due to its opportunistic predation behavior on fish eggs, mollusks, crustaceans, and benthic invertebrates, this species disrupts native food webs and reduces biodiversity, causing disturbances to sediment structure and aquatic vegetation and affecting nursery habitats for native species. Due to its competition for food and space with native crabs and other benthic organisms, C. sapidus can cause significant shifts in local communities and habitat alteration. The species’ high reproductive rate and adaptive ability to colonize estuaries, lagoons, and coastal waters make the containment of the spread of C. sapidus increasingly difficult in temperate environments. C. sapidus negatively impacts both tourism and fisheries, directly threatening the economy of coastal communities. It is therefore necessary to implement effective strategies for the monitoring and control of this invasive species. One viable route is to valorize C. sapidus catches beyond their use in the food sector, fully exploiting the species’ economic potential [14]. This approach is sustainable in the long term, as it simultaneously supports population control and strengthens the economy of coastal communities, enhancing their resilience to environmental pressures such as climate change and the growing presence of invasive species. Recent advances in biorefinery processes have underscored the importance of sustainably extracting value‐added compounds from biological feedstocks to improve the resource efficiency of industrial processes in the pharmaceutical, nutraceutical, and cosmetic fields [15]. In the fishery sector, the biorefinery approach aims at valorizing all components of fishery products to produce additional value and reduce the volume of production waste to near zero [16]. This approach is particularly impactful if applied to residual biological sources and production waste within the circular bioeconomy framework [8, 17]. In this work, we explore the structural and functional properties of the carotenoproteins present in the shells of H. gammarus and C. sapidus. We also evaluated the activity of astaxanthin, the carotenoid complexed and delivered by the decapod carotenoproteins, as a putative antioxidant agent, targeting Keap1. For the analysis of protein molecular structures, we applied molecular dynamics (MD) and molecular docking to provide a quantitative structural analysis of the carotenoproteins from the two species. To obtain a residue‐level characterization of the functional properties of the shell carotenoproteins from both species and of the human Keap1 forms—in their unbound state and in complex with astaxanthin—we employed the protein contact network (PCN) approach. This method has been extensively validated as a powerful framework for probing the structural and functional role of individual residues within protein architectures [18] and has proven particularly effective in elucidating mechanisms of allosteric regulation [19, 20]. We also explored in silico the activity of astaxanthin as a therapeutic agent toward oxidative stress by molecular docking against Keap1. Astaxanthin is a potent antioxidant due to direct physical and chemical quenching of reactive oxygen species [21]. Beyond its direct radical‐scavenging capacity, astaxanthin exerts indirect antioxidant activity through the transcriptional activation of cytoprotective machinery, most notably the nuclear factor erythroid 2‐related factor 2/heme oxygenase‐1 (Nrf2/HO‐1) signaling pathway [22], which is also implicated in cancer development and response to therapy. Therefore, based on the astaxanthin structure, we identified the kelch domain of Keap1 as a putative pharmacological target, which is a negative regulator of Nrf2 activity and a recognized molecular node in oxidative stress– and cancer‐related signaling. Any compound able to interfere with the Keap1–Nrf2 interaction is an Nrf2 inducer, thus promoting activation of the cellular antioxidant machinery [23–26]. Astaxanthin cannot induce Nrf2 through nucleophilic addition with Cys151 of the Keap1‐BTB domain [27], because it lacks an sp2 leaving group for any nucleophilic addition, such as Michael addition. Astaxanthin would likely be a protein–protein interaction disruptor by interfering with the interaction between the Keap1 kelch domain (Keap1‐DC) and Nrf2 [28]. In this study, we predict the binding of astaxanthin to the Keap1 kelch domain by molecular docking and analyze the Keap1–astaxanthin complex by means of MD simulations, comparing the results with those for the unbound Keap1 form. 2. Materials and Methods 2.1. Structural Information of Molecules We introduced different computational methodologies to clarify the molecular function of shell carotenoproteins through quantitative descriptors. The computational methods were applied to the following molecular structures provided as PDB files:– β‐Crustacyanin of H. gammarus, PDB code 1GKA [21]: the molecule is a heterodimer (see Figure 1A);–Carotenoprotein of C. sapidus: the molecular structure shown in Figure 1B was designed by applying the following protocol: first, we identified the gene of the C. sapidus apolipoprotein D‐like protein (https://www.ncbi.nlm.nih.gov/protein/2091619298) by means of BLAST using as a template the crustacyanin A2 subunit of H. gammarus (P80007.1, Chain B in PDB: 1GKA). We did not find in C. sapidus any homologous protein to the crustacyanin A1 subunit of H. gammarus (Chain A in PDB: 1GKA), so we supposed that the biologically active form of C. sapidus is a homodimer. The monomer model for the C. sapidus apolipoprotein was created by means of the SwissModel platform (https://swissmodel.expasy.org/) using the abovementioned sequence and the structure of H. gammarus as a template; the dimer bound to two molecules of astaxanthin was then created in the PyMOL environment using H. gammarus crustacyanin as a guide for superposition. All structures underwent energy minimization before the MD simulation. In Figure 1, the two astaxanthin molecules are in the inner cavity of the two heterodimers, made up of two moieties belonging to the two chains. In H. gammarus, β‐crustacyanin is a heterodimer belonging to the lipocalin family [29], binding two astaxanthin molecules (orange and yellow, represented as VdW spheres in Figure 1A). – β‐Crustacyanin of H. gammarus, PDB code 1GKA [21]: the molecule is a heterodimer (see Figure 1A); β‐Crustacyanin of H. gammarus, PDB code 1GKA [21]: the molecule is a heterodimer (see Figure 1A); –Carotenoprotein of C. sapidus: the molecular structure shown in Figure 1B was designed by applying the following protocol: first, we identified the gene of the C. sapidus apolipoprotein D‐like protein (https://www.ncbi.nlm.nih.gov/protein/2091619298) by means of BLAST using as a template the crustacyanin A2 subunit of H. gammarus (P80007.1, Chain B in PDB: 1GKA). We did not find in C. sapidus any homologous protein to the crustacyanin A1 subunit of H. gammarus (Chain A in PDB: 1GKA), so we supposed that the biologically active form of C. sapidus is a homodimer. The monomer model for the C. sapidus apolipoprotein was created by means of the SwissModel platform (https://swissmodel.expasy.org/) using the abovementioned sequence and the structure of H. gammarus as a template; the dimer bound to two molecules of astaxanthin was then created in the PyMOL environment using H. gammarus crustacyanin as a guide for superposition. All structures underwent energy minimization before the MD simulation. In Figure 1, the two astaxanthin molecules are in the inner cavity of the two heterodimers, made up of two moieties belonging to the two chains. In H. gammarus, β‐crustacyanin is a heterodimer belonging to the lipocalin family [29], binding two astaxanthin molecules (orange and yellow, represented as VdW spheres in Figure 1A). Carotenoprotein of C. sapidus: the molecular structure shown in Figure 1B was designed by applying the following protocol: first, we identified the gene of the C. sapidus apolipoprotein D‐like protein (https://www.ncbi.nlm.nih.gov/protein/2091619298) by means of BLAST using as a template the crustacyanin A2 subunit of H. gammarus (P80007.1, Chain B in PDB: 1GKA). We did not find in C. sapidus any homologous protein to the crustacyanin A1 subunit of H. gammarus (Chain A in PDB: 1GKA), so we supposed that the biologically active form of C. sapidus is a homodimer. The monomer model for the C. sapidus apolipoprotein was created by means of the SwissModel platform (https://swissmodel.expasy.org/) using the abovementioned sequence and the structure of H. gammarus as a template; the dimer bound to two molecules of astaxanthin was then created in the PyMOL environment using H. gammarus crustacyanin as a guide for superposition. All structures underwent energy minimization before the MD simulation. In Figure 1, the two astaxanthin molecules are in the inner cavity of the two heterodimers, made up of two moieties belonging to the two chains. In H. gammarus, β‐crustacyanin is a heterodimer belonging to the lipocalin family [29], binding two astaxanthin molecules (orange and yellow, represented as VdW spheres in Figure 1A). Molecular structures of interest. (A) H. gammarus β‐crustacyanin (PDB: 1GKA). (B) C. sapidus apolipoprotein D‐like protein (model built as reported in Section 2). In green (Chain A) and blue (Chain B) are the two chains binding two molecules of astaxanthin (yellow and orange spheres); the residues in contact with the two molecules of astaxanthin are highlighted in red. (C) Chemical structure of astaxanthin (3,3 ′‐dihydroxy‐β,β‐carotene‐4,4 ′‐dione). (D) Keap1 kelch domain structure: the six blades (domains) are shown in different colors. (E) Electrostatic surface potential representation of the kelch domain. Blue indicates negatively charged regions, red corresponds to positively charged areas, and white highlights hydrophobic surface patches. (F) Keap1 kelch domain bound to the Nrf2 peptide (in magenta mesh in the figure). In the model of the C. sapidus complex, the two molecules of astaxanthin appear more twisted (see Figure 1B). Figure 1 also shows the residues in contact with the astaxanthin molecules (in red): the distribution of these residues is quite different in the two conformations, being more extended in the crustacyanin of H. gammarus than in the apolipoprotein of C. sapidus. For both conformations, the distribution of residues in contact with astaxanthin is uneven: in H. gammarus, 33 residues in Chain A are in contact with astaxanthin and eight in Chain B, while in the C. sapidus complex, 32 residues are in contact with astaxanthin in Chain A and 21 in Chain B.–Astaxanthin (its molecular structure is shown in Figure 1C): it has gained attention for its beneficial health effects, including its potential role in preventing oxidative stress–related diseases [21, 30]. Astaxanthin (3,3 ′‐dihydroxy‐β,β‐carotene‐4,4 ′‐dione) is a major pigment in decapod shells: it is a carotenoid xanthophyll pigment with the chemical formula C40H52O4, a molecular mass of 596.85 Da, and a density of 1.081 g/L (see Figure 1C).–Keap1, the kelch domain of human Keap1, PDB code 1U6D [26]: it is shown in Figure 1D (sequence segment 322–609). The kelch domain of human Keap1 comprises six kelch repeats forming a symmetric, six‐bladed β‐propeller structure (50 Å in diameter and 35 Å in height, with a small channel—around 6 Å in diameter) through the center of the propeller. In Figure 1E, the narrowest part of the kelch domain is negatively charged and interacts directly with the LDEETGEFA peptide. Table 1 reports the sequence distribution of the six blades in different colors from blue to red in Figure 1E. –Astaxanthin (its molecular structure is shown in Figure 1C): it has gained attention for its beneficial health effects, including its potential role in preventing oxidative stress–related diseases [21, 30]. Astaxanthin (3,3 ′‐dihydroxy‐β,β‐carotene‐4,4 ′‐dione) is a major pigment in decapod shells: it is a carotenoid xanthophyll pigment with the chemical formula C40H52O4, a molecular mass of 596.85 Da, and a density of 1.081 g/L (see Figure 1C). Astaxanthin (its molecular structure is shown in Figure 1C): it has gained attention for its beneficial health effects, including its potential role in preventing oxidative stress–related diseases [21, 30]. Astaxanthin (3,3 ′‐dihydroxy‐β,β‐carotene‐4,4 ′‐dione) is a major pigment in decapod shells: it is a carotenoid xanthophyll pigment with the chemical formula C40H52O4, a molecular mass of 596.85 Da, and a density of 1.081 g/L (see Figure 1C). –Keap1, the kelch domain of human Keap1, PDB code 1U6D [26]: it is shown in Figure 1D (sequence segment 322–609). The kelch domain of human Keap1 comprises six kelch repeats forming a symmetric, six‐bladed β‐propeller structure (50 Å in diameter and 35 Å in height, with a small channel—around 6 Å in diameter) through the center of the propeller. In Figure 1E, the narrowest part of the kelch domain is negatively charged and interacts directly with the LDEETGEFA peptide. Table 1 reports the sequence distribution of the six blades in different colors from blue to red in Figure 1E. Keap1, the kelch domain of human Keap1, PDB code 1U6D [26]: it is shown in Figure 1D (sequence segment 322–609). The kelch domain of human Keap1 comprises six kelch repeats forming a symmetric, six‐bladed β‐propeller structure (50 Å in diameter and 35 Å in height, with a small channel—around 6 Å in diameter) through the center of the propeller. In Figure 1E, the narrowest part of the kelch domain is negatively charged and interacts directly with the LDEETGEFA peptide. Table 1 reports the sequence distribution of the six blades in different colors from blue to red in Figure 1E. Table 1Distribution along the sequence of the six blades of the kelch domain of Keap1.BladesSequenceI322–365; 597–609II366–416III417–463IV464–510V511–557VI558–596 Distribution along the sequence of the six blades of the kelch domain of Keap1. Figure 1F shows the complex of Keap1 with the LDEETGEFA Nrf2 peptide (PDB code 7K2A): the binding site is on the narrowest side of the kelch domain, at the center of the propeller. 2.2. Computational Analysis of Molecular Structures 2.2.1. MD and Molecular Docking Simulations Classical MD simulations were performed by means of the open‐source software GROMACS v. 2024.4 [31] with the AMBER99SB‐ILDN force field [32]. We performed MD simulations of 1 ns for all structures described above on the Neurosnap platform (https://neurosnap.ai/); each simulation was replicated three times. Each system was placed in a cubic box with periodic boundary conditions. The box size was automatically adjusted in each case to optimize the MD simulation. Each box was filled with water molecules and an appropriate number of counterions (Na+ and Cl−) necessary for system neutrality (0.15 M). The temperature was kept in the NPT ensemble at 300 K and the pressure at 1 bar. We applied a canonical analysis of motion to the MD frames to identify concerted motions of protein regions, according to a procedure explained extensively elsewhere [33]. Briefly, once the displacement matrix (DISPL) is computed from the MD trajectory, it is possible to compute the dynamic cross‐correlation (DCC) matrix, whose generic elementDCC i j is the Pearson correlation coefficient between the displacement vectors of the i‐th andj‐th residues. The DCC allows one to highlight the concerted motions between distal residues [33]: high values of the DCC matrix highlight the high probability of concerted motions of residue pairs. In this work, we introduced a correction to the displacement evaluation for each residue to get rid of effects due to rototranslation and to keep only motion contributions that are independent of rigid body movements of the whole molecular structure [34]. We performed molecular docking of astaxanthin with Keap1 by means of the DiffDock‐L molecular docking software [35] on the Neurosnap platform. The receptor–ligand complex was then annealed by means of MD simulations, as previously described. 2.2.2. PCN Analysis We applied the method of PCNs to analyze the protein molecular structures of carotenoproteins and both forms of the kelch domain of Keap1. For all structures, we analyzed the last frame after MD relaxation, as detailed above. The PCN method is based on the representation of the protein molecular structure as a network, whose nodes are the single residues and whose links between nodes are the active noncovalent contacts (between residues separated by 4–8 Å) [36]. The network is described by means of an undirected unweighted graph, whose mathematical representation is provided by the adjacency matrix A, defined as follows: (1) A=148 if A° 0). Finally, the negative values of Δ i G indicate that the interface between the two chains in both complexes is hydrophobic (positive protein affinity). Table 4 reports the protein stability for all the Keap1 forms (computation was applied to the final frame after the relaxation over 1 ns): the bound forms appear very similar in stability, confirming the possible role of astaxanthin as a competitive inhibitor of the binding to the Nrf2 peptide. Table 4Protein stability of Keap1 forms.Keap1 form Δ G_unfoldingRMSD (Å)Unbound14.474—Kelch–Nrf214.2011.147Kelch–AXT14.4570.930 Protein stability of Keap1 forms. Additionally, the RMSD value for the Nrf2 apo form, with respect to the bound form, is slightly larger than that for the Keap1–AXT complex, further promoting the latter as a possible complex if astaxanthin molecules are present in the Keap1 microenvironment. 4. Conclusions In this study, we analyze the structural properties of two crustacean carotenoproteins that complex astaxanthin. We also demonstrate in silico the biological activity of astaxanthin as an antioxidant agent, targeting the human Keap1 kelch domain. These findings provide molecular‐level evidence of how seafood processing by‐products, often considered environmental burdens, can be transformed into sustainable sources of bioactive molecules. We hereby produce a computational model for the C. sapidus apolipoprotein D‐like protein, homologous to H. gammarus crustacyanin. The stable carotenoprotein complex protects astaxanthin from photooxidation; thus, this complex represents a new, more stable source of astaxanthin, with improved pharmacokinetic and possibly pharmacodynamic properties. Therefore, we hereby provide insight into an innovative nutraceutical formulation of astaxanthin complexed with crustacean carotenoproteins isolated from shell waste. Currently, the European Commission considers astaxanthin to be a food dye, but in Italy, hundreds of nutraceuticals for human consumption contain astaxanthin as one of the active principles (https://www.salute.gov.it/new/it/tema/alimenti-fini-medici-speciali-ed-integratori/registro-degli-integratori-alimentari/), usually extracted from the microalga Haematococcus pluvialis [57]. Therefore, carotenoproteins complexing astaxanthin would be not only an alternative and sustainable source of astaxanthin but also a complex that serves as an alternative food supplement, improving the safety and efficacy of astaxanthin compared to the free form [58]. We also identify the Keap1‐DC domain as a molecular target for the astaxanthin antioxidant agent, competing with Nrf2 binding and promoting its induction, highlighting a pathway of considerable interest in oncology. All in all, our results highlight the relevance of the valorization of crustacean shell waste to foster circular economy principles, contributing simultaneously to waste reduction, environmental protection, and innovation in food and health sciences. The use of invasive species such as C. sapidus in the Mediterranean Sea area further underscores the dual benefit of mitigating their ecological impact while unlocking the biotechnological potential of these fishery resources. While additional experimental validation and translational studies are needed, this work paves the way for the development of seafood‐derived functional ingredients and therapeutic products. In doing so, it reinforces the concept that marine waste and by‐products represent a valuable reservoir for sustainable biorefinery processes, bridging food science, biotechnology, and human health. Funding Funding was provided by the Next Generation EU Fund, Hub Rome Technopole (Spoke 2). Open access publishing was facilitated by Università Campus Bio‐Medico di Roma as part of the Wiley‐CRUI‐CARE agreement. Conflicts of Interest The authors declare no conflicts of interest.