Structural and evolutionary analysis of non-specific lipid transfer proteins (nsLTPs) in Cereus (Cactaceae)

Structural and evolutionary analysis of non-specific lipid transfer proteins (nsLTPs) in Cereus (Cactaceae) | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Structural and evolutionary analysis of non-specific lipid transfer proteins (nsLTPs) in Cereus (Cactaceae) Maria Izadora Oliveira Cardoso, João Alfredo Teodoro, Tales Alexandre Costa-Silva, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9269266/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The genus Cereus (Cactaceae), widely distributed across arid and semi-arid environments of South America, exhibits adaptive capacity that is partly associated with proteins involved in lipid metabolism and stress responses, such as non-specific lipid transfer proteins (nsLTPs). Although well-characterized in model plants, the structural diversity and functional roles of these proteins in cacti remain largely unexplored. In this study, we characterized the nsLTP repertoire of Cereus by integrating transcriptomic data, phylogenetic analyses, structural modeling, and ligand-binding site mapping. Structural comparisons revealed conservation of the typical nsLTP fold, with variations mainly localized to loop regions surrounding the binding cavity, suggesting functional plasticity without compromising scaffold stability. Phylogenetic analyses revealed frequent duplication and loss events, as well as species-specific paralogs. Several lineage-specific copies exhibited distinctive loop architectures and pocket geometries relative to their closest orthologs, suggesting rapid structural diversification following duplication. This pattern is consistent with functional fine-tuning after gene duplication. Positive selection analyses identified 18 codons under episodic diversifying selection, many in spatial proximity to predicted functional regions, supporting the hypothesis that selective pressures have shaped key interaction interfaces. Together, these results provide the first integrated characterization of nsLTPs in Cereus , supporting a gate-modulation mechanistic model in which loop-level variation reshapes tunnel access and local physicochemical compatibility while preserving the conserved four-helix scaffold. Rather than asserting adaptation directly, our findings provide a structural basis and testable predictions for how nsLTP diversification may influence lipid trafficking processes relevant to cuticle/membrane function under abiotic stress. Cereus gene duplication lipid transfer proteins protein structure stress response Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1. Introduction The genus Cereus Mill. (Cactaceae, tribe Cereeae) represents a diverse and widely distributed group within the cactus family (Franco et al., 2017). Comprising approximately 31 recognized species, Cereus occurs across a variety of South American biomes, ranging from coastal regions to Andean habitats above 3,000 meters in altitude (Hunt et al., 2006). This broad distribution reflects a remarkable adaptive capacity to contrasting environmental conditions, including drought and temperature fluctuations typical of xeric and semi-arid ecosystems (Amaral et al., 2021; 2025a). The phylogenetic framework proposed for Cereus was later refined by recent phylogenetic analyses employing nuclear orthologous genes and a coalescent-based approach (Bombonato et al., 2020; Taylor et al., 2023). These studies support the genus as monophyletic and, although previous analyses structured it into major clades (Bombonato et al., 2020), the new phylogeny, which tested the preexisting subgeneric classification from 1992, partially recovered three of the four subgenera. The adaptability of Cereus to xeric and semi-arid ecosystems, an adaptation observed at anatomical, physiological, and biochemical levels, is reflected in the production of a wealth of secondary metabolites. In traditional medicine, the cladode, root, and seeds of Cereus jamacaru D.C. are used to treat various ailments, including urinary tract infections, kidney inflammation, syphilis, gastritis, and cardiovascular and respiratory disorders (Andrade et al., 2006; Lucena et al., 2013; Palheta et al., 2017; da Silva et al., 2019; Rodrigues Almeida and Gonzaga Fernandez, 2025; Teodoro et al., 2025; Cardoso et al., 2026). The phytochemical composition of C. jamacaru cladodes reveals the presence of phenethylamine alkaloids, such as hordenine, tyramine, and N-methyltyramine, as well as tyrosine. Notably, tyramine exhibits sympathomimetic activity and a probable cardiotonic effect. Other bioactive compounds identified include flavonoids and tannins, which confer crucial therapeutic properties, including anti-inflammatory, antifungal, antioxidant, and wound-healing activities (Rodrigues Almeida and Gonzaga Fernandez, 2025). Moreover, exploring Cereus at the molecular level, particularly through the analysis of its proteins, may uncover key mechanisms underlying the evolutionary history and adaptive radiation of the genus. Among these proteins, non-specific lipid transfer proteins (nsLTPs) are of particular interest due to their structural diversity and functional versatility across plant lineages (Missaoui et al., 2022). These are small (around 9 kDa) and basic proteins broadly distributed across higher plants, typically composed of four to five \(\:\alpha\:\) -helices stabilized by a conserved motif of eight cysteine residues forming four disulfide bonds (D’Agostino et al., 2018). This compact fold creates a hydrophobic tunnel-like cavity that allows nsLTP to bind and transport a wide range of lipid molecules; structural features that not only stabilize under stress conditions but also enable participation in multiple biological roles, including membrane remodelling, cell wall organization, and signaling pathways (Egger et al., 2010; Missaoui et al., 2022). Further investigation is needed to clarify their physiological roles, particularly in non-model taxa like Cereus . Although recent phylogenomic studies have refined the evolutionary framework of Cereus , the molecular mechanisms underlying its remarkable ecological adaptability and biochemical versatility remain poorly understood. In particular, the processes related to lipid interaction and transport, which are central to stress adaptation and membrane homeostasis in xerophytic plants, have not been investigated in this genus. While nsLTPs are well characterized in model plants, their diversity, structural features, and functional roles in non-model taxa such as Cereus and Cipocereus are virtually unknown. This study aims to characterize the nsLTP repertoire in Cereus by integrating comparative transcriptomics, structural modeling, pocket/hot-spot mapping, and evolutionary inference. We test the central mechanistic hypothesis that Cereus nsLTPs exhibit a conserved structural scaffold coupled to evolutionarily labile ‘gate’ regions (loop rims), such that variation at the tunnel entrance can modulate ligand access and local compatibility with lipid-like molecules without disrupting fold stability. This framework yields explicit, testable predictions (e.g., species- and paralog-specific differences in tunnel geometry and surface chemistry) that can be evaluated in future functional assays, thereby bridging molecular structure to physiological processes relevant to stress-associated lipid trafficking. 2. Materials and Methods 2.1. Data collection and transcriptome assembly Transcriptomic datasets from Cereus species were obtained from previous studies of our group (Amaral et al., 2025a; 2025b). Redundant contigs were clustered with CD-HIT v4.8.1 (Li and Godzik, 2006) at a 90% identity threshold. Protein-coding sequences were inferred and used for orthology assignment in OrthoFinder v2.4.0 (Emms and Kelly, 2019), generating the orthogroup matrix employed in the present analyses. The orthogroups were annotated against the nsLTP identified for C. jamacaru (Cardoso et al., 2026). The isolated orthogroups with the nsLTP function were used for downstream analysis. 2.2. Sequence collection and structural modeling Structural predictions of Cereus nsLTP homologs were generated with Modeller v.10.6 (Webb and Salli, 2016) and AlphaFold3 (Jumper et al., 2023) using default settings in both cases. The standard AlphaFold confidence output pLDDT (per-residue confidence) was recorded to guide downstream interpretation. Although these approaches yield static structural models, they are known to capture evolutionarily conserved and functionally relevant conformational states, particularly for small, disulfide-stabilized proteins such as nsLTPs. For cross-target comparability, we used the full polypeptide as predicted; no manual trimming or domain excision was applied before alignment. For structural comparisons, we designated nsLTP from C. jamacaru as the reference scaffold. The choice of a single reference ensures consistent interpretation of TM-scores and RMSDs across the dataset while avoiding ambiguity introduced by all-vs-all normalization. We performed structure-only alignments using TM-align. For each mobile model vs the reference, TM-align was executed with the -o flag to write superposed coordinates. TM-align returns three key metrics that we recorded for every pair: (i) TM-score (dimensionless, 0–1), (ii) RMSD (Å) over the aligned Cα subset; and (iii) aligned length. We accepted alignments when TM-align successfully returned all three metrics and produced a superposed PDB, and then, they were inspected in PyMOL. Additionally, structural stereochemical quality was evaluated with PROCHECK (Laskowski et al., 1993), using the standard Ramachandran plot assessment and geometry checks to ensure that modeled structures met expected quality thresholds. 2.3. Mapping and Characterization of Binding Regions Identification of potential ligand-binding sites was performed using FTSite (Ngan et al., 2012; https://ftsite.bu.edu/ ) and FTMap (Jones et al., 2022; Kozakov et al., 2015; Ngan et al., 2012; Brenke et al., 2009; https://ftmap.bu.edu/ ) webservers. We used FTSite and FTMap to identify ligand-binding regions on protein surfaces. Both methods probe the surface with small organic fragments and define consensus binding sites where multiple probe clusters overlap. FTMap further characterizes these clusters as energetically favorable “hot spots.” We used the default FTMap probe set and server settings (see Table S1 ). Hot-spots identified by FTMap were ranked by cluster strength and mapped onto predicted cavities. Across species and paralogs, high-density probe clusters systematically co-localized with large, high-ligandability pockets, indicating concordance between independent surface-mapping and cavity-detection approaches. This convergence supports the robustness of predicted binding regions. The FPocket web server (Kochnev and Durrant, 2022; https://durrantlab.pitt.edu/fpocketweb/ ) was employed to characterize the proteins further using default parameters (minimum alpha-spheres = 35; minimum cavity volume = 50 ų), and pockets were ranked by their ligandability score, which was subsequently interpreted in the context of compatibility with lipid-like ligands relevant to plant physiology. Only pockets with ligandability score > 0.05 or volume > 200 ų were retained for comparative discussion. This tool is designed with three main objectives: (i) the identification of pockets to delimit cavities on the protein surface with potential to bind small compounds; (ii) the classification of these pockets according to their likelihood of accommodating ligand-like molecules; and (iii) consideration of conformational adaptability, known as induced fit, whereby the pocket geometry may undergo rearrangements during complex formation with a ligand. 2.4. Phylogenetic reconstruction and evolutionary analysis Protein sequences from Cereus spp. were aligned using MAFFT v7.520 (Katoh and Standley, 2013) with the L-INS-i strategy, which provides high accuracy for sequences with conserved motifs and variable loop regions. The resulting alignment was visually inspected in Bioedit 5.0.9 (Hall et al., 2011) to remove ambiguously aligned positions and terminal gaps. Phylogenetic inference using the aminoacids sequences aligned was performed with IQ-TREE2 v2.3.5 (Minh et al., 2020), using the best-fitting substitution model automatically selected by the software. Branch support was assessed with 1,000 ultrafast bootstrap replicates and 1,000 SH-aLRT replicates. Gene duplication events were inferred by reconciling the inferred gene tree with a species tree using Notung v2.9 (Stolzer et al., 2012), allowing identification and visualization of duplication and loss events within the phylogeny. 2.5. Individual Sites Subject to Episodic Diversifying Selection The coding sequences (CDS) of nsLTPs were retrieved from the published genome of Cereus (Amaral et al., 2025a). To detect signatures of adaptive evolution, the Mixed Effects Model of Evolution (MEME) was applied through the Datamonkey web server (Weaver et al. , 2018; https://www.datamonkey.org ) , which implements the MEME (Murrell et al , 2012) method of the HyPhy software package. MEME estimates the ratio of nonsynonymous to synonymous substitution rates (ω = dN/dS), allowing ω to vary both across sites (fixed effects) and across branches at a given site (random effects). This approach enables the identification of episodic positive selection, meaning adaptive events that occur only on a subset of branches at specific codon positions. Because such variation may reflect adaptive, compensatory, or selectively neutral processes, MEME results were interpreted in conjunction with structural context. The analysis was conducted using a significance threshold of p < 0.05. Sites showing evidence of episodic diversifying selection were identified based on the likelihood ratio test (LRT) and the empirical Bayes posterior probabilities reported by Datamonkey. 3. Results 3.1. Structural assessment of nsLTP across Cereus species TM-scores consistently fell within the range typically interpreted as “same topology” (≈ 0.5–0.7), with aligned cores of ~ 60–80 residues and backbone RMSDs around 2–3 Å. Importantly, models displaying higher loop divergence also tended to exhibit increased pocket-volume variability and altered rim polarity (Table S1 ), indicating a coupling between peripheral structural variation and tunnel geometry. This pattern supports a conserved-core/variable-rim organization. In other words, the Cereus protein is not a new fold, it is a moderately diverged representative of the family (Yeats and Rose, 2008; Edstam et al., 2011). The fraction of residues that fail to align one-to-one mainly reflects peripheral segments, rather than wholesale rearrangements of the structural core. It is also important to note that structural comparisons were performed relative to a single reference structure. While this standardizes TM-scores, it may underestimate structural divergence among non-reference homologs. Nevertheless, the conserved fold-level architecture observed across all alignments suggests that reference-based comparisons are adequate for capturing the shared nsLTP topology. When the superposed models are inspected, the largest deviations concentrate in solvent-exposed loops and turns that rim the putative binding groove, while the inner β/α framework that shapes the pocket is comparatively conserved. This pattern indicates that static models consistently resolve a structurally constrained core that defines tunnel geometry, while allowing peripheral regions to vary in ways that are functionally meaningful for ligand access. In practical terms, the Cereus nsLTP likely preserves the overall pocket architecture, but varies the local geometry and side-chain presentation at the entrance (loop lengths, insertions/deletions, and substitutions; Table S1 ). Such changes are precisely the ones that modulate ligand preference and kinetic fine-tuning without compromising fold stability (Kader, 1996; Yeats and Rose, 2008). The structural signal points to evolutionary conservation of the scaffold with meaningful variation at the binding interface. The Cereus homolog is therefore best described as conformationally compatible but functionally tunable relative to its relatives: it aligns well enough to support homology-based inference, yet it carries loop-level alterations consistent with shifts in specificity. The stereochemical analysis of the nsLTP_modeller and nsLTP_AF3 models using PROCHECK shows that both present good structural quality (Fig. 1 A, B), with most residues located in favored regions of the Ramachandran plot (93.3% for nsLTP_modeller and 92.3% for nsLTP_AF3). However, nsLTP_AF3 (Fig. 1 A) shows no bad contacts and slightly smaller deviations in bond lengths and angles. This suggests a higher predicted geometric stability, although these differences must be interpreted cautiously, as static predictions do not capture conformational flexibility. Both models have only three residues flagged in the Ramachandran plots and satisfactory side-chain parameters, but nsLTP_AF3 exhibits positive overall G-factors (0.23), suggesting a more reliable global conformation. Thus, while both models are suitable for structural and docking analyses, nsLTP_AF3 presents marginally better geometric scores. Still, these differences are within the expected range for in silico models, and do not guarantee functional superiority without experimental validation. 3.2. Identification of Pockets and Functional Regions 3.2.1. Identification and Characterization of Binding Sites Given the predictive nature of structural modelling and the potential conformational variations between different methods, we evaluated the protein using two distinct topologies: one generated by Modeller v.10.6 (nsLTP_Modeller) and the other by AlphaFold (nsLTP_AF3) (Fig. 2 A,B). This approach not only allows for a more robust identification of binding sites and hot spots but also enables the assessment of the consistency and reliability of the detected regions, providing a solid basis for subsequent analysis. The structures of nsLTP_modeller and nsLTP_AF3 were submitted to the online servers FTSite and FTMap to locate potential binding sites and determine their main characteristics. Additionally, both structures were analyzed using FPocket to identify and quantify pockets based on geometric and physicochemical properties, providing complementary information on pocket volume, depth, hydrophobicity, and ligandability. Hydrogen bond interactions for individual residues were further assessed using FTMAP, and the results are provided in Supplementary Figure S1 , offering detailed insight into residue participation within the predicted binding sites. For nsLTP_Modeller (Fig. 2 A), three distinct hydrophobic cavities were identified, each forming a single cluster of overlapping probes (Table 1 ), suggesting well-defined binding regions with limited conformational flexibility. The main interacting residues are predominantly hydrophobic (e.g., Ile, Leu, Val, Ala, and Phe), consistent with the protein’s lipid-binding nature (Kader, 1996; Yeats and Rose, 2008). The hydrophobic side chains are located predominantly in the interior of a protein and this arrangement stabilizes the folded polypeptide backbone, since unfolding it or extending it would expose the hydrophobic side chains to the solvent (Gowder et al., 2014). Table 1 Comparative Summary of Binding Sites and FTMap Probe Clusters in nsLTP_modeller and nsLTP_AF3. Binding site Residues involved FTMap probe clusters Clusters Probe nsLTP_Modeller Structure Site #1 Ile61, Leu64, Ser65, Ala68, Arg74, Val77, Val105, Ile107 Single cluster BEN, BUT, EOL, THS, BDY, AMN, ACD, ADY, DFO, ETH, ACT, DME Site #2 Leu38, Ile61, Leu81, Ala96, Gly97, Met99, Pro100, Ile111 Single cluster ACN, EOL, ETH, THS, ACT, ADY, DFO, DME, AMN, ACD, URE Site #3 Met19, Ala20, Leu21, Ala22, Thr70, Ile71, Ala72, Asp73 Single cluster BDY, CHX, BUT, BEN, PHN nsLTP_AF3 Structure Site #1 Pro63, Tyr64, Lys65, Thr70, Asp71, Cys72, Lys74, Val75, Gln76, Lys77 Cluster 1 CHX, ACD, ACT, BDY, BEN, BUT, ETH, ACN, DME, URE, ADY, DFO, EOL, PHN, THS, AMN Cluster 2 BUT, BDY, CHX, BEN, ETH, THS, ACT, ADY, DFO, DME, PHN Cluster 3 BUT, ACT, BDY Site #2 Ile28, Thr32, Val33, Ser36, Leu37, Asp58, Gly59, Ile60, Asn61, Ile62, Ser78, Ser80, Ser81, Thr82, Ala83, Val92, Leu96 Cluster 1 ACD, ACN, ACT, ADY, AMN, BEN, DFO, DME, EOL, ETH, PHN, THS, URE Cluster 2 URE, DFO, PHN, ACN Site #3 Pro63, Tyr64, Lys65, Gly112, Ser113, Met114, Pro115, Thr116, Val120 Single cluster BDY, CHX, ACD, ACT, ADY, BEN, DFO, ETH, PHN, THS, DME, BUT, EOL, URE, ACN In contrast, nsLTP_AF3 (Fig. 2 B) exhibited three corresponding binding regions, but with a greater number of clusters and higher probe diversity, especially in site #1 (Table 1 ), where sixteen probes were mapped across three overlapping clusters. This indicates a more accessible and flexible binding environment, potentially enabling the accommodation of ligands with different physicochemical properties. Sites #2 and #3 in nsLTP_AF3 (Fig. 2 B) also showed expanded residue participation and overlap among probes, reinforcing the idea of a structurally dynamic pocket architecture. Overall, the comparative analysis highlights that while both models preserve the characteristic hydrophobic tunnel of nsLTPs, the AlphaFold predicted structure displays increased binding plasticity and chemical diversity, which may reflect a more realistic representation of the protein’s functional conformational ensemble. 3.2.2. Structural Pocket Analysis Pocket analysis of the nsLTP_Modeller model revealed a total of nine potential binding sites with distinct physicochemical properties (Fig. 3 A, B). Most pockets exhibited low ligandability scores (< 0.01), except for pocket #9, the largest site, combined a significant apolar contribution with substantial polar SASA, a high number of alpha spheres (263), and the largest volume (1865.5 ų). These features resulted in a high pocket score (0.756), which is interpreted here as a measure of geometric and physicochemical ligandability for lipid-like molecules. This cavity therefore represents a structurally accessible region compatible with physiologically relevant hydrophobic and amphipathic ligands, consistent with the known lipid-binding function of nsLTPs. These results highlight pocket #9 as the most promising target for docking studies, whereas the other sites might serve secondary or more specialized binding roles. When compared to the nsLTP_modeller structure, the nsLTP_AF3 model exhibited fewer binding pockets but with improved physicochemical properties. Pocket analysis of the nsLTP_AF3 model identified five potential binding cavities (Fig. 3 A) with varying physicochemical properties. Among these, pocket #1 stood out with the highest pocket score (0.368) and the largest volume (452.3 ų), indicating enhanced geometric suitability for lipid-like ligand accommodation, and suggesting a versatile and accessible region for physiologically relevant hydrophobic interactions. Pockets #2 and #5 were predominantly hydrophobic (apolar SASA > 84%), suggesting potential affinity for lipid or hydrocarbon ligands, whereas pockets #3 and #4 exhibited higher polarity and moderate volumes (259–391 ų), possibly accommodating amphipathic molecules. Overall, nsLTP_AF3 displayed a smaller number of binding sites compared to nsLTP_modeller but with generally higher hydrophobicity and ligandability, particularly in pocket #1, which appears to represent the principal ligand-binding region of this model. The main cavity of nsLTP_AF3 (pocket #1) presented a higher ligandability index, indicating a geometrically favorable and accessible region for ligand interaction. In contrast, the nsLTP_Modeller pockets were generally smaller, less ligandable, and displayed higher polarity, suggesting weaker affinity toward hydrophobic ligands. The reduction in the number of predicted pockets in nsLTP_AF3, together with the larger primary site volume, suggests a more consolidated binding region within this model. However, such differences may also arise from model-dependent structural rearrangements, especially in loops, and should be interpreted cautiously. Previous work by our group (Cardoso et al., 2026) has demonstrated the feasibility of combining large-scale molecular docking with molecular dynamics simulations using triacylglycerols and related lipid substrates in Cereus proteins. These analyses highlighted both the potential and the methodological challenges associated with modeling long-chain and highly flexible lipid molecules in dynamic environments. 3.3. Phylogeny and Evolutionary Context of the Gene/Protein The phylogeny constructed for the genus shows clear diversification of the studied proteins, with several clades well supported by bootstrap values. Many nodes exhibit high support (≥ 95), indicating that certain groupings are highly reliable, as observed in subclades including sequences from Cipocereus , C. pierrebraunianus , and C. mirabella (Fig. 4 ). Some branches, however, display intermediate or low bootstrap values (~ 36), suggesting lower confidence in these groupings, possibly due to rapid divergence or less conserved regions of the proteins. Interestingly, species-specific duplications are evident, with multiple copies of sequences within the same species forming distinct subclades, suggesting paralogous individuals and potential functional diversification. Additionally, some sequences form clades that span multiple species, indicating ancestral duplications predating the divergence of species within the genus. Phylogenetic analyses of nsLTP proteins in cactus species, such as C. jamacaru , reveal multiple gene duplication events (Fig. 4 ). However, many of these duplications are non-persistent, meaning they were not maintained throughout the evolution of different lineages. This pattern indicates that some duplicated gene copies were lost in descendant species, suggesting that not all provided sufficient adaptive advantage to be fixed. Thus, nsLTPs in cacti display an evolutionary dynamic characterized by frequent duplication and loss events, a pattern commonly observed in gene families associated with defense processes and responses to environmental stress (Edstam et al., 2011). This behavior reinforces the notion that only some recent duplications were functionally relevant, possibly linked to adaptation to arid conditions and the formation of a protective cuticle. Species-specific paralogs were also detected. In particular, some nsLTP copies were found exclusively in C. calcirupicola or C. pierrebraunianus , pointing to recent duplications or species-specific retention. These unique paralogs are potential candidates for neofunctionalization, possibly related to ecological adaptation. Finally, cases of single-copy retention in nsLTPs were observed, where certain species maintained only one exclusive representative. 3.4. Identification of Coding Sites under Positive Selection Across the Phylogeny The analysis using MEME (Mixed Effects Model of Evolution) on Datamonkey was conducted with 39 aligned sequences, totaling 133 codon sites. The model employed approximately 69 branches per partition in the test and 100 bootstrap replications, providing statistical robustness to the inference. From the 133 sites evaluated, 18 showed evidence of episodic positive selection (Fig. 5B), corresponding to about 13.5% of the total. On average, each codon under positive selection was detected in 1 to 2 branches of the phylogenetic tree (Table S2), suggesting that selective pressure does not act uniformly across the entire gene but rather locally in specific clades or lineages. This pattern is consistent with episodic functional fine-tuning, although alternative explanations such as compensatory evolution or lineage-specific constraint relaxation cannot be excluded. This result indicates the occurrence of lineage-specific adaptations throughout the evolution of the studied group, possibly related to functional changes, ecological specialization, or diversification in interactions with partner molecules (substrates, ligands, or other proteins) (Fig. 5C). The fact that no sites were identified with variable ω (dN/dS) across all branches reinforces (Table S2) the idea that the observed positive selection is episodic and restricted rather than pervasive adaptive pressure. From a functional perspective, these 18 sites represent relevant candidates for further investigation. Several codons under episodic rate variation mapped to loop rims and pocket-adjacent surfaces. Relative to randomly sampled residues, these sites showed a higher-than-expected spatial proximity to tunnel entrance regions, indicating non-random localization of evolutionary rate variation within functionally sensitive structural contexts. Their localization supports a potential role in modulating local conformational or interaction properties, but does not, by itself, demonstrate adaptive functional divergence. This structural context strengthens the mechanistic plausibility of these substitutions. The positive selection analysis using the MEME model identified 18 codons (2, 6, 9, 12, 20, 34, 51, 53, 56, 65, 88, 100, 110, 112, 116, 131, 132, and 133) under episodic diversifying selection throughout the phylogeny (Table S2). These sites showed p-values ≤ 0.05 (Fig. 5A, B, C), indicating that they were targets of adaptive events in at least one evolutionary branch of the tested tree (Murrell et al., 2012). Among the 133 analyzed codons, approximately 13.5% exhibited signs of positive selection. On average, each selected codon showed support in 1 to 3 branches, suggesting that selective pressure occurred in a localized manner, reflecting specific adaptations to certain evolutionary contexts rather than a constant and widespread pressure across the entire lineage. The distribution of sites under selection was not concentrated in a single region of the gene but occurred relatively dispersed along the sequence. This pattern may be associated with the maintenance of multiple molecular functions, modulation of protein–protein interactions, or the need to respond to diverse environmental or functional pressures. 4. Discussion Our analyses support a gate-modulation mechanistic model for Cereus nsLTPs, in which a conserved four-helix scaffold maintains the hydrophobic tunnel, while loop-level diversification at rim/entrance regions tunes tunnel accessibility and local physicochemical properties. By integrating structural comparisons, pocket/hot-spot mapping, phylogenetic context, and episodic selection signals, we show that functional tunability is concentrated in peripheral, flexible regions rather than in the structural core. This provides a mechanistic bridge between molecular variation and physiological roles commonly attributed to nsLTPs (e.g., lipid trafficking linked to cuticle/membrane function under stress), without implying direct ecological causality in the absence of functional assays. 4.1. Structure modelling quality and characterization for Cereus nsLTP The comparative analysis of nsLTP_Modeller and nsLTP_AF3 highlights significant differences in pocket organization and binding plasticity, offering insights into the molecular determinants of ligand recognition in nsLTPs. Additionally, this comparison provides a means to assess the extent to which predictions from distinct modeling approaches capture the protein’s true structural features, facilitating evaluation of binding site stability and the reliability of identified functional regions (Krokidis et al., 2025). Furthermore, understanding the variations in pocket accessibility and flexibility between models can inform the selection of suitable binding sites for docking studies, ligand design, and functional characterization, ultimately enhancing the predictive power and biological relevance of computational analyses. Throughout this study, pocket and cavity scores are used as structural descriptors of ligandability for lipid-like molecules involved in plant physiology. In nsLTP_modeller, the main binding residues (Table 1 ) are predominantly hydrophobic or small polar side chains, consistent with the classical description of the hydrophobic cavity found in plant nsLTPs. These residues correspond to positions typically forming the central cleft between helices H1–H4 and the C-terminal tail, stabilized by disulfide bridges in the canonical 8CM motif. The single-cluster organization observed in all Modeller sites indicates a more restricted cavity topology, with limited conformational rearrangement and lower accessibility for diverse ligands (Salminen et al., 2016). Conversely, nsLTP_AF3 displayed three binding sites with a more complex cluster distribution (Table 1 ), particularly in Site #1 (Fig. 1 B), which contained three overlapping FTMap probe clusters. This pattern suggests an expanded or more solvent-exposed cavity, consistent with a higher dynamic range of side-chain orientations. The presence of polar residues such as Lys65, Asp71, and Gln76, adjacent to hydrophobic residues (Val75, Pro63), points to a mixed environment that may facilitate interaction with amphipathic ligands or transient solvent molecules. As reported for TaLTP1.1 (Simorre et al., 1991; Gincel et al., 1994; Salminen et al., 2016; PDB ID: 1GH1), the hydrophobic core is formed by residues distributed across multiple helices and small conformational changes can significantly alter cavity volume and connectivity. Therefore, the AlphaFold3 structure may represent one plausible accessible conformation of the same fold, potentially capturing states that differ from the Modeller-derived model. However, given that loop conformations are intrinsically flexible and AlphaFold confidence decreases in these regions, these differences should be interpreted as model-dependent rather than definitive. The functional implications of these structural differences are particularly relevant in light of the multifunctional nature of nsLTPs. These proteins are known to participate in diverse physiological processes, from cuticle formation and pathogen defense to lipid signaling and membrane remodeling (Shenkarev et al., 2017; Madni et al., 2020). However, it remains unclear whether such functional diversity arises from the existence of multiple isoforms or from the intrinsic ability of nsLTPs to accommodate a broad range of lipid molecules. Many of the positively selected codons mapped by MEME fall within or adjacent to flexible loop regions, which typically define the entrances, rims, or dynamic gating elements of lipid-binding pockets in nsLTPs (Fig. 5B, C). This spatial correspondence is reinforced by the overlap between selected residues and cavities identified by FPocket/FTMap, particularly at the edges of predicted binding pockets and putative interface regions. Such a pattern suggests that episodic selection has likely modulated the physicochemical properties of these structural rims, altering local charge, hydrophobicity, or conformational flexibility, to fine-tune ligand affinity or specificity across different Cereus lineages. These adjustments may reflect ecological divergence in lipid composition or cuticular properties, but this remains a working hypothesis requiring empirical validation. However, because MEME does not distinguish between ecological, structural, or neutral drivers of substitution, the connection to ligand specificity must be treated as a working hypothesis rather than a demonstrated mechanism. 4.2. Structural comparison Comparative structural analyses indicate that the nsLTP homolog preserves the same global architecture observed across the homologous set (Fleury et al., 2019; Fig. 3 B). The combination of high topological agreement with modest geometric spread supports the view that the Cereus protein is not a new fold among them, but rather a moderately diverged member of the family (Santos-Silva et al., 2023). The residues that fail to align one-to-one are largely confined to peripheral segments, whereas the inner β/α framework that defines the pocket is retained. Despite this fold-level conservation, the largest deviations cluster in solvent-exposed loops and turns that rim the binding groove (Malinina et al., 2017). These insertions/deletions and side-chain substitutions reshape the local geometry and electrostatics at the pocket entrance without disrupting the core cavity. Functionally, such “rim remodeling” is a well-known route to tune specificity and affinity while preserving catalytic or binding competence (Hamaï and Drin, 2024). Thus, the most parsimonious interpretation is binding-site plasticity superimposed on a conserved scaffold (Fig. 3 B). The Cereus nsLTP likely recognizes similar chemical motifs as its relatives but with altered rank order of preferences and distinct kinetic parameters (e.g., shifts in K D , K M , or K cat depending on the biochemical role of nsLTP). Across Cereus species, this architectural pattern helps reconcile conservation and divergence. A conserved fold constrains the baseline function (the class of ligands/substrates accommodated and the overall reaction/interaction mechanism), whereas loop-level variability affords ecologically meaningful differentiation (Hamaï and Drin, 2024; Melnikova et al., 2022). In practice, different Cereus lineages may exploit the same scaffold to optimize binding under distinct microenvironmental conditions (pH, ionic strength, cofactor availability), to adjust on/off rates for related ligand families, or to gate interactions with species-specific partners via motif gain/loss at flexible rims. We therefore anticipate conserved pocket cores across species, accompanied by species-specific loop chemistries that modulate recognition and regulation rather than abolish function. These structural signals lead to clear, testable predictions. First, a common ligand panel should bind across homologs, but affinities and preferred poses will vary in ways that track loop differences (Brissos et al., 2024). Docking guided by the aligned cores is expected to reproduce this behavior, with poses diverging at the pocket entrance (Velesinović and Nikolić, 2021). Second, environmental sensitivity should shift between species: pH-dependence, ionic modulation (e.g., Na+/Ca2+/Mg2+), or allosteric responses are likely to differ if loop mutations alter local protonation or cation-π networks (Yuan et al., 2025). Third, post-translational regulation and protein–protein contacts are plausible divergence points, because loop rims are typical carriers of PTM sites and linear interaction motifs (Yuan et al., 2025). Finally, loop-swap mutagenesis (transplanting variable rims between species on the same core) should transfer part of the specificity phenotype, directly linking structural variability at the interface to functional outcomes. Two caveats temper these conclusions. Most observed differences arise in low-confidence, high-flexibility regions by their very nature (loops), and single static models cannot capture their full conformational ensembles (Srinivasan et al., 2024). Nevertheless, the consistent localization of structural variation at tunnel rims across independent modeling approaches indicates that these regions represent constrained functional states rather than modeling artifacts (Hasegawa and Holm, 2009). Static structures therefore provide a reliable framework for identifying evolutionarily stable features and functionally tunable interfaces in nsLTPs. Thus, the Cereus nsLTP is best described as conformationally compatible but functionally tunable relative to its homologs. It aligns well enough to warrant homology-based inference of mechanism, yet carries loop-level alterations consistent with fine-grained shifts in specificity, affinity, and regulation. This scaffold-conserved, rim-variable organization offers a coherent evolutionary route for species-level adaptation without requiring innovation at the fold level. Structural divergence, pocket geometry, surface hot-spot distribution, phylogenetic history, and episodic rate variation converge on a coherent pattern: evolutionary diversification in Cereus nsLTPs is concentrated at tunnel rims and entrance regions, where modest sequence changes are sufficient to modulate ligand compatibility without destabilizing the conserved scaffold. This integrative signal moves the present analysis beyond descriptive cataloguing toward a mechanistic interpretation of structure–function–evolution relationships 4.3. Ecological and Adaptive Significance A possible functional implication of our structural analyses is that loop flexibility and binding-pocket plasticity in Cereus nsLTPs could influence lipid activation under abiotic stress. The conserved β/α framework provides a stable scaffold for maintaining the hydrophobic tunnel, while the more variable loops and rim residues that shape the entrance of the pocket are ideally positioned to modulate ligand access, residence time, and selectivity (Fig. 1 ). Under desiccation, heat, or high irradiance, cuticular lipids and membrane components undergo remodeling, and nsLTPs must accommodate changes in chain length, saturation, and head-group chemistry (Tapia et al., 2013; Xiao et al., 2023). In this context, a structurally core combined with a conformationally permissive rim allows the protein to bind a broader spectrum of hydrophobic ligands without compromising global stability, thus supporting dynamic lipid transport and membrane/cuticle protection in fluctuating xeric environments (Missaoui et al., 2022). This pattern of a conserved scaffold with localized plasticity is consistent with nsLTP adaptations described in other xerophytic or drought-tolerant systems, such as cacti, agaves, and drought-adapted Arabidopsis thaliana ecotypes (Deeken et al., 2016; Rojas et al., 2019), where nsLTPs are frequently associated with cuticle deposition, wax accumulation, and stress-induced lipid trafficking. In those species, nsLTP upregulation under water deficit and heat stress has been linked to reinforced barrier formation and improved maintenance of cell integrity, suggesting that subtle structural adjustments at the binding interface are sufficient to tune ligand preference to stress-related lipids (Xue et al., 2022; Xiao et al., 2023). Our results, combining a canonical hydrophobic tunnel with model-dependent differences in pocket size, hydrophobicity, and probe diversity, corroborate this broader pattern of xerophyte nsLTPs acting as flexible lipid shuttles that buffer membranes and the extracellular matrix against environmental extremes. The binding-site plasticity, species-specific paralogs, and codons under episodic rate variation support a scenario in which nsLTP diversification may contribute to functional modulation of lipid-binding properties, particularly through loop-mediated effects on tunnel accessibility. However, these evolutionary signals are interpreted here as indicators of structural and functional fine-tuning rather than direct evidence of ecological adaptation. Experimental validation will be required to determine whether these molecular differences translate into measurable physiological advantages. Lineage- and species-specific nsLTP variants, particularly those with altered loop architectures and re-shaped pockets, may underlie fine-scale differences in lipid composition of the cuticle and membranes, influencing traits such as transpirational control, thermal tolerance, and UV shielding (Tapia et al., 2013; Liu et al., 2022). Such molecular flexibility is consistent with the hypothesis that nsLTP diversification may contribute to the ecological breadth of Cereus , although this connection remains to be tested experimentally. This could be a mechanistic explanation for the ability of Cereus species to occupy a wide biogeographic range, spanning distinct arid and semi-arid habitats, and suggests that nsLTPs are key components of the adaptive toolkit that enabled the genus to persist and diversify under harsh environmental conditions. However, no causal inference can be made without functional assays or comparative physiological data, and experimental analysis will be necessary to confirm it. 4.4. Structural–Functional–Ecological Framework Across Cereus , the β-sheets and α-helical core remain conserved, consistent with the canonical four-helix fold that stabilizes the internal hydrophobic tunnel (Missaoui et al., 2022). This conserved core scaffold ensures functional stability by preserving the protein's ability to maintain a hydrophobic cavity optimized for lipid encapsulation and transfer (Xiao et al., 2023). Superimposed on this stable core, variable loop chemistry, comprising insertions, deletions, and charged or polar residue substitutions, creates the basis for adaptive specificity. Loops at the rim of the binding pocket show sequence variability and are under episodic positive selection in Cereus , a pattern also observed in drought-tolerant Arabidopsis thaliana ecotypes and Opuntia streptacantha , where nsLTPs modulate cuticular lipid deposition and membrane remodeling under desiccation stress (Rojas et al., 2019; Xiao et al., 2023). These flexible loops act as adaptive gates, allowing differential access of lipid ligands that vary in chain length or polarity, thereby fine-tuning transport and protection functions under environmental extremes. The positive selection observed at specific codons (≈ 13% of the total) indicates localized adaptive fine-tuning rather than wholesale innovation. This pattern implies that Cereus nsLTPs evolve through small, targeted amino acid changes that alter binding-site electrostatics or dynamics without compromising the global fold (Fig. 4 B, C; McLaughlin and Tumer, 2025). Such diversification may reflect selective responses to environmental gradients in humidity, temperature, and UV exposure across the genus’ wide geographic range. Thus, these findings show that nsLTPs exemplify a principle of molecular adaptation through constrained innovation, where they maintain structural conservation of the core for stability and folding efficiency, while exploiting loop-level plasticity and selective diversification to achieve ecological specialization. 4.5. Biotechnological potential The identification of pocket #1 in the nsLTP_AF3 model as a high-ligandability cavity for lipid-like molecules highlights the functional versatility of nsLTPs. Here, pocket scores are interpreted as descriptors of structural compatibility with physiologically relevant hydrophobic ligand. This consolidated cavity reflects a structurally optimized region for lipid binding and transport, consistent with the primary biological roles of plant nsLTPs. These findings align with previous crystallographic studies showing that surface binding of myristic acid near the N-terminal cavity induces conformational changes in loop-3, facilitating lipid internalization into the hydrophobic cavity, while binding near the C-terminal end induces only minor structural deviations (Lev, 2010; Madni et al., 2020). These studies also support a model of largely unidirectional lipid transfer, with lipids entering from the N-terminal end, moving through the hydrophobic cavity, and exiting at the C-terminal end, coordinated by loop-3 and C-terminal loop movements. Similar to the unexpected lipolytic activity reported for Cor a 8 in C. avellana , which operates without a classical catalytic triad (Fissore et al., 2025), the presence of a structurally compact and accessible pocket indicates that nsLTPs may accommodate diverse ligands and potentially perform catalytic or transport functions that remain underexplored. While such findings expand the functional landscape of nsLTPs, the present study is focused on structural, evolutionary, and ligand-binding features of Cereus nsLTPs and does not provide experimental or computational evidence for catalytic activity. Therefore, any potential relationship between Cereus nsLTPs and lipolytic functions should be regarded as a working hypothesis, to be addressed in future studies combining biochemical assays and dynamic simulations. Key residues such as Arg44 and Tyr79 in nsLTP1s (Finkina et al., 2016; Melnikova et al., 2018; Missaoui et al., 2022) participate in ligand binding, with additional residues stabilizing interactions with polar heads and aliphatic chains. Regulatory features, including calmodulin-binding regions and phosphorylation sites, may further modulate lipid uptake, linking nsLTP function to stress responses and plant adaptation (Chiu et al., 2020). These structural insights, combined with high stability and compact folding of nsLTPs, make them attractive scaffolds for in silico docking studies, molecular dynamics simulations, and potential synthetic biology or biomaterial applications. Functionally, nsLTPs contribute to both biotic and abiotic stress tolerance. Transgenic plants overexpressing nsLTPs demonstrate increased resistance to fungi, bacteria, and oomycetes, and show enhanced tolerance to drought, salinity, cold, and oxidative stress (McLaughlin et al., 2021; Xu et al., 2018). Their ability to act synergistically with other antimicrobial peptides, mediate cutin and wax deposition, and regulate trichome development underlines their versatile roles in plant defense and adaptation (Meng et al., 2018; Song et al., 2020). Taken together, the structural, mechanistic, and physiological insights into nsLTPs suggest multiple avenues for future research. High-ligandability pockets, particularly in nsLTP_AF3, can be explored for ligand docking and bioactive compound screening. Molecular dynamics simulations could clarify the lipid transfer pathway, while precise gene-editing approaches may reveal the amino acids responsible for ligand specificity or stress-related functions. Such integrated studies could enable the development of nsLTP-based tools for synthetic biology, crop improvement, and novel biotechnological applications, including stress-resilient plants and targeted delivery systems. 5. Conclusion This study provides the first comprehensive structural and evolutionary characterization of nsLTPs in the genus Cereus , revealing how these proteins may contribute to adaptation in arid and semi-arid environments. Our integrated approach combining comparative transcriptomics, homology-based modeling, binding-pocket analysis, and evolutionary inference demonstrates that Cereus nsLTPs preserve the canonical four-helix hydrophobic-tunnel architecture typical of the family, while exhibiting local flexibility at loop regions surrounding the binding pocket. This conserved-core yet adaptive-rim pattern supports functional tunability: loop variation and episodic positive selection act as fine-scale mechanisms to modulate ligand affinity, lipid specificity, and potentially stress-related signaling. Such structural–functional plasticity likely enhances membrane protection and lipid trafficking under desiccation, heat, and UV stress, contributing to the ecological resilience and biogeographical breadth of the genus. Eighteen codon sites under episodic diversifying selection, together with species-specific paralogs, indicate ongoing adaptive refinement rather than radical innovation. These features suggest that nsLTP diversification in Cereus followed a model of constrained innovation, where small structural adjustments at the interface level generate functional diversity without compromising scaffold stability. Moreover, the identification of highly ligandable pockets points to potential biotechnological applications, as nsLTPs may serve as molecular scaffolds for lipid-binding, delivery systems, or bioactive compound transport. Declarations Funding This work was supported by Fundação de Amparao à Pesquisa do Estado de São Paulo (FAPESP/ n. 2024/19266-5 to MIOC, n. 2025/17270-8 to JAT, and n. 2023/05589-4 to DTA) Competing Interests The authors have no relevant financial or non-financial interests to disclose. Author contributions The idea for this study was conceived by MIOC and DTA. Data collection and analyses were performed by MIOC, JAT, TACS, and DTA. MIOC led the writing of the paper, while JAT and DTA contributed numerous conceptions and writing. All authors contributed to the intellectual development of the paper, made multiple revisions, and approved the final draft. Acknowledgements The authors acknowledge Dr. Livia Seno Ferreira Camargo for her valuable scientific input and guidance. Financial support from CNPq and FAPESP is gratefully acknowledged. Declaration of generative AI and AI-assisted technologies in the writing process During the preparation of this work, the author(s) used ChatGPT 5.0 to improve English grammar. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. References Amaral, D.T., Minhós-Yano, I., Oliveira, J.V.M., Romeiro-Brito, M., Bonatelli, I.A.S., Taylor, N.P., Zappi, D.C., Moraes, E.M., Eaton, D.A.R., Franco, F.F., 2021. Tracking the xeric biomes of South America: the spatiotemporal diversification of mandacaru cactus. J. Biogeogr. 48 (12), 3085–3103. https://doi.org/10.1111/jbi.14265. Amaral, D.T., Trier, C., Copetti, D., Franco, F.F., Moraes, E.M., 2025a. Chromosome-level genome assembly of the iconic South American mandacaru cactus (Cereus jamacaru DC, Tribe Cereeae, Cactaceae). Planta 262 (2). https://doi.org/10.1007/s00425-025-04762-6. Amaral, D.T., Oliveira, J.V.M., Moraes, E.M., Zappi, D.C., Taylor, N.P., Franco, F.F., 2025b. The potential distribution of Cereus (Cactaceae) species in scenarios of climate crises. J. Arid Environ. 226. https://doi.org/10.1016/j.jaridenv.2024.105285. Andrade, C.T. da S., Marques, J.G.W., Zappi, D.C., 2006. The use of Cactaceae by sertanejos in Bahia, Brazil: connexive types to define utilitarian categories. Sitientibus Sér. Ciênc. Biol. 6 (Especial), 3–12. https://doi.org/10.13102/scb8142. Bombonato, J.R., do Amaral, D.T., Silva, G.A.R., Khan, G., Moraes, E.M., da Silva Andrade, S.C., Eaton, D.A.R., Alonso, D.P., Ribolla, P.E.M., Taylor, N., Zappi, D., Franco, F.F., 2020. The potential of genome-wide RAD sequences for resolving rapid radiations: a case study in Cactaceae. Mol. Phylogenet. Evol. 151, 106896. https://doi.org/10.1016/j.ympev.2020.106896. Brenke, R., Kozakov, D., Chuang, G.Y., Beglov, D., Hall, D., Landon, M.R., Mattos, C., Vajda, S., 2009. Fragment-based identification of druggable “hot spots” of proteins using Fourier domain correlation techniques. Bioinformatics 25 (5), 621–627. https://doi.org/10.1093/bioinformatics/btp036. Brissos, V., Borges, P.T., Sancho, F., Lucas, M.F., Frazão, C., Conzuelo, F., Martins, L.O., 2024. Flexible active-site loops fine-tune substrate specificity of hyperthermophilic metallo-oxidases. J. Biol. Inorg. Chem. 29, 339–351. https://doi.org/10.1007/s00775-023-02040-y. Cardoso, M.I.O., Oliveira, J., Bassaneze, V., Amaral, D.T., 2026. Discovery of triacylglycerol-binding proteins in Cereus jamacaru DC. through integrative in silico methods. J. Chem. 2026, 5448761. https://doi.org/10.1155/joch/5448761. Chiu, L.-Y., Chen, I.-H., Hsu, Y.-H., Tsai, C.-H., 2020. The lipid transfer protein 1 from Nicotiana benthamiana assists Bamboo mosaic virus accumulation. Viruses 12, 1361. https://doi.org/10.3390/v12121361. da Silva, N.F., Hanazaki, N., Albuquerque, U.P., Almeida Campos, J.L., Feitosa, I.S., Araújo, E.L., 2019. Local knowledge and conservation priorities of medicinal plants near a protected area in Brazil. Evid.-Based Complement. Altern. Med. 2019, 8275084. https://doi.org/10.1155/2019/8275084. D’Agostino, N., Buonanno, M., Ayoub, J., Barone, A., Monti, S.M., Rigano, M.M., 2019. Identification of non-specific lipid transfer protein gene family members in Solanum lycopersicum and insights into the features of Sola l 3 protein. Sci. Rep. 9, 1–12. https://doi.org/10.1038/s41598-018-38301-z. Deeken, R., Saupe, S., Klinkenberg, J., Riedel, M., Leide, J., Hedrich, R., Mueller, T.D., 2016. The nonspecific lipid transfer protein AtLtpI-4 is involved in suberin formation of Arabidopsis thaliana crown galls. Plant Physiology 172(3), 1911–1927. https://doi.org/10.1104/pp.16.01486. Edstam, M. M., Viitanen, L., Salminen, T. A., Edqvist, J., 2011. Evolutionary history of the non-specific lipid transfer proteins. In Molecular Plant (Vol. 4, Issue 6, pp. 947–964). Oxford University Press. https://doi.org/10.1093/mp/ssr019. Egger, M., Hauser, M., Mari, A., Ferreira, F., Gadermaier, G., 2010. The role of lipid transfer proteins in allergic diseases. Curr. Allergy Asthma Rep. 10 (5), 326–335. https://doi.org/10.1007/s11882-010-0128-9. Emms, D.M., Kelly, S., 2019. OrthoFinder: phylogenetic orthology inference for comparative genomics. Genome Biol. 20, 238. https://doi.org/10.1186/s13059-019-1832-y. Finkina, E.I., Melnikova, D.N., Bogdanov, I.V., Ovchinnikova, T.V., 2016. Lipid transfer proteins as components of the plant innate immune system: Structure, functions, and applications. Acta Naturae 8(2), 47–61. https://doi.org/10.32607/20758251-2016-8-2-47-61. Fissore, A., di Napoli, G., de Sciscio, M.L., Santoro, V., Salladini, E., Marengo, M., Oliaro-Bosso, S., Vanzetti, G., Caratti, A., Dal Piaz, F., Piazza, F., Manzoli, M., Genova, G., Barbiroli, A., Iametti, S., Fraternali, F., Adinolfi, S., 2025. Unravelling a new enzymatic activity of non-specific lipid transfer protein Cor a 8: A novel lipase from hazelnuts. Int. J. Biol. Macromol. 322, 146537. https://doi.org/10.1016/j.ijbiomac.2025.146537. Fleury, C., Gracy, J., Gautier, M.F., Pons, J.L., Dufayard, J.F., Labesse, G., Ruiz, M., de Lamotte, F., 2019. Comprehensive classification of the plant non-specific lipid transfer protein superfamily towards its sequence-structure-function analysis. PeerJ 7, e7504. https://doi.org/10.7717/peerj.7504. Franco, F.F., Jojima, C.L., Perez, M.F., Zappi, D.C., Taylor, N., Moraes, E.M., 2017. The xeric side of the Brazilian Atlantic Forest: The forces shaping phylogeographic structure of cacti. Ecology and Evolution 7(22), 9281–9293. https://doi.org/10.1002/ece3.3458. Gincel, E., Simorre, J.P., Caille, A., Marion, D., Ptak, M., Vovelle, F., 1994. Three-dimensional structure in solution of a wheat lipid-transfer protein from multidimensional ¹H-NMR data: a new folding for lipid carriers. Eur. J. Biochem. 226, 413–422. https://doi.org/10.1111/j.1432-1033.1994.tb20066.x. Gowder, S.M., Chatterjee, J., Chaudhuri, T., Paul, K., 2014. Prediction and analysis of surface hydrophobic residues in tertiary structure of proteins. Sci. World J. 2014, 971258. https://doi.org/10.1155/2014/971258. Hall, T., Biosciences, I., & Carlsbad, C. J. G. B. B. (2011). BioEdit: an important software for molecular biology. GERF Bull Biosci, 2(1), 60-61. Hamaï, A., Drin, G., 2024. Specificity of lipid transfer proteins: An in vitro story. Biochimie 227, 85–110. https://doi.org/10.1016/j.biochi.2024.09.007. Hasegawa, H., Holm, L., 2009. Advances and pitfalls of protein structural alignment. Curr. Opin. Struct. Biol. 19, 341–348. https://doi.org/10.1016/j.sbi.2009.04.003. Hunt, D.R., Taylor, N.P., Charles, G., 2006. The New Cactus Lexicon. dh Books. Jones, G., Jindal, A., Ghani, U., Kotelnikov, S., Egbert, M., Hashemi, N., Vajda, S., Padhorny, D., Kozakov, D., 2022. Elucidation of protein function using computational docking and hotspot analysis by ClusPro and FTMap. Acta Crystallogr. D Struct. Biol. 78 (6), 690–697. https://doi.org/10.1107/S2059798322002741. Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al., 2021. Highly accurate protein structure prediction with AlphaFold. Nature 596 (7873), 583–589. https://doi.org/10.1038/s41586-021-03819-2. Kader, J.-C., 1996. Lipid-transfer proteins in plants. Annual Review of Plant Biology 47, 627–654. https://doi.org/10.1146/annurev.arplant.47.1.627. Katoh, K., Standley, D.M., 2013. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol. Biol. Evol. 30 (4), 772–780. https://doi.org/10.1093/molbev/mst010. Kochnev, Y., Durrant, J.D., 2022. FPocketWeb: protein pocket hunting in a web browser. J. Cheminform. 14, 58. https://doi.org/10.1186/s13321-022-00637-0. Kozakov, D., Grove, L.E., Hall, D.R., Bohnuud, T., Mottarella, S.E., Luo, L., Xia, B., Beglov, D., Vajda, S., 2015. The FTMap family of web servers for determining and characterizing ligand-binding hot spots of proteins. Nat. Protoc. 10 (5), 733–755. https://doi.org/10.1038/nprot.2015.043. Krokidis, M.G., Koumadorakis, D.E., Lazaros, K., Ivantsik, O., Exarchos, T.P., Vrahatis, A.G., Kotsiantis, S., Vlamos, P., 2025. AlphaFold3: An overview of applications and performance insights. Int. J. Mol. Sci. 26, 3671. https://doi.org/10.3390/ijms26083671. Laskowski, R.A., MacArthur, M.W., Moss, D.S., Thornton, J.M., 1993. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26 (2), 283–291. https://doi.org/10.1107/s0021889892009944. Lev, S., 2010. Non-vesicular lipid transport by lipid-transfer proteins and beyond. Nat. Rev. Mol. Cell Biol. 11, 739–750. https://doi.org/10.1038/nrm2971. Li, W., Godzik, A., 2006. Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics 22 (13), 1658–1659. https://doi.org/10.1093/bioinformatics/btl158. Liu, L., Wang, X., Chang, C., 2022. Toward a smart skin: Harnessing cuticle biosynthesis for crop adaptation to drought, salinity, temperature, and ultraviolet stress. Front. Plant Sci. 13, 961829. https://doi.org/10.3389/fpls.2022.961829. Lucena, C.M., Lucena, R.F., Costa, G.M., Carvalho, T.K., Costa, G.G., Alves, R.R., Pereira, D.D., Ribeiro, J.E., Alves, C.A., Quirino, Z.G., Nunes, E.N., 2013. Use and knowledge of Cactaceae in Northeastern Brazil. J. Ethnobiol. Ethnomed. 9 (1), 62. https://doi.org/10.1186/1746-4269-9-62. Malinina, L., Patel, D.J., Brown, R.E., 2017. How α-Helical Motifs Form Functionally Diverse Lipid-Binding Compartments. Annu. Rev. Biochem. 86, 609–636. https://doi.org/10.1146/annurev-biochem-061516-044445. Madni, Z.K., Tripathi, S.K., Salunke, D.M., 2020. Structural insights into the lipid transfer mechanism of a non-specific lipid transfer protein. Plant J. 102, 340–352. https://doi.org/10.1111/tpj.14627. McLaughlin, J.E., Darwish, N.I., Garcia-Sanchez, J., Tyagi, N., Trick, H.N., McCormick, S., Dill-Macky, R., Tumer, N.E., 2021. A lipid transfer protein has antifungal and antioxidant activity and suppresses fusarium head blight disease and DON accumulation in transgenic wheat. Phytopathology 111, 671–683. https://doi.org/10.1094/PHYTO-04-20-0153-R. McLaughlin, J.E., Tumer, N.E., 2025. Roles of non-specific lipid transfer proteins in plant defense: structural and functional perspectives. Front. Fungal Biol. 6, 1640465. https://doi.org/10.3389/ffunb.2025.1640465. Melnikova, D.N., Finkina, E.I., Bogdanov, I.V., et al., 2018. Plant pathogenesis-related proteins binding lipids and other hydrophobic ligands. Russian Journal of Bioorganic Chemistry 44, 586–594. https://doi.org/10.1134/S1068162018060055. Melnikova, D.N., Finkina, E.I., Bogdanov, I.V., Tagaev, A.A., Ovchinnikova, T.V., 2022. Features and possible applications of plant lipid-binding and transfer proteins. Membranes 13, 2. https://doi.org/10.3390/membranes13010002. Meng, C., Yan, Y., Liu, Z., Chen, L., Zhang, Y., Li, X., Wu, L., Zhang, G., Wang, X., Ma, Z., 2018. Systematic analysis of cotton non-specific lipid transfer protein family revealed a special group that is involved in fiber elongation. Frontiers in Plant Science 9, 1285. https://doi.org/10.3389/fpls.2018.01285. Minh, B.Q., Schmidt, H.A., Chernomor, O., Schrempf, D., Woodhams, M.D., von Haeseler, A., Lanfear, R., 2020. IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era. Mol. Biol. Evol. 37 (5), 1530–1534. https://doi.org/10.1093/molbev/msaa015. Missaoui, K., Gonzalez-Klein, Z., Pazos-Castro, D., Hernandez-Ramirez, G., Garrido-Arandia, M., Brini, F., Diaz-Perales, A., Tome-Amat, J., 2022. Plant non-specific lipid transfer proteins: An overview. Plant Physiol. Biochem. 171, 115–127. https://doi.org/10.1016/j.plaphy.2021.12.026. Murrell, B., Wertheim, J.O., Moola, S., Weighill, T., Scheffler, K., Kosakovsky Pond, S.L., 2012. Detecting individual sites subject to episodic diversifying selection. PLoS Genet. 8 (7), e1002764. https://doi.org/10.1371/journal.pgen.1002764. Ngan, C.H., Hall, D.R., Zerbe, B.S., Grove, L.E., Kozakov, D., Vajda, S., 2012. FTSite: high accuracy detection of ligand binding sites on unbound protein structures. Bioinformatics 28 (2), 286–287. https://doi.org/10.1093/bioinformatics/btr651. Palheta, I.C., Tavares-Martins, A.C.C., Lucas, F.C.A., Jardim, M.A.G., 2017. Ethnobotanical study of medicinal plants in urban home gardens in the city of Abaetetuba, Pará state, Brazil. Bol. Latinoam. Caribe Plant. Med. Aromat. 16 (3), 206–262. Rodrigues Almeida, L.M., Gonzaga Fernandez, L., 2025. Ethnomedical uses, biocompounds and biological properties of Cereus jamacaru DC. (Cactaceae): a comprehensive review. Nat. Prod. Res. 39 (4), 961–975. https://doi.org/10.1080/14786419.2024.2330541. Rojas, M., Jimenez-Bremont, F., Villicaña, C., Carreón-Palau, L., Arredondo-Vega, B.O., Gómez-Anduro, G., 2019. Involvement of OpsLTP1 from Opuntia streptacantha in abiotic stress adaptation and lipid metabolism. Funct. Plant Biol. 46 (9), 816–829. https://doi.org/10.1071/FP18280. Salminen, T.A., Blomqvist, K., Edqvist, J., 2016. Lipid transfer proteins: classification, nomenclature, structure, and function. Planta 244, 971–997. https://doi.org/10.1007/s00425-016-2585-4. Santos-Silva, C.A.d., Ferreira-Neto, J.R.C., Amador, V.C., Bezerra-Neto, J.P., Vilela, L.M.B., Binneck, E., Rêgo, M.d.S., da Silva, M.D., Mangueira de Melo, A.L.T., da Silva, R.H., Benko-Iseppon, A.M., 2023. From Gene to Transcript and Peptide: A Deep Overview on Non-Specific Lipid Transfer Proteins (nsLTPs). Antibiotics 12, 939. https://doi.org/10.3390/antibiotics12050939. Shenkarev, Z.O., Melnikova, D.N., Finkina, E.I., Sukhanov, S.V., Boldyrev, I.A., Gizatullina, A.K., Mineev, K.S., Arseniev, A.S., Ovchinnikova, T.V., 2017. Ligand binding properties of the lentil lipid transfer protein: molecular insight into the possible mechanism of lipid uptake. Biochemistry 56, 1785–1796. https://doi.org/10.1021/acs.biochem.6b01079. Simorre, J.P., Caille, A., Marion, D., Marion, D., Ptak, M., 1991. Two- and three-dimensional proton NMR studies of a wheat phospholipid transfer protein: sequential resonance assignments and secondary structure. Biochemistry 30, 11600–11608. https://doi.org/10.1021/bi00113a016. Song, X., Li, E., Song, H., Du, G., Li, S., Zhu, H., Chen, G., Zhao, C., Qiao, L., Wang, J., Yu, S., Sui, J.M., 2020. Genome-wide identification and characterization of nonspecific lipid transfer protein (nsLTP) genes in Arachis duranensis. Genomics 112(6), 4332–4341. https://doi.org/10.1016/j.ygeno.2020.07.034. Srinivasan, S., Di Luca, A., Álvarez, D., John Peter, A.T., Gehin, C., Lone, M.A., Hornemann, T., D'Angelo, G., Vanni, S., 2024. The conformational plasticity of structurally unrelated lipid transport proteins correlates with their mode of action. PLoS Biol. 22, e3002737. https://doi.org/10.1371/journal.pbio.3002737. Stolzer, M., Lai, H., Xu, M., Sathaye, D., Vernot, B., & Durand, D. (2012). Inferring duplications, losses, transfers and incomplete lineage sorting with nonbinary species trees. Bioinformatics (Oxford, England), 28(18), i409–i415. https://doi.org/10.1093/bioinformatics/bts386. Tapia, G., Morales-Quintana, L., Parra, C., Berbel, A., Alcorta, M., 2013. Study of nsLTPs in Lotus japonicus genome reveal a specific epidermal cell member (LjLTP10) regulated by drought stress in aerial organs with a putative role in cutin formation. Plant Mol. Biol. 82, 485–501. https://doi.org/10.1007/s11103-013-0080-x. Teodoro, J.A., Senra, M.V.X., Amaral, D.T., 2025. In silico bioprospecting of the Neotropical plant mandacaru (Cereus) for antimicrobial properties. Probiotics Antimicrob. Proteins. https://doi.org/10.1007/s12602-025-10580-9. Velesinović, A., Nikolić, G., 2021. Protein-protein interaction networks and protein-ligand docking: Contemporary insights and future perspectives. Acta Fac. Med. Naissensis 38, 5–17. https://doi.org/10.5937/AFMNAI38-28322. Weaver, S., Shank, S.D., Spielman, S.J., Li, M., Muse, S.V., Kosakovsky Pond, S.L., 2018. Datamonkey 2.0: a modern web application for characterizing selective and other evolutionary processes. Mol. Biol. Evol. 35 (3), 773–777. https://doi.org/10.1093/molbev/msx335. Webb, B., Sali, A., 2016. Comparative protein structure modeling using MODELLER. Curr. Protoc. Protein Sci. 86, 2.9.1–2.9.37. https://doi.org/10.1002/cpps.20. Xiao, Y., Xiao, C., He, X., Yang, X., Tong, Z., Wang, Z., Sun, Z., Qiu, W., 2023. A novel non-specific lipid transfer protein gene, CmnsLTP6.9, enhanced osmotic and drought tolerance by regulating ROS scavenging and remodeling lipid profiles in Chinese chestnut (Castanea mollissima Blume). Plants 12, 3916. https://doi.org/10.3390/plants12223916. Xu, Y., Zheng, X., Song, Y., Zhu, L., Yu, Z., Gan, L., Zhou, S., Liu, H., Wen, F., Zhu, C., 2018. NtLTP4, a lipid transfer protein that enhances salt and drought stresses tolerance in Nicotiana tabacum. Scientific Reports 8, 8873. https://doi.org/10.1038/s41598-018-27274-8. Xue, Y., Zhang, C., Shan, R., Li, X., Tseke Inkabanga, A., Li, L., Jiang, H., Chai, Y., 2022. Genome-wide identification and expression analysis of nsLTP gene family in rapeseed (Brassica napus) reveals their critical roles in biotic and abiotic stress responses. Int. J. Mol. Sci. 23, 8372. https://doi.org/10.3390/ijms23158372. Yeats, T. H., Rose, J. K. C., 2008. The biochemistry and biology of extracellular plant lipid‐transfer proteins (LTPs). Protein Science , 17 (2), 191–198. https://doi.org/10.1110/ps.073300108. Yuan, R., Zhang, J., Zhou, J., Cong, Q., 2025. Recent progress and future challenges in structure-based protein-protein interaction prediction. Mol. Ther. 33, 2252–2268. https://doi.org/10.1016/j.ymthe.2025.04.003. Additional Declarations No competing interests reported. Supplementary Files JMESMArticlensLTP.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9269266","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":622454963,"identity":"d5dc0e41-11be-4f5c-9a2b-c83dead1bebf","order_by":0,"name":"Maria Izadora Oliveira Cardoso","email":"","orcid":"","institution":"Universidade Federal do ABC (UFABC)","correspondingAuthor":false,"prefix":"","firstName":"Maria","middleName":"Izadora Oliveira","lastName":"Cardoso","suffix":""},{"id":622454964,"identity":"3b69205f-ecf7-4114-a8fc-9771655c5e73","order_by":1,"name":"João Alfredo Teodoro","email":"","orcid":"","institution":"Universidade Federal do ABC (UFABC)","correspondingAuthor":false,"prefix":"","firstName":"João","middleName":"Alfredo","lastName":"Teodoro","suffix":""},{"id":622454965,"identity":"c45ca847-79d0-499c-85bb-f970533fadf8","order_by":2,"name":"Tales Alexandre Costa-Silva","email":"","orcid":"","institution":"Universidade Federal do ABC (UFABC)","correspondingAuthor":false,"prefix":"","firstName":"Tales","middleName":"Alexandre","lastName":"Costa-Silva","suffix":""},{"id":622454966,"identity":"4faeeb3e-a857-4999-8638-78872619476f","order_by":3,"name":"Danilo Trabuco Amaral","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCElEQVRIie2RsWrDMBRFnyk0i6nXZ+z8g8HQUBL6LRIGZ3E9ayhFk7sodO3fVCbgLHK9GrLkA0LRmKlUNimUorodO+ggiQvS4QoJwOH4j0gzDmPyh2UFMz6GCwCcUMi4PZ7MwZd/UOCLsv1dudq1tSQMHhaPbYMn1lExa2sNbEl5tDnYlFCVRBIFGKsyD4XaU+GXGYJaUx7vEpuSyCKRtAJEKK4jr9rTFxPAq7aUY269WNIdz0pwHJRXKkwA731C6T9bcGyRVODQwn9Wwt60EIXhM75lN0JlqTABSbNOq7ixv1hXpFqzVYDBXd2f2O1cmKD1/XL+FFVW5cz3LyBmXk4JDofD4ZjkA7e9YY7NmsU9AAAAAElFTkSuQmCC","orcid":"","institution":"Universidade Federal do ABC (UFABC)","correspondingAuthor":true,"prefix":"","firstName":"Danilo","middleName":"Trabuco","lastName":"Amaral","suffix":""}],"badges":[],"createdAt":"2026-03-30 15:59:35","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9269266/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9269266/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107140927,"identity":"a5e1cac8-da0e-429b-89bb-3d1be5318324","added_by":"auto","created_at":"2026-04-17 08:59:06","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":457774,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eProtein structural quality of AlphaFold (A) and Modeller (B) models. \u003c/strong\u003eRamachandran plots indicate high reliability for both structures, with most residues in favored regions and none in disallowed regions.\u003c/p\u003e","description":"","filename":"Binder11.png","url":"https://assets-eu.researchsquare.com/files/rs-9269266/v1/9a5a7cdd68f353dbaad629f0.png"},{"id":107140958,"identity":"f2f57870-f651-4f67-ba44-be5edb4e12ad","added_by":"auto","created_at":"2026-04-17 08:59:24","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1132333,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBinding sites and probe clusters in nsLTP structures.\u003c/strong\u003e (A) Distribution of hydrogen bonds across residues in the nsLTP_AF3 model. (B) Binding site organization and probe clusters identified by FTMap in the nsLTP_AF3 structure. (C) Distribution of hydrogen bonds across residues in the nsLTP_Modeller model. (D) Binding site organization and probe clusters identified by FTMap in the nsLTP_Modeller structure. Binding regions are highlighted on the protein surface, and clusters are represented by colored probes corresponding to different chemical fragments. The comparison reveals variations in pocket topology and ligand accessibility between the two modeled structures.\u003c/p\u003e","description":"","filename":"Binder12.png","url":"https://assets-eu.researchsquare.com/files/rs-9269266/v1/78819a6b42eebdfa125d8cb6.png"},{"id":107140959,"identity":"17c91ba0-5e79-4050-8d0e-ba7125c2dc05","added_by":"auto","created_at":"2026-04-17 08:59:24","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":390717,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStructural and ligandability features of Modeller and AlphaFold protein models. \u003c/strong\u003e(A) Relationship between binding-site volume (ų) and predicted ligandability scores (log scale) for cavities identified in the Modeller (blue) and AlphaFold (green) models, highlighting differences in pocket size and ligand-binding potential between the two modeling approaches. (B) Structural superposition of the corresponding three-dimensional models, with Modeller shown in blue and AlphaFold in green, illustrating global and local conformational differences between the predicted structures.\u003c/p\u003e","description":"","filename":"Binder13.png","url":"https://assets-eu.researchsquare.com/files/rs-9269266/v1/a726d208c81d88d981d7860f.png"},{"id":107140943,"identity":"c7e145ba-c8e9-4605-95e6-e787b541a5da","added_by":"auto","created_at":"2026-04-17 08:59:15","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":93561,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEvidence of positive selection and structural–functional mapping of the analyzed protein\u003c/strong\u003e\u003cem\u003e. \u003c/em\u003eA) Codon sites under positive selection detected by the MEME (Mixed Effects Model of Evolution) analysis. Each point represents a codon with a statistically significant signal of diversifying selection (p-value). B) Predicted 3D structure generated by AlphaFold, colored according to the pLDDT confidence score (blue = high confidence; green/yellow = lower confidence). Codon positions under positive selection are highlighted as orange spheres. C) Structural overlap between positively selected residues (red spheres) and three predicted functional binding sites: \u003cem\u003eSite 1\u003c/em\u003e (light red), \u003cem\u003eSite 2\u003c/em\u003e(green), and \u003cem\u003eSite 3\u003c/em\u003e (violet). Several positively selected residues are spatially adjacent to functional regions, suggesting potential adaptive relevance for binding or molecular interaction interfaces.\u003c/p\u003e","description":"","filename":"Binder14.png","url":"https://assets-eu.researchsquare.com/files/rs-9269266/v1/1afe86bc9968180619c9f4e5.png"},{"id":107140954,"identity":"c0ff57cb-1ae4-4c09-b487-e571ab5ec30e","added_by":"auto","created_at":"2026-04-17 08:59:22","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":475845,"visible":true,"origin":"","legend":"\u003cp\u003e\u0026nbsp;Legend not included with this version.\u003c/p\u003e","description":"","filename":"Binder15.png","url":"https://assets-eu.researchsquare.com/files/rs-9269266/v1/8fb66eb98607ee5761240be2.png"},{"id":107141144,"identity":"13719462-3b4c-4ede-94d3-429a420863c4","added_by":"auto","created_at":"2026-04-17 08:59:37","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3058665,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9269266/v1/9313b7ae-939d-4049-ae3a-7b06a4250105.pdf"},{"id":107140956,"identity":"7fc0a11d-9a2e-482f-b26c-1ac19eea2613","added_by":"auto","created_at":"2026-04-17 08:59:23","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":712832,"visible":true,"origin":"","legend":"","description":"","filename":"JMESMArticlensLTP.docx","url":"https://assets-eu.researchsquare.com/files/rs-9269266/v1/a0ed3db4abd315c3f24f487f.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Structural and evolutionary analysis of non-specific lipid transfer proteins (nsLTPs) in Cereus (Cactaceae)","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe genus \u003cem\u003eCereus\u003c/em\u003e Mill. (Cactaceae, tribe Cereeae) represents a diverse and widely distributed group within the cactus family (Franco et al., 2017). Comprising approximately 31 recognized species, \u003cem\u003eCereus\u003c/em\u003e occurs across a variety of South American biomes, ranging from coastal regions to Andean habitats above 3,000 meters in altitude (Hunt et al., 2006). This broad distribution reflects a remarkable adaptive capacity to contrasting environmental conditions, including drought and temperature fluctuations typical of xeric and semi-arid ecosystems (Amaral et al., 2021; 2025a). The phylogenetic framework proposed for \u003cem\u003eCereus\u003c/em\u003e was later refined by recent phylogenetic analyses employing nuclear orthologous genes and a coalescent-based approach (Bombonato et al., 2020; Taylor et al., 2023). These studies support the genus as monophyletic and, although previous analyses structured it into major clades (Bombonato et al., 2020), the new phylogeny, which tested the preexisting subgeneric classification from 1992, partially recovered three of the four subgenera.\u003c/p\u003e \u003cp\u003eThe adaptability of \u003cem\u003eCereus\u003c/em\u003e to xeric and semi-arid ecosystems, an adaptation observed at anatomical, physiological, and biochemical levels, is reflected in the production of a wealth of secondary metabolites. In traditional medicine, the cladode, root, and seeds of \u003cem\u003eCereus jamacaru\u003c/em\u003e D.C. are used to treat various ailments, including urinary tract infections, kidney inflammation, syphilis, gastritis, and cardiovascular and respiratory disorders (Andrade et al., 2006; Lucena et al., 2013; Palheta et al., 2017; da Silva et al., 2019; Rodrigues Almeida and Gonzaga Fernandez, 2025; Teodoro et al., 2025; Cardoso et al., 2026). The phytochemical composition of \u003cem\u003eC. jamacaru\u003c/em\u003e cladodes reveals the presence of phenethylamine alkaloids, such as hordenine, tyramine, and N-methyltyramine, as well as tyrosine. Notably, tyramine exhibits sympathomimetic activity and a probable cardiotonic effect. Other bioactive compounds identified include flavonoids and tannins, which confer crucial therapeutic properties, including anti-inflammatory, antifungal, antioxidant, and wound-healing activities (Rodrigues Almeida and Gonzaga Fernandez, 2025).\u003c/p\u003e \u003cp\u003eMoreover, exploring \u003cem\u003eCereus\u003c/em\u003e at the molecular level, particularly through the analysis of its proteins, may uncover key mechanisms underlying the evolutionary history and adaptive radiation of the genus. Among these proteins, non-specific lipid transfer proteins (nsLTPs) are of particular interest due to their structural diversity and functional versatility across plant lineages (Missaoui et al., 2022). These are small (around 9 kDa) and basic proteins broadly distributed across higher plants, typically composed of four to five \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\alpha\\:\\)\u003c/span\u003e\u003c/span\u003e-helices stabilized by a conserved motif of eight cysteine residues forming four disulfide bonds (D\u0026rsquo;Agostino et al., 2018). This compact fold creates a hydrophobic tunnel-like cavity that allows nsLTP to bind and transport a wide range of lipid molecules; structural features that not only stabilize under stress conditions but also enable participation in multiple biological roles, including membrane remodelling, cell wall organization, and signaling pathways (Egger et al., 2010; Missaoui et al., 2022). Further investigation is needed to clarify their physiological roles, particularly in non-model taxa like \u003cem\u003eCereus\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eAlthough recent phylogenomic studies have refined the evolutionary framework of \u003cem\u003eCereus\u003c/em\u003e, the molecular mechanisms underlying its remarkable ecological adaptability and biochemical versatility remain poorly understood. In particular, the processes related to lipid interaction and transport, which are central to stress adaptation and membrane homeostasis in xerophytic plants, have not been investigated in this genus. While nsLTPs are well characterized in model plants, their diversity, structural features, and functional roles in non-model taxa such as \u003cem\u003eCereus\u003c/em\u003e and \u003cem\u003eCipocereus\u003c/em\u003e are virtually unknown. This study aims to characterize the nsLTP repertoire in \u003cem\u003eCereus\u003c/em\u003e by integrating comparative transcriptomics, structural modeling, pocket/hot-spot mapping, and evolutionary inference. We test the central mechanistic hypothesis that \u003cem\u003eCereus\u003c/em\u003e nsLTPs exhibit a conserved structural scaffold coupled to evolutionarily labile \u0026lsquo;gate\u0026rsquo; regions (loop rims), such that variation at the tunnel entrance can modulate ligand access and local compatibility with lipid-like molecules without disrupting fold stability. This framework yields explicit, testable predictions (e.g., species- and paralog-specific differences in tunnel geometry and surface chemistry) that can be evaluated in future functional assays, thereby bridging molecular structure to physiological processes relevant to stress-associated lipid trafficking.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Data collection and transcriptome assembly\u003c/h2\u003e \u003cp\u003eTranscriptomic datasets from \u003cem\u003eCereus\u003c/em\u003e species were obtained from previous studies of our group (Amaral et al., 2025a; 2025b). Redundant contigs were clustered with CD-HIT v4.8.1 (Li and Godzik, 2006) at a 90% identity threshold. Protein-coding sequences were inferred and used for orthology assignment in OrthoFinder v2.4.0 (Emms and Kelly, 2019), generating the orthogroup matrix employed in the present analyses. The orthogroups were annotated against the nsLTP identified for \u003cem\u003eC. jamacaru\u003c/em\u003e (Cardoso et al., 2026). The isolated orthogroups with the nsLTP function were used for downstream analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Sequence collection and structural modeling\u003c/h2\u003e \u003cp\u003eStructural predictions of \u003cem\u003eCereus\u003c/em\u003e nsLTP homologs were generated with Modeller v.10.6 (Webb and Salli, 2016) and AlphaFold3 (Jumper et al., 2023) using default settings in both cases. The standard AlphaFold confidence output pLDDT (per-residue confidence) was recorded to guide downstream interpretation. Although these approaches yield static structural models, they are known to capture evolutionarily conserved and functionally relevant conformational states, particularly for small, disulfide-stabilized proteins such as nsLTPs. For cross-target comparability, we used the full polypeptide as predicted; no manual trimming or domain excision was applied before alignment.\u003c/p\u003e \u003cp\u003eFor structural comparisons, we designated nsLTP from \u003cem\u003eC. jamacaru\u003c/em\u003e as the reference scaffold. The choice of a single reference ensures consistent interpretation of TM-scores and RMSDs across the dataset while avoiding ambiguity introduced by all-vs-all normalization. We performed structure-only alignments using TM-align. For each mobile model vs the reference, TM-align was executed with the -o flag to write superposed coordinates. TM-align returns three key metrics that we recorded for every pair: (i) TM-score (dimensionless, 0\u0026ndash;1), (ii) RMSD (\u0026Aring;) over the aligned Cα subset; and (iii) aligned length. We accepted alignments when TM-align successfully returned all three metrics and produced a superposed PDB, and then, they were inspected in PyMOL. Additionally, structural stereochemical quality was evaluated with PROCHECK (Laskowski et al., 1993), using the standard Ramachandran plot assessment and geometry checks to ensure that modeled structures met expected quality thresholds.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Mapping and Characterization of Binding Regions\u003c/h2\u003e \u003cp\u003eIdentification of potential ligand-binding sites was performed using FTSite (Ngan et al., 2012; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ftsite.bu.edu/\u003c/span\u003e\u003cspan address=\"https://ftsite.bu.edu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e and FTMap (Jones et al., 2022; Kozakov et al., 2015; Ngan et al., 2012; Brenke et al., 2009; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ftmap.bu.edu/\u003c/span\u003e\u003cspan address=\"https://ftmap.bu.edu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e webservers. We used FTSite and FTMap to identify ligand-binding regions on protein surfaces. Both methods probe the surface with small organic fragments and define consensus binding sites where multiple probe clusters overlap. FTMap further characterizes these clusters as energetically favorable \u0026ldquo;hot spots.\u0026rdquo; We used the default FTMap probe set and server settings (see Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Hot-spots identified by FTMap were ranked by cluster strength and mapped onto predicted cavities. Across species and paralogs, high-density probe clusters systematically co-localized with large, high-ligandability pockets, indicating concordance between independent surface-mapping and cavity-detection approaches. This convergence supports the robustness of predicted binding regions.\u003c/p\u003e \u003cp\u003eThe FPocket web server (Kochnev and Durrant, 2022; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://durrantlab.pitt.edu/fpocketweb/\u003c/span\u003e\u003cspan address=\"https://durrantlab.pitt.edu/fpocketweb/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e was employed to characterize the proteins further using default parameters (minimum alpha-spheres\u0026thinsp;=\u0026thinsp;35; minimum cavity volume\u0026thinsp;=\u0026thinsp;50 \u0026Aring;\u0026sup3;), and pockets were ranked by their ligandability score, which was subsequently interpreted in the context of compatibility with lipid-like ligands relevant to plant physiology. Only pockets with ligandability score\u0026thinsp;\u0026gt;\u0026thinsp;0.05 or volume\u0026thinsp;\u0026gt;\u0026thinsp;200 \u0026Aring;\u0026sup3; were retained for comparative discussion. This tool is designed with three main objectives: (i) the identification of pockets to delimit cavities on the protein surface with potential to bind small compounds; (ii) the classification of these pockets according to their likelihood of accommodating ligand-like molecules; and (iii) consideration of conformational adaptability, known as induced fit, whereby the pocket geometry may undergo rearrangements during complex formation with a ligand.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Phylogenetic reconstruction and evolutionary analysis\u003c/h2\u003e \u003cp\u003eProtein sequences from \u003cem\u003eCereus\u003c/em\u003e spp. were aligned using MAFFT v7.520 (Katoh and Standley, 2013) with the L-INS-i strategy, which provides high accuracy for sequences with conserved motifs and variable loop regions. The resulting alignment was visually inspected in Bioedit 5.0.9 (Hall et al., 2011) to remove ambiguously aligned positions and terminal gaps. Phylogenetic inference using the aminoacids sequences aligned was performed with IQ-TREE2 v2.3.5 (Minh et al., 2020), using the best-fitting substitution model automatically selected by the software. Branch support was assessed with 1,000 ultrafast bootstrap replicates and 1,000 SH-aLRT replicates. Gene duplication events were inferred by reconciling the inferred gene tree with a species tree using Notung v2.9 (Stolzer et al., 2012), allowing identification and visualization of duplication and loss events within the phylogeny.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Individual Sites Subject to Episodic Diversifying Selection\u003c/h2\u003e \u003cp\u003eThe coding sequences (CDS) of nsLTPs were retrieved from the published genome of \u003cem\u003eCereus\u003c/em\u003e (Amaral et al., 2025a). To detect signatures of adaptive evolution, the Mixed Effects Model of Evolution (MEME) was applied through the Datamonkey web server (Weaver \u003cem\u003eet al.\u003c/em\u003e, 2018; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.datamonkey.org\u003c/span\u003e\u003cspan address=\"https://www.datamonkey.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e, which implements the MEME (Murrell \u003cem\u003eet al\u003c/em\u003e, 2012) method of the HyPhy software package. MEME estimates the ratio of nonsynonymous to synonymous substitution rates (ω\u0026thinsp;=\u0026thinsp;dN/dS), allowing ω to vary both across sites (fixed effects) and across branches at a given site (random effects). This approach enables the identification of episodic positive selection, meaning adaptive events that occur only on a subset of branches at specific codon positions. Because such variation may reflect adaptive, compensatory, or selectively neutral processes, MEME results were interpreted in conjunction with structural context. The analysis was conducted using a significance threshold of p\u0026thinsp;\u0026lt;\u0026thinsp;0.05. Sites showing evidence of episodic diversifying selection were identified based on the likelihood ratio test (LRT) and the empirical Bayes posterior probabilities reported by Datamonkey.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Structural assessment of nsLTP across \u003cem\u003eCereus\u003c/em\u003e species\u003c/h2\u003e \u003cp\u003eTM-scores consistently fell within the range typically interpreted as \u0026ldquo;same topology\u0026rdquo; (\u0026asymp;\u0026thinsp;0.5\u0026ndash;0.7), with aligned cores of ~\u0026thinsp;60\u0026ndash;80 residues and backbone RMSDs around 2\u0026ndash;3 \u0026Aring;. Importantly, models displaying higher loop divergence also tended to exhibit increased pocket-volume variability and altered rim polarity (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e), indicating a coupling between peripheral structural variation and tunnel geometry. This pattern supports a conserved-core/variable-rim organization. In other words, the \u003cem\u003eCereus\u003c/em\u003e protein is not a new fold, it is a moderately diverged representative of the family (Yeats and Rose, 2008; Edstam et al., 2011). The fraction of residues that fail to align one-to-one mainly reflects peripheral segments, rather than wholesale rearrangements of the structural core. It is also important to note that structural comparisons were performed relative to a single reference structure. While this standardizes TM-scores, it may underestimate structural divergence among non-reference homologs. Nevertheless, the conserved fold-level architecture observed across all alignments suggests that reference-based comparisons are adequate for capturing the shared nsLTP topology.\u003c/p\u003e \u003cp\u003eWhen the superposed models are inspected, the largest deviations concentrate in solvent-exposed loops and turns that rim the putative binding groove, while the inner β/α framework that shapes the pocket is comparatively conserved. This pattern indicates that static models consistently resolve a structurally constrained core that defines tunnel geometry, while allowing peripheral regions to vary in ways that are functionally meaningful for ligand access. In practical terms, the \u003cem\u003eCereus\u003c/em\u003e nsLTP likely preserves the overall pocket architecture, but varies the local geometry and side-chain presentation at the entrance (loop lengths, insertions/deletions, and substitutions; Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Such changes are precisely the ones that modulate ligand preference and kinetic fine-tuning without compromising fold stability (Kader, 1996; Yeats and Rose, 2008).\u003c/p\u003e \u003cp\u003eThe structural signal points to evolutionary conservation of the scaffold with meaningful variation at the binding interface. The \u003cem\u003eCereus\u003c/em\u003e homolog is therefore best described as conformationally compatible but functionally tunable relative to its relatives: it aligns well enough to support homology-based inference, yet it carries loop-level alterations consistent with shifts in specificity.\u003c/p\u003e \u003cp\u003eThe stereochemical analysis of the nsLTP_modeller and nsLTP_AF3 models using PROCHECK shows that both present good structural quality (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA, B), with most residues located in favored regions of the Ramachandran plot (93.3% for nsLTP_modeller and 92.3% for nsLTP_AF3). However, nsLTP_AF3 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA) shows no bad contacts and slightly smaller deviations in bond lengths and angles. This suggests a higher predicted geometric stability, although these differences must be interpreted cautiously, as static predictions do not capture conformational flexibility. Both models have only three residues flagged in the Ramachandran plots and satisfactory side-chain parameters, but nsLTP_AF3 exhibits positive overall G-factors (0.23), suggesting a more reliable global conformation. Thus, while both models are suitable for structural and docking analyses, nsLTP_AF3 presents marginally better geometric scores. Still, these differences are within the expected range for \u003cem\u003ein silico\u003c/em\u003e models, and do not guarantee functional superiority without experimental validation.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Identification of Pockets and Functional Regions\u003c/h2\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e3.2.1. Identification and Characterization of Binding Sites\u003c/h2\u003e \u003cp\u003eGiven the predictive nature of structural modelling and the potential conformational variations between different methods, we evaluated the protein using two distinct topologies: one generated by Modeller v.10.6 (nsLTP_Modeller) and the other by AlphaFold (nsLTP_AF3) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA,B). This approach not only allows for a more robust identification of binding sites and hot spots but also enables the assessment of the consistency and reliability of the detected regions, providing a solid basis for subsequent analysis.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe structures of nsLTP_modeller and nsLTP_AF3 were submitted to the online servers FTSite and FTMap to locate potential binding sites and determine their main characteristics. Additionally, both structures were analyzed using FPocket to identify and quantify pockets based on geometric and physicochemical properties, providing complementary information on pocket volume, depth, hydrophobicity, and ligandability. Hydrogen bond interactions for individual residues were further assessed using FTMAP, and the results are provided in Supplementary Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e, offering detailed insight into residue participation within the predicted binding sites.\u003c/p\u003e \u003cp\u003eFor nsLTP_Modeller (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA), three distinct hydrophobic cavities were identified, each forming a single cluster of overlapping probes (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), suggesting well-defined binding regions with limited conformational flexibility. The main interacting residues are predominantly hydrophobic (e.g., Ile, Leu, Val, Ala, and Phe), consistent with the protein\u0026rsquo;s lipid-binding nature (Kader, 1996; Yeats and Rose, 2008). The hydrophobic side chains are located predominantly in the interior of a protein and this arrangement stabilizes the folded polypeptide backbone, since unfolding it or extending it would expose the hydrophobic side chains to the solvent (Gowder et al., 2014).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComparative Summary of Binding Sites and FTMap Probe Clusters in nsLTP_modeller and nsLTP_AF3.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eBinding site\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eResidues involved\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003eFTMap probe clusters\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eClusters\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eProbe\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003ensLTP_Modeller Structure\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSite #1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIle61, Leu64, Ser65, Ala68, Arg74, Val77, Val105, Ile107\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSingle cluster\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBEN, BUT, EOL, THS, BDY, AMN, ACD, ADY,\u003c/p\u003e \u003cp\u003eDFO, ETH, ACT, DME\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSite #2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLeu38, Ile61, Leu81, Ala96, Gly97, Met99, Pro100, Ile111\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSingle cluster\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eACN, EOL, ETH, THS, ACT, ADY, DFO, DME,\u003c/p\u003e \u003cp\u003eAMN, ACD, URE\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSite #3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMet19, Ala20, Leu21, Ala22, Thr70, Ile71, Ala72, Asp73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSingle cluster\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBDY, CHX, BUT, BEN, PHN\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003ensLTP_AF3 Structure\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eSite #1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003ePro63, Tyr64, Lys65, Thr70, Asp71, Cys72, Lys74, Val75, Gln76, Lys77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCluster 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCHX, ACD, ACT, BDY, BEN, BUT, ETH, ACN, DME, URE, ADY, DFO, EOL, PHN, THS, AMN\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCluster 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBUT, BDY, CHX, BEN, ETH, THS, ACT, ADY, DFO, DME, PHN\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCluster 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBUT, ACT, BDY\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSite #2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eIle28, Thr32, Val33, Ser36, Leu37, Asp58, Gly59, Ile60, Asn61, Ile62, Ser78, Ser80, Ser81, Thr82, Ala83, Val92, Leu96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCluster 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eACD, ACN, ACT, ADY, AMN, BEN, DFO, DME, EOL, ETH, PHN, THS, URE\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCluster 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eURE, DFO, PHN, ACN\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSite #3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePro63, Tyr64, Lys65, Gly112, Ser113, Met114, Pro115, Thr116, Val120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSingle cluster\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBDY, CHX, ACD, ACT, ADY, BEN, DFO, ETH, PHN, THS, DME, BUT, EOL, URE, ACN\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eIn contrast, nsLTP_AF3 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB) exhibited three corresponding binding regions, but with a greater number of clusters and higher probe diversity, especially in site #1 (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), where sixteen probes were mapped across three overlapping clusters. This indicates a more accessible and flexible binding environment, potentially enabling the accommodation of ligands with different physicochemical properties. Sites #2 and #3 in nsLTP_AF3 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB) also showed expanded residue participation and overlap among probes, reinforcing the idea of a structurally dynamic pocket architecture.\u003c/p\u003e \u003cp\u003eOverall, the comparative analysis highlights that while both models preserve the characteristic hydrophobic tunnel of nsLTPs, the AlphaFold predicted structure displays increased binding plasticity and chemical diversity, which may reflect a more realistic representation of the protein\u0026rsquo;s functional conformational ensemble.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e3.2.2. Structural Pocket Analysis\u003c/h2\u003e \u003cp\u003ePocket analysis of the nsLTP_Modeller model revealed a total of nine potential binding sites with distinct physicochemical properties (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA, B). Most pockets exhibited low ligandability scores (\u0026lt;\u0026thinsp;0.01), except for pocket #9, the largest site, combined a significant apolar contribution with substantial polar SASA, a high number of alpha spheres (263), and the largest volume (1865.5 \u0026Aring;\u0026sup3;). These features resulted in a high pocket score (0.756), which is interpreted here as a measure of geometric and physicochemical ligandability for lipid-like molecules. This cavity therefore represents a structurally accessible region compatible with physiologically relevant hydrophobic and amphipathic ligands, consistent with the known lipid-binding function of nsLTPs. These results highlight pocket #9 as the most promising target for docking studies, whereas the other sites might serve secondary or more specialized binding roles.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWhen compared to the nsLTP_modeller structure, the nsLTP_AF3 model exhibited fewer binding pockets but with improved physicochemical properties. Pocket analysis of the nsLTP_AF3 model identified five potential binding cavities (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA) with varying physicochemical properties. Among these, pocket #1 stood out with the highest pocket score (0.368) and the largest volume (452.3 \u0026Aring;\u0026sup3;), indicating enhanced geometric suitability for lipid-like ligand accommodation, and suggesting a versatile and accessible region for physiologically relevant hydrophobic interactions. Pockets #2 and #5 were predominantly hydrophobic (apolar SASA\u0026thinsp;\u0026gt;\u0026thinsp;84%), suggesting potential affinity for lipid or hydrocarbon ligands, whereas pockets #3 and #4 exhibited higher polarity and moderate volumes (259\u0026ndash;391 \u0026Aring;\u0026sup3;), possibly accommodating amphipathic molecules. Overall, nsLTP_AF3 displayed a smaller number of binding sites compared to nsLTP_modeller but with generally higher hydrophobicity and ligandability, particularly in pocket #1, which appears to represent the principal ligand-binding region of this model.\u003c/p\u003e \u003cp\u003eThe main cavity of nsLTP_AF3 (pocket #1) presented a higher ligandability index, indicating a geometrically favorable and accessible region for ligand interaction. In contrast, the nsLTP_Modeller pockets were generally smaller, less ligandable, and displayed higher polarity, suggesting weaker affinity toward hydrophobic ligands. The reduction in the number of predicted pockets in nsLTP_AF3, together with the larger primary site volume, suggests a more consolidated binding region within this model. However, such differences may also arise from model-dependent structural rearrangements, especially in loops, and should be interpreted cautiously.\u003c/p\u003e \u003cp\u003ePrevious work by our group (Cardoso et al., 2026) has demonstrated the feasibility of combining large-scale molecular docking with molecular dynamics simulations using triacylglycerols and related lipid substrates in \u003cem\u003eCereus\u003c/em\u003e proteins. These analyses highlighted both the potential and the methodological challenges associated with modeling long-chain and highly flexible lipid molecules in dynamic environments.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.3. Phylogeny and Evolutionary Context of the Gene/Protein\u003c/h2\u003e \u003cp\u003eThe phylogeny constructed for the genus shows clear diversification of the studied proteins, with several clades well supported by bootstrap values. Many nodes exhibit high support (\u0026ge;\u0026thinsp;95), indicating that certain groupings are highly reliable, as observed in subclades including sequences from \u003cem\u003eCipocereus\u003c/em\u003e, \u003cem\u003eC. pierrebraunianus\u003c/em\u003e, and \u003cem\u003eC. mirabella\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Some branches, however, display intermediate or low bootstrap values (~\u0026thinsp;36), suggesting lower confidence in these groupings, possibly due to rapid divergence or less conserved regions of the proteins. Interestingly, species-specific duplications are evident, with multiple copies of sequences within the same species forming distinct subclades, suggesting paralogous individuals and potential functional diversification. Additionally, some sequences form clades that span multiple species, indicating ancestral duplications predating the divergence of species within the genus.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003ePhylogenetic analyses of nsLTP proteins in cactus species, such as \u003cem\u003eC. jamacaru\u003c/em\u003e, reveal multiple gene duplication events (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). However, many of these duplications are non-persistent, meaning they were not maintained throughout the evolution of different lineages. This pattern indicates that some duplicated gene copies were lost in descendant species, suggesting that not all provided sufficient adaptive advantage to be fixed. Thus, nsLTPs in cacti display an evolutionary dynamic characterized by frequent duplication and loss events, a pattern commonly observed in gene families associated with defense processes and responses to environmental stress (Edstam et al., 2011). This behavior reinforces the notion that only some recent duplications were functionally relevant, possibly linked to adaptation to arid conditions and the formation of a protective cuticle.\u003c/p\u003e \u003cp\u003eSpecies-specific paralogs were also detected. In particular, some nsLTP copies were found exclusively in \u003cem\u003eC. calcirupicola\u003c/em\u003e or \u003cem\u003eC. pierrebraunianus\u003c/em\u003e, pointing to recent duplications or species-specific retention. These unique paralogs are potential candidates for neofunctionalization, possibly related to ecological adaptation. Finally, cases of single-copy retention in nsLTPs were observed, where certain species maintained only one exclusive representative.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.4. Identification of Coding Sites under Positive Selection Across the Phylogeny\u003c/h2\u003e \u003cp\u003eThe analysis using MEME (Mixed Effects Model of Evolution) on Datamonkey was conducted with 39 aligned sequences, totaling 133 codon sites. The model employed approximately 69 branches per partition in the test and 100 bootstrap replications, providing statistical robustness to the inference. From the 133 sites evaluated, 18 showed evidence of episodic positive selection (Fig.\u0026nbsp;5B), corresponding to about 13.5% of the total. On average, each codon under positive selection was detected in 1 to 2 branches of the phylogenetic tree (Table S2), suggesting that selective pressure does not act uniformly across the entire gene but rather locally in specific clades or lineages. This pattern is consistent with episodic functional fine-tuning, although alternative explanations such as compensatory evolution or lineage-specific constraint relaxation cannot be excluded. This result indicates the occurrence of lineage-specific adaptations throughout the evolution of the studied group, possibly related to functional changes, ecological specialization, or diversification in interactions with partner molecules (substrates, ligands, or other proteins) (Fig.\u0026nbsp;5C). The fact that no sites were identified with variable ω (dN/dS) across all branches reinforces (Table S2) the idea that the observed positive selection is episodic and restricted rather than pervasive adaptive pressure.\u003c/p\u003e \u003cp\u003eFrom a functional perspective, these 18 sites represent relevant candidates for further investigation. Several codons under episodic rate variation mapped to loop rims and pocket-adjacent surfaces. Relative to randomly sampled residues, these sites showed a higher-than-expected spatial proximity to tunnel entrance regions, indicating non-random localization of evolutionary rate variation within functionally sensitive structural contexts. Their localization supports a potential role in modulating local conformational or interaction properties, but does not, by itself, demonstrate adaptive functional divergence. This structural context strengthens the mechanistic plausibility of these substitutions. The positive selection analysis using the MEME model identified 18 codons (2, 6, 9, 12, 20, 34, 51, 53, 56, 65, 88, 100, 110, 112, 116, 131, 132, and 133) under episodic diversifying selection throughout the phylogeny (Table S2). These sites showed p-values\u0026thinsp;\u0026le;\u0026thinsp;0.05 (Fig.\u0026nbsp;5A, B, C), indicating that they were targets of adaptive events in at least one evolutionary branch of the tested tree (Murrell et al., 2012).\u003c/p\u003e \u003cp\u003eAmong the 133 analyzed codons, approximately 13.5% exhibited signs of positive selection. On average, each selected codon showed support in 1 to 3 branches, suggesting that selective pressure occurred in a localized manner, reflecting specific adaptations to certain evolutionary contexts rather than a constant and widespread pressure across the entire lineage. The distribution of sites under selection was not concentrated in a single region of the gene but occurred relatively dispersed along the sequence. This pattern may be associated with the maintenance of multiple molecular functions, modulation of protein\u0026ndash;protein interactions, or the need to respond to diverse environmental or functional pressures.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eOur analyses support a gate-modulation mechanistic model for \u003cem\u003eCereus\u003c/em\u003e nsLTPs, in which a conserved four-helix scaffold maintains the hydrophobic tunnel, while loop-level diversification at rim/entrance regions tunes tunnel accessibility and local physicochemical properties. By integrating structural comparisons, pocket/hot-spot mapping, phylogenetic context, and episodic selection signals, we show that functional tunability is concentrated in peripheral, flexible regions rather than in the structural core. This provides a mechanistic bridge between molecular variation and physiological roles commonly attributed to nsLTPs (e.g., lipid trafficking linked to cuticle/membrane function under stress), without implying direct ecological causality in the absence of functional assays.\u003c/p\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Structure modelling quality and characterization for \u003cem\u003eCereus\u003c/em\u003e nsLTP\u003c/h2\u003e \u003cp\u003eThe comparative analysis of nsLTP_Modeller and nsLTP_AF3 highlights significant differences in pocket organization and binding plasticity, offering insights into the molecular determinants of ligand recognition in nsLTPs. Additionally, this comparison provides a means to assess the extent to which predictions from distinct modeling approaches capture the protein\u0026rsquo;s true structural features, facilitating evaluation of binding site stability and the reliability of identified functional regions (Krokidis et al., 2025). Furthermore, understanding the variations in pocket accessibility and flexibility between models can inform the selection of suitable binding sites for docking studies, ligand design, and functional characterization, ultimately enhancing the predictive power and biological relevance of computational analyses. Throughout this study, pocket and cavity scores are used as structural descriptors of ligandability for lipid-like molecules involved in plant physiology.\u003c/p\u003e \u003cp\u003eIn nsLTP_modeller, the main binding residues (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) are predominantly hydrophobic or small polar side chains, consistent with the classical description of the hydrophobic cavity found in plant nsLTPs. These residues correspond to positions typically forming the central cleft between helices H1\u0026ndash;H4 and the C-terminal tail, stabilized by disulfide bridges in the canonical 8CM motif. The single-cluster organization observed in all Modeller sites indicates a more restricted cavity topology, with limited conformational rearrangement and lower accessibility for diverse ligands (Salminen et al., 2016).\u003c/p\u003e \u003cp\u003eConversely, nsLTP_AF3 displayed three binding sites with a more complex cluster distribution (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), particularly in Site #1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB), which contained three overlapping FTMap probe clusters. This pattern suggests an expanded or more solvent-exposed cavity, consistent with a higher dynamic range of side-chain orientations. The presence of polar residues such as Lys65, Asp71, and Gln76, adjacent to hydrophobic residues (Val75, Pro63), points to a mixed environment that may facilitate interaction with amphipathic ligands or transient solvent molecules. As reported for TaLTP1.1 (Simorre et al., 1991; Gincel et al., 1994; Salminen et al., 2016; PDB ID: 1GH1), the hydrophobic core is formed by residues distributed across multiple helices and small conformational changes can significantly alter cavity volume and connectivity. Therefore, the AlphaFold3 structure may represent one plausible accessible conformation of the same fold, potentially capturing states that differ from the Modeller-derived model. However, given that loop conformations are intrinsically flexible and AlphaFold confidence decreases in these regions, these differences should be interpreted as model-dependent rather than definitive.\u003c/p\u003e \u003cp\u003eThe functional implications of these structural differences are particularly relevant in light of the multifunctional nature of nsLTPs. These proteins are known to participate in diverse physiological processes, from cuticle formation and pathogen defense to lipid signaling and membrane remodeling (Shenkarev et al., 2017; Madni et al., 2020). However, it remains unclear whether such functional diversity arises from the existence of multiple isoforms or from the intrinsic ability of nsLTPs to accommodate a broad range of lipid molecules.\u003c/p\u003e \u003cp\u003eMany of the positively selected codons mapped by MEME fall within or adjacent to flexible loop regions, which typically define the entrances, rims, or dynamic gating elements of lipid-binding pockets in nsLTPs (Fig.\u0026nbsp;5B, C). This spatial correspondence is reinforced by the overlap between selected residues and cavities identified by FPocket/FTMap, particularly at the edges of predicted binding pockets and putative interface regions. Such a pattern suggests that episodic selection has likely modulated the physicochemical properties of these structural rims, altering local charge, hydrophobicity, or conformational flexibility, to fine-tune ligand affinity or specificity across different \u003cem\u003eCereus\u003c/em\u003e lineages. These adjustments may reflect ecological divergence in lipid composition or cuticular properties, but this remains a working hypothesis requiring empirical validation. However, because MEME does not distinguish between ecological, structural, or neutral drivers of substitution, the connection to ligand specificity must be treated as a working hypothesis rather than a demonstrated mechanism.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Structural comparison\u003c/h2\u003e \u003cp\u003eComparative structural analyses indicate that the nsLTP homolog preserves the same global architecture observed across the homologous set (Fleury et al., 2019; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). The combination of high topological agreement with modest geometric spread supports the view that the \u003cem\u003eCereus\u003c/em\u003e protein is not a new fold among them, but rather a moderately diverged member of the family (Santos-Silva et al., 2023). The residues that fail to align one-to-one are largely confined to peripheral segments, whereas the inner β/α framework that defines the pocket is retained.\u003c/p\u003e \u003cp\u003eDespite this fold-level conservation, the largest deviations cluster in solvent-exposed loops and turns that rim the binding groove (Malinina et al., 2017). These insertions/deletions and side-chain substitutions reshape the local geometry and electrostatics at the pocket entrance without disrupting the core cavity. Functionally, such \u0026ldquo;rim remodeling\u0026rdquo; is a well-known route to tune specificity and affinity while preserving catalytic or binding competence (Hama\u0026iuml; and Drin, 2024). Thus, the most parsimonious interpretation is binding-site plasticity superimposed on a conserved scaffold (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). The \u003cem\u003eCereus\u003c/em\u003e nsLTP likely recognizes similar chemical motifs as its relatives but with altered rank order of preferences and distinct kinetic parameters (e.g., shifts in K\u003csub\u003eD\u003c/sub\u003e, K\u003csub\u003eM\u003c/sub\u003e, or K\u003csub\u003ecat\u003c/sub\u003e depending on the biochemical role of nsLTP).\u003c/p\u003e \u003cp\u003eAcross \u003cem\u003eCereus\u003c/em\u003e species, this architectural pattern helps reconcile conservation and divergence. A conserved fold constrains the baseline function (the class of ligands/substrates accommodated and the overall reaction/interaction mechanism), whereas loop-level variability affords ecologically meaningful differentiation (Hama\u0026iuml; and Drin, 2024; Melnikova et al., 2022). In practice, different \u003cem\u003eCereus\u003c/em\u003e lineages may exploit the same scaffold to optimize binding under distinct microenvironmental conditions (pH, ionic strength, cofactor availability), to adjust on/off rates for related ligand families, or to gate interactions with species-specific partners via motif gain/loss at flexible rims. We therefore anticipate conserved pocket cores across species, accompanied by species-specific loop chemistries that modulate recognition and regulation rather than abolish function.\u003c/p\u003e \u003cp\u003eThese structural signals lead to clear, testable predictions. First, a common ligand panel should bind across homologs, but affinities and preferred poses will vary in ways that track loop differences (Brissos et al., 2024). Docking guided by the aligned cores is expected to reproduce this behavior, with poses diverging at the pocket entrance (Velesinović and Nikolić, 2021). Second, environmental sensitivity should shift between species: pH-dependence, ionic modulation (e.g., Na+/Ca2+/Mg2+), or allosteric responses are likely to differ if loop mutations alter local protonation or cation-π networks (Yuan et al., 2025). Third, post-translational regulation and protein\u0026ndash;protein contacts are plausible divergence points, because loop rims are typical carriers of PTM sites and linear interaction motifs (Yuan et al., 2025). Finally, loop-swap mutagenesis (transplanting variable rims between species on the same core) should transfer part of the specificity phenotype, directly linking structural variability at the interface to functional outcomes.\u003c/p\u003e \u003cp\u003eTwo caveats temper these conclusions. Most observed differences arise in low-confidence, high-flexibility regions by their very nature (loops), and single static models cannot capture their full conformational ensembles (Srinivasan et al., 2024). Nevertheless, the consistent localization of structural variation at tunnel rims across independent modeling approaches indicates that these regions represent constrained functional states rather than modeling artifacts (Hasegawa and Holm, 2009). Static structures therefore provide a reliable framework for identifying evolutionarily stable features and functionally tunable interfaces in nsLTPs. Thus, the \u003cem\u003eCereus\u003c/em\u003e nsLTP is best described as conformationally compatible but functionally tunable relative to its homologs. It aligns well enough to warrant homology-based inference of mechanism, yet carries loop-level alterations consistent with fine-grained shifts in specificity, affinity, and regulation. This scaffold-conserved, rim-variable organization offers a coherent evolutionary route for species-level adaptation without requiring innovation at the fold level.\u003c/p\u003e \u003cp\u003eStructural divergence, pocket geometry, surface hot-spot distribution, phylogenetic history, and episodic rate variation converge on a coherent pattern: evolutionary diversification in \u003cem\u003eCereus\u003c/em\u003e nsLTPs is concentrated at tunnel rims and entrance regions, where modest sequence changes are sufficient to modulate ligand compatibility without destabilizing the conserved scaffold. This integrative signal moves the present analysis beyond descriptive cataloguing toward a mechanistic interpretation of structure\u0026ndash;function\u0026ndash;evolution relationships\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Ecological and Adaptive Significance\u003c/h2\u003e \u003cp\u003eA possible functional implication of our structural analyses is that loop flexibility and binding-pocket plasticity in \u003cem\u003eCereus\u003c/em\u003e nsLTPs could influence lipid activation under abiotic stress. The conserved β/α framework provides a stable scaffold for maintaining the hydrophobic tunnel, while the more variable loops and rim residues that shape the entrance of the pocket are ideally positioned to modulate ligand access, residence time, and selectivity (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Under desiccation, heat, or high irradiance, cuticular lipids and membrane components undergo remodeling, and nsLTPs must accommodate changes in chain length, saturation, and head-group chemistry (Tapia et al., 2013; Xiao et al., 2023). In this context, a structurally core combined with a conformationally permissive rim allows the protein to bind a broader spectrum of hydrophobic ligands without compromising global stability, thus supporting dynamic lipid transport and membrane/cuticle protection in fluctuating xeric environments (Missaoui et al., 2022).\u003c/p\u003e \u003cp\u003eThis pattern of a conserved scaffold with localized plasticity is consistent with nsLTP adaptations described in other xerophytic or drought-tolerant systems, such as cacti, agaves, and drought-adapted \u003cem\u003eArabidopsis thaliana\u003c/em\u003e ecotypes (Deeken et al., 2016; Rojas et al., 2019), where nsLTPs are frequently associated with cuticle deposition, wax accumulation, and stress-induced lipid trafficking. In those species, nsLTP upregulation under water deficit and heat stress has been linked to reinforced barrier formation and improved maintenance of cell integrity, suggesting that subtle structural adjustments at the binding interface are sufficient to tune ligand preference to stress-related lipids (Xue et al., 2022; Xiao et al., 2023). Our results, combining a canonical hydrophobic tunnel with model-dependent differences in pocket size, hydrophobicity, and probe diversity, corroborate this broader pattern of xerophyte nsLTPs acting as flexible lipid shuttles that buffer membranes and the extracellular matrix against environmental extremes.\u003c/p\u003e \u003cp\u003eThe binding-site plasticity, species-specific paralogs, and codons under episodic rate variation support a scenario in which nsLTP diversification may contribute to functional modulation of lipid-binding properties, particularly through loop-mediated effects on tunnel accessibility. However, these evolutionary signals are interpreted here as indicators of structural and functional fine-tuning rather than direct evidence of ecological adaptation. Experimental validation will be required to determine whether these molecular differences translate into measurable physiological advantages. Lineage- and species-specific nsLTP variants, particularly those with altered loop architectures and re-shaped pockets, may underlie fine-scale differences in lipid composition of the cuticle and membranes, influencing traits such as transpirational control, thermal tolerance, and UV shielding (Tapia et al., 2013; Liu et al., 2022). Such molecular flexibility is consistent with the hypothesis that nsLTP diversification may contribute to the ecological breadth of \u003cem\u003eCereus\u003c/em\u003e, although this connection remains to be tested experimentally. This could be a mechanistic explanation for the ability of \u003cem\u003eCereus\u003c/em\u003e species to occupy a wide biogeographic range, spanning distinct arid and semi-arid habitats, and suggests that nsLTPs are key components of the adaptive toolkit that enabled the genus to persist and diversify under harsh environmental conditions. However, no causal inference can be made without functional assays or comparative physiological data, and experimental analysis will be necessary to confirm it.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e4.4. Structural\u0026ndash;Functional\u0026ndash;Ecological Framework\u003c/h2\u003e \u003cp\u003eAcross \u003cem\u003eCereus\u003c/em\u003e, the β-sheets and α-helical core remain conserved, consistent with the canonical four-helix fold that stabilizes the internal hydrophobic tunnel (Missaoui et al., 2022). This conserved core scaffold ensures functional stability by preserving the protein's ability to maintain a hydrophobic cavity optimized for lipid encapsulation and transfer (Xiao et al., 2023). Superimposed on this stable core, variable loop chemistry, comprising insertions, deletions, and charged or polar residue substitutions, creates the basis for adaptive specificity. Loops at the rim of the binding pocket show sequence variability and are under episodic positive selection in \u003cem\u003eCereus\u003c/em\u003e, a pattern also observed in drought-tolerant \u003cem\u003eArabidopsis\u003c/em\u003e thaliana ecotypes and \u003cem\u003eOpuntia streptacantha\u003c/em\u003e, where nsLTPs modulate cuticular lipid deposition and membrane remodeling under desiccation stress (Rojas et al., 2019; Xiao et al., 2023). These flexible loops act as adaptive gates, allowing differential access of lipid ligands that vary in chain length or polarity, thereby fine-tuning transport and protection functions under environmental extremes.\u003c/p\u003e \u003cp\u003eThe positive selection observed at specific codons (\u0026asymp;\u0026thinsp;13% of the total) indicates localized adaptive fine-tuning rather than wholesale innovation. This pattern implies that \u003cem\u003eCereus\u003c/em\u003e nsLTPs evolve through small, targeted amino acid changes that alter binding-site electrostatics or dynamics without compromising the global fold (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB, C; McLaughlin and Tumer, 2025). Such diversification may reflect selective responses to environmental gradients in humidity, temperature, and UV exposure across the genus\u0026rsquo; wide geographic range. Thus, these findings show that nsLTPs exemplify a principle of molecular adaptation through constrained innovation, where they maintain structural conservation of the core for stability and folding efficiency, while exploiting loop-level plasticity and selective diversification to achieve ecological specialization.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e4.5. Biotechnological potential\u003c/h2\u003e \u003cp\u003eThe identification of pocket #1 in the nsLTP_AF3 model as a high-ligandability cavity for lipid-like molecules highlights the functional versatility of nsLTPs. Here, pocket scores are interpreted as descriptors of structural compatibility with physiologically relevant hydrophobic ligand. This consolidated cavity reflects a structurally optimized region for lipid binding and transport, consistent with the primary biological roles of plant nsLTPs.\u003c/p\u003e \u003cp\u003eThese findings align with previous crystallographic studies showing that surface binding of myristic acid near the N-terminal cavity induces conformational changes in loop-3, facilitating lipid internalization into the hydrophobic cavity, while binding near the C-terminal end induces only minor structural deviations (Lev, 2010; Madni et al., 2020). These studies also support a model of largely unidirectional lipid transfer, with lipids entering from the N-terminal end, moving through the hydrophobic cavity, and exiting at the C-terminal end, coordinated by loop-3 and C-terminal loop movements. Similar to the unexpected lipolytic activity reported for Cor a 8 in \u003cem\u003eC. avellana\u003c/em\u003e, which operates without a classical catalytic triad (Fissore et al., 2025), the presence of a structurally compact and accessible pocket indicates that nsLTPs may accommodate diverse ligands and potentially perform catalytic or transport functions that remain underexplored. While such findings expand the functional landscape of nsLTPs, the present study is focused on structural, evolutionary, and ligand-binding features of \u003cem\u003eCereus\u003c/em\u003e nsLTPs and does not provide experimental or computational evidence for catalytic activity. Therefore, any potential relationship between \u003cem\u003eCereus\u003c/em\u003e nsLTPs and lipolytic functions should be regarded as a working hypothesis, to be addressed in future studies combining biochemical assays and dynamic simulations.\u003c/p\u003e \u003cp\u003eKey residues such as Arg44 and Tyr79 in nsLTP1s (Finkina et al., 2016; Melnikova et al., 2018; Missaoui et al., 2022) participate in ligand binding, with additional residues stabilizing interactions with polar heads and aliphatic chains. Regulatory features, including calmodulin-binding regions and phosphorylation sites, may further modulate lipid uptake, linking nsLTP function to stress responses and plant adaptation (Chiu et al., 2020). These structural insights, combined with high stability and compact folding of nsLTPs, make them attractive scaffolds for in silico docking studies, molecular dynamics simulations, and potential synthetic biology or biomaterial applications.\u003c/p\u003e \u003cp\u003eFunctionally, nsLTPs contribute to both biotic and abiotic stress tolerance. Transgenic plants overexpressing nsLTPs demonstrate increased resistance to fungi, bacteria, and oomycetes, and show enhanced tolerance to drought, salinity, cold, and oxidative stress (McLaughlin et al., 2021; Xu et al., 2018). Their ability to act synergistically with other antimicrobial peptides, mediate cutin and wax deposition, and regulate trichome development underlines their versatile roles in plant defense and adaptation (Meng et al., 2018; Song et al., 2020).\u003c/p\u003e \u003cp\u003eTaken together, the structural, mechanistic, and physiological insights into nsLTPs suggest multiple avenues for future research. High-ligandability pockets, particularly in nsLTP_AF3, can be explored for ligand docking and bioactive compound screening. Molecular dynamics simulations could clarify the lipid transfer pathway, while precise gene-editing approaches may reveal the amino acids responsible for ligand specificity or stress-related functions. Such integrated studies could enable the development of nsLTP-based tools for synthetic biology, crop improvement, and novel biotechnological applications, including stress-resilient plants and targeted delivery systems.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eThis study provides the first comprehensive structural and evolutionary characterization of nsLTPs in the genus \u003cem\u003eCereus\u003c/em\u003e, revealing how these proteins may contribute to adaptation in arid and semi-arid environments. Our integrated approach combining comparative transcriptomics, homology-based modeling, binding-pocket analysis, and evolutionary inference demonstrates that \u003cem\u003eCereus\u003c/em\u003e nsLTPs preserve the canonical four-helix hydrophobic-tunnel architecture typical of the family, while exhibiting local flexibility at loop regions surrounding the binding pocket. This conserved-core yet adaptive-rim pattern supports functional tunability: loop variation and episodic positive selection act as fine-scale mechanisms to modulate ligand affinity, lipid specificity, and potentially stress-related signaling. Such structural\u0026ndash;functional plasticity likely enhances membrane protection and lipid trafficking under desiccation, heat, and UV stress, contributing to the ecological resilience and biogeographical breadth of the genus.\u003c/p\u003e \u003cp\u003eEighteen codon sites under episodic diversifying selection, together with species-specific paralogs, indicate ongoing adaptive refinement rather than radical innovation. These features suggest that nsLTP diversification in \u003cem\u003eCereus\u003c/em\u003e followed a model of constrained innovation, where small structural adjustments at the interface level generate functional diversity without compromising scaffold stability. Moreover, the identification of highly ligandable pockets points to potential biotechnological applications, as nsLTPs may serve as molecular scaffolds for lipid-binding, delivery systems, or bioactive compound transport.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by Funda\u0026ccedil;\u0026atilde;o de Amparao \u0026agrave; Pesquisa do Estado de S\u0026atilde;o Paulo (FAPESP/ n. 2024/19266-5 to MIOC, n. 2025/17270-8 to JAT, and n. 2023/05589-4 to DTA)\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors have no relevant financial or non-financial interests to disclose.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe idea for this study was conceived by MIOC and DTA. Data collection and analyses were performed by MIOC, JAT, TACS, and DTA. MIOC led the writing of the paper, while JAT and DTA contributed numerous conceptions and writing. All authors contributed to the intellectual development of the paper, made multiple revisions, and approved the final draft.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors acknowledge Dr. Livia Seno Ferreira Camargo for her valuable scientific input and guidance. Financial support from CNPq and FAPESP is gratefully acknowledged.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of generative AI and AI-assisted technologies in the writing process\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDuring the preparation of this work, the author(s) used ChatGPT 5.0 to improve English grammar. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAmaral, D.T., Minh\u0026oacute;s-Yano, I., Oliveira, J.V.M., Romeiro-Brito, M., Bonatelli, I.A.S., Taylor, N.P., Zappi, D.C., Moraes, E.M., Eaton, D.A.R., Franco, F.F., 2021. Tracking the xeric biomes of South America: the spatiotemporal diversification of mandacaru cactus. J. Biogeogr. 48 (12), 3085\u0026ndash;3103. https://doi.org/10.1111/jbi.14265.\u003c/li\u003e\n\u003cli\u003eAmaral, D.T., Trier, C., Copetti, D., Franco, F.F., Moraes, E.M., 2025a. Chromosome-level genome assembly of the iconic South American mandacaru cactus (Cereus jamacaru DC, Tribe Cereeae, Cactaceae). Planta 262 (2). https://doi.org/10.1007/s00425-025-04762-6.\u003c/li\u003e\n\u003cli\u003eAmaral, D.T., Oliveira, J.V.M., Moraes, E.M., Zappi, D.C., Taylor, N.P., Franco, F.F., 2025b. The potential distribution of Cereus (Cactaceae) species in scenarios of climate crises. J. Arid Environ. 226. https://doi.org/10.1016/j.jaridenv.2024.105285.\u003c/li\u003e\n\u003cli\u003eAndrade, C.T. da S., Marques, J.G.W., Zappi, D.C., 2006. The use of Cactaceae by sertanejos in Bahia, Brazil: connexive types to define utilitarian categories. Sitientibus S\u0026eacute;r. Ci\u0026ecirc;nc. Biol. 6 (Especial), 3\u0026ndash;12. https://doi.org/10.13102/scb8142.\u003c/li\u003e\n\u003cli\u003eBombonato, J.R., do Amaral, D.T., Silva, G.A.R., Khan, G., Moraes, E.M., da Silva Andrade, S.C., Eaton, D.A.R., Alonso, D.P., Ribolla, P.E.M., Taylor, N., Zappi, D., Franco, F.F., 2020. The potential of genome-wide RAD sequences for resolving rapid radiations: a case study in Cactaceae. Mol. Phylogenet. Evol. 151, 106896. https://doi.org/10.1016/j.ympev.2020.106896.\u003c/li\u003e\n\u003cli\u003eBrenke, R., Kozakov, D., Chuang, G.Y., Beglov, D., Hall, D., Landon, M.R., Mattos, C., Vajda, S., 2009. Fragment-based identification of druggable \u0026ldquo;hot spots\u0026rdquo; of proteins using Fourier domain correlation techniques. Bioinformatics 25 (5), 621\u0026ndash;627. https://doi.org/10.1093/bioinformatics/btp036.\u003c/li\u003e\n\u003cli\u003eBrissos, V., Borges, P.T., Sancho, F., Lucas, M.F., Fraz\u0026atilde;o, C., Conzuelo, F., Martins, L.O., 2024. Flexible active-site loops fine-tune substrate specificity of hyperthermophilic metallo-oxidases. J. Biol. Inorg. Chem. 29, 339\u0026ndash;351. https://doi.org/10.1007/s00775-023-02040-y.\u003c/li\u003e\n\u003cli\u003eCardoso, M.I.O., Oliveira, J., Bassaneze, V., Amaral, D.T., 2026. Discovery of triacylglycerol-binding proteins in Cereus jamacaru DC. through integrative in silico methods. J. Chem. 2026, 5448761. https://doi.org/10.1155/joch/5448761.\u003c/li\u003e\n\u003cli\u003eChiu, L.-Y., Chen, I.-H., Hsu, Y.-H., Tsai, C.-H., 2020. The lipid transfer protein 1 from Nicotiana benthamiana assists Bamboo mosaic virus accumulation. Viruses 12, 1361. https://doi.org/10.3390/v12121361.\u003c/li\u003e\n\u003cli\u003eda Silva, N.F., Hanazaki, N., Albuquerque, U.P., Almeida Campos, J.L., Feitosa, I.S., Ara\u0026uacute;jo, E.L., 2019. Local knowledge and conservation priorities of medicinal plants near a protected area in Brazil. Evid.-Based Complement. Altern. Med. 2019, 8275084. https://doi.org/10.1155/2019/8275084.\u003c/li\u003e\n\u003cli\u003eD\u0026rsquo;Agostino, N., Buonanno, M., Ayoub, J., Barone, A., Monti, S.M., Rigano, M.M., 2019. Identification of non-specific lipid transfer protein gene family members in Solanum lycopersicum and insights into the features of Sola l 3 protein. Sci. Rep. 9, 1\u0026ndash;12. https://doi.org/10.1038/s41598-018-38301-z.\u003c/li\u003e\n\u003cli\u003eDeeken, R., Saupe, S., Klinkenberg, J., Riedel, M., Leide, J., Hedrich, R., Mueller, T.D., 2016. The nonspecific lipid transfer protein AtLtpI-4 is involved in suberin formation of Arabidopsis thaliana crown galls. Plant Physiology 172(3), 1911\u0026ndash;1927. https://doi.org/10.1104/pp.16.01486.\u003c/li\u003e\n\u003cli\u003eEdstam, M. M., Viitanen, L., Salminen, T. A., Edqvist, J., 2011. Evolutionary history of the non-specific lipid transfer proteins. In Molecular Plant (Vol. 4, Issue 6, pp. 947\u0026ndash;964). Oxford University Press. https://doi.org/10.1093/mp/ssr019. \u003c/li\u003e\n\u003cli\u003eEgger, M., Hauser, M., Mari, A., Ferreira, F., Gadermaier, G., 2010. The role of lipid transfer proteins in allergic diseases. Curr. Allergy Asthma Rep. 10 (5), 326\u0026ndash;335. https://doi.org/10.1007/s11882-010-0128-9.\u003c/li\u003e\n\u003cli\u003eEmms, D.M., Kelly, S., 2019. OrthoFinder: phylogenetic orthology inference for comparative genomics. Genome Biol. 20, 238. https://doi.org/10.1186/s13059-019-1832-y.\u003c/li\u003e\n\u003cli\u003eFinkina, E.I., Melnikova, D.N., Bogdanov, I.V., Ovchinnikova, T.V., 2016. Lipid transfer proteins as components of the plant innate immune system: Structure, functions, and applications. Acta Naturae 8(2), 47\u0026ndash;61. https://doi.org/10.32607/20758251-2016-8-2-47-61.\u003c/li\u003e\n\u003cli\u003eFissore, A., di Napoli, G., de Sciscio, M.L., Santoro, V., Salladini, E., Marengo, M., Oliaro-Bosso, S., Vanzetti, G., Caratti, A., Dal Piaz, F., Piazza, F., Manzoli, M., Genova, G., Barbiroli, A., Iametti, S., Fraternali, F., Adinolfi, S., 2025. Unravelling a new enzymatic activity of non-specific lipid transfer protein Cor a 8: A novel lipase from hazelnuts. Int. J. Biol. Macromol. 322, 146537. https://doi.org/10.1016/j.ijbiomac.2025.146537.\u003c/li\u003e\n\u003cli\u003eFleury, C., Gracy, J., Gautier, M.F., Pons, J.L., Dufayard, J.F., Labesse, G., Ruiz, M., de Lamotte, F., 2019. Comprehensive classification of the plant non-specific lipid transfer protein superfamily towards its sequence-structure-function analysis. PeerJ 7, e7504. https://doi.org/10.7717/peerj.7504.\u003c/li\u003e\n\u003cli\u003eFranco, F.F., Jojima, C.L., Perez, M.F., Zappi, D.C., Taylor, N., Moraes, E.M., 2017. The xeric side of the Brazilian Atlantic Forest: The forces shaping phylogeographic structure of cacti. Ecology and Evolution 7(22), 9281\u0026ndash;9293. https://doi.org/10.1002/ece3.3458.\u003c/li\u003e\n\u003cli\u003eGincel, E., Simorre, J.P., Caille, A., Marion, D., Ptak, M., Vovelle, F., 1994. Three-dimensional structure in solution of a wheat lipid-transfer protein from multidimensional \u0026sup1;H-NMR data: a new folding for lipid carriers. Eur. J. Biochem. 226, 413\u0026ndash;422. https://doi.org/10.1111/j.1432-1033.1994.tb20066.x.\u003c/li\u003e\n\u003cli\u003eGowder, S.M., Chatterjee, J., Chaudhuri, T., Paul, K., 2014. Prediction and analysis of surface hydrophobic residues in tertiary structure of proteins. Sci. World J. 2014, 971258. https://doi.org/10.1155/2014/971258.\u003c/li\u003e\n\u003cli\u003eHall, T., Biosciences, I., \u0026amp; Carlsbad, C. J. G. B. B. (2011). BioEdit: an important software for molecular biology. GERF Bull Biosci, 2(1), 60-61.\u003c/li\u003e\n\u003cli\u003eHama\u0026iuml;, A., Drin, G., 2024. Specificity of lipid transfer proteins: An in vitro story. Biochimie 227, 85\u0026ndash;110. https://doi.org/10.1016/j.biochi.2024.09.007.\u003c/li\u003e\n\u003cli\u003eHasegawa, H., Holm, L., 2009. Advances and pitfalls of protein structural alignment. Curr. Opin. Struct. Biol. 19, 341\u0026ndash;348. https://doi.org/10.1016/j.sbi.2009.04.003.\u003c/li\u003e\n\u003cli\u003eHunt, D.R., Taylor, N.P., Charles, G., 2006. The New Cactus Lexicon. dh Books.\u003c/li\u003e\n\u003cli\u003eJones, G., Jindal, A., Ghani, U., Kotelnikov, S., Egbert, M., Hashemi, N., Vajda, S., Padhorny, D., Kozakov, D., 2022. Elucidation of protein function using computational docking and hotspot analysis by ClusPro and FTMap. Acta Crystallogr. D Struct. Biol. 78 (6), 690\u0026ndash;697. https://doi.org/10.1107/S2059798322002741.\u003c/li\u003e\n\u003cli\u003eJumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Ž\u0026iacute;dek, A., Potapenko, A., et al., 2021. Highly accurate protein structure prediction with AlphaFold. Nature 596 (7873), 583\u0026ndash;589. https://doi.org/10.1038/s41586-021-03819-2.\u003c/li\u003e\n\u003cli\u003eKader, J.-C., 1996. Lipid-transfer proteins in plants. \u003cem\u003eAnnual Review of Plant Biology\u003c/em\u003e 47, 627\u0026ndash;654. https://doi.org/10.1146/annurev.arplant.47.1.627. \u003c/li\u003e\n\u003cli\u003eKatoh, K., Standley, D.M., 2013. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol. Biol. Evol. 30 (4), 772\u0026ndash;780. https://doi.org/10.1093/molbev/mst010.\u003c/li\u003e\n\u003cli\u003eKochnev, Y., Durrant, J.D., 2022. FPocketWeb: protein pocket hunting in a web browser. J. Cheminform. 14, 58. https://doi.org/10.1186/s13321-022-00637-0.\u003c/li\u003e\n\u003cli\u003eKozakov, D., Grove, L.E., Hall, D.R., Bohnuud, T., Mottarella, S.E., Luo, L., Xia, B., Beglov, D., Vajda, S., 2015. The FTMap family of web servers for determining and characterizing ligand-binding hot spots of proteins. Nat. Protoc. 10 (5), 733\u0026ndash;755. https://doi.org/10.1038/nprot.2015.043.\u003c/li\u003e\n\u003cli\u003eKrokidis, M.G., Koumadorakis, D.E., Lazaros, K., Ivantsik, O., Exarchos, T.P., Vrahatis, A.G., Kotsiantis, S., Vlamos, P., 2025. AlphaFold3: An overview of applications and performance insights. Int. J. Mol. Sci. 26, 3671. https://doi.org/10.3390/ijms26083671.\u003c/li\u003e\n\u003cli\u003eLaskowski, R.A., MacArthur, M.W., Moss, D.S., Thornton, J.M., 1993. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26 (2), 283\u0026ndash;291. https://doi.org/10.1107/s0021889892009944.\u003c/li\u003e\n\u003cli\u003eLev, S., 2010. Non-vesicular lipid transport by lipid-transfer proteins and beyond. Nat. Rev. Mol. Cell Biol. 11, 739\u0026ndash;750. https://doi.org/10.1038/nrm2971.\u003c/li\u003e\n\u003cli\u003eLi, W., Godzik, A., 2006. Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics 22 (13), 1658\u0026ndash;1659. https://doi.org/10.1093/bioinformatics/btl158.\u003c/li\u003e\n\u003cli\u003eLiu, L., Wang, X., Chang, C., 2022. Toward a smart skin: Harnessing cuticle biosynthesis for crop adaptation to drought, salinity, temperature, and ultraviolet stress. Front. Plant Sci. 13, 961829. https://doi.org/10.3389/fpls.2022.961829.\u003c/li\u003e\n\u003cli\u003eLucena, C.M., Lucena, R.F., Costa, G.M., Carvalho, T.K., Costa, G.G., Alves, R.R., Pereira, D.D., Ribeiro, J.E., Alves, C.A., Quirino, Z.G., Nunes, E.N., 2013. Use and knowledge of Cactaceae in Northeastern Brazil. J. Ethnobiol. Ethnomed. 9 (1), 62. https://doi.org/10.1186/1746-4269-9-62.\u003c/li\u003e\n\u003cli\u003eMalinina, L., Patel, D.J., Brown, R.E., 2017. How \u0026alpha;-Helical Motifs Form Functionally Diverse Lipid-Binding Compartments. Annu. Rev. Biochem. 86, 609\u0026ndash;636. https://doi.org/10.1146/annurev-biochem-061516-044445.\u003c/li\u003e\n\u003cli\u003eMadni, Z.K., Tripathi, S.K., Salunke, D.M., 2020. Structural insights into the lipid transfer mechanism of a non-specific lipid transfer protein. Plant J. 102, 340\u0026ndash;352. https://doi.org/10.1111/tpj.14627.\u003c/li\u003e\n\u003cli\u003eMcLaughlin, J.E., Darwish, N.I., Garcia-Sanchez, J., Tyagi, N., Trick, H.N., McCormick, S., Dill-Macky, R., Tumer, N.E., 2021. A lipid transfer protein has antifungal and antioxidant activity and suppresses fusarium head blight disease and DON accumulation in transgenic wheat. Phytopathology 111, 671\u0026ndash;683. https://doi.org/10.1094/PHYTO-04-20-0153-R.\u003c/li\u003e\n\u003cli\u003eMcLaughlin, J.E., Tumer, N.E., 2025. Roles of non-specific lipid transfer proteins in plant defense: structural and functional perspectives. Front. Fungal Biol. 6, 1640465. https://doi.org/10.3389/ffunb.2025.1640465.\u003c/li\u003e\n\u003cli\u003eMelnikova, D.N., Finkina, E.I., Bogdanov, I.V., et al., 2018. Plant pathogenesis-related proteins binding lipids and other hydrophobic ligands. Russian Journal of Bioorganic Chemistry 44, 586\u0026ndash;594. https://doi.org/10.1134/S1068162018060055.\u003c/li\u003e\n\u003cli\u003eMelnikova, D.N., Finkina, E.I., Bogdanov, I.V., Tagaev, A.A., Ovchinnikova, T.V., 2022. Features and possible applications of plant lipid-binding and transfer proteins. Membranes 13, 2. https://doi.org/10.3390/membranes13010002.\u003c/li\u003e\n\u003cli\u003eMeng, C., Yan, Y., Liu, Z., Chen, L., Zhang, Y., Li, X., Wu, L., Zhang, G., Wang, X., Ma, Z., 2018. Systematic analysis of cotton non-specific lipid transfer protein family revealed a special group that is involved in fiber elongation. Frontiers in Plant Science 9, 1285. https://doi.org/10.3389/fpls.2018.01285.\u003c/li\u003e\n\u003cli\u003eMinh, B.Q., Schmidt, H.A., Chernomor, O., Schrempf, D., Woodhams, M.D., von Haeseler, A., Lanfear, R., 2020. IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era. Mol. Biol. Evol. 37 (5), 1530\u0026ndash;1534. https://doi.org/10.1093/molbev/msaa015.\u003c/li\u003e\n\u003cli\u003eMissaoui, K., Gonzalez-Klein, Z., Pazos-Castro, D., Hernandez-Ramirez, G., Garrido-Arandia, M., Brini, F., Diaz-Perales, A., Tome-Amat, J., 2022. Plant non-specific lipid transfer proteins: An overview. Plant Physiol. Biochem. 171, 115\u0026ndash;127. https://doi.org/10.1016/j.plaphy.2021.12.026.\u003c/li\u003e\n\u003cli\u003eMurrell, B., Wertheim, J.O., Moola, S., Weighill, T., Scheffler, K., Kosakovsky Pond, S.L., 2012. Detecting individual sites subject to episodic diversifying selection. PLoS Genet. 8 (7), e1002764. https://doi.org/10.1371/journal.pgen.1002764.\u003c/li\u003e\n\u003cli\u003eNgan, C.H., Hall, D.R., Zerbe, B.S., Grove, L.E., Kozakov, D., Vajda, S., 2012. FTSite: high accuracy detection of ligand binding sites on unbound protein structures. Bioinformatics 28 (2), 286\u0026ndash;287. https://doi.org/10.1093/bioinformatics/btr651.\u003c/li\u003e\n\u003cli\u003ePalheta, I.C., Tavares-Martins, A.C.C., Lucas, F.C.A., Jardim, M.A.G., 2017. Ethnobotanical study of medicinal plants in urban home gardens in the city of Abaetetuba, Par\u0026aacute; state, Brazil. Bol. Latinoam. Caribe Plant. Med. Aromat. 16 (3), 206\u0026ndash;262.\u003c/li\u003e\n\u003cli\u003eRodrigues Almeida, L.M., Gonzaga Fernandez, L., 2025. Ethnomedical uses, biocompounds and biological properties of Cereus jamacaru DC. (Cactaceae): a comprehensive review. Nat. Prod. Res. 39 (4), 961\u0026ndash;975. https://doi.org/10.1080/14786419.2024.2330541.\u003c/li\u003e\n\u003cli\u003eRojas, M., Jimenez-Bremont, F., Villica\u0026ntilde;a, C., Carre\u0026oacute;n-Palau, L., Arredondo-Vega, B.O., G\u0026oacute;mez-Anduro, G., 2019. Involvement of OpsLTP1 from Opuntia streptacantha in abiotic stress adaptation and lipid metabolism. Funct. Plant Biol. 46 (9), 816\u0026ndash;829. https://doi.org/10.1071/FP18280.\u003c/li\u003e\n\u003cli\u003eSalminen, T.A., Blomqvist, K., Edqvist, J., 2016. Lipid transfer proteins: classification, nomenclature, structure, and function. Planta 244, 971\u0026ndash;997. https://doi.org/10.1007/s00425-016-2585-4.\u003c/li\u003e\n\u003cli\u003eSantos-Silva, C.A.d., Ferreira-Neto, J.R.C., Amador, V.C., Bezerra-Neto, J.P., Vilela, L.M.B., Binneck, E., R\u0026ecirc;go, M.d.S., da Silva, M.D., Mangueira de Melo, A.L.T., da Silva, R.H., Benko-Iseppon, A.M., 2023. From Gene to Transcript and Peptide: A Deep Overview on Non-Specific Lipid Transfer Proteins (nsLTPs). Antibiotics 12, 939. https://doi.org/10.3390/antibiotics12050939.\u003c/li\u003e\n\u003cli\u003eShenkarev, Z.O., Melnikova, D.N., Finkina, E.I., Sukhanov, S.V., Boldyrev, I.A., Gizatullina, A.K., Mineev, K.S., Arseniev, A.S., Ovchinnikova, T.V., 2017. Ligand binding properties of the lentil lipid transfer protein: molecular insight into the possible mechanism of lipid uptake. Biochemistry 56, 1785\u0026ndash;1796. https://doi.org/10.1021/acs.biochem.6b01079.\u003c/li\u003e\n\u003cli\u003eSimorre, J.P., Caille, A., Marion, D., Marion, D., Ptak, M., 1991. Two- and three-dimensional proton NMR studies of a wheat phospholipid transfer protein: sequential resonance assignments and secondary structure. Biochemistry 30, 11600\u0026ndash;11608. https://doi.org/10.1021/bi00113a016.\u003c/li\u003e\n\u003cli\u003eSong, X., Li, E., Song, H., Du, G., Li, S., Zhu, H., Chen, G., Zhao, C., Qiao, L., Wang, J., Yu, S., Sui, J.M., 2020. Genome-wide identification and characterization of nonspecific lipid transfer protein (nsLTP) genes in Arachis duranensis. Genomics 112(6), 4332\u0026ndash;4341. https://doi.org/10.1016/j.ygeno.2020.07.034.\u003c/li\u003e\n\u003cli\u003eSrinivasan, S., Di Luca, A., \u0026Aacute;lvarez, D., John Peter, A.T., Gehin, C., Lone, M.A., Hornemann, T., D\u0026apos;Angelo, G., Vanni, S., 2024. The conformational plasticity of structurally unrelated lipid transport proteins correlates with their mode of action. PLoS Biol. 22, e3002737. https://doi.org/10.1371/journal.pbio.3002737.\u003c/li\u003e\n\u003cli\u003eStolzer, M., Lai, H., Xu, M., Sathaye, D., Vernot, B., \u0026amp; Durand, D. (2012). Inferring duplications, losses, transfers and incomplete lineage sorting with nonbinary species trees. Bioinformatics (Oxford, England), 28(18), i409\u0026ndash;i415. https://doi.org/10.1093/bioinformatics/bts386. \u003c/li\u003e\n\u003cli\u003eTapia, G., Morales-Quintana, L., Parra, C., Berbel, A., Alcorta, M., 2013. Study of nsLTPs in Lotus japonicus genome reveal a specific epidermal cell member (LjLTP10) regulated by drought stress in aerial organs with a putative role in cutin formation. Plant Mol. Biol. 82, 485\u0026ndash;501. https://doi.org/10.1007/s11103-013-0080-x.\u003c/li\u003e\n\u003cli\u003eTeodoro, J.A., Senra, M.V.X., Amaral, D.T., 2025. In silico bioprospecting of the Neotropical plant mandacaru (Cereus) for antimicrobial properties. Probiotics Antimicrob. Proteins. https://doi.org/10.1007/s12602-025-10580-9.\u003c/li\u003e\n\u003cli\u003eVelesinović, A., Nikolić, G., 2021. Protein-protein interaction networks and protein-ligand docking: Contemporary insights and future perspectives. Acta Fac. Med. Naissensis 38, 5\u0026ndash;17. https://doi.org/10.5937/AFMNAI38-28322.\u003c/li\u003e\n\u003cli\u003eWeaver, S., Shank, S.D., Spielman, S.J., Li, M., Muse, S.V., Kosakovsky Pond, S.L., 2018. Datamonkey 2.0: a modern web application for characterizing selective and other evolutionary processes. Mol. Biol. Evol. 35 (3), 773\u0026ndash;777. https://doi.org/10.1093/molbev/msx335.\u003c/li\u003e\n\u003cli\u003eWebb, B., Sali, A., 2016. Comparative protein structure modeling using MODELLER. Curr. Protoc. Protein Sci. 86, 2.9.1\u0026ndash;2.9.37. https://doi.org/10.1002/cpps.20.\u003c/li\u003e\n\u003cli\u003eXiao, Y., Xiao, C., He, X., Yang, X., Tong, Z., Wang, Z., Sun, Z., Qiu, W., 2023. A novel non-specific lipid transfer protein gene, CmnsLTP6.9, enhanced osmotic and drought tolerance by regulating ROS scavenging and remodeling lipid profiles in Chinese chestnut (Castanea mollissima Blume). Plants 12, 3916. https://doi.org/10.3390/plants12223916.\u003c/li\u003e\n\u003cli\u003eXu, Y., Zheng, X., Song, Y., Zhu, L., Yu, Z., Gan, L., Zhou, S., Liu, H., Wen, F., Zhu, C., 2018. NtLTP4, a lipid transfer protein that enhances salt and drought stresses tolerance in Nicotiana tabacum. Scientific Reports 8, 8873. https://doi.org/10.1038/s41598-018-27274-8.\u003c/li\u003e\n\u003cli\u003eXue, Y., Zhang, C., Shan, R., Li, X., Tseke Inkabanga, A., Li, L., Jiang, H., Chai, Y., 2022. Genome-wide identification and expression analysis of nsLTP gene family in rapeseed (Brassica napus) reveals their critical roles in biotic and abiotic stress responses. Int. J. Mol. Sci. 23, 8372. https://doi.org/10.3390/ijms23158372.\u003c/li\u003e\n\u003cli\u003eYeats, T. H., Rose, J. K. C., 2008. The biochemistry and biology of extracellular plant lipid‐transfer proteins (LTPs). \u003cem\u003eProtein Science\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(2), 191\u0026ndash;198. https://doi.org/10.1110/ps.073300108. \u003c/li\u003e\n\u003cli\u003eYuan, R., Zhang, J., Zhou, J., Cong, Q., 2025. Recent progress and future challenges in structure-based protein-protein interaction prediction. Mol. Ther. 33, 2252\u0026ndash;2268. https://doi.org/10.1016/j.ymthe.2025.04.003.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Cereus, gene duplication, lipid transfer proteins, protein structure, stress response","lastPublishedDoi":"10.21203/rs.3.rs-9269266/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9269266/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe genus \u003cem\u003eCereus\u003c/em\u003e (Cactaceae), widely distributed across arid and semi-arid environments of South America, exhibits adaptive capacity that is partly associated with proteins involved in lipid metabolism and stress responses, such as non-specific lipid transfer proteins (nsLTPs). Although well-characterized in model plants, the structural diversity and functional roles of these proteins in cacti remain largely unexplored. In this study, we characterized the nsLTP repertoire of \u003cem\u003eCereus\u003c/em\u003e by integrating transcriptomic data, phylogenetic analyses, structural modeling, and ligand-binding site mapping. Structural comparisons revealed conservation of the typical nsLTP fold, with variations mainly localized to loop regions surrounding the binding cavity, suggesting functional plasticity without compromising scaffold stability. Phylogenetic analyses revealed frequent duplication and loss events, as well as species-specific paralogs. Several lineage-specific copies exhibited distinctive loop architectures and pocket geometries relative to their closest orthologs, suggesting rapid structural diversification following duplication. This pattern is consistent with functional fine-tuning after gene duplication. Positive selection analyses identified 18 codons under episodic diversifying selection, many in spatial proximity to predicted functional regions, supporting the hypothesis that selective pressures have shaped key interaction interfaces. Together, these results provide the first integrated characterization of nsLTPs in \u003cem\u003eCereus\u003c/em\u003e, supporting a gate-modulation mechanistic model in which loop-level variation reshapes tunnel access and local physicochemical compatibility while preserving the conserved four-helix scaffold. Rather than asserting adaptation directly, our findings provide a structural basis and testable predictions for how nsLTP diversification may influence lipid trafficking processes relevant to cuticle/membrane function under abiotic stress.\u003c/p\u003e","manuscriptTitle":"Structural and evolutionary analysis of non-specific lipid transfer proteins (nsLTPs) in Cereus (Cactaceae)","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-17 08:57:42","doi":"10.21203/rs.3.rs-9269266/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"1e3cf0a4-7166-4ebb-8f62-53f907d6fb42","owner":[],"postedDate":"April 17th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-17T08:57:43+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-17 08:57:42","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9269266","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9269266","identity":"rs-9269266","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}