(Tyr0)-fibrinopeptide A is a variant of fibrinopeptide A, which is a small polypeptide derived from the cleavage of fibrinogen by thrombin. Fibrinogen, a glycoprotein produced primarily in the liver, plays a crucial role in blood coagulation. The molecular structure of (Tyr0)-fibrinopeptide A includes a phenylalanine residue at the N-terminus, which distinguishes it from the standard fibrinopeptide A. This compound is significant in the context of hemostasis and thrombosis, as it serves as a sensitive biochemical marker for these processes .
(Tyr0)-Fibrinopeptide A itself doesn't have a specific mechanism of action. It serves as a marker for the enzymatic activity of thrombin, which plays a crucial role in blood clotting. By measuring its concentration, researchers can indirectly assess clot formation.
This reaction is essential for initiating the fibrin formation process during blood clotting, where fibrinogen is converted into fibrin strands that stabilize blood clots .
(Tyr0)-fibrinopeptide A exhibits biological activity primarily through its role in coagulation. It acts as a marker for thrombin activity and reflects the state of coagulation in vivo. Elevated levels of (Tyr0)-fibrinopeptide A can indicate thrombotic disorders or ongoing clotting processes. Additionally, it has been studied for its potential involvement in inflammatory responses and vascular biology, although these roles require further investigation .
The synthesis of (Tyr0)-fibrinopeptide A can be achieved through several methods:
These methods allow for the production of (Tyr0)-fibrinopeptide A with high purity and specificity for research and clinical applications .
(Tyr0)-fibrinopeptide A has several applications in both clinical and research settings:
Studies on the interactions involving (Tyr0)-fibrinopeptide A focus on its binding with various proteins involved in hemostasis. For instance:
(Tyr0)-fibrinopeptide A shares structural and functional similarities with several other peptides derived from fibrinogen. Here are some comparable compounds:
| Compound Name | Structure/Function | Unique Features |
|---|---|---|
| Fibrinopeptide B | Cleaved from fibrinogen by thrombin | Longer chain than Fibrinopeptide A |
| Des-fibrinopeptide A | Lacks one or more amino acids from FPA | Less active than Fibrinopeptide A |
| Fibrinogen | Precursor protein for fibrin formation | Larger glycoprotein that contains both FPA and FPB |
Uniqueness of (Tyr0)-fibrinopeptide A: Its specific amino acid composition allows it to serve as a more sensitive indicator of thrombin activity compared to other variants. Additionally, its unique interaction profile with thrombin sets it apart in studies related to coagulation dynamics .
Solid-phase peptide synthesis represents the established methodology for producing (Tyr0)-fibrinopeptide A and related tyrosine-modified analogues [10] [11]. The synthesis strategy employs the fluorenylmethyloxycarbonyl/tert-butyl (Fmoc/tBu) protection scheme, which has become the most widely utilized approach for peptide synthesis due to its compatibility with a broad range of amino acid derivatives [11] [12]. The synthesis process involves successive coupling of protected amino acid derivatives to a growing peptide chain attached to an insoluble polymeric resin support [10] [16].
For tyrosine-modified analogues, specific considerations must be addressed during the synthesis process [13] [15]. The incorporation of tyrosine residues requires careful selection of protecting group strategies to prevent unwanted side reactions during chain assembly [15] [17]. Traditional tyrosine protection employs the tert-butyl group (Fmoc-Tyr(tBu)-OH), which provides adequate stability under standard synthesis conditions [10] [15]. However, for specialized applications requiring post-translational modifications, alternative protection schemes such as fluorosulfate esters (Fmoc-Tyr(OSO2F)-OH) have been developed [17].
The coupling efficiency for tyrosine-containing sequences can be enhanced through optimization of coupling reagents and reaction conditions [11] [12]. Aminium-derived coupling reagents such as HBTU, HATU, and PyBOP demonstrate superior performance for challenging sequences, with HATU showing particular effectiveness for tyrosine incorporation [12] [15]. The use of 2-methylpiperidine for Fmoc deprotection has been shown to minimize side reactions with certain tyrosine derivatives compared to conventional piperidine [17].
Table 1: Solid-Phase Peptide Synthesis Optimization Parameters for Tyrosine-Modified Analogues
| Parameter | Standard SPPS | Tyrosine-Modified Protocol | Optimization Range |
|---|---|---|---|
| Resin Type | Wang Resin | Rink Amide MBHA | Various linkers tested |
| Coupling Reagent | HBTU/HOBt | HATU/DIPEA | COMU, PyBOP alternatives |
| Deprotection Reagent | 20% Piperidine/DMF | 2-Methylpiperidine/DMF | 10-25% concentration |
| Tyrosine Protection Strategy | Fmoc-Tyr(tBu)-OH | Fmoc-Tyr(OSO2F)-OH | Multiple protection schemes |
| Coupling Time (min) | 60 | 90-120 | 30-180 |
| Deprotection Time (min) | 20 | 25 | 10-30 |
| Coupling Efficiency (%) | 95-97 | 98-99 | 95-99.5 |
| Amino Acid Excess (equiv.) | 3-5 | 4-6 | 2-8 |
| Washing Protocol (cycles) | 6-8 | 8-10 | 4-12 |
| Temperature (°C) | 25 | 25 | 20-40 |
Aggregation represents a significant challenge in peptide synthesis, particularly for sequences containing aromatic amino acids like tyrosine [11] [12]. The hydrophobic interactions between tyrosine residues can lead to secondary structure formation during chain assembly, resulting in reduced coupling efficiency and formation of deletion sequences [12] [44]. Strategies to minimize aggregation include the use of pseudoproline and depsipeptide units at strategic positions, elevated synthesis temperatures, and chaotropic solvents [12].
The theoretical yield calculations for peptide synthesis demonstrate the critical importance of maintaining high coupling efficiency at each step [44]. For a 17-residue peptide such as (Tyr0)-fibrinopeptide A, a coupling efficiency of 97% at each step would result in an overall yield of approximately 59%, while increasing efficiency to 99% dramatically improves the overall yield to 84% [44]. This relationship underscores the necessity for optimization of synthesis parameters, particularly for tyrosine-containing sequences that may exhibit reduced coupling efficiency [11] [44].
High-performance liquid chromatography represents the primary methodology for purification and purity assessment of (Tyr0)-fibrinopeptide A [18] [19]. Reverse-phase high-performance liquid chromatography (RP-HPLC) demonstrates particular effectiveness for peptide separations due to its ability to resolve compounds based on hydrophobic interactions with the stationary phase [21] [42]. The separation mechanism relies on the differential partitioning of peptides between a hydrophobic stationary phase, typically C18-bonded silica, and a polar mobile phase consisting of water and acetonitrile [21] [25].
The optimization of high-performance liquid chromatography methods for (Tyr0)-fibrinopeptide A purification requires systematic evaluation of multiple parameters [21] [25]. Column selection plays a crucial role, with C18 reverse-phase columns demonstrating superior performance for peptide separations [21] [42]. Particle size significantly influences resolution and analysis time, with smaller particles (3-5 micrometers) providing enhanced resolution for analytical applications, while larger particles (10-15 micrometers) are preferred for preparative-scale purifications [25].
Mobile phase composition and gradient optimization represent critical factors in achieving effective separations [18] [22]. The standard mobile phase system employs 0.1% trifluoroacetic acid in water as solvent A and 0.1% trifluoroacetic acid in acetonitrile as solvent B [19] [21]. Trifluoroacetic acid serves as an ion-pairing reagent that improves peak symmetry and maximizes retention of acidic peptides [21]. Gradient optimization typically involves shallow gradients with slopes of 1-4% organic solvent per minute to achieve adequate resolution of closely related impurities [21] [25].
Table 2: High-Performance Liquid Chromatography Purification Parameters and Optimization Data
| Parameter | Analytical Scale | Semi-Preparative | Preparative Scale |
|---|---|---|---|
| Column Type | C18 reversed-phase | C18 reversed-phase | C18 reversed-phase |
| Particle Size (μm) | 3-5 | 5-10 | 10-15 |
| Mobile Phase A | 0.1% TFA in H2O | 0.1% TFA in H2O | 0.1% TFA in H2O |
| Mobile Phase B | 0.1% TFA in ACN | 0.1% TFA in ACN | 0.1% TFA in ACN |
| Flow Rate (mL/min) | 1.0 | 5-10 | 20-50 |
| Gradient Slope (%B/min) | 1-2 | 0.5-1 | 0.2-0.5 |
| Column Temperature (°C) | 30 | 30 | 30 |
| Detection Wavelength (nm) | 214, 280 | 214, 280 | 214, 280 |
| Injection Volume (μL) | 10-20 | 100-500 | 1000-5000 |
| Run Time (min) | 30-60 | 60-120 | 120-180 |
Purity assessment of (Tyr0)-fibrinopeptide A requires multiple analytical approaches to ensure comprehensive characterization [23] [24]. High-performance liquid chromatography provides quantitative purity determination through ultraviolet detection at 214 nanometers, which corresponds to the peptide bond absorption [20] [23]. The purity calculation is based on peak area integration, with the target peptide peak area expressed as a percentage of the total integrated area [23]. Commercial preparations of (Tyr0)-fibrinopeptide A typically achieve purities greater than 95% by high-performance liquid chromatography [1] [2].
Mass spectrometry serves as a complementary technique for molecular weight confirmation and structural verification [20] [23]. Electrospray ionization mass spectrometry provides precise molecular weight determination, allowing confirmation of the expected mass of 1699.76 daltons for (Tyr0)-fibrinopeptide A [1] [5]. This technique also enables detection of potential impurities arising from incomplete deprotection, side reactions, or degradation processes [23].
Table 3: Peptide Purity Assessment Methods and Criteria
| Method | Detection Principle | Purity Range (%) | Typical Application | Detection Limit |
|---|---|---|---|---|
| RP-HPLC | UV absorption at 214 nm | 85-99.5 | Routine purity analysis | 0.1-1% |
| Mass Spectrometry (ESI-MS) | Mass-to-charge ratio | 90-99.9 | Molecular weight confirmation | 0.01-0.1% |
| Amino Acid Analysis | Ninhydrin derivatization | 80-95 | Quantitative analysis | 1-5% |
| Capillary Electrophoresis | Charge-based separation | 90-99 | Charge variant analysis | 0.5-2% |
| Analytical Ultracentrifugation | Sedimentation behavior | 85-98 | Aggregation studies | 1-3% |
| NMR Spectroscopy | Nuclear magnetic resonance | 95-99.9 | Structural confirmation | 0.1-1% |
Scale-up considerations for preparative high-performance liquid chromatography require careful attention to maintain separation quality while increasing throughput [25]. Linear scale-up principles dictate that flow rates, injection volumes, and gradient slopes must be adjusted proportionally to column dimensions [25]. The relationship between analytical and preparative conditions follows established scaling factors, with flow rates typically increasing by the square of the column diameter ratio [25].
Surface plasmon resonance spectroscopy has emerged as a pivotal technique for quantifying the binding interactions between (Tyrosine 0)-fibrinopeptide A and thrombin, providing real-time kinetic data essential for understanding the molecular recognition mechanisms underlying fibrinogen conversion [1] [2] [3]. The application of surface plasmon resonance to thrombin-fibrinopeptide interactions has revealed critical insights into the binding dynamics that govern the specificity and efficiency of fibrinolytic processes.
Comprehensive surface plasmon resonance investigations have demonstrated that fibrinogen-derived fragments exhibit varying binding affinities depending on their structural organization and orientation. Studies utilizing desaminocarboxypeptidase A-treated N-terminal disulfide knot fragments revealed dissociation constants of 5.8 ± 1.1 microMolar for desA-NDSK and 3.7 ± 0.7 microMolar for desAB-NDSK when binding to fibrinogen surfaces [1]. These measurements indicate that the presence of both A and B knobs enhances binding affinity, though the improvement is modest, suggesting that A-a interactions predominate in the binding mechanism.
The surface-dependent nature of fibrinopeptide accessibility has been extensively characterized through surface plasmon resonance methodologies. When fibrinogen is immobilized on sensor chip surfaces, the binding affinity for plasminogen activator inhibitor-1 reaches 0.26 microMolar, demonstrating high-affinity interactions that are not observed with fragment E alone [3]. This differential binding behavior indicates that surface adsorption exposes cryptic binding sites within the amino-terminal region of the alpha chain, specifically the alpha(20-88) segment, which becomes accessible upon fibrinogen immobilization or fibrin formation.
Surface plasmon resonance experiments investigating thrombin binding to fibrin degradation products have identified the central E domain as the primary binding site, with fragment E and (DD)E complexes demonstrating strong interactions while d-dimers show no binding capacity [4]. These findings confirm that the thrombin-binding site resides within the N-terminal disulfide knot region, consistent with the localization of (Tyrosine 0)-fibrinopeptide A within this structural domain.
Aptamer-based surface plasmon resonance studies have provided additional insights into thrombin recognition mechanisms. The 15-mer aptamer, which targets the fibrinogen-recognition exosite, demonstrates enhanced binding capacity when modified with DNA spacers, indicating that molecular flexibility facilitates optimal protein-ligand interactions [2]. Conversely, the 29-mer aptamer, specific for the heparin-binding exosite, exhibits lower apparent affinity when immobilized compared to solution-phase measurements, highlighting the importance of aptamer folding and accessibility in surface-based assays.
Kinetic analysis through surface plasmon resonance has revealed that association rates are significantly influenced by structural modifications in fibrinopeptide analogs. Studies employing synthetic knob peptides demonstrated that the presence of additional proline residues and charged amino acids in positions six through seven profoundly affects association kinetics, with optimal knob mimics displaying 10-fold higher association rates than conventional analogs [5]. These structure-activity relationships provide critical design parameters for developing enhanced fibrinolytic modulators.
The temperature and ionic strength dependence of thrombin-fibrinopeptide interactions, as measured by surface plasmon resonance, reveals the predominant role of hydrophobic interactions in complex stabilization. Despite the highly charged nature of the interacting surfaces, the binding energy derives primarily from hydrophobic contacts, with electrostatic interactions providing specificity and orientation [6]. This energetic profile explains the remarkable selectivity of thrombin for fibrinogen among the numerous potential substrates in the coagulation cascade.
| Study/Fragment | Binding Partner | Kd (μM) | Reference |
|---|---|---|---|
| desA-NDSK | Fibrinogen | 5.8 | [1] |
| desAB-NDSK | Fibrinogen | 3.7 | [1] |
| Fibrinogen (immobilized) | PAI-1 | 0.26 | [3] |
| Fibrin (immobilized) | PAI-1 | 0.26 | [3] |
| Fragment E (immobilized) | PAI-1 | No binding | [3] |
| Fragment D1 | PMSF-thrombin | Strong binding | [7] |
| 15-mer aptamer | Thrombin | Higher affinity | [2] |
| 29-mer aptamer | Thrombin | Lower affinity (surface) | [2] |
The sequential cleavage of fibrinopeptides A and B represents a fundamental regulatory mechanism in fibrin assembly, with (Tyrosine 0)-fibrinopeptide A serving as a critical intermediate in this highly orchestrated process [8] [9] [10]. The kinetic analysis of fibrinopeptide release has revealed intricate allosteric mechanisms that ensure proper temporal coordination of fibrin polymerization through differential substrate recognition and cleavage rates.
Kinetic studies utilizing bovine fibrinogen have established fundamental parameters for fibrinopeptide A release, with catalytic efficiency (kcat) measured at 6.3 × 10⁻¹⁰ M s⁻¹ per thrombin unit and Michaelis constant (KM) of 11.0 ± 3.0 microMolar under physiological conditions [11]. These measurements demonstrate that fibrinopeptide A release follows strict first-order kinetics across a wide range of fibrinogen concentrations, despite potential product inhibition effects. The consistency of these kinetic parameters with literature values validates the mechanistic model wherein thrombin exhibits high specificity for the Arginine 16-Glycine 17 cleavage site.
The sequential nature of fibrinopeptide release becomes evident through comparative analysis of fibrinopeptide A and B cleavage rates. Under normal physiological conditions, fibrinopeptide A release precedes fibrinopeptide B release, with the ratio of cleavage rates (k₂/k₁) measuring 1.2 ± 0.3 for human fibrinogen [10]. This temporal separation ensures that initial fibrin protofibril formation occurs through A-a interactions before lateral aggregation mediated by B-b interactions, thereby controlling clot architecture and mechanical properties.
Allosteric regulation of sequential cleavage has been demonstrated through sodium ion-dependent modulation of thrombin activity. In the presence of sodium, thrombin adopts a fast-acting conformation that promotes rapid release of the second fibrinopeptide A from partially cleaved fibrinogen molecules [10]. When sodium is replaced with choline, transforming thrombin to its slow form, the maximal levels of alpha-profibrin (fibrinogen lacking one fibrinopeptide A) increase to levels expected for independent cleavage, indicating that fast thrombin specifically enhances the release rate of the second fibrinopeptide A through cooperative mechanisms.
Surface-mediated fibrinopeptide release exhibits distinct kinetic profiles compared to solution-phase reactions. When fibrinogen adsorbs to surfaces in an end-on orientation, both fibrinopeptides are released rapidly with similar initial rates, but the total amount of fibrinopeptide B released (30 pmol) exceeds that of fibrinopeptide A (17 pmol) [12]. This reversal of the normal cleavage hierarchy suggests that surface adsorption alters the accessibility and positioning of cleavage sites, potentially through conformational changes in the central E domain.
The kinetic analysis of variant fibrinogens has provided mechanistic insights into the structural requirements for efficient cleavage. Fibrinogens London I and Ashford, which exhibit primary polymerization defects, demonstrate decreased rates of fibrinopeptide B release compared to normal fibrinogen [13]. These variants maintain the sequential cleavage pattern but with reduced overall efficiency, indicating that proper D-domain interactions are necessary for optimal substrate presentation to thrombin.
Pharmacological modulation of cleavage kinetics has been achieved through polymerization inhibitors. Glycyl-Prolyl-Arginyl-Proline reduces fibrinopeptide B release rates by a factor of five while leaving fibrinopeptide A release unaffected [13]. This selective inhibition demonstrates that fibrin polymerization state influences the kinetics of subsequent cleavage events, supporting the model wherein small fibrin oligomers serve as optimal substrates for fibrinopeptide B removal.
The concentration dependence of fibrinopeptide release reveals interesting mechanistic features. At high fibrinogen concentrations, thrombin can undergo transformation from slow to fast forms even in the absence of sodium, suggesting that fibrinogen itself can induce allosteric changes in thrombin [10]. This concentration-dependent activation provides a mechanism for enhanced cleavage efficiency at sites of vascular injury where fibrinogen concentrations are locally elevated.
| Substrate | FpA Release Rate (k1) | FpB Release Rate (k2) | KM (μM) | Reference |
|---|---|---|---|---|
| Bovine fibrinogen | 6.3 × 10⁻¹⁰ M s⁻¹ | Complex kinetics | 11.0 | [11] |
| Human fibrinogen | Higher than FpB | Delayed, accelerated in polymers | Not specified | [8] [9] |
| Fibrinogen (end-on) | Rapid | Higher than FpA (30 vs 17 pmol) | Not specified | [12] |
| Fibrinogen (side-on) | Higher than FpB | Approaches FpA level | Not specified | [12] |
| Normal fibrinogen | Precedes FpB | k2/k1 = 1.2 ± 0.3 | Not specified | [10] |
| Fibrinogen London I | Decreased rate | Decreased rate | Not specified | [13] |
| Fibrinogen Ashford | Decreased rate | Decreased rate | Not specified | [13] |
| A'β fibrinogen | Similar to normal | Similar to FpA | Similar to normal | [9] |
X-ray crystallographic analysis has provided unprecedented atomic-level resolution of thrombin-peptide interactions, revealing the precise molecular mechanisms underlying the recognition and cleavage of (Tyrosine 0)-fibrinopeptide A [6] [14] [15] [16]. These structural studies have elucidated the spatial organization of binding sites, the conformational changes accompanying complex formation, and the catalytic mechanisms governing fibrinopeptide release.
The crystal structure of human alpha-thrombin complexed with fibrinopeptide A(7-16) at 2.5 Angström resolution has revealed the detailed binding mode of the substrate within the active site [14]. The peptide adopts a compact conformation that brings together hydrophobic residues, which insert into the apolar binding site of thrombin. Specifically, Phenylalanine 8 occupies the aryl-binding site adjacent to Leucine 9 and Valine 15 in the S2 subsite, while the peptide backbone forms critical contacts with the catalytic machinery. This structural arrangement explains the exceptional specificity of thrombin for the Arginine 16-Glycine 17 scissile bond among the numerous potential cleavage sites in fibrinogen.
The crystal structure of the thrombin-central E region complex, solved at 3.65 Angström resolution, has provided insights into the non-substrate interactions that modulate thrombin activity [6] [16]. The complex consists of two thrombin molecules bound symmetrically to opposite sides of the central E domain, creating an X-shaped structure with dimensions of approximately 127 × 88 Angströms. Each thrombin molecule buries approximately 1,200 Angström² of surface area through interactions with exosite I, positioning the active sites optimally for fibrinopeptide cleavage.
The molecular interface between thrombin and the central E region involves seven key thrombin residues: Phenylalanine 34, Serine 36A, Leucine 65, Tyrosine 76, Arginine 77A, Isoleucine 82, and Lysine 110 [16]. From the fibrin side, the interface includes Tryptophan 33, Phenylalanine 35, Aspartate 38, and Glutamate 39 from the alpha chain, Alanine 68 and Aspartate 69 from the beta chain, and Aspartate 27 and Serine 30 from the gamma chain. Despite the highly charged nature of the interacting surfaces, the binding energy derives primarily from hydrophobic contacts surrounding a well-defined hydrophobic core.
Comparative crystallographic analysis of thrombin complexed with different substrates has revealed the structural basis for substrate specificity. The thrombin-PPACK complex, determined at 1.92 Angström resolution, shows the inhibitor occupying a hydrophobic cage formed by Isoleucine 174, Tryptophan 215, Leucine 99, Histidine 57, Tyrosine 60A, and Tryptophan 60D [17]. The exceptional specificity of this interaction arises from the complementary shape and electrostatic properties of the binding site, which accommodate the specific side chain requirements of fibrinogen-derived peptides.
The structural analysis of thrombin complexed with protease-activated receptor fragments has provided additional insights into recognition mechanisms. The crystal structures reveal that receptor peptides can bind simultaneously to the active site and exosite I, creating bidentate interactions that enhance specificity and binding affinity [18]. However, the inability to capture both binding modes simultaneously in single crystals suggests that the peptides possess dynamic conformations that favor one binding mode over another depending on solution conditions.
Molecular modeling studies based on crystallographic data have revealed the conformational requirements for optimal peptide binding. The thrombin-bound conformation of fibrinopeptide A exhibits a strand-turn-strand motif with a beta-turn centered at residues Glutamate 11 and Glycine 12 [19] [20]. This structural arrangement positions critical residues for optimal interactions with the S1, S2, and S3 subsites of thrombin, explaining the high catalytic efficiency observed for fibrinopeptide A cleavage.
The temperature factors and electron density quality in thrombin-peptide crystal structures indicate that certain regions of the peptide exhibit high flexibility while others are rigidly constrained by protein contacts. The N-terminal portion of fibrinopeptide A (residues 1-5) shows limited interaction with thrombin and high mobility, while the C-terminal region (residues 7-16) is well-ordered and makes extensive contacts with the enzyme [21]. This differential binding pattern suggests that the N-terminal region may function primarily as a recognition element rather than contributing significantly to binding energy.
Comparative analysis of thrombin crystal structures in different ligation states has revealed the conformational changes accompanying substrate binding and catalysis. The binding of fibrinopeptide A induces subtle but significant changes in the active site geometry, particularly in the oxyanion hole and the S1 specificity pocket [15]. These conformational adjustments optimize the catalytic machinery for efficient peptide bond hydrolysis while maintaining the structural integrity required for product release.
| Complex | Resolution (Å) | Key Binding Residues (Thrombin) | Binding Site | PDB ID | Reference |
|---|---|---|---|---|---|
| Thrombin-Fibrinopeptide A(7-16) | 2.50 | Phe-34, Tyr-76, Ile-82, Trp-215 | Active site | 1UCY | [14] |
| Thrombin-Central E region | 3.65 | Phe-34, Ser-36A, Leu-65, Tyr-76, Arg-77A, Ile-82, Lys-110 | Exosite I | 2A45 | [6] [22] |
| Thrombin-PPACK | 1.92 | Ile-174, Trp-215, Leu-99, His-57, Tyr-60A, Trp-60D | Active site | 1PPB | [17] |
| Thrombin-Hirudin | 2.95 | Exosite I residues | Exosite I | 4HTC | [23] |
| Thrombin-PAR1 fragment | 1.80 | Active site and exosite I | Dual binding | 3LU9 | [24] |
| Thrombin-γ' peptide | 2.40 | Exosite II residues | Exosite II | 2HWL | [25] |