Esta referencia está disponible actualmente en inglés. La versión en español está en preparación.

Platelet-Rich Fibrin and Fibrin Matrix Biology

Por Benn Bluestein-Veyra, M.Sc. en Química Orgánica · Actualizado el 21 de octubre de 2026

Platelet-rich fibrin (PRF) is a fibrin clot that forms when blood collected without anticoagulant is centrifuged, gathering platelets and leukocytes into the fibrin matrix.^[1][2] Its biology is the biology of fibrin itself: thrombin converts fibrinogen into monomers that polymerize into a branched, cross-linked network.^[3]

This reference follows that sequence from the fibrinogen molecule to the finished network, then describes how preparation variables shape a platelet-rich fibrin clot and how the matrix holds and releases platelet proteins in vitro. Dohan Ehrenfest and colleagues proposed classifying platelet concentrates by their leukocyte and fibrin content, and that framework anchors the terms used here.^[4]

Understanding the molecular architecture of fibrinogen

Fibrinogen is a large, 340 kDa hexameric homodimer that circulates in plasma and is secreted by the liver.^[5] Each half carries one A-alpha, one B-beta and one gamma chain, and disulfide bridges hold the two halves together.^[6]

The three chains are encoded by neighboring genes: FGA and FGB at 4q31.3 and FGG at 4q32.1.^[7][8][9] Mosesson describes the folded molecule as two outer D domains joined to a central E domain by coiled-coil segments.^[6]

Kollman and colleagues solved the crystal structure of human fibrinogen purified from plasma, resolving the twisted coiled-coil regions and the carbohydrate groups on the beta chains, with all 11 sugar residues positioned (Fig. 1).^[10] As in earlier structures, the flexible alphaC domains gave no resolvable electron density.^[10]

Fibrinogen also binds partners that matter for a platelet-rich matrix. Mosesson lists binding to the platelet integrin alphaIIb-beta3 and to fibronectin, and notes that bound growth factors such as FGF-2 and VEGF enhance the activities of the fibrin beta15-42 sequence.^[6]

Thrombin converts fibrinogen into fibrin through ordered knob-hole binding

Thrombin cleaves fibrinopeptide A from the A-alpha chains, which starts polymerization.^[6] Weisel and Litvinov describe an ordered sequence of fibrinopeptide release, A first and B later, that drives the knob-hole interactions behind oligomer formation (Fig. 2).^[11]

Removing the fibrinopeptides exposes knobs A and B in the central region, and these bind holes a and b that are always exposed at the ends of other molecules.^[3] Kononova and colleagues modeled the A:a and B:b bonds with the knob mimetic peptides GPRP and GHRP and found the two complexes of roughly equal strength, with dissociation kinetics sensitive to pH and temperature.^[12]

The bound monomers form half-staggered, double-stranded oligomers that lengthen into protofibrils. Protofibrils then aggregate laterally into fibers that branch into a three-dimensional network.^[3]

Lateral aggregation is the least understood step. The alphaC region and B:b knob-hole binding appear to enhance it, and calcium ions bound to fibrin(ogen) at strong and weak sites have been located.^[11]

Factor XIIIa cross-links the assembled network

Factor XIII is a protransglutaminase activated in the last stage of coagulation, and it works on an insoluble substrate.^[13] Its A subunit is encoded by F13A1 at 6p25.1, and the same subunit occurs in the cytoplasm of platelets and monocytes.^[13][14]

Fibrin assembly aligns the C-terminal ends of gamma chains from neighboring molecules in antiparallel pairs, and factor XIIIa joins them covalently into gamma dimers.^[6] Ryan and colleagues found that cross-linking changed network morphology very little and raised clot rigidity, most likely by stiffening existing fibers.^[15]

Thrombin concentration shapes fiber thickness and network density

Wolberg reviews in vitro work showing that the thrombin concentration present at gelation strongly influences fibrin clot structure. Clots formed at low thrombin consist of thick fibers, and clots formed at high thrombin consist of thin fibers (Fig. 3).^[16]

Ryan and colleagues varied fibrinogen, thrombin and calcium and measured fiber length, diameter and branching from stereo scanning electron micrographs. Large fiber diameters and lengths appeared only when branching was minimal, and three-way junctions were the dominant branch point.^[15]

In the same study, maximal rigidity came from a balance between large fibers and extensive branching, and higher fibrinogen raised rigidity by increasing fiber and branch point density.^[15] Pieters and Wolberg note that genetic, biochemical and environmental factors all regulate fibrin formation, structure and stability.^[5]

Fibrin combines high elasticity with very low protein content

Litvinov and Weisel describe fibrin as both elastic and viscous, with remarkable elasticity and stability for a gel of very low protein content. Like many filamentous protein polymers, it stiffens under shear, tension or compression.^[17]

The mechanics of the whole gel follow from its single fibers, through changes in their orientation, stretching, bending and buckling. Fiber properties in turn depend on how many protofibrils each fiber holds and how they pack.^[17]

Platelets and fibrin together contract the clot

Platelets bind fibrinogen through alphaIIb-beta3, which incorporates them into a developing clot.^[6] Cines and colleagues found that contracted blood clots carry a meshwork of fibrin and platelet aggregates on the outside and a close-packed array of compressed, polyhedral erythrocytes inside.^[18]

Contraction required platelets, with their cytoskeletal motility proteins, and fibrin(ogen) as the substrate bridging them. The same structure formed with several clotting activators and in reconstituted human blood.^[18] The authors also produced the polyhedral erythrocyte shape with the forces of centrifugation alone, in unclotted blood.^[18]

Platelet-rich fibrin forms from blood collected without anticoagulant

Dohan and colleagues described PRF in 2006 as a second-generation platelet concentrate prepared without biochemical blood handling (Fig. 4). They traced it from fibrin glues and concentrated platelet-rich plasma, and proposed that slow polymerization during PRF preparation gives a fibrin network close to the natural one.^[1]

The two preparation routes differ in the anticoagulant and in when fibrin forms. Platelet-rich plasma is collected as a liquid from anticoagulated blood, and PRF clots during the spin itself.^[1][19]

The Dohan Ehrenfest classification sorts platelet concentrates into four families by leukocyte and fibrin content.^[4] The sibling reference on study reporting compares this framework with later reporting systems.

• Pure platelet-rich plasma (P-PRP): a liquid platelet concentrate without leukocytes.^[4]
• Leukocyte- and platelet-rich plasma (L-PRP): a liquid platelet concentrate that keeps leukocytes.^[4]
• Pure platelet-rich fibrin (P-PRF): a fibrin matrix without leukocytes.^[4]
• Leukocyte- and platelet-rich fibrin (L-PRF): a fibrin matrix that holds platelets and leukocytes, the family that includes the original PRF protocol.^[4]

Dohan Ehrenfest and colleagues analyzed the residual plasma left after collecting PRF clots and estimated that about 97% of platelets and more than 50% of leukocytes were concentrated in the clot. Platelets and fibrin formed large clusters in the first millimeters of the clot above the red cell base, within a mature, dense fibrin network.^[2]

In that study, glass and glass-coated plastic tubes, and forcible or soft compression, gave clots of similar architecture.^[2]

Relative centrifugal force and time set where cells settle in the clot

Miron and colleagues distinguish revolutions per minute from relative centrifugal force, the g-force a sample experiences, as the starting point for describing a protocol.^[20] Their overview also covers fixed-angle and horizontal rotors, tube surface chemistry and temperature control of clotting.^[20]

Ghanaati and colleagues compared a standard protocol with a lower-speed, longer-time protocol in clots from four donors. Platelets appeared throughout both clots, more of them in the distal part with the lower-speed protocol, and neutrophils shifted from the red cell interface toward the distal clot.^[19]

Miron and colleagues introduced a method that pipettes sequential 1-milliliter layers from the top of a centrifuged tube and counts each with a hematology analyzer. Standard L-PRF settings concentrated platelets and leukocytes within the buffy coat, and slower settings distributed platelets more evenly through the upper layers.^[21]

In the same comparison, horizontal centrifugation yielded higher numbers and concentrations of platelets and leukocytes than fixed-angle rotors.^[21] The group then compared 24 horizontal-rotor protocols across six forces, 100 to 1,200 g, and four times, 3 to 12 min, using 960 complete blood counts.^[22]

Spin time appeared to affect layer separation more than force, and protocols at or below 200 g accumulated few platelets and leukocytes in the upper layers. Baseline cell counts and hematocrit of each donor sample also influenced separation, most clearly at low forces.^[22]

Liquid platelet-rich fibrin is collected before the fibrinogen clots

Very short, low-force spins yield a liquid upper fraction that can be drawn off before it clots. Miron and colleagues prepared such a liquid PRF at about 60 g for 3 min, with no anticoagulant.^[23]

Layer counts showed that liquid protocols reached the highest concentrations of platelets and leukocytes in the collected fraction, with lower total cell numbers because the collected volume was smaller.^[21]

The fibrin matrix binds growth factors and releases them gradually in vitro

Fibrin binds growth factors directly. Sahni and colleagues measured specific, saturable binding of basic fibroblast growth factor to fibrinogen and to fibrin, with high-affinity dissociation constants between 0.13 and 1.3 nM and two to four FGF molecules per fibrinogen at saturation.^[24]

Dohan Ehrenfest and colleagues measured release from PRF membranes in culture over seven days. The membranes released TGF-beta1, PDGF-AB, VEGF and thrombospondin-1 throughout that time, and the authors proposed that trapped leukocytes went on producing TGF-beta1 and VEGF.^[25]

Kobayashi and colleagues compared PRP, PRF and advanced PRF (A-PRF) from six donors by ELISA at 15 min, 60 min, 8 h and days 1, 3 and 10.^[26] PDGF-AA was the most abundant factor released, followed by PDGF-BB, TGF-beta1, VEGF and PDGF-AB.^[26]

In that study PRP released more protein in the first hour, and the PRF preparations released growth factors steadily through day 10, with A-PRF releasing the largest totals.^[26] Fujioka-Kobayashi and colleagues reported that lowering centrifugal force raised the release of PDGF, TGF-beta1, EGF and IGF from PRF clots in culture.^[27]

The receptors that read these factors, including the PDGF, EGF, VEGF and TGF-beta receptors, are described in the sibling growth factor receptors reference.

Plasmin turns fibrin over through regulated fibrinolysis

Fibrin is both the main product of coagulation and the substrate of fibrinolysis.^[28] The zymogen plasminogen, encoded by PLG at 6q26, binds fibrin together with tissue-type plasminogen activator, and the activated enzyme plasmin cleaves fibrin at specific lysine residues.^[3][29]

Mosesson describes this as a ternary complex of tPA, plasminogen and fibrin that stimulates plasminogen activation, with alpha2-antiplasmin and plasminogen activator inhibitor-2 among the binding inhibitors.^[6] Longstaff and Kolev review how fibrin structure, clot composition, cells and flow localize and regulate these reactions.^[30]

Network structure sets the pace. Wolberg notes that thick-fiber clots formed at low thrombin are highly susceptible to fibrinolysis, and thin-fiber clots formed at high thrombin are relatively resistant.^[16]

Risman and colleagues added protofibril packing as a further variable. Tissue factor clots with dense networks and loosely packed fibers lysed faster at first than thrombin clots with looser networks and densely packed fibers.^[31] Cines and colleagues proposed that clot contraction itself slows fibrinolysis.^[18]

Scanning electron microscopy, turbidity and rheology measure fibrin structure

Each method reads a different scale of the network. Scanning electron microscopy images fibers and cells directly, and Dohan Ehrenfest and colleagues used it with light microscopy to map the architecture of PRF clots.^[2][15]

Turbidity reports average fiber size in a hydrated gel. Carr and Hermans showed that the wavelength dependence of turbidity gives the mass-length ratio of fibrin fibers and, for thicker fibers, their radius.^[32]

Rheometers measure stiffness and viscous behavior, which Ryan and colleagues correlated with fiber dimensions and branching.^[15] Risman and colleagues combined turbidimetry with confocal and scanning electron microscopy to follow clot formation, protofibril packing and lysis together.^[31]

For preparations, cell counts and release assays complete the picture. The sequential layer method counts cells in each fraction of a centrifuged tube, and ELISA follows growth factors released into the medium.^[21][26]

Where to go from here

The platelet granule biology reference covers the alpha granule proteins that a fibrin matrix holds, and the PRP preparation variables reference follows anticoagulant, force and time through liquid concentrates. The growth factor receptors reference explains how PDGF, TGF-beta, VEGF and EGF signal once released.

The study reporting reference shows how classification systems describe a preparation, the research methods library explains study designs, and the BMAC and SVF reference turns to marrow and adipose preparations. The MSC secretome reference, the HCT/P regulatory reference and the quality references round out the library, and our team is glad to discuss the methods on this page.