Fibrin Fiber Stiffness Is Strongly Affected by Fiber Diameter, but Not by Fibrinogen Glycation

Abstract

The major structural component of a blood clot is a mesh of fibrin fibers. Our goal was to determine whether fibrinogen glycation and fibrin fiber diameter have an effect on the mechanical properties of single fibrin fibers. We used a combined atomic force microscopy/fluorescence microscopy technique to determine the mechanical properties of individual fibrin fibers formed from blood plasma. Blood samples were taken from uncontrolled diabetic patients as well as age-, gender-, and body-mass-index-matched healthy individuals. The patients then underwent treatment to control blood glucose levels before end blood samples were taken. The fibrinogen glycation of the diabetic patients was reduced from 8.8 to 5.0 mol glucose/mol fibrinogen, and the healthy individuals had a mean fibrinogen glycation of 4.0 mol glucose/mol fibrinogen. We found that fibrinogen glycation had no significant systematic effect on single-fiber modulus, extensibility, or stress relaxation times. However, we did find that the fiber modulus, Y, strongly decreases with increasing fiber diameter, D, as Y∝D(-1.6). Thin fibers can be 100 times stiffer than thick fibers. This is unusual because the modulus is a material constant and should not depend on the sample dimensions (diameter) for homogeneous materials. Our finding, therefore, implies that fibrin fibers do not have a homogeneous cross section of uniformly connected protofibrils, as is commonly thought. Instead, the density of protofibril connections, ρPb, strongly decreases with increasing diameter, as ρPb∝D(-1.6). Thin fibers are denser and/or have more strongly connected protofibrils than thick fibers. This implies that it is easier to dissolve clots that consist of fewer thick fibers than those that consist of many thin fibers, which is consistent with experimental and clinical observations.

Figure 1

Fibrin fiber assembly. (A) Crystal structure of human fibrinogen (42). Fibrinogen has a nearly centrosymmetric structure consisting of two Aα-chains (yellow, 610 amino acids), two Bβ-chains (magenta, 461 amino acids), and two γ-chains (cyan, 411 amino acids). Not resolved in the crystal structure are residues α1–26 and α201–610, β1–57 and β459–461, and γ1–13 and γ395–411. The central E region, which contains the N-termini of all chains, including fibrinopeptides A (α1–16) and B (β1–14), is connected by two triple-helical coiled coils to the two distal D regions. The α-C regions (α221–610), which are drawn in by hand, consist of the unstructured, 61-nm-long αC connector (α221–391; drawn as a yellow line) and the folded αC domain (α392–610; drawn as two yellow squares). The small blue, red, and black spheres indicate plasmin lysis sites in the coiled coils; numerous additional lysis sites (not shown) are in the α-C region. (B) Fibrin interactions. The half-staggered assembly of fibrin monomers into double-stranded protofibrils is mainly mediated by A:a knob-hole interactions and D:D interface interactions, and to a lesser extent by B:b knob-hole interactions (not shown). Lateral (radial) assembly of protofibrils into mature fibers is thought to be mostly mediated by interactions of the α-C regions, resulting in a dense, complex network of α-C regions between protofibrils. For clarity, not all α-C regions are shown in the space between protofibrils. (C) Schematic, longitudinal cross section of a mature, ∼130-nm-wide fibrin fiber, showing a dense fiber core and a less dense fiber periphery. This decrease in fiber density (or more specifically, bond density) with increasing fiber diameter is one of the key findings of this study. The lateral structure is mostly held together by the network of connected α-C regions (partially shown in B, but for clarity not shown in C). (D) A false-color scanning electron micrograph of a blood clot conveys the central role of fibrin fibers in providing mechanical and structural support to a blood clot. Green, fibrin fibers; purple, platelets; red, red blood cell (image courtesy of Y. Veklich and J. W. Weisel, University of Pennsylvania School of Medicine). To see this figure in color, go online.

Figure 2

Fibrin fiber manipulation. (A) Schematic of fibrin fiber manipulation. The fiber is suspended over the grooves in a striated substrate. The AFM tip, located above the sample, pulls on the fiber while the optical microscope, located below the sample, acquires images and movies of the manipulation. (B) Top view schematic of fiber manipulation. L_initial is half the initial length of the fiber, L′ is half the length of the stretched fiber, and s is the distance the tip has traveled. L′ can be found trigonometrically from L_initial and s, and the strain can be calculated from these quantities (see text). Schematics (A) and (B) were adapted from (11). (C) Optical microscopy movie frames of a fiber being stretched and broken. The movie was recorded from underneath the sample. The large dark object (rectangle plus triangle shape) is the AFM cantilever and the AFM tip is marked by an asterisk. The fiber broke at a strain of 200%. Scale bar, 10 μm. To see this figure in color, go online.

Figure 3

Fibrin fiber mechanical properties as a function of glycation. (A) Extensibility. (B) Fast and slow stress relaxation times. (C) Total modulus. (D) Diameter-normalized total modulus for an average 130 nm fibrin fiber. No significant, systematic trend is observed between these mechanical properties and glycation. Error bars are standard error of the mean.

Figure 4

Fibrin fiber properties as a function of fibrinogen concentration. (A) Extensibility. (B) Fast and slow stress relaxation time. (C) Total modulus. (D) Diameter-normalized total modulus for an average 130 nm fibrin fiber. No significant, systematic trend is observed between these mechanical properties and fibrinogen concentration. Error bars are standard error of the mean.

Figure 5

Fibrin fiber modulus as a function of fiber diameter. (A and B) Log-log plot of the total stretch modulus as a function of diameter for plasma samples (A) and purified fibrinogen samples (B). The modulus strongly depends on the fiber diameter, as $\mathit{Y}\propto {\mathit{D}}^{-1.6}$ (plasma) and $\mathit{Y}\propto {\mathit{D}}^{-1.4}$ (purified fibrinogen).

Figure 6

Fibrin fiber models and their corresponding stretch modulus. (A) A fiber with a cross section of uniformly connected protofibrils will have a stretch modulus that is independent of diameter, D. (B) A fiber with a bicycle-spokes-like cross section will have a stretch modulus that decreases as D⁻¹. (C) In our experiments, the stretch modulus scales as D^−1.6, indicative of a cross section in which the density of connected protofibrils likewise decreases strongly with increasing D, as D^−1.6. To see this figure in color, go online.