Hemostatic properties of glucosamine-based materials



Glucosamine- and N-acetyl glucosamine-containing polymers are being used in an increasing number of biomedical applications, including in products for surface (topical) hemostasis. The studies presented here investigate the relationship between the structure (conformation) and function (activation of hemostasis) of glucosamine-based materials. Several polymer systems were studied, including fibers isolated from a microalgal source containing poly-N-acetyl glucosamine polymers that are organized in a parallel, hydrogen-bonded tertiary structure and can be chemically modified to an antiparallel orientation; and gel formulation derivatives of the microalgal fibers consisting of partially deacetylated (F2 gel) and fully deacetylated (F3 gel) polymers. Comparison of the properties of the poly-N-acetyl glucosamine fiber-derived materials with chitin, chitosan, and commercial chitosan-based products are presented. Several studies were performed with the glucosamine-based materials, including (1) an analysis of the ability of materials to activate platelets and turnover of the intrinsic coagulation cascade, (2) an examination of the viscoelastic properties of mixtures of platelet-rich plasma and the glucosamine-based materials via thromboelastography, and (3) scanning electron microscopic studies to examine the morphology of the glucosamine-based materials. The results presented demonstrate that hemostatic responses to the glucosamine-based materials studied are highly dependent on their chemical nature and tertiary/quaternary structure. The unique natural microalgal fibers were found to have strongly prohemostatic activity compared to the other materials studied. © 2006 Wiley Periodicals, Inc. J Biomed Mater Res, 2007


The development of glucosamine containing polymers for modulating hemostatic systems is an advancing area of medical product development. A new polymeric fiber material, poly-N-acetyl glucosamine (p-GlcNAc), formulated as patches and lyophilized pads, has been shown to be effective in achieving hemostasis in surgical procedures and trauma.1–7 The p-GlcNAc fiber material is isolated and purified from large scale cultures of a marine microalga/diatom.8 An investigation of the physical and chemical properties of the p-GlcNAc fibers was carried out to determine how they structurally and functionally differ from chitin, chitosan, and other N-acetyl glucosamine containing compounds. These observations allow us to better understand prior observations of differences in hemostatic function.5

N-acetyl glucosamine polymers occur in natural products such as chitin, which can be deacetylated chemically or enzymatically to produce glucosamine polymers referred to as chitosan. Chitosan, which is a chemically heterogeneous mixture of polymers with 10–20% N-acetyl residues and 80–90% glucosamine residues,3 can also be extracted from crustacean exoskeletons and fungal mycelial mats. Chitin and chitosan are found in nature complexed with minerals and protein contaminants in a wide variety of microcrystalline forms.9 The supramolecular structure of naturally occurring chitin–mineral–protein complexes are generally amorphous, reflecting their physiological barrier function.

Chitosans are commonly purified from natural sources of chitin by processes that include acid/base extraction.10 Chitosans can exist in many structural conformations, depending on a variety of factors that include the degree of hydration, the counterion mixture, and the complexity of the original chitin mixture.11 Chitin- and chitosan-based products have been advanced for a wide variety of applications, including wound healing,12 pharmaceutical formulation,12–16 tissue engineering,17–21 and surface hemostasis.22, 23 However, medical product development with these materials has been hampered by the chemical and physical heterogeneity of the polymer products and contamination of preparations by proteins and other components.3

Recently discovered nanofibers isolated from a marine microalgal source3 consist of high molecular weigh (MW = 2.8 × 106 Da) linear polymers of fully-acetylated poly-N-acetyl glucosamine. The individual polymers in the fibers are tightly bound to one another by interchain hydrogen bonding in a parallel (β-structure) orientation. P-GlcNAc fibers are generally longer than 100 μm3 with 2–4 nm diameters, and consist of ∼80 polymer molecules per fiber. The β-pGlcNAc fibers are isolated from microalgal cultures in a highly pure chemical form3 with preservation of the native supramolecular structure. Microalgal pGlcNAc fibers are the single and unique component of several FDA cleared medical device products currently in the topical hemostasis market.24, 25 β-pGlcNAc fibers can be disassociated into their individual poly-N-acetyl glucosamine polymer components with strongly chaotropic (hydrogen-bond breaking) solvents, and can be reassembled with the individual polymers in an antiparallel orientation to generate the α-pGlcNAc structural form. Also, microalgal-derived β-pGlcNAc fibers can be deacetylated to varying extents to obtain soluble, partially deacetylated or fully deacetylated cationic polymers that can be formulated as hydrogels (F2 and F3 gels, respectively).

The goal of the studies presented in this article was to understand how differences in chemical, ternary and quaternary structure of N-acetylglucosamine and glucosamine containing materials affect their interaction with hemostatic systems. The effect of a panel of materials on platelet activation responses, intrinsic (contact) coagulation cascade turnover, the chemical/physical surface adsorption of plasma proteins and the visco-elastic properties of fibrin clots was analyzed. The principle result is that the structural properties of β-pGlcNAc fibers, when formulated into materials, are significantly prohemostatic in comparison to the other materials studied.



β-pGlcNAc, α-pGlcNAc microalgal-derived fiber materials, and partially deacetylated and fully deacetylated F2 and F3 gel formulations made from microalgal pGlcNAc fibers, were obtained from Marine Polymer Technologies (Danvers, MA). Chitin and chitosan were procured from Sigma-Aldrich (St. Louis, MO) and Vanson (Redmond, WA), respectively. The vascular closure medical products Closure™ and Chitoseal™ were purchased respectively from Medtronic (Minneapolis, MN) and Abbott Vascular (Redwood City, CA). The thrombin substrate D-phe-pro-arg-ANSNH and corn trypsin inhibitor were obtained from Haematologic Technologies (Essex Junction, VT). Infusion grade Integrilin™ (eptifibatide) was obtained from COR Therapeutics (San Francisco, CA). Lovenox™ (low molecular weight heparin or enoxaparin sodium) was from Sanofi-Aventis (Bridgewater, NJ).

Platelet plasma solutions isolation

Platelet rich plasma and platelet free plasma were isolated from freshly drawn citrate anticoagulated whole blood with differential centrifugation as detailed elsewhere.26 The platelet concentration in the platelet rich plasma was measured with a Hiska haematological analyzer, and the platelet concentration was adjusted to 150,000 platelets/uL by diluting the sample with platelet free plasma.

Thrombin generation kinetics

The effect of glucosamine-based materials on the kinetics of thrombin generation in platelet rich plasma (at 150,000 platelets/uL) and platelet-free plasma was investigated by following the hydrolysis of the thrombin substrate D-phe-pro-arg-ANSNH to yield a fluorescent reaction product. Experiments were performed in 96 well Costar tissue culture platelets that had been incubated for 24 h at 37°C with 5% human serum albumin in citrated saline to block sites of spontaneous contact activation initiation. The serum albumin solution was removed, then 300 ug of each glucosamine-based material was added to triplicate wells in 10 uL citrated saline. D-phe-pro-arg-ANSNH was added to platelet rich plasma or platelet free plasma samples by performing a 200/1 dilution of a DMSO stock solution for a final concentration of 50 uM. The time-course for thrombin generation was initiated by adding CaCl2 for 10 mM to plasma or platelet samples, then immediately transferring 90 uL to each well that contains the polyglucosamine samples at 23°C. The D-phe-pro-arg-ANSNH hydrolysis was followed by measuring the fluorescence at 480 nm every minute in a Fluorstar Galaxy fluorescent plate reader (BMG Labtechnologies, Durham, NC). Triplicate relative fluorescence values were averaged to obtain the time-course curves presented in this article. The lag time for thrombin generation was defined as the time point at which the fluorescence increased 10% over the initial baseline value.


Thromboelastographic measurements were performed with a TEG-5000 thrombelastograph hemostasis analyzer (Haemoscope Corporation, Niles, IL). The assays were initiated by adding CaCl2 to 10 mM to platelet rich plasma (at 150,000 platelets/uL) and then immediately transferring 327 uL of the calcified platelet rich plasma to the thromboelastography chamber that contained the glucosamine-based materials in 33 uL citrated saline. The final glucosamine-based material concentration was 3.0 mg/mL. Measurements were performed for 1 h in quadruplicate 37°C, and then relevant parameters were extracted from the “stiffness” curve.

Scanning electron microscopy

SEM analysis on glucosamine-based materials was performed as detailed elsewhere.5


A series of experiments were performed to better understand how hemostatic systems interact with the glucosamine-based materials described earlier. The experiments examined the ability of the various materials to affect intrinsic coagulation cascade turnover, platelet activation responses, fibrin clot visco-elastic properties, Vroman protein layer formation, and platelet morphology.

The effect of glucosamine-based materials on thrombin generation

The ability of glucosamine-based materials to affect intrinsic coagulation cascade turnover for thrombin generation was ascertained by following cleavage of the fluorogenic thrombin substrate D-phe-pro-arg-ANSNH. Eight glucosamine-based material formulations were tested in platelet-rich and platelet-free plasma. As depicted with representative samples in Figure 1, after a lag time that is a function of intrinsic coagulation cascade turnover rates, thrombin begins to cleave D-phe-pro-arg-ANSNH to yield the fluorescent cleavage product. The three representative thrombin substrate hydrolysis curves show that the lag time for thrombin generation can be very different for glucosamine-based materials. A comparison of all eight materials studied is depicted in Figure 2. The β-pGlcNAc material was the most potent activator of intrinsic coagulation in both the presence and absence of platelets. The F2 and F3 gels, chitin, chitosan and α-pGlcNAc also had a statistically significant acceleration effect on thrombin generation as compared to the saline control, with a trend towards chitin being the most thrombogenic of this group of materials. In contrast, Chitoseal™ and Closure™ did not have a significant effect on thrombin generation in plasma and actually inhibited the ability of platelets to accelerate turnover of the coagulation cascade in comparison to the saline control.

Figure 1.

Representative thrombin substrate hydrolysis time courses. The ability of glucosamine-based materials to generate thrombin in platelet rich plasma and plasma alone was measured by following the hydrolysis of a thrombin fluorogenic substrate as detailed in the Material and Methods section. Representative curves for β-pGlcNAc, α-pGlcNAc, and saline (control) samples in platelet rich plasma are shown. Arrows indicate the positions of the “lag time” for thrombin generation.

Figure 2.

The effect of glucosamine-based materials on thrombin generation. The glucosamine-based materials were analyzed with the fluorogenic assay as in Figure 1. Each material was examined in PRP (black bars) and plasma alone (grey bars). Error bars are standard deviations for triplicate samples. Lag time average values and standard deviations are tabulated below the horizontal chart labels.

Mechanisms for platelet activation by prohemostatic glucosamine-based materials

The six glucosamine-based materials that accelerated thrombin generation in the presence of platelets shown in Figure 2 (β-pGlcNAc, F2 and F3 gels, chitin, chitosan and α-pGlcNAc) were studied to determine the extent to which the platelet-dependent acceleration was due to integrin outside-in signaling and/or to the surface catalysis of factor XIIa activation for turnover of the intrinsic coagulation cascade. The relative contribution of these two mechanisms was judged by measuring the inhibition of lag times for thrombin generation in the presence of Integrilin™, which inhibits integrin functions, and of corn trypsin inhibitor, which inhibits factor XIIa. Figure 3 shows that the prothrombotic effects of the six prothrombotic glucosamine-based materials were sensitive to factor XIIa inhibition, whereas only β-pGlcNAc mediated thrombin generation in PRP was sensitive to both factor XIIa inhibition and integrin function inhibition. This result suggests that only the β-pGlcNAc fiber material functions via at least two separate mechanisms in activating hemostasis; a unique observation in comparison with all of the other materials tested.

Figure 3.

The effect of inhibitors on thrombin generation kinetics. The ability of the glucosamine-based materials to accelerate thrombin generation was measured in the presence of Integrilin™ to inhibit platelet integrin function and corn trypsin inhibitor to inhibit factor XIIa as described in the Materials and Methods section. Error bars are standard deviations for triplicate samples.

The effect of glucosamine-based materials on fibrin clot visco-elastic properties

Thromboelastographic (TEG) methods were used to estimate the “stiffness” of clots formed in platelet-rich plasma mixtures with glucosamine-based materials as fibrin polymerization occurs. Representative thromboelastograms for PRP with β-pGlcNAc, α-pGlcNAc, or carrier citrated saline (as a negative control) are presented in Figure 4. As anticipated by the acceleration of thrombin generation times reported in Figure 2, the onset of fibrin polymerization (see arrows in Fig. 4) was earliest with β-pGlcNAc fiber material. The β-pGlcNAc fiber slurry with platelet rich plasma formed a clot that is significantly stiffer (as judged by the maximal amplitude (MA value) of the separation of the TEG curves) than the α-pGlcNAc slurry or with platelet rich plasma alone (saline control). Figure 6 summarizes the “stiffness” (MA parameter) values that were obtained with each of the glucosamine-based materials studied. β-pGlcNAc fiber materials, F2 gel, F3 gel, Chitoseal™ and Closure™ formed a fibrin matrix that was stiffer than the corresponding matrix that was formed with α-pGlcNAc, chitin, chitosan or saline (control without added polyglucosamines). These differences were in part a result of the intrinsic viscosity of test materials. Other differences were not statistically significant. Relative onset times for fibrin polymerization in the TEG analysis (Fig. 5) paralleled the kinetics of thrombin generation as measured with the fluorogenic assay (Fig. 2).

Figure 4.

Representative thromboelastograms. The glucosamine-based materials were analyzed with Thromboelastography as detailed in the Materials and Methods with representative curves for β-pGlcNAc, α-pGlcNAc, and saline (control) presented. The dashed arrows indicate the maximal amplitude (MA) value while the solid arrows point to the time for initiation of fibrin polymerization.

Figure 5.

The effect of glucosamine-based materials on initial thrombin generation time by thromboelastogram analysis. The time for initiation of fibrin polymerization with each glucosamine-based material was measured as shown in Figure 4. Error bars indicate the standard deviation with four evaluations with each material.

Figure 6.

The effect of glucosamine-based materials on fibrin clot stiffness. The maximal amplitude parameter for each glucosamine-based material was measured as indicated Figure 3. Error bars indicate the standard deviation with four evaluations with each material.

Ultrastructural analysis of the interaction of platelets with glucosamine-based materials

Scanning electron microscopic analysis of the glucosamine-based materials (Fig. 7) showed that β-pGlcNAc material has a unique ultrastructure in comparison to the other materials; a fine (∼50 nm diameter, ∼100 um length) fibrous structure. Chitoseal™ has a tubular structure, and the diameters of the tubes are much larger (∼15 μm in diameter) than the fine fibers that make up the β-pGlcNAc. These tube-like entities are apparently the backing onto which a very thin layer of chitosan is coated to make up the final product. All of the other materials are amorphic aggregates with irregular shapes and varying degrees of surface roughness. The β-pGlcNAc fibers organized in nonwoven fabric patches resemble the dimensions of natural fibrin networks. A scanning electron micrograph of platelets interacting with fibrin under native conditions is presented in the lower-right hand panel. This analysis indicates blood components may interact more favorably in accelerating clot formation and forming strong clots because of the fibrin network-like structural dimensions of β-pGlcNAc fiber-based materials.

Figure 7.

Scanning electron micrographs of glucosamine-based materials. Scanning electron microscopy of glucosamine-based materials in saline was performed as detailed in the Materials and Methods section. The lower right hand panel depicts platelets (white arrows) interacting with fibrin under native conditions.


Several polymeric classes of materials have been used to develop hemostatic materials for use in medical applications, including collagen (actifoam™, avitene™), oxidized regenerated cellulose (Surgicel™), thrombin (D-Stat™), fibrin sealants (Tisseel™), and fibrin bandages (Tachocomb™), Chitan (Hemecon™), and Chitosan (Closure™, Chitoseal™). Many of these products have been studied in comparison with β-pGlcNAc fiber-based materials.4, 27 In general, these animal model and clinical studies have demonstrated that the β-pGlcNAc fiber-based materials are superior to the collagen, thrombin, chitin, chitosan oxidized cellulose, and fibrin sealant/bandage products in activating rapid hemostasis with strong clot structure. The studies presented in the examination presented here explain some of the mechanistic basis for these observations.

The results presented here demonstrate important differences in the interaction of hemostatic systems with glucosamine-based materials that vary with respect to chemical composition and conformational form. The β-conformer of the microalgal pGlcNAc fibers have several properties that distinguish it from the other glucosamine-based materials. β-pGlcNAc fiber material was the most effective for accelerating the turnover of the intrinsic coagulation cascade for fibrin clot formation in both the presence and absence of platelets. α-pGlcNAc, which has the same chemical composition as β-pGlcNac but is organized with the polymer chains arranged in an antiparallel instead of a parallel orientation, is less effective at accelerating platelet-dependent or independent fibrin polymerization. Chemical modification (deacetylation) of β-pGlcNAc to produce F2 and F3 gels also reduces prothrombotic potential. Chitin, which contains a heterogeneous structure and is complexed with minerals and proteins, was less prothrombotic than microalgal β-pGlcNAc. This result indicates that β-pGlcNAc fiber structure results in a more prothrombotic material than any of the other materials tested. The deacetylation of chitin to chitosan further reduces the ability of the chitin material to accelerate intrinsic coagulation. As we have noted elsewhere,5 reformulation of chitin and chitosan into the commercial products Chitoseal™ and Closure™ yields materials that do not accelerate platelet dependent or independent intrinsic coagulation. The SEM data presented in Figure 7 show a dramatic and significant structural difference between the β-pGlcNAc fibers and the other seven materials.

We have previously shown28 that the procoagulative nature of microalgal β-pGlcNAc fibers is reduced when factor XIIa is inhibited (with corn trypsin inhibitor) or when platelet integrin outside-in signaling is inhibited (with Integrilin). The current study revealed that the microalgal β-pGlcNAc fiber structure is unique with respect to the ability of the tested materials to strongly elicit platelet integrin outside-in signaling. We have previously shown that platelets are activated upon contact with β-pGlcNAc. This finding is extended by showing that the interaction of β-pGlcNAc fibers with platelets for generation of a catalytic surface for thrombin generation is more pronounced than with the other materials.

Current theories concerning how foreign materials interact with hemostatic systems point to two independent factors as being important: prothrombotic potential per unit surface area and the surface area per unit material mass. The degree of material prothrombogenicity per unit surface area is in part driven by the specific properties of the “Vroman” layer29 of chemically and physically adsorbed plasma proteins. SDS-PAGE studies (unpublished data) show that there are significant differences between the profile of plasma proteins that tightly adsorb to each glucosamine-based material. Platelet integrins have previously been shown to tightly associate with β-pGlcNAc fibers.26 Similarly, the tested materials might have differential abilities to bind the N-terminus of factor XII for activation and turnover of the intrinsic coagulation cascade.30–32 Two pieces of information indicate that factor XII activation is an important mechanistic element in the enhanced hemostatic properties of β-pGlcNAc. First, thrombin generation times in both the presence and absence of platelets are sensitive to factor XIIa inhibition by corn trypsin inhibitor (Fig. 3). Second, factor XIIa and integrin mediated signaling are synergistic mechanistic elements of the platelet activation response to β-pGlcNAc26 These findings point to the importance of platelet bound factor XII in mediating the response of materials to hemostatic systems.

The SEM analysis of the glucosamine polymers (Fig. 7) revealed important differences in the degree of surface area (on a per weight basis) exposed for each material. Microalgal β-pGlcNAc forms very fine fibers with an average diameter of ∼50 nm, similar to the size of native fibrinogen. In contrast, the other materials exist as solid phases with amorphous nonfibrous structures and with lower surface to mass ratios. Microalgal β-pGlcNAc exists as fine fibers that platelets and fibrin strands can interact with on a nanometer scale. In contrast, the other materials present, to varying extents, lower surface area per unit mass ratios that are geometrically less favorable for a submicron scale interaction with platelets and fibrinogen.

In summary, these studies demonstrate that different glucosamine polymers can have very different abilities to activate hemostatic systems. Microalgal-derived β-pGlcNAc was found to be the most effective polymeric material for accelerating fibrin polymerization. Our results suggest that the high surface area and specific β-tertiary conformation may be key factors in determining how microalgal β-pGlcNAc interacts with factor XII and platelet integrin complexes to accelerate coagulation.


The authors would like to thank the Microscopy Services staff of the UNC Department of Pathology and Laboratory Medicine for their technical assistance.