Apolipoprotein(a) and plasminogen interactions with fibrin: a study with recombinant apolipoprotein(a) and isolated plasminogen fragments.

Lipoprotein(a) [Lp(a)], but not low-density lipoprotein (LDL), was previously shown to impair the generation of fibrin-bound plasmin [Rouy et al. (1991) Arterioscler. Thromb. 11, 629-638] by a mechanism involving binding of Lp(a) to fibrin. It was therefore suggested that the binding was mediated by apolipoprotein(a) [apo(a)], a glycoprotein absent from LDL which has a high degree of homology with plasminogen, the precursor of the fibrinolytic enzyme plasmin. Here we have evaluated this hypothesis by performing comparative fibrin binding studies using a recombinant form of apo(a) containing 17 copies of the apo(a) domain resembling kringle 4 of plasminogen, native Lp(a), and Glu-plasminogen (Glu1-Asn791). Attempts were also made to identify the kringle domains involved in such interactions using isolated elastase-derived plasminogen fragments. The binding experiments were performed using a well-characterized model of an intact and of a plasmin-digested fibrin surface as described by Fleury and Anglés-Cano [(1991) Biochemistry 30, 7630-7638]. Binding of r-apo(a) to the fibrin surfaces was of high affinity (Kd = 26 +/- 8.4 nM for intact fibrin and 7.7 +/- 4.6 nM for plasmin-degraded fibrin) and obeyed the Langmuir equation for adsorption at interfaces. The binding to both surfaces was inhibited by the lysine analogue AMCHA and was completely abolished upon treatment of the degraded surface with carboxypeptidase B, indicating that r-apo(a) binds to both the intrachain lysines of intact fibrin and the carboxy-terminal lysines of degraded fibrin. As expected from these results, both r-apo(a) and native Lp(a) inhibited the binding of Glu-plasminogen to the fibrin surfaces.(ABSTRACT TRUNCATED AT 250 WORDS)     

Incorporation of vitronectin into fibrin clots. Evidence for a binding interaction between vitronectin and gamma A/gamma' fibrinogen.

Vitronectin is an abundant plasma protein that regulates coagulation, fibrinolysis, complement activation, and cell adhesion. Recently, we demonstrated that plasma vitronectin inhibits fibrinolysis by mediating the interaction of type 1 plasminogen activator inhibitor with fibrin (Podor, T. J., Peterson, C. B., Lawrence, D. A., Stefansson, S., Shaughnessy, S. G., Foulon, D. M., Butcher, M., and Weitz, J. I. (2000) J. Biol. Chem. 275, 19788-19794). The current studies were undertaken to further examine the interactions between vitronectin and fibrin(ogen). Comparison of vitronectin levels in plasma with those in serum indicates that approximately 20% of plasma vitronectin is incorporated into the clot. When the time course of biotinylated-vitronectin incorporation into clots formed from (125)I-fibrinogen is monitored, vitronectin incorporation into the clot parallels that of fibrinogen in the absence or presence of activated factor XIII. Vitronectin binds specifically to fibrin matrices with an estimated K(d) of approximately 0.6 microm. Additional vitronectin subunits are assembled on fibrin-bound vitronectin multimers through self-association. Confocal microscopy of fibrin clots reveals the globular vitronectin aggregates anchored at intervals along the fibrin fibrils. This periodicity raised the possibility that vitronectin interacts with the gamma A/gamma' variant of fibrin(ogen) that represents about 10% of total fibrinogen. In support of this concept, the vitronectin which contaminates fibrinogen preparations co-purifies with the gamma A/gamma' fibrinogen fraction, and clots formed from gamma A/gamma' fibrinogen preferentially bind vitronectin. These studies reveal that vitronectin associates with fibrin during coagulation, and may thereby modulate hemostasis and inflammation.     

Truncated vitronectins: binding to immobilized fibrin and to fibrin clots, and their subsequent interaction with cells.

The plasminogen activator inhibitor-1 (PAI-1) is stabilized in its inhibitory conformation by binding to Vitronectin (Vn). The anchorage of PAI-1 to the fibrin fibers was recently shown to be mediated by Vn, and as such to modulate fibrinolysis. Here we report the mapping of the fibrin binding sites in Vn using truncated recombinant Vns, and show that two segments of Vn are involved: one at its carboxyl terminus (within residues 348-459) and one at its amino terminus (within residues 1-44). This mapping sets the stage for (i) the design of specific inhibitors for the Vn-fibrin interaction; (ii) for studying the role of this interaction in the anchoring of endothelial cells and platelets onto the fibrin clot; and (iii) for getting a deeper insight into the mechanism of the Vn-fibrin interaction in fibrinolysis. (c)2002 Elsevier Science.     

Electron microscope structural study of modified fibrin and a related modified fibrinogen aggregate

The structure of proteolytically modified fibrin and a closely related modified fibrinogen aggregate have been studied by analysis of electron microscope images. For both structures, we propose a model that consists of double-stranded, 2-fold helical protofibrils, which are associated laterally to form ordered fibrils, with a C222 space group: a = 44.0 nm, b = c = 9.4 nm. Each fibril is 80 nm or less in diameter, and twists along its length in a right-handed sense, with a pitch from 7 to 12 times the molecular length. The fibrils associate laterally to form bundles, which tend to twist in a left-handed sense, with a pitch of the order of 40 times the molecular length. The specific volume of modified fibrin calculated from this model is 3.9 A3 per dalton, which is comparable to the specific volume of 3.6 A3 per dalton for modified fibrinogen crystals but is lower than the 6 A3 per dalton determined for fibrin from light-scattering experiments. Comparison of our electron microscope results with X-ray and neutron diffraction data suggest a similar, but less well-ordered, structure for native fibrin, with a smaller fibril, approximately 18.4 nm wide, consisting of eight protofibrils.     

Inhibition of plasminogen activation by lipoprotein(a): critical domains in apolipoprotein(a) and mechanism of inhibition on fibrin and degraded fibrin surfaces

Similarity between the apolipoprotein(a) (apo(a)) moiety of lipoprotein(a) (Lp(a)) and plasminogen suggests a potentially important link between atherosclerosis and thrombosis. Lp(a) may interfere with tissue plasminogen activator (tPA)-mediated plasminogen activation in fibrinolysis, thereby generating a hypercoagulable state in vivo. A fluorescence-based system was employed to study the effect of apo(a) on plasminogen activation in the presence of native fibrin and degraded fibrin cofactors and in the absence of positive feedback reactions catalyzed by plasmin. Human Lp(a) and a physiologically relevant, 17-kringle recombinant apo(a) species exhibited strong inhibition with both cofactors. A variant lacking the protease domain also exhibited strong inhibition, indicating that the apo(a)-plasminogen binding interaction mediated by the apo(a) protease domain does not ultimately inhibit plasminogen activation. A variant in which the strong lysine-binding site in kringle IV type 10 had been abolished exhibited substantially reduced inhibition whereas another lacking the kringle V domain showed no inhibition. Amino-terminal truncation mutants of apo(a) also revealed that additional sequences within kringle IV types 1-4 are required for maximal inhibition. To investigate the inhibition mechanism, the concentrations of plasminogen, cofactor, and a 12-kringle recombinant apo(a) species were systematically varied. Kinetics for both cofactors conformed to a single, equilibrium template model in which apo(a) can interact with all three fibrinolytic components and predicts the formation of ternary (cofactor, tPA, and plasminogen) and quaternary (cofactor, tPA, plasminogen, and apo(a)) catalytic complexes. The latter complex exhibits a reduced turnover number, thereby accounting for inhibition of plasminogen activation in the presence of apo(a)/Lp(a).     

Mapping of a minimal apolipoprotein(a) interaction motif conserved in fibrin(ogen) beta - and gamma -chains

Lipoprotein(a) (Lp(a)) is a major independent risk factor for atherothrombotic disease in humans. The physiological function(s) of Lp(a) as well as the precise mechanism(s) by which high plasma levels of Lp(a) increase risk are unknown. Binding of apolipoprotein(a) (apo(a)) to fibrin(ogen) and other components of the blood clotting cascade has been demonstrated in vitro, but the domains in fibrin(ogen) critical for interaction are undefined. We used apo(a) kringle IV subtypes to screen a human liver cDNA library by the yeast GAL4 two-hybrid interaction trap system. Among positive clones that emerged from the screen, clones were identified as fibrinogen beta- and gamma-chains. Peptide-based pull-down experiments confirmed that the emerging peptide motif, conserved in the carboxyl-terminal globular domains of the fibrinogen beta and gamma modules specifically interacts with apo(a)/Lp(a) in human plasma as well as in cell culture supernatants of HepG2 and Chinese hamster ovary cells, ectopically expressing apo(a)/Lp(a). The influence of lysine in the fibrinogen peptides and of lysine binding sites in apo(a) for the interaction was evaluated by binding experiments with apo(a) mutants and a mutated fibrin(ogen) peptid. This confirmed the lysine binding sites in kringle IV type 10 of apo(a) as the major fibrin(ogen) binding site but also demonstrated lysine-independent interactions.     

The elastic modulus of fibrin clots and fibrinogen gels: the effect of fibronectin and dithiothreitol

The elastic modulus (G') of factor XIIIa induced fibrinogen gels was found to be substantially lower than the G' of fibrin gels that were formed by clotting fibrinogen with thrombin. The addition of fibronectin and/or the reducing reagent dithiothreitol (DTT) to the factor XIIIa coagulation mixture led to the formation of a weaker gel structure, while the rigidity of thrombin induced clots was not appreciably affected by the inclusion of the DTT but increased somewhat in the presence of fibronectin. The reasons for the differing clot rigidities are discussed in terms of biochemical mechanisms.     

Interaction of fibrinogen-binding tetrapeptides with fibrin oligomers and fine fibrin clots

The polymerization of fibrin, at pH 8.5 and ionic strength 0.45, and under conditions where the action of thrombin on fibrinogen was the rate-determining step, was interrupted by inactivating thrombin with p-nitrophenyl-p′-guanidinobenzoate (NPGB). Addition of the tetrapeptide Gly-Pro-Arg-Pro (GPRP) partially dissociated the fibrin oligomers as shown by subsequent ligation with Factor XIIIa and calcium ion followed by denaturation and gel electrophoresis; polyacrylamide gel electrophoresis with reduction showed a decrease in the proportion of γ-γ ligation compared with controls untreated by GPRP, and agarose gel electrophoresis showed a shift in the distribution of oligomer sizes. The dissociation was accomplished within 15 min and its extent was consistent with establishment of an equilibrium in which two molecules of GPRP react to sever an oligomer. When GPRP was introduced into fine unligated fibrin clots by diffusion, there was some dissociation as shown by differences in the degree of γ-γ ligation after treatment by Factor XIIIa; but the action of GPRP was much slower and less complete than on soluble oligomers. However, even a small amount of dissociation affected the mechanical properties of fine clots profoundly. The shear modulus (measured 25 s after application of stress) decreased progressively with increasing concentration of GPRP introduced by diffusion. The rate of shear creep under constant stress and the proportion of irrecoverable deformation also increased enormously. If the steadystate creep rate is interpreted in terms of an effective viscosity, the latter is decreased by up to three orders of magnitude by the presence of GPRP. In terms of transient network theories of viscoelasticity, the average lifetime of a network strand is greatly diminished. However, the total density of strands remains constant during creep and creep recovery as shown by constancy of the differential modulus or compliance. Removal of GPRP by diffusion only partially restores the original shear modulus and creep behavior of the original clot. Some limited data on the effect of the tetrapeptide Gly-His-Arg-Pro are also reported.     

Characterization of the inhibition of fibrin assembly by fibrinogen fragment D.

Fragment D (Mr 100 000) prepared from a terminal plasmin digest of fibrinogen was isolated and used to study its effect on fibrin formation. Increasing amounts of fragment D added to a solution of fibrinogen and thrombin decrease the rigidity of the resultant gel (10% of control at 2 mol of fragment D/mol of fibrinogen). Half-maximal inhibition is achieved at 1 mol of fragment D/mol of fibrinogen for non-cross-linked clots and at 1/2 mol of fragment D/mol of fibrinogen for cross-linked clots. "Clottability' decreases concomitantly with the rigidity. Only small amounts of fragment D (less than 10% for non-cross-linked gels) are incorporated into the gel. Light-scattering shows an increase in the final fibre thickness at fragment D concentrations up to 2 mol of fragment D/mol of fibrinogen, from 60 molecules/cross-section for the control to 120 molecules/cross-section. Higher fragment D concentrations lead to a decrease in the final fibre thickness. The limit fibre thickness is 8 nm, with a length of 80 nm, which is equivalent to a fibrin trimer. On the basis of results of synthetic-substrate and fibrinopeptide-release assays, it is clear that thrombin inactivation is not responsible for this effect. These data suggest that fragment D may inhibit fibrin formation by blocking the bimolecular polymerization of activated fibrin monomer molecules to form protofibrils, although additional effects on subsequent assembly steps may also be involved.     

Inhibited thrombins. Interactions with fibrinogen and fibrin.

Fibrin-monomer-Sepharose was used to study thrombin binding to fibrin and the role of the enzyme active centre in this interaction. Binding properties of preformed enzyme-inhibitor complexes, as well as inhibition of thrombin already adsorbed to fibrin monomer, were investigated. No apparent difference was found in binding properties of phenylmethanesulphonyl fluoride-, D-Phe-Pro-Arg-CH2Cl- and dansylarginine NN-(3-ethylpentane-1,5-diyl)amide-inhibited thrombins. Also, the elution profile of phenylmethane-sulphonyl fluoride-inhibited thrombin from fibrinogen-Sepharose was identical with that of active thrombin from fibrin-monomer-Sepharose. Thus far the only low-Mr inhibitor that prevents thrombin from binding to fibrin monomer is pyridoxal 5'-phosphate. Preformed hirudin-thrombin complexes do not interact with fibrin. The extent to which the active centre of thrombin associated with fibrin is still accessible to substrates and inhibitors was also studied. Thrombin bound to fibrin hydrolyses a synthetic substrate at the same rate as the free enzyme. Water-soluble low-Mr inhibitors such as D-Phe-Pro-Arg-CH2Cl and dansylarginine NN-(3-ethylpentane-1,5-diyl)amide can readily modify the active centre of the fibrin-associated enzyme, and the active centre is exposed to the degree that displacement of dansylarginine NN-(3-ethylpentane-1,5-diyl)amide by D-Phe-Pro-Arg-CH2Cl is possible without disturbing the binding. Hirudin disrupts the affinity between thrombin and fibrin. These data indicate that the active centre of thrombin associated with fibrin through extended binding is fully exposed and freely accessible. It is possible that extended binding may play a regulatory role in the activation of Factor XIII by thrombin, as well as inactivation of this enzyme by antithrombin III.     

An analysis of the interaction between mouse apolipoprotein B100 and apolipoprotein(a)

The assembly of lipoprotein(a) (Lp(a)) involves an initial noncovalent interaction between apolipoprotein (apo) B100 and apo(a), followed by the formation of a disulfide bond between apoB100 cysteine 4326 and apo(a) cysteine 4057. The structural features of apoB100 that are required for its noncovalent interaction with apo(a) have not been fully defined. To analyze that initial interaction, we tested whether apo(a) could bind noncovalently to two apoB proteins that lack cysteine 4326: mouse apoB100 and human apoB100-C4326G. Our experiments demonstrated that both mouse apoB and the human apoB100-C4326G bind noncovalently to apo(a). We next sought to gain insights into the apoB amino acid sequences required for the interaction between apoB100 and apo(a). Previous studies of truncated human apoB proteins indicated that the carboxyl terminus of human apoB100 (amino acids 4330-4397) is important for Lp(a) assembly. To determine whether the carboxyl terminus of mouse apoB100 can interact with apo(a), transgenic mice were produced with a mutant human apoB gene construct in which human apoB100 amino acids 4279-4536 were replaced with the corresponding mouse apoB100 sequences and tyrosine 4326 was changed to a cysteine. The mutant apoB100 bound to apo(a) and formed bona fide disulfide-linked Lp(a), but Lp(a) assembly was less efficient than with wild-type human apoB100. The fact that Lp(a) assembly was less efficient with the mouse apoB sequences provides additional support for the notion that sequences in the carboxyl terminus of apoB100 are important for Lp(a) assembly.     

Specificity determinants in the interaction of apolipoprotein(a) kringles with tetranectin and LDL

Lipoprotein(a) is composed of low density lipoprotein and apolipoprotein(a). Apolipoprotein(a) has evolved from plasminogen and contains 10 different plasminogen kringle 4 homologous domains [KIV(1-110)]. Previous studies indicated that lipoprotein(a) non-covalently binds the N-terminal region of lipoprotein B100 and the plasminogen kringle 4 binding plasma protein tetranectin. In this study recombinant KIV(2), KIV(7) and KIV(10) derived from apolipoprotein(a) were produced in E. coli and the binding to tetranectin and low density lipoprotein was examined. Only KIV(10) bound to tetranectin and binding was similar to that of plasminogen kringle 4 to tetranectin. Only KIV(7) bound to LDL. In order to identify the residues responsible for the difference in specificity between KIV(7) and KIV(10), a number of surface-exposed residues located around the lysine binding clefts were exchanged. Ligand binding analysis of these derivatives showed that Y62, and to a minor extent W32 and E56, of KIV(7) are important for LDL binding to KIV(7), whereas R32 and D56 of KIV(10) are required for tetranectin binding of KIV(10).     

Interaction of fibrin(ogen) with apolipoprotein(a): further characterization and identification of a novel lysine-dependent apolipoprotein(a)-binding site within the gamma chain 287-411 region

Interaction of lipoprotein(a) with fibrin associated with atherosclerotic lesions promotes its accumulation in the lesions, thereby contributing to the development of atherothrombosis. Numerous studies revealed that this interaction occurs through the apolipoprotein(a) [apo(a)] component of lipoprotein(a) and COOH-terminal Lys residues generated by partial degradation of fibrin with plasmin (a COOH-Lys-dependent mechanism). At the same time, the mechanism of the interaction of apo(a) with intact fibrin(ogen) remained unclear. Our recent study identified the Lys-independent apo(a)-binding sites within the fibrin(ogen) alphaC domains which contribute to an alternative Lys-independent mechanism. In this study, we performed direct measurements of the interaction between apo(a) and various fibrin(ogen) fragments representing the whole fibrin(ogen) molecule except the alphaC regions. The experiments revealed that the apo(a)-binding site, identified previously within fibrinogen gamma chain residues 207-235 [Klose, R., et al. (2000) J. Biol. Chem. 275, 38206-38212], is a high-affinity site and mainly Lys-independent, suggesting that it should also contribute to the Lys-independent mechanism. The experiments also identified a novel Lys-dependent high-affinity apo(a)-binding site within the sequence of gamma chain residues 287-411. This site may provide interaction of apo(a) with intact fibrin(ogen) through another alternative mechanism, which depends on internal Lys residues. Thus, apo(a) may interact with intact fibrin through the Lys-independent and Lys-dependent mechanisms, while the COOH-Lys-dependent mechanism may prevail in the presence of fibrinolytic activity.     

The interaction of tumour-localizing porphyrins with collagen, elastin, gelatin, fibrin and fibrinogen

We have already reported in Balb C mouse transplantable mammary carcinoma, that uroporphyrin I and III are superior as tumour localizers when compared to hematoporphyrin derivative and a derivative thereof, photofrin II. This study compares the binding of porphyrins to proteins which may be found in tumour cells or stroma to investigate whether there is a common binding determinant. Coproporphyrin III and deuteroporphyrin IX which are non-tumour localizing porphyrins, were also part of the comparative study. The interaction of these porphyrins with acid soluble collagen and acid insoluble collagen, elastin, and fibrin was evaluated, and the binding of uroporphyrin isomers I and III and deuteroporphyrin IX to gelatin and fibrinogen, was also determined. The results suggest that collagen, especially the acid soluble form, and gelatin preferentially bind the four porphyrins which localize in mammary carcinoma tissue. The well reported observations that malignant epithelial cells, including breast cancer, produce collagen and contain a rate-limiting enzyme in collagen biosynthesis would support the notion that de novo synthesis of this protein may in part govern the tumour uptake and retention of porphyrins. Elastin, fibrinogen and fibrin showed non-discriminant binding to the porphyrins under study.     

Identification and characterization of novel lysine-independent apolipoprotein(a)-binding sites in fibrin(ogen) alphaC-domains

Accumulation of lipoprotein(a) (Lp(a)) in atherosclerotic plaques is mediated through interaction of fibrin-(ogen) deposits with the apolipoprotein(a) (apo(a)) moiety of Lp(a). It was suggested that because apo(a) competes with plasminogen for binding to fibrin, causing inhibition of fibrinolysis, it could also promote atherothrombosis. Because the fibrin(ogen) alphaC-domains bind plasminogen and tissue-type plasminogen activator with high affinity in a Lys-dependent manner, we hypothesized that they could also bind apo(a). To test this hypothesis, we studied the interaction between the recombinant apo(a) A10 isoform and the recombinant alphaC-fragment (Aalpha-(221-610)) corresponding to the alphaC-domain by enzyme-linked immunosorbent assay and surface plasmon resonance. Both methods revealed a high affinity interaction (Kd = 19-21 nm) between the immobilized alphaC-fragment and apo(a), indicating that the former contains an apo(a)-binding site. This affinity was comparable to that of apo(a) for fibrin. At the same time, no interaction was observed between soluble fibrinogen and immobilized apo(a), suggesting that, in the former, this and other apo(a)-binding sites are cryptic. Further experiments with truncated recombinant variants of the alphaC-fragment allowed localization of the apo(a)-binding site to the Aalpha-(392-610) region. The presence of epsilon-aminocaproic acid only slightly inhibited binding of apo(a) to the alphaC-fragment, indicating the Lys-independent nature of their interaction. In agreement, the influence of plasminogen or tissue-type plasminogen activator on binding of apo(a) to the alphaC-fragment was minimal. These results indicate that the alphaC-domains contain novel high affinity apo(a)-binding sites that may provide a Lys-independent mechanism for bringing Lp(a) to places of fibrin deposition such as injured vessels or atherosclerotic lesions.     

Interaction of fibrinogen and its derivatives with fibrin

The binding between complementary polymerization sites of fibrin monomers plays an essential role in the formation of the fibrin clot. One set of polymerization sites involved in the interaction of fibrin monomers is believed to pre-exist in fibrinogen, while the complementary set of binding sites is exposed after the cleavage of fibrinopeptides from fibrinogen. The polymerization sites present in fibrinogen and its derivatives mediate their binding to fibrin. Although the binding of fibrinogen and its derivatives to fibrin have been qualitatively studied, there has been no systematic, quantitative investigation of their interaction with forming or preformed clots. In the present study, the binding of fibrinogen and fragments DD, D1, and E1 was measured using a sonicated suspension of plasminogen- and thrombin-free human cross-linked fibrin as a model of a preformed clot. Dissociation constants of 0.056, 0.19, and 2.44 microM, and the number of binding sites corresponding to 0.10, 0.21, and 0.13/fibrin monomer unit of fibrin polymer were found for fibrinogen, fragment DD, and fragment D1, respectively. Fragment E1 did not bind to sonicated noncross-linked or cross-linked fibrin suspensions. However, it was bound to forming fibrin clots as well as to fibrin-Celite, suggesting that the binding sites on fibrin involved in the interaction with fragment E1 may have been altered upon sonication. Affinity chromatography of various fibrinogen derivatives on a fibrin-Celite column showed that only part of the bound fragment DD was displaced by arginine, whereas fragments D1 and E1 were completely eluted under the same conditions. The results indicate that interaction of fibrinogen with the preformed fibrin clots is characterized by affinity in the nanomolar range and that binding between fibrin monomers, in the process of clot formation, could be characterized by even a higher affinity.     

Rheology of fibrin clots. IV. Darcy constants and fiber thickness.

Measurements of small oscillatory deformations of a fibrin clot by axial motion of a rod in a closed tube reveal an anomalous mechanical loss due to permeation of fluid through the clot structure. The Darcy constant for permeation can be calculated from data at the frequency where the apparent storage and loss shear moduli are equal, without the necessity of measurements at much lower frequencies as previously employed. From the Darcy constant, the average number of fibrin monomer units (v) per cross-section of a fibrous element of the clot can be calculated; it ranges from 4 to several hundred. In the range of fibrin concentration(c) from 3 to 14 milligrams, v is approximately proportional to c-2 for clots of coarse structure and to c-0.5 for clots of fine structure.     

Rheology of fibrin clots. v. shear modulus,creep, and creep recovery of fine unligated clots

Creep and creep recovery in small shearing deformations have been studied in fibrin clots at pH 8.5 and ionic strength 0.45, where the fine, transparent clot is formed with very little lateral aggregation of protofibrils. The initial shear modulus G₁ was measured 25 s after deformation on clots aged long enough for complete development of structure. For both human and bovine fibrin, the data were approximately described by log G₁ = 1.45 + 1.90 log c, where c is concentration in $\text{g}\text{l}$ and G₁ is in $\text{dyn}\text{cm}{}^2$, over a range of c from 4 to 13 $\text{g}\text{l}$. For bovine clots with completely developed structure, creep and creep recovery showed substantial irrecoverable deformation but the differential modulus G_Δ measured at intervals agreed with G₁ and did not change during the course of the experiment; it also agreed with the value calculated from the initial recovery after removal of stress. Moreover, several tests showed that the course of recovery conformed closely to the Boltzmann superposition principle. Thus the irrecoverable strain was associated with a structural rearrangement which caused no permanent damage. The irrecoverable deformation relative to the initial deformation was proportional to the elapsed time during creep in the early stages with a proportionality constant that decreased somewhat with increasing clot age prior to imposition of stress; it corresponded to a pseudo-viscosity of the order of 10⁷ poise. However, the irrecoverable deformation does not represent viscous flow and appears to approach a limiting value at long times. Experiments on clots without completely developed structure, i.e., with imposition of stress at an earlier clot age, showed an increase in the differential modulus G_Δ during creep. The irrecoverable deformation was greater and a portion of it could be attributed to the balance between two structures formed in the unstrained and strained states. However, unlike the case of ligated clots strained before complete development of structure, where the irrecoverable deformation is entirely due to a two-structure balance, there is also a contribution from structural rearrangement. Experiments with reverse creep and creep recovery showed that the structural rearrangement is symmetrical with respect to direction of deformation. The interpretation of these results in terms of clot structure and internal motions of protofibrils is discussed.