ThromboelasTographic sTudy of fibrin cloT and molecular basis of maximum cloT firmness

Maximum clot firmness (MCF) is the main parameter of thromboelastography (TEG) reflecting the stability of a clot. In this work, we looked for markers that can influence the enhancement of MCF and detec ted molecular markers and blood clotting parameters that can be involved in such mechanisms. Blood samples of pregnant women with placental disorders were collected in the Kyiv Perinatal Center. TEG was performed on whole blood in EXTEM and INTEM tests. APTT, INR, fibrinogen concentration and platelet aggregation were measured using traditional laboratory approaches. D-dimer was detected in sandwich ELISA using monoclo nal antibodies III-3B and II-4D. The relative cross-linking activity of factor XIIIa was measured by the direct quantification of the cross-linked γ-chain of fibrin using Western-Blotting with monoclonal antibody II-4D. D-dimer and fibrinogen concentrations, clotting time in the APTT test, INR and rate of platelet aggregation did not correlate with the MCF. However, we found positive correlations of MCF with factor XIIIa activity: 0.51 and 0.87 for EXTEM and INTEM, respectively. These data indicate that for normal and slightly increased fibrinogen concentrations, fibrin clot firmness will depend mostly on the activity of factor XIIIa. Thus the di rect determination of factor XIIIa activity in blood plasma of patients can be relevant for predicting the risk of intravascular coagulation. evaluation of the content and activity of individual clotting factors or other components of the coagulation system can be useful additions to the TEG diagnostics and should not be neglected.


Maximum clot firmness (MCF) is the main parameter of thromboelastography (TEG) reflecting the stability of a clot.
In this work, we looked for markers that can influence the enhancement of MCF and detec ted molecular markers and blood clotting parameters that can be involved in such mechanisms. Blood samples of pregnant women with placental disorders were collected in the Kyiv Perinatal Center. TEG was performed on whole blood in EXTEM and INTEM tests. APTT, INR, fibrinogen concentration and platelet aggregation were measured using traditional laboratory approaches. D-dimer was detected in sandwich ELISA using monoclonal antibodies III-3B and II-4D. The relative cross-linking activity of factor XIIIa was measured by the direct quantification of the cross-linked γ-chain of fibrin using Western-Blotting with monoclonal antibody II-4D. D-dimer and fibrinogen concentrations, clotting time in the APTT test, INR and rate of platelet aggregation did not correlate with the MCF. However, we found positive correlations of MCF with factor XIIIa activity: 0.51 and 0.87 for EXTEM and INTEM, respectively. These data indicate that for normal and slightly increased fibrinogen concentrations, fibrin clot firmness will depend mostly on the activity of factor XIIIa. Thus the direct determination of factor XIIIa activity in blood plasma of patients can be relevant for predicting the risk of intravascular coagulation. evaluation of the content and activity of individual clotting factors or other components of the coagulation system can be useful additions to the TEG diagnostics and should not be neglected.
K e y w o r d s: thromboelastography, factor XIIIa, fibrinogen, maximum clot firmness, thrombosis.
introduction Thromboelastography (TEG) is a widely used and useful approach that allows determination of overall hemostasis potential and clot properties in the whole blood [1]. Its results mainly provide gene ral information about the risk of bleeding [2,3].
The use of TEG for the prediction of intravascular thrombus formation is under discussion [4].TEG was shown to be informative for the detection of dissemi nated intravascular coagulation in patients with sep sis [5]. It has also been applied for the characteriza tion of blood coagulation potential in patients with abbreviations: PRP -platelet rich plasma; PPP -platelet poor plasma; APTT -activated partial thromboplastin time; INR -international normalized ratio; TEG -thromboelastography; MCF -maximum clot firmness; PT -prothrombin time. chronic kidney disease [6]. Currently, its application is widely discussed for patients with severe conse quences of COVID19 [7,8].
TEG provides a series of coagulation parame ters for the curve of thrombus formation and de struction in vitro, and each of these parameters is associated with molecular markers of blood coagula tion and fibrinolysis that can be evaluated by other methods [9].
In this work, we focused on the possible use of TEG for the prediction of intravascular thrombus formation. We assumed that maximum clot firmness (MCF) is the main TEG parameter reflecting the stability of clot, thus indicating the danger of throm bosis [10,11]. We examined the correlation of MCF with the concentration of fibrinogen, D-dimer, plate let aggregation, APTT, INR and factor XIIIa activity in order to select those molecular markers that most affect fibrin clot firmness.

materials and methods
ethical statement. This clinical trial was ap proved by experts of the Ethics Commission of the Shupyk National Medical Academy of Postgraduate Education (protocol № 14, 07.12.2020) and the Ethics Commission of the Kyiv Perinatal Center (protocol No 3, 05.05.2020) and complies with current legisla tion of Ukraine, modern ethical standards and prin ciples of scientific clinical trials.
reagents. ADP, acrylamide, bisacrylamide and thrombin (50 NIH/ml) were purchased from SigmaAldrich (St Louis, Missouri, USA). Molecular weight markers were from ThermoFisher (Waltham, Massachusetts, USA). APTTreagent, thrombo plastin, protein C activator and control donor blood plasma were from Siemens (Munich, Germany). β-mercaptoethanol was purchased from Bio-Rad Laboratories (Hercules, California, USA). All other reagents were of chemical grade and provided by lo cal suppliers.
Monoclonal antibodies. Mouse IgG monoclonal antibodies ІІ-4D and III-3B both specific to the dif ferent parts of NH 2 -terminal portion of the γ-chain in the fibrin D-domain were obtained at the Palladin Institute of Biochemistry of NAS of Ukraine using a hybridoma technique [12].
Thrombin-like enzyme. Thrombinlike enzyme was purified from the venom of agkistrodon halys halys using Blue-Sepharose (St Louis, Missouri, SigmaAldrich, USA) at the Palladin Institute of bio chemistry of NAS of Ukraine [13].
Blood samples. Somatically healthy women (n = 42) of 1842 years with single spontaneous preg nancies with placental disorders (fetal growth retar dation and/or impaired blood flow in the arteries of the umbilical cord іn the second or third trimesters) were enrolled in the study. These placental dysfunc tion patients did not receive specific antithrombotic treatment. Blood samples of pregnant women were collected at the Kyiv Perinatal Center and analyzed immediately. Venous blood sampling for testing was collected from a peripheral vein using vacuum sys tems into sterile plastic 4 ml tubes, containing 3.8% sodium citrate solution (Eximlab, Kyiv, Ukraine). All women gave oral and written informed consent to be included in the study. This prospective cohort study was approved by the Ethics Commission of the Shupyk National Medical Academy of Postgraduate Education and the Ethics Commission of the Kyiv Perinatal Center (# 3 from 05/05/2020).
Platelet rich plasma (PRP). For the aggregom etry study, PRP was obtained from whole blood by centrifuging at 160 g for 30 min at 25°C.
Platelet-poor plasma (PPP). PRP was centri fuged at 300 g for 15 min at 25°C. PPP was collected above the platelet pellet and frozen at 35°C. PPP was thawed prior to the measurements at 37°C during a period of at least 30 min.
teG. TEG was performed on whole blood using Rotem Delta (Tem Innovations GmbH, Mu nich, Germany). EXTEM and INTEM modes that reflect the formation of fibrin clots initiated by the extrinsic and intrinsic pathways, respectively were measured.
Fibrinogen concentration. Fibrinogen con centration in blood plasma was determined by the modified spectrophotometric method. Blood plas ma (0.2 ml) and phosphate-buffered saline (PBS, 1.7 ml) were mixed in a glass tube. Coagulation was initiated by the addition of 0.1 ml of thrombinlike enzyme from the venom of agkistrodon halys halys (1 NIH/ml) that prevented fibrin cross-linking. The mixture was incubated for 30 min at 37°C. The fi brin clot was removed and dissolved in 5 ml of 1.5% аcetic acid. The concentration of protein was meas ured using a POP spectrophotometer (Optizen, Dae jeon, Korea) at 280 nm (ε = 1,5) [14].
D-dimer concentration measurements in sandwich ELISA. DD-specific monoclonal antibody III-3B was used as the catch-agent. 100 µl of analyzed blood plasma was dissolved 1:10 in PBS with 5% of milk and 0.1% of tween-20. Another DD-specific monoclonal antibody (II4D) was used as the tag agent. Concentration of Ddimer was measured using the calibration curve obtained for purified Ddimer [15]. Normal parameters of Ddimer content in the developed test were calculated as 80 ng/ml. Activated partial thromboplastin time (APTT). APTT was measured according to the following procedure: 0.1 ml of blood plasma was mixed with an equal volume of APTTreagent and incubated for 3 min at 37°C. Then the coagulation was initiated by adding 0.1 ml of a 0.025 M solution of CaCl 2 . The clotting time was monitored using the Coagulometer Solar (Solar, Minsk, Belarus).
Prothrombin time (PT). PT was measured as follows: clotting was initiated by mixing 0.1 ml of blood plasma with 0.1 ml of 0.025 M CaCl 2 and 0.1 ml of thromboplastin reagent (Siemens, Munich, Germany). The clotting time was monitored. Throm boplastin acts through the tissue factor pathway of coagulation and activates only carboxylated and uncleaved forms of prothrombin. Results were pre sented as international normalized ratio (INR) that was calculated by the formula: INR = (CTp/CTd) ISI ; where CTp -patient's blood plasma clotting time, CTd -donor's blood plasma clotting time, ISI -In ternational Sensitivity Index of thromboplastin.
Platelet aggregation. Platelet aggregation was measured based on changes in the turbidity of hu man PRP [16]. In a typical experiment, 250 μl of PRP was incubated with 25 μl of 0.025 M CaCl 2 and 25 μl of 12.5 μM ADP at 37°C. Aggregation was monitored for 10 min using the aggregometer Solar 2110 (Solar, Belarus).
SDS-PAGE/Western-Blotting. The poly peptide chains of fibrin formed in blood plasma were separated using SDSPAGE in 6% gels ac cording to Laemmli [17] in the presence of 0.2% β-mercaptoethanol. The separated bands were trans ferred to a nitrocellulose membrane to specify the γ-chain by immunoprobing. The membrane was blocked with 5% milk in PBS for one hour, incubated with a mouse monoclonal antibody ІІ-4D (specific to the NH 2 -terminal portion of the γ-chain in fibrin Ddomain) for an additional hour and then developed with a HRPlabelled goat antimouse secondary antibody. The bands were visualized using 0.001 M 4chloro1naphtol solution in 0.5 M Tris pH 7.5 and 0.03% H 2 O 2 . Concentrations of the cross linked γ-chains were calculated using densitometry of scanned electrophoregrams with TotalLab TL100 software (Nonlinear Dynamics, NewcastleUpon Tyne, UK) [18].
Statistical analysis. Statistical analysis was performed using Microsoft Excel (Microsoft Corpo ration, Redmond, WA, USA). All assays were per formed in a series of three replicates. The data were fitted with standard errors using Statistica 7 (StatSoft Power Solutions, Round Rock, Texas, USA). Pear son's correlation coefficient was calculated using the Pearson Correlation Coefficient Calculator provided by Social Science Statistics (https://www.socscista tistics.com/).

results
Measuring MCF in patients' blood. Routine co agulation tests do not provide any information on the kinetics of clot formation, clot strength or interac tions between the coagulation components. Applica tion of TEG allowed us to select the group of patients with an increased value of MCF. This parameter was estimated in EXTEM and in INTEM systems when blood coagulation in the test was initiated by the thromboplastin or APTTreagent, respectively. MCF in EXTEM and INTEM tests is dependent on platelet concentration and function as well as plate let-fibrin interactions. Concentrating on only the fibrin properties, we expected not much difference between these two modes. However, the results in dicated that 38% and 20% of patients had increased MCF (> 72 mm) in the INTEM and EXTEM, re spectively, indicating an increase in total coagulation potential in pregnant women (Fig. 1).
Interestingly, we did not find many patients that had a decreased value of MCF (< 50 mm) in either mode. However, this very parameter had to be im portant for predicting an increased risk of intravas cular coagulation. Thus, we studied the correlation of other blood coagulation parameters with MCF using Pearson's correlation coefficient (Table).
Correlation of MCF with basic coagulation markers. Analysis of blood coagulation parameters in patients with placental disorders showed they generally had a higher fibrinogen level compared to the healthy nonpregnant donors, in accord with a previous report [19]. All other studied parameters including platelet aggregation induced by ADP, as well as APTT and INR were not changed (Table). Also, we did not find an increase in D-dimer concen tration, as only 30% of patients had accumulation of the Ddimer above 100 ng/ml. In addition, none of the studied parameters were correlated with MCF. We detected weak or moderate correlation only between MCF in the EXTEM mode and the fibrinogen concentration (r = 0.46) and between MCF in the INTEM mode and the INR (r = -0.46). These findings emphasize once again that the division of the external and in ternal blood clotting pathways is only a conditional distribution. Since all these links are quite intercon nected and work almost simultaneously, they cannot be considered separately from each other. Factor XIIIa cross-linking activity. Factor XIIIa crosslinking activity was evaluated according to the method that was developed by Gryshchuk and coauthors [20]. This method is based on the direct evaluation of crosslinked polypeptide chains of po lymerized fibrin using SDS-PAGE. To adapt it to the study in blood plasma we added the immunodetec tion of cross-linked chains in Western-Blotting.

Basic parameters of the blood coagulation system of patients with placental disorders and their correlation with maximum clot firmness (MCF)
Commercially available methods are based on measuring the ammonium released during fibrin sta bilization using an ammoniumsensitive electrode or alternatively using the NAD(P)Hdependent gluta mate dehydrogenase indicator reaction [21,22]. The main advantage of our immunodetection method is the direct measurement of the amount of cross Fibrin γ-chain-specific monoclonal antibody ІІ-4D was used for detection of cross-linked γ-chain oligomers. Densitometry of protein bands detected with monoclonal antibody allowed us to estimate the relative crosslinking activity of factor XIIIa in each sample of blood plasma.
The Western-Blotting analysis of blood plasma samples from patients with normal or high levels of MCF is presented in Fig. 2. As it is seen in Fig. 2, the amount of cross-linked γ-chain oligomers was sub stantially bigger in the blood plasma samples of pa tients with high level of MCF (> 72 mm) in compari son to those with normal level (5072 mm). Such a difference indicated the differentiation in the activity of crosslinking factor XIIIa. We applied the Pear son's correlation analysis to quantify this difference .
Pearson's correlation coefficients were deter mined for the relative crosslinking activity of fac tor XIIIa and MCF in INTEM or EXTEM. Strong (r = 0.87) and moderate (r = 0.51) positive correlations were found for INTEM and EXTEM, respectively.

discussion
In this study, we examined the blood coagu lation system parameters in patients with placental disorders, focusing mainly on thrombotic compli cations that can be extremely dangerous for mother and fetus [23]. Аctivation of intravascular coagula tion is currently considered as one of the reasons for placental disorders with placentalassociated complications such as fetal growth retardation, pre mature birth, preeclampsia, fetal growth retardation and premature detachment of the normally located placenta [24,25].
In our analysis of the blood coagulation system of patients with placental disorders, we focused on those factors that could influence blood clot firm ness as reflected in changes of the MCF parameter. The most important factors should be fibrinogen and platelets. As the core of a clot, fibrinogen can obvi ously influence mechanical properties of a clot [26]. Furthermore, platelets are important for the filling of a clot 'body' and make it stiffer due to retraction [27]. However, only the fibrinogen level weakly cor related with the MCF. The platelet aggregation level showed no correlation with the MCF.
The correlation of MCF with the APTT and INR was predictable as far as these two tests can be assumed as another way to study intrinsic and extrinsic pathways of the coagulation cascade [28]. The only difference is that in these tests, the coagu lometer monitors the time of clot formation in blood plasma and the TEG monitors the maximum rigidity of the blood clot. However, again we did not find any correlations. We found the same situation with the Ddimer, which is assumed to be the most important molecular marker for indicating intravascular throm bosis [29].
Our results indicated strong positive correla tions of MCF with factor XIIIa activity: 0.51 and 0.87 for EXTEM and INTEM, respectively. Thus, our findings indicated a noteworthy contribution of factor XIIIa crosslinking to the MCF (Fig. 3). Neither the platelet aggregation level nor the total fibrino gen concentration influenced this parameter. Also, the accumulation of Ddimer that appears as a result of plasmin hydrolysis of covalently cross linked fibrin does not allow the prediction of the me chanical properties of a clot.
Summarizing the obtained data, we can pre dict that in some range of fibrinogen concentrations covering normal and slightly increased values, the MCF of fibrin will depend only on the activity of factor XIIIa. Being proved in vitro in model systems this conclusion would open a new direction in the search for ways to regulate thrombus formation and digestion. Also, our findings indicate the potential rele vance of determining factor XIIIa activity in blood plasma of patients when intravascular coagu lation is expected. The study of parameters of clot firmness by TEG and factor XIIIa crosslinking activity in pa tients with high risk for placental disorders during pregnancy is a promising strategy for predicting perinatal complications and making timely decisions about anticoagulant therapy. Our findings that sug gest the impact of factor XIIIa crosslinking activity on fibrin clot firmness correspond to data indicating the importance of this parameter as a part of TEG studies [30].
conclusion. Factor XIIIa crosslinking activity had the most influence on the MCF. Platelet aggrega tion and fibrinogen concentration had minor effects on this parameter. Also MCF did not correlate with molecular markers of intravascular thrombus forma tion (Ddimer) and with basic clotting tests (APTT, PT). Factor XIIIa crosslinking activity should be evaluated for the accurate analysis of mechanical properties of fibrin clots during pathologies accom panied by intravascular thrombosis.

Conflict of interest. Authors have completed
the Unified Conflicts of Interest form at http://ukr biochemjournal.org/wpcontent/uploads/2018/12/ coi_disclosure.pdf and declare no conflict of interest.