The thrombomodulin analog Solulin promotes reperfusion and reduces infarct volume in a thrombotic stroke model

See also Andreou AP, Crawley JTB. Thrombomodulin analogues for the treatment of ischemic stroke. This issue, pp 1171–3. Summary Background: Currently there is no approved anticoagulant for treating acute stroke. This is largely because of concern for hemorrhagic complications, and suggests a critical need for safer anticoagulants. Solulin is a soluble analog of the endothelial cell receptor thrombomodulin, able to bind free thrombin and convert it to an activator of the anticoagulant, protein C. Objective: Solulin was tested for its ability to inhibit middle cerebral artery occlusion (MCAO) induced by photothrombosis, and to restore MCA patency after establishment of stable occlusion. Methods: Cerebral blood flow (CBF) was monitored by laser Doppler for 1.5 h after occlusion and again 72 h later. Results: Solulin treatment 30 min before thrombosis resulted in an approximately 50% increase in time to form a stable occlusion. When administered 30 or 60 min after MCAO, Solulin significantly improved CBF within 90 min of treatment. In contrast, none of the vehicle-treated mice showed restoration of CBF in the first 90 min and only 17% did so by 72 h. Solulin treatment was associated with a significant reduction in infarct volume, and was well tolerated with no overt hemorrhage observed in any treatment group. Mechanistic studies in mice homozygous for the factor (F)V Leiden mutation, suggest that Solulin’s efficacy derives primarily from the anticoagulant activity of the thrombin–Solulin complex and not from direct anti-inflammatory or neuroprotective effects of Solulin or activated protein C. Conclusions: Our data indicate that Solulin is a safe and effective anticoagulant that is able to antagonize active thrombosis in acute ischemic stroke, and to reduce infarct volume.


Structure of Solulin
Solulin is a recombinant soluble analogue of human thrombomodulin (UniProt Database: http://www.uniprot.org/uniprot/P07204) produced from the genetically engineered Chinese Hamster Ovary (CHO) cell line DXB11 carrying a recombinant plasmid with the Solulin gene. Consisting of the extracellular domains of thrombomodulin, Solulin contains 487 amino acids in the mature secreted form. It is distinguished from other available forms of soluble thrombomodulin by several directed mutations: deletion of the first three amino acids of the amino terminus and the last 7 amino acids of the carboxyl terminus, and four single amino acid exchanges: Met 388  Leu, Arg 456  Gly, His 457  Gln, and Ser 474  Ala [1].
The deletion at the N terminus generates a strongly favored signal sequence cleavage site and abolishes the N terminal heterogeneity arising from two common cleavage sites in the wild type sequence. Similarly, removal of a portion of the Cterminal region of the polypeptide, which is not critical to biological function, leads to a polypeptide that terminates in a Pro-Pro sequence at amino acids 489-490, a sequence especially resistant to exocarboxypeptidase activity [2].
For the sake of enhanced protease stability and oxidation resistance, the trypsin cleavage site Arg 456 -His 457 has been replaced with Gly -Gln [3] and the Met 388 with Leu [4]. A final point mutation is Ser 474  Ala, an exchange known to block chondroitin sulphate attachment to the serine/threonine-rich domain [5]. The efficacy of these point mutations has been verified for Solulin [1,3,6]. Experiments using soluble thrombomodulin (extracellular domains expressed in four different cell lines, no further modifications) indicated that the Met 388  Leu mutation conferred a substantial increase in the ability to activate protein C (1.4-2.2 fold in molecules lacking the chondroitin sulphate moiety like Solulin [7]. There are minor differences in domain definitions used for Solulin and the related molecule, thrombomodulin alfa (ART-123). Solulin consists of amino acids 4 to 490, of mature thrombomodulin. This represents a slightly truncated form of the extracellular domains of thrombomodulin, which have been defined as either residues 1-497 [8] or 1-496 [9]. In contrast, ART-123 is comprised of amino acids 1-498, and therefore contains 11 additional residues including either one or two amino acids from the transmembrane domain.

Relevant Pharmacodynamics and Pharmacokinetics
Solulin causes a concentration-dependent activation of TAFI [8,10] as well as of protein C [8,[11][12][13]. In saline containing the minimum constituents only (thrombin and Solulin plus either PC or TAFI), half-maximal rates of activation were reached at ~ 5 nM for TAFI and ~ 20 nM for PC [8]. There are no published studies in normal plasma which show that Solulin shares with rabbit lung thrombomodulin the concentration-dependent transition from prolongation to shortening of clot lysis time due to a change from activation of TAFI to activation of PC [14,15]. However, preliminary data in haemophilic blood suggest very similar effects of Solulin (Data not shown).
Moreover, APC generation is predominant at higher concentrations of Solulin. In human volunteers, a plasma concentration of about 14 nM Solulin was associated with suppression of the endogenous thrombin potential by approximately 90 % [16]. This effect strongly suggests that TAFI activation is hardly possible in this setting, neither by high concentrations of thrombin (which are averted by Solulin) nor by the Solulin/thrombin complex (whose formation will be attenuated by down-regulation of prothrombin activation). Pharmacokinetic studies in mice yield C max  100 nM, and elimination half life  4,5 hr, after a single dose of 0.5 mg/kg of Solulin (Data not shown). These data are important for the present study since they indicate that Solulin concentrations produced at relevant times (0 -60 post-administration) in mice by a dose of 1 mg/kg should clearly result in Solulin concentrations that are by far sufficient for efficient activation of PC.

Solulin and APC
The findings of intracerebal hemorrhage with APC are consistent with a general bleeding risk that is part of its clinical safety profile [17]. If proven in clinical practice, a lower bleeding propensity of Solulin could be explained by dissimilar concentration profiles of APC, being sizable throughout the circulation during direct infusion of APC but very low to nil with infusion of Solulin and only locally increasing upon incidental thrombin generation. Thus, APC will be included in the growing clot and readily antagonize clot growth, whereas Solulin first needs to react with thrombin to activate PC. This difference may also help to explain why APC, unlike solulin, prolongs the lag time of the thrombin burst [18][19][20], a gross measure of the initial clot appearance, which coincides with aPTT.

Ischemic stroke model and STAIR recommendations.
All animal experiments followed the STAIR recommendations for preclinical studies [21]. Sample sizes were calculated based on power analysis of previous studies with this model [22]. Animals were randomized to each treatment group, and treatments were given by the technician with no knowledge of the specific treatments, which were prepared beforehand by a second person. The only animals excluded from analysis were animals that died before the experiments were terminated. The overall mortality of the study was approximately 3% and was evenly spread between groups. All data analysis including stroke volumes, times to occlusion, CBF measurements, haemoglobin analysis, and TUNEL staining was performed by a blinded investigator with no knowledge of the treatment groups, and results were decoded by the senior investigator after analysis.