Preserved Expression of mRNA Coding von Willebrand Factor-Cleaving Protease ADAMTS13 by Selenite and Activated Protein C
- 18 Downloads
In sepsis, the severity-dependent decrease of von Willebrand factor (VWF)-inactivating protease, a disintegrin and metalloproteinase with thrombospondin motifs 13 (ADAMTS13), results in platelet aggregation and consumption, leading to sepsis-associated thrombotic microangiopathy (TMA) and organ failure. Previous reports assessing its functional deficiency have pinpointed involvement of autoantibodies or mutations to propagate thrombotic thrombocytopenic purpura (TTP). However, mechanisms of acquired ADAMTS13 deficiency during host response remain unclear. To enhance understanding of ADAMTS13 deficiency in sepsis, we evaluated changes in expression of mRNA coding ADAMTS13 during septic conditions using primary cellular sources of the protease. We hypothesized that proinflammatory cytokines and constituents of serum from septic patients affect the transcriptional level of ADAMTS13 in vitro, and previously recommended therapeutic agents as adjunctive therapy for sepsis interact therewith. Cultured hepatic stellate cells (HSCs), endothelial cells (HMEC) and human precision-cut liver slices as an ex vivo model were stimulated with sepsis prototypic cytokines, bacterial endotoxin and pooled serum obtained from septic patients. Stimulation resulted in a significant decrease in ADAMTS13 mRNA between 10% and 80% of basal transcriptional rates. Costimulation of selenite or recombinant activated protein C (APC) with serum prevented ADAMTS13 decrease in HSCs and increased ADAMTS13 transcripts in HMEC. In archived clinical samples, the activity of ADAMTS13 in septic patients treated with APC (n = 5) increased with an accompanying decrease in VWF propeptide as surrogate for improved endothelial function. In conclusion, proinflammatory conditions of sepsis repress mRNA coding ADAMTS13 and the ameliorating effect by selenite and APC may support the concept for identification of beneficial mechanisms triggered by these drugs at a molecular level.
Despite therapy with appropriate antibiotics and intensive supportive care, sepsis carries a persistently high morbidity and mortality (1, 2, 3, 4). Sepsis is a progressive injurious process aggravating from a systemic inflammatory response to infection (5). The continuum of sepsis pathophysiology involves a complex integrated response involving immune cell activation, inflammatory mediators and the coagulation system. Central to these responses are alterations in endothelial cell function and a shift to a proinflammatory and procoagulant surface, which amplifies the initial signal intensity (6,7). Alterations in microcirculation and microvascular permeability precede early tissue injury and organ failure (8). However, pathogenic mechanisms and messengers involved in processes resulting in remote organ injury require further investigation.
During sepsis, von Willebrand factor (VWF), a multimeric acute phase protein, is abundantly secreted into plasma by activated endothelial cells and platelets (9,10). VWF in its native, ultralarge isoform (ulVWF) is only detectable in conditions of an activated endothelium (11) and functions as a determinant of platelet activation and aggregation (12,13). ADAMTS13 is a plasma protease primarily synthesized and secreted from hepatic stellate cells (HSCs) and is the only enzyme reported to inactivate ulVWF, thereby reducing its thrombogenicity (14,15). Congenital or immunologically induced deficiency of ADAMTS13 has been identified to result in thrombotic thrombocytopenic purpura (TTP) (16). In sepsis, the term thrombotic microangiopathy (TMA) was introduced to characterize thrombotic occlusion of arterioles less than 100 µm diameter independent of the underlying pathophysiology (17,18). Hence, thrombocytopenia is a risk factor for mortality in critically ill patients (19) and is also an independent predictor of multiple organ failure and death (20, 21, 22).
Further evidence suggests ADAMTS13 deficiency in patients with systemic inflammation and severe sepsis (9,23, 24, 25). This deficiency is associated with an increase in ulVWF and is postulated to cause TMA in patients (9,23,26). The impact of ADAMTS13 deficiency in thrombocytopenia and organ failure is supported by the observation that plasma exchange reversed organ dysfunction in thrombocytopenia-associated multiple organ failure in children (27) and renal deposition of thrombi enriched from activated platelets and VWF in a porcine sepsis model (28). Also, an inverse relationship between severity of inflammation and ADAMTS13 activity is described in clinical studies (29,30). However, the actual cause for ADAMTS13 deficiency is controversial, ranging from its consumption (31), secretion defects (32) or proteolytic degradation by enzymes circulating in plasma of critically ill patients (33,34). Furthermore, an unbalanced expression of ADAMTS13 and VWF was reported in mouse endotoxemia (35) and in the cecal ligation and puncture model of sepsis (36). However, VWF secretion was found to function as the major determinant of sepsis-induced mortality because the complete absence of its regulating protease had only a minor impact in the murine sepsis model.
To gain insight into mechanisms of ADAMTS13 deficiency in systemic inflammation and sepsis, we investigated the effects of proinflammatory cytokines, endotoxins and constituents of serum obtained from septic patients on ADAMTS13 expression on a transcriptional level. Furthermore, we evaluated the contribution of selenite and APC as adjuvant sepsis therapeutics in regulating inflammation-induced ADAMTS13 changes in transcription. Finally, we compared data obtained from in vitro experiments to an observational study cohort for consistency of our findings.
Materials and Methods
Institutional ethical approval by the local ethics committee of the medical faculty of Friedrich-Schiller University Hospital in Jena (A-1827-07/06 and A-1840-08/06) and written informed consent of each patient were obtained.
Cell Culture Experiments
Using cell culture models, an immortalized hepatic stellate cell line (LX-2) (37,38) and a human microvascular endothelial cell line (HMEC) (39), both of human origin, were cultured under standard conditions (LX-2: 1% heat inactivated fetal calf serum [FCS] in Dulbecco’s Modified Eagle Medium (DMEM) media, 5% CO2; HMEC: 10% heat inactivated FCS in DMEM media, 5% CO2). Cells (1 × 106) were seeded in media supplemented with pooled human AB-serum (1%), equilibrated for 2 h and stimulated for 6 or 24 h with human recombinant cytokines (tumor necrosis factor-α [TNFα], interferon γ [IFNγ], interleukin-1β [IL-1β], IL-2 or IL-6; reconstituted in saline supplemented with 0.1% human albumin) to final concentration of 100 ng/mL, with endotoxin (200 ng/mL E. coli, serotype B4; saline supplemented with 0.1% AB-serum), and a mixture of cytokines (TNFα, IFNγ and IL-1β [10 ng/mL] each combined with 100 ng/mL of endotoxin as described previously) (40). The stimulation period was limited to a maximum of 24 h because ADAMTS13 mRNA transcripts remain stable during this period, therefore providing a threshold to identify significant changes in ADAMTS13 transcripts. Longer stimulation times may provoke mRNA changes due to artifacts resulting from cellular stress induced by starvation and accumulation of cell debris and not the intended effect observed by the inflammatory stimuli. A concentration of 5% pooled serum from septic patients (n = 12) was used for stimulation where indicated. Serum samples were obtained within 24 h after the patients met the criteria for sepsis or severe sepsis (5). Pivotal parameters of the serum pool were determined as follows (normal values are given in brackets): creatinine 121 µmol/L [44–106], blood urea nitrogen 37.2 mg/dL [2.6–6.4], γ-glutamyltransferase 80 U/L [6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28], glutamatoxalacetate-transaminase 209 U/L [<18], lactate dehydrogenase 463 U/L [135–223], ammonia 128 µg/dL [19–94], creatine-phosphokinase 640 U/L [<174], albumin 2.6 g/dL [3.5–5.5], glucose 6.9 mmol/L [3.9–5.6], IL-6 2282.75 pg/mL, TNFα below detection limit (10 pg/mL).
Cells were also coincubated with one of the following substances: noradrenalin (NA), protein C-activated (Eli Lilly, Indianapolis, IN, USA), arginine vasopressin (AVP) or desmopressin (DDAVP), sodium selenite (biosyn Arzneimittel GmbH, Fellbach, Germany) and tolvaptan (Sigma-Aldrich, St. Louis, MO, USA).
Human High Precision-Cut Liver Slices (HPLS)
As an established ex vivo model, HPLS from patients who underwent liver resection spatially distant to metastases controlled by macroscopic inspection by two independent and experienced scientists were obtained. Specimens were precision-cut into 200-µm slices (5 mm diameter) with a Krumdieck slicer (Alabama Research and Development, Munford, AL, USA). Slices were then incubated into primary cultures for 24 h before stimulation as described previously (41).
RNA Preparation and Determination of Transcript Level
RNA was isolated with RNeasy kit (Qiagen, Hilden, Germany), reverse transcribed (Bio-Rad, Berkeley, CA, USA) and ADAMTS13 transcripts were quantified using an iCycler. Data were normalized to unstimulated controls with unvaried transcripts (hypoxanthine-guanine phosphoribosyltransferase [HPRT])/disintegrin and metalloproteinase domain-containing protein 9 [ADAM9] for HSCs-HPRT/porphobilinogen deaminase [PBGD] for HMEC) and ratios analyzed according to Pfaffl (42). ADAM9 (stability value [M] < 0.340), GUSB (M < 0.305), PBGD (M < 0.303) and HPRT (M < 0.554) were most suitable for data normalization.
Characterization of Septic Patients and Determination of Laboratory Parameters
Blood was collected in sodium citrate anticoagulant. ADAMTS13 activity was determined using a von Willebrand factor fluorescent resonance energy transfer 73 (VWF-FRETS73) substrate as previously described (43,44). von Willebrand factor antigen (VWF:Ag), von Willebrand collagen binding activity (VWF:CB), von Willebrand ristocetin cofactor assay (VWF:RCo), VWF propeptides were determined as described previously (9,30,45).
Statistical analyses were performed using SPSS 22 (SPSS [IBM, Armonk, NY, USA]). The Student t test or MannWhitney test was used to calculate statistical significance of data sets (P < 0.05). Data are shown as means ± standard deviation of at least four independent experiments. Based on normalized expression data, a log2-threshold value of +1.0 or −1.0 signifying a two-fold variation was used to define transcripts differentially expressed.
Proinflammatory Cytokines and Serum from Septic Patients Suppresses ADAMTS13 Transcription in Primary Cellular Sources
Incubation with Adjuvant Septic Therapeutics Regulates ADAMTS13 Transcription
Proinflammatory Conditions Affect ADAMTS13 Transcription in a Human Ex Vivo Liver Model
To investigate the effects of proinflammatory stimulation on ADAMTS13 transcription in a human ex vivo liver model, we evaluated the expression value of disintegrin and metalloproteinase domain-containing protein 12 (ADAM12) as a marker specific for the activation state of HSCs. We observed an increase of ADAM12 expression values at 6 h after stimulation with TNFα, endotoxin, IL-1β, IL-2 and cytokine mixture (see Figure 3). IFNγ and IL-6 displayed moderate effects that were completely absent following incubation with AVP and DDAVP. Furthermore, stimulation with serum obtained from patients with sepsis also exhibited an increase in ADAM12 expression compared with unstimulated controls. Additionally, in this model, proinflammatory stimulation resulted in a prominently decreased expression of ADAMTS13 by DDAVP and IL-1β (see Figure 3). Stimulation with serum from septic patients resulted in a decreased expression value of ADAMTS13 as compared with unstimulated controls.
Selenite and APC Prevent ADAMTS13 Suppression during Proinflammatory Conditions
APC Administration Increased ADAMTS13 Activity in Patients with Sepsis
Patient characteristics at the day of enrollment.
No adjunctive APC-treatmentb
Number of patients
Sex distribution, male
Length of stay at ICU, days
Length of stay at hospital, days
Mortality rate, %
In this study, we demonstrate a mechanism of ADAMTS13 suppression originating from a decrease in mRNA transcription encoding ADAMTS13. This suppression of transcription was induced by proinflammatory cytokines, which are abundantly secreted in the circulation of septic patients due to an overwhelming immune response during infection (3,4). HSCs and endothelial cells are primary sources of ADAMTS13, but it is unclear to what extent they regulate ADAMTS13 activity. Therefore, we explored the effect of proinflammatory stimulation in HSCs and endothelial cells as the primary cellular sources of ADAMTS13 synthesis and secretion in circulation. The transcriptional suppression of ADAMTS13 mRNA by proinflammatory cytokines may explain the observed loss of proteolytic activity because ADAMTS13 is secreted directly from the Golgi to the cell exterior without storage (50). In addition, secreted ADAMTS13 is inactivated by plasma proteases such as elastase, plasmin or thrombin (29,34) abundantly present in the circulation of septic patients. A significant reduction in ADAMTS13 mRNA in this study is consistent with data obtained from primary HSCs in rats and endothelial cells in humans (32). ADAMTS13 activity is reduced markedly during apoptosis of HSC following treatment with dimethylnitrosamine in rats, but not in animals treated with tetrachloromethane (CCl4) or thioacetamide, where tissue injury is rather characterized by hepatic stellate cell activation and fibrogenesis (51). Transactivation of rat HSC, triggered by cultivation on a plastic surface after isolation or in vivo by administration of CCl4, only controlled the concordant activity of plasma-secreted ADAMTS13 activity marginally (52). We also demonstrate the regulation of ADAMTS13 transcription by vasoactive agents (DDAVP and AVP). At 6 h and 24 h post-stimulation, HSCs and endothelial cells expressed less ADAMTS13 mRNA. However, ADAMTS13 mRNA deficiency should not significantly influence proteolysis of circulating or endothelium-bound VWF because an increase in shear stress during administration of the vasoconstrictors results in the exposure of selective epitopes of VWF, thereby enhancing proteolysis by residual ADAMTS13 as a compensatory event (53). In our study, ADAMTS13 transcript levels and proteolytic activity are downregulated by vasopressin in vitro. Because it is reported that DDAVP action is mediated via the V2 receptor, we determined the role of the V2 receptor using a potent V2-receptor antagonist. From our study, we identified that inhibition of the V2 receptor prevented ADAMTS13 mRNA suppression, suggesting a potential role of the V2 receptor in mediating ADAMTS13 mRNA down-regulation. However, the predefined two-fold threshold value for definition of differentially expressed genes was not attained in HPLS. This may lead to dilution of HSC-specific transcripts in a mixed tissue sample with the reference transcripts simultaneously expressed in hepatocytes. Nevertheless, the variation of ADAM12 upregulation as a marker and ADAMTS13 downregulation in our study supports our proposed mechanism. Given that DDAVP is mediated via the arginine vasopressin receptor 2 (V2 receptor) (54) and suppressed ADAMTS13 mRNA, it could be speculated that a V2-receptor subtype might be present and functionally active in HSCs. Because increased VWF:Ag levels following DDAVP administration are paralleled by a mild decrease of plasma ADAMTS13 activity in human volunteers (55,56), our current observations provide first evidence that the decline might be caused by transcriptional repression.
There is evidence of splice variants of ADAMTS13 in a hepatoma cell line resulting in synthesis of an intracellularly accumulated protease (57). To the best of our knowledge, there was no systematic report on splice variants of ADAMTS13 in human hepatic tumor tissue. Very recent reports (58) found a maintained balance of platelet activation status in patients with cirrhosis and cancer compared with age-related controls.
We also demonstrate that the administration of human recombinant APC resulted in preserved basal transcriptional rate of ADAMTS13 in vitro as well as an increased plasma activity. APC treatment was accompanied by an improved clinical condition of patients, as reflected by enhanced levels of VWF-related parameters and of endothelial dysfunction such as VWF propeptide (59,60). Following APC administration, the mechanism resulting in a decrease in collagen-binding activity, which is the measure of decreased VWF-adhesive activity, is unclear. It might be speculated that an increase in activity of ADAMTS13 by the presence of APC should have induced this phenomenon. The interaction between APC and human HSCs showed an expression of EPCR and an abrogation of thrombin-induced TNFα expression (61). As shown by our blocking experiments, the mechanism of action is strongly associated to the activity of EPCR-receptor. It might be speculated by which pathway the protective effect of APC is triggered along intracellular signaling events, resulting in a rather antiinflammatory condition. Otherwise, events of oxidative stress might be involved in the regulation of ADAMTS13 transcription, while the antioxidative properties of selenite are capable of interrupting the unfavorable effect with respect to ADAMTS13. There is mounting evidence that the prevention of oxidative damage and the attenuation of inflammatory response by selenite is reflected by either a decrease of the transcriptional rate of endothelial adhesion molecules (62) and cytokine-induced cellular activation (63) or the restoration of antioxidative selenoenzymes (64). Strikingly, costimulation with APC and selenite exert at least an additive effect on rescue of ADAMTS13 transcription. The mechanisms involved warrant further investigation. Furthermore, a direct interaction between ADAMTS13 activity and inflammation is provided by the evidence that ADAMTS13 deficiency results in increased leukocyte rolling and adhesion, as well as enhanced extravasation of neutrophils, where both processes are dependent on the presence of ulVWF and platelets (65). In accordance with these observations, our findings suggest that ADAMTS13 deficiency during systemic inflammation and infection accelerate inflammatory responses by slowing down leukocytes and facilitating their extravasation, which functions intrinsically as a protective mechanism, but becomes destructive due to an overwhelming systemic response which might be limited by APC.
In septic patients, the role of selenite supplementation to modulate reactive oxygen species generation, lower endogenous antioxidative capacity and improve clinical outcome is controversial (66, 67, 68). Here, we demonstrate that selenite abrogates the inhibitory effect of serum constituents from septic patients on cultivated HSCs and endothelial cells. It could be speculated that this mechanism involves genes regulating cell cycle, apoptosis and intermediary metabolism (69).
In this study, we obtained a data set with respect to ADAMTS13 regulation from surgical waste material, which is not representative for healthy, naïve tissue, but clearly reflects the comorbid conditions which are decisive for a patient cohort which is susceptible for generalized infection compared with healthy controls. For evaluation of the response generated to inflammatory stimuli, we determined the expression of ADAM9, which is a well-established and validated approach. ADAM9 was decreased in the presence of inflammatory stimuli. The use of HPLS with preserved tissue architecture was to overcome limitations of cell culture experiments. This model is an alternative to mouse models where VWF plasma concentration during the complete absence of ADAMTS13 activity is the major determinant of sepsis-induced mortality, which is unrepresentative of the clinical situation in humans (36). As observed in the mouse endotoxemia model, LPS administration affects both VWF and ADAMTS13 transcription in an opposite manner in hepatic and renal tissue accompanied by variations in plasma concentration (35).
In conclusion, we provide consistent data from in vitro experiments and an observational patient cohort implicating a possible benefit of selenite and APC in mediating ADAMTS13 regulation during sepsis. We also report that the dysbalance ratio between VWF and ADAMTS13 might be governed by treatment of endothelial dysfunction with APC. Proinflammatory cytokines highly abundant in plasma during sepsis, endotoxin and serum from septic patients suppress ADAMTS13 transcripts in primary production sources. These observations may reflect the decline of proteolytic activity of ADAMTS13 in sepsis and sepsis-associated TMA. Moreover, selenite and APC are able to overcome ADAMTS13 suppression on RNA-level in vitro. Suppression of ADAMTS13 also was overcome in patients treated with APC alone. Despite the fact that neither selenite nor APC is currently recommended by the international guidelines for treatment of severe sepsis or septic shock (70), we provide evidence of a direct manipulation of ADAMTS13 transcript level by use of pharmacological agents in sepsis, which may help to characterize and further evaluate the relevance of this therapeutic agents in the management of sepsis. The beneficial action of APC is nearly completely EPC-receptor driven. Furthermore, the interaction of APC with the receptor is also significant for abrogation of effects triggered by proinflammatory cytokines as present in our serum obtained from septic patients.
The authors declare they have no competing interests as defined by Molecular Medicine, or other interests that might be perceived to influence the results and discussion reported in this paper.
The authors thank Gordon Philipp Otto, Sina Coldewey, Brigitte Specht, Barbara Schmidt and Edith Walther for their technical support and helpful comments. The authors acknowledge Utz Settmacher for providing surgical waste material and Dieter Muller for assistance in handling HPLS. This study was supported by grants from the German Research Foundation to RA Claus (DFG CL 173/4–1); German Federal Ministry of Education and Research within the Center for Sepsis Control and Care (grant 01 EO 1002, Project D1.9; PhD fellowship to ML Ekaney); and by the United States National Institutes of Health (DK56621 to SL Friedman).
- 25.Peigne V, et al. (2013) The prognostic value of ADAMTS13 (a disintegrin and metalloprotease with thrombospondin type 1 repeats, member 13) deficiency in septic shock patients involves interleukin-6 and is not dependent on disseminated intravascular coagulation. Crit. Care. 17:R273.CrossRefGoogle Scholar
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, and provide a link to the Creative Commons license. You do not have permission under this license to share adapted material derived from this article or parts of it.
The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this license, visit (https://doi.org/creativecommons.org/licenses/by-nc-nd/4.0/)