Staphylococcal superantigen-like protein 5 induces thrombotic and bleeding complications in vivo: inhibition by an anti-SSL5 antibody and the glycan Bimosiamose


Karlheinz Peter, Baker IDI Heart and Diabetes Institute, PO Box 6492, St Kilda Road Central, Victoria 8008, Australia.
Tel.: +61 38532 1490; fax: +61 38532 1100.


Armstrong PCJ, Hu H, Rivera J, Rigby S, Chen Y-C, Howden BP, Gardiner E, Peter K. Staphylococcal superantigen-like protein 5 induces thrombotic and bleeding complications in vivo: inhibition by an anti-SSL5 antibody and the glycan Bimosiamose. J Thromb Haemost 2012; 10: 2607–9.

Systemic Staphylococcus (S.) aureus infections are associated with disseminated intravascular coagulation (DIC), a disorder characterized by simultaneous thrombosis and bleeding, which ultimately leads to multiple organ failure and thrombocytopenia [1]. S. aureus secretes pathogenic proteins, including the Staphylococcal superantigen-like proteins (SSLs), a protein family structurally homologous to the superantigens (SAg). Generally considered immunomodulatory, SSLs prevent the adhesion and migration of neutrophils to sites of infection [2] and leukocyte activation by chemokines and anaphylatoxins [3].

We and others have shown that one member, SSL5, directly binds to and activates platelets via interaction with glycoproteins (GP)Ibα and GPVI [4,5]. Here, we investigated the pathophysiological relevance of SSL5 as an important mediator, and thus potential therapeutic target, of S. aureus-induced thromboembolic complications in vivo.

To examine whether SSL5 induces acute thromboembolism, C57BL/6 mice were intravenously administered SSL5 (10 μg g−1 body weight (BW)) for 15 min prior to analysis. Immunohistochemical analyses of lungs (Fig. 1A and Fig. S3) found evidence of platelet-rich (CD41+) thrombi (10 ± 1.8 thrombi), which were absent in vehicle-treated mice (Fig. 1B). This finding reflects the major symptom of DIC [6] and provided us with the rationale to pharmacologically target SSL5 interaction with platelets in vivo.

Figure 1.

 (A) Thrombus formation in the lungs of C57BL/6 mice induced by SSL5 (4 μg g−1 BW), as shown in representative hematoxylin and eosin stained sections from mice administered vehicle (top left), SSL5 (top right) and SSL5 co-administered with 5E3 (8 μg g−1 BW; bottom left) or Bimosiamose (3.5 μg g−1 BW; bottom right), magnification × 100. (B) Quantification of SSL5-induced lung thrombi and inhibition by Bimosiamose (3.5 μg g−1BW) and 5E3 (8 μg g−1 BW), n = 4–5 per group. (C) Reduction of platelet adhesion by Bimosiamose (500 μm) in flow chamber experiments with reconstituted blood, flowing (300 s−1) through a SSL5-coated (1 μm) capillary for 5 min, n = 5 independent experiments. (D) Inhibition of SSL5-induced platelet activation by 5E3 in a concentration-dependent manner as measured by P-selectin expression, n = 3 independent experiments. (E) Effect of SSL5 (10 μg mL−1) on platelet adhesion following static incubation in VWF-coated wells with or without pre-activation with ADP (20 μm), n = 7 independent experiments. (F) SSL5 (10 μg g−1 BW) prolonged mouse tail bleeding time compared with vehicle; this was completely ablated when co-administered with Bimosiamose (3.5 μg g−1 BW) or 5E3 (8 μg g−1 BW), n = 7–15 per group. Data presented as mean ± SEM, *< 0.05; **< 0.01; ***< 0.001, by Student’s t-test or one-way anova with Bonferroni post-test as appropriate.

Our identification of key glycan motifs mediating SSL5-receptor interactions led us to test a glycan-based therapeutic, Bimosiamose (Revotar, Hennigsdorf, Germany), which we hypothesized would compete with glycoproteins for SSL5 binding. We have demonstrated that Bimosiamose prevents SSL5-induced platelet activation during static incubation [4]. DeHaas et al. [5] have shown adhesion of platelets to SSL5 under flow in reconstituted blood and we used this model to show that Bimosiamose (500 μm) inhibits platelets adhering to SSL5 (Fig. 1C) whilst under flow (5 min, 300 s−1).

As an alternative therapeutic approach, we developed a monoclonal anti-SSL5 antibody (5E3), which displays strong SSL5 specificity over SSL7 and SSL11 (data not shown), SSL proteins with the highest structural homogeneity to SSL5 [7]. 5E3 blocked SSL5-induced P-selectin expression, a marker of platelet activation, in a concentration-dependent manner (Fig. 1D), but did not inhibit ADP-induced platelet activation (Fig. S1). Flow cytometric analysis confirmed that Bimosiamose and 5E3 directly prevented SSL5 binding to platelets (Fig. S2).

We next tested Bimosiamose and 5E3 as therapeutic approaches in vivo. Compared with SSL5 alone (4 μg g−1 BW; 10 ± 1.8), co-administration with Bimosiamose (3.5 μg g−1 BW; 1.5 ± 0.3; < 0.001) or 5E3 (8 μg g−1 BW; 2.9 ± 0.6; < 0.01) significantly reduced the number of thrombi (Fig. 1B).

Bleeding complications are a major characteristic feature of DIC. Von Willebrand factor (VWF) is a key ligand mediating platelet rolling and contact adhesion via the platelet glycoprotein GPIbα, part of the GPIb-IX-V complex [8]. This interaction is especially important in the initial stages of the thrombotic response and disruption may cause increased bleeding. SSL5 binds to the sulfated tyrosine region of GPIbα [4], a key VWF-binding motif, and thus in addition to activating platelets, may inhibit GPIbα-vWF binding.

SSL5 activates platelets and therefore on its own, can increase adhesion of platelets to VWF under static conditions (Fig. 1E). However, at the same time, SSL5 can have an inhibitory effect on the adhesion of ADP-activated platelets (Fig. 1E; ∼48%; < 0.05). This indicates that when platelets are already activated, SSL5 has an independent inhibiting effect on platelet adhesion. We used a mouse tail-cut bleeding assay to test the potential functional consequence of this inhibition in vivo. SSL5 (10 μg g−1 BW) administered 10 min prior to analysis increased bleeding time (588 ± 72 s) compared with vehicle (147 ± 23 s; < 0.05; Fig. 1F). Both Bimosiamose (224 ± 64 s) and 5E3 (269 ± 39 s) significantly reduced bleeding times compared with SSL5 alone (Fig. 1F).

This study is the first to implicate SSL5 as a mediator of thrombotic and bleeding complications in vivo. Indeed, anti-SSL5 antibodies have been detected in human sera [9], highlighting exposure to SSL5 during S. aureus infection in humans. Moreover, our findings show that two structurally different and novel inhibitory approaches, one glycan based the other antibody based, reverse the effects of SSL5 in vivo. Future investigations using an animal model of S. aureus infection are warranted to validate whether these two compounds have an inhibitory effect in a clinically relevant setting, and may provide insight into additional mechanisms by which SSL5 may contribute to sepsis-induced DIC (e.g. platelet-neutrophil aggregate formation and organ dysfunction).

Mortality rates in S. aureus sepsis are around 20%, and increase above 40% upon developing DIC [10]. Therefore, there is an enormous clinical need to prevent S. aureus-induced DIC by identifying novel therapeutic strategies. Antibody-based therapeutics currently represents the majority of new drug development programs [11], and thus 5E3 is a promising template for further drug development. The concentrations of Bimosiamose used in our study compare well with doses well tolerated in animal studies [12]. Bimosiamose is currently undergoing phase II clinical trials as an anti-inflammatory drug [13], indicating that this glycan-based therapeutic is highly translational and may be rapidly available for clinical use.

In summary, our studies indicate that SSL5 may play a causative role in thrombotic and bleeding complications associated with S. aureus infections and represents a potential therapeutic target. We also demonstrate the therapeutic potential of Bimosiamose and our 5E3 monoclonal antibody, which warrants further examination.


We gratefully acknowledge the contribution of G. Krippner in obtaining Bimosiamose.

Disclosure of Conflict of Interests

The authors state that they have no conflict of interests.