Two-step labeling of Staphylococcus aureus with Lysostaphin-Azide and DIBO-Alexa using click chemistry

Abstract

Specific bacteria imaging is highly desirable in clinical diagnostics. Probes enabling rapid and specific diagnostics of bacteria are limited. Current clinical infection diagnostics is time consuming and invasive, relies on microbiological cultures. We investigated the potential of Lysostaphin as a specific probe to label staphylococci in a new labeling protocol. We used azido (N₃)-modified Lysostaphin-N₃ and DIBO-dye in a two-step bacteria-labeling protocol. N₃ and DIBO (di-benzocyclooctyne) are the counterparts of the “click” chemistry. In the first step, Lysostaphin-N₃ binds specifically to Staphylococcus aureus. In the second step, N₃ clicks to DIBO thus achieving the selective for S. aureus labeling. Such a two-step approach effectively distinguishes S. aureus from Escherichia coli; non-toxic and was proven to work in vivo. The two-step labeling protocol is a promising approach for diagnostic imaging of staphylococcal infections in clinical settings.

Introduction

Implant associated bone infections are serious clinical complications with devastating outcomes (Miclau et al., 2010, Zimmerli, 2011). Currently, multiple time consuming and invasive tests are required to confirm the infections. Specific and rapid diagnosis of the infections is highly desirable in order to apply a correct therapy. The main obstacle of non-invasive infection diagnostics is that infection symptoms are often similar to non-infectious pathologies, such as aseptic loosening and non-unions, delayed wound healing, allergic and immune reactions to implants (Love et al., 2009, Parvizi et al., 2009). Diagnostic imaging of infection is based on accumulation of infection markers at the infection site. The most commonly used markers, such as radiolabeled WBC (White Blood Cells) (Palestro et al., 2006), antibodies (Richter et al., 2011) and FDG (Fluorodeoxyglucose) (Gemmel et al., 2010) have drawbacks. These probes lack specificity and sensitivity to infection, risk possible immune reaction, and often display poorly controllable pharmacokinetics (Gemmel et al., 2009, Palestro, 2009, Signore et al., 2010, Signore and Glaudemans, 2011). Furthermore, WBC and antibodies target immune factors rather than bacteria and therefore detect non-infectious inflammatory processes as well as true infections. In the case of FDG the increased metabolic activity of cells in any inflammatory process, infectious or non-infectious, is detected. Infection probes specifically targeting bacteria have potential to overcome those drawbacks. For instance, ubiquicidin, an antimicrobial peptide, was found to be an excellent infection probe (Akhtar et al., 2005, Brouwer et al., 2010), which specifically targets bacterial surface and therefore distinguishes septic and aseptic conditions. However radiolabeled ubiquicidin displayed a low TNT (target to non-target) ratio in preclinical imaging which is a potential limiting factor for its use in human patients (Baldoni, 2009).

Here, we present the new concept of a two-step imaging protocol based on the recognition of bacterial infection with an antimicrobial peptide followed by its conjugation with a detection probe via bioorthogonal copper free “click” chemistry. The copper free click concept allows detection of bacteria in vivo (Baskin et al., 2007, Chang et al., 2010). This concept may be useful in clinical imaging for the careful assessment of infection, especially for the detection of antibiotic resistant, low grade and biofilm related infections. The hypothesis is that a clickable recognition probe such as Lysostaphin-Azide shall selectively target staphylococci. Then a small detection probe such as DIBO-Alexa applied in the second step shall specifically click azides of the bacteria associated Lysostaphin-Azide. This separation of the recognition probe from the detection probe promises an improvement in the efficacy and specificity of bacteria detection.

Lysostaphin was chosen as the recognition probe as it is an enzyme produced by Staphylococcus simulans which actively hydrolyzes coagulase-positive and coagulase-negative staphylococci (Bastos et al., 2010, Kloos and Schleifer, 1975, Kumar, 2008). In particular, it is active against MRSA (Methicillin Resistant Staphylococcus aureus) (Desbois et al., 2010a, Desbois et al., 2010b) and S. aureus and epidermidis biofilms (Wu et al., 2003). It was shown that Lysostaphin has antimicrobial activity against non-dividing and dividing cells (Bastos et al., 2010). Also it has very low to no toxicity and it does not have direct enzymatic activity on eukaryotic cells (Kokai-Kun et al., 2009). Lysostaphin is stable in vivo when conjugated to polyethylene glycol (PEG) (Walsh et al., 2003). Lysostaphin-PEG conjugates have reduced antimicrobial activity. However they have improved pharmacokinetics and reduced human antibody reactivity to Lysostaphin (Walsh et al., 2003) making them useful for a human therapy and imaging diagnostics. Therefore, we believe that Lysostaphin is a good candidate as a staphylococci imaging probe.

Here, we tested feasibility of selective staphylococci detection with modified Lysostaphin using one and two-step labeling protocols in vitro and in vivo. In the two-step labeling protocol Lysostaphin was modified with azide (N₃) groups and applied to two different bacterial species, the Gram positive bacterium S. aureus and the Gram negative bacterium Escherichia coli. Both species were then clicked to a DIBO containing fluorescent probe. Scheme 1 shows the principle of the novel two-step labeling of S. aureus with Lysostaphin-Azide and DIBO-Alexa. In addition cytotoxicity of the click components on eukaryotic cells was performed.