Help is on the way: Monoclonal antibody therapy for multi-drug resistant bacteria

KEYWORDS:

• antibiotic resistance
• drug resistant
• Fc receptor
• Gram negative organisms
• Klebsiella pneumoniae
• monoclonal antibody therapy

This article refers to:

Endotoxin neutralization by an O-antigen specific monoclonal antibody: A potential novel therapeutic approach against Klebsiella pneumoniae ST258

Antibiotic resistance is a substantial global threat to human health. Antibiotic use, overuse, and misuse has led to the emergence of multi-drug resistant bacterial strains that continue to present a major therapeutic challenge in the clinic. A recent mortality caused by a strain of Klebsiella pneumoniae that was resistant to all 26 antibiotics currently available to US clinicians¹ exemplifies the urgent need for potent treatment modalities for drug resistant bacteria that do not increase the risk of antibiotic resistance. One promising alternative to antibiotics are pathogen-specific monoclonal antibodies (mAbs). The concept of mAb therapy originates from the successful development and use of serum therapy for bacterial infections. The efficacy of this modality, which consisted of specific antisera that were the inaugural antimicrobial agents, was validated in clinical trials and used in clinical practice from the early part of the 20th century until the 1940s.² Serum therapy was abandoned with the arrival of antibiotics, in part due to numerous toxicities and the inability to purify or produce antibodies to single determinants at the time. However, today, advances in molecular biology, technology and antibody engineering make it possible to generate defined, homogenous, fully human and/or humanized mAbs with a single antigen-specificity to target a pathogen of interest.³ In fact, mAb generation only requires an antigen (immunogen) and an immune or immunized individual or immunization platform. To date, mouse models, as exemplified by the study of Szijártó and colleagues⁴ that served as the basis for their work reported in this issue of Virulence, are the most tried and true. Due to their potency, specificity, and safety profiles, mAbs have entered the oncology armamentarium and are increasingly used for treatment of rheumatologic and inflammatory diseases.^5,6 Although at present, only one mAb, Palivizumab, for treatment of respiratory syncytial virus (RSV) in high risk infants, is licensed in the US,⁷ several candidate mAbs are now in advanced clinical trials for other infectious diseases such as HIV, Clostridium difficile, rabies prophylaxis and Staphylococcus aureus.^8-11

In this issue of Virulence, Szijártó and colleagues⁴ describe a candidate mAb for treatment of multi-drug resistant Klebsiella pneumoniae carbapenamase (KPC)-producing K. pneumoniae. Globally, most KPC-producing K. pneumoniae isolates are associated with the multilocus sequence type ST258.^12,13 In previous work, the authors established that most of K. pneumoniae ST258 (hereafter, ST258) isolates express the LPS O-antigen, D-galactan-III (gal-III). Based on this finding, they generated gal-III-specific mouse mAbs¹⁴ with the hope of identifying mAbs that might serve as a therapeutic agent for the main ST258 clades. The current study reports the biologic activity in vitro and in vivo efficacy in experimental models of ST258 infection of A1102, a humanized mouse gal-III mAb that expresses human kappa and IgG1 constant regions. The data demonstrate that passive immunization with A1102 before lethal challenge with ST258 whole bacteria or ST258-derived LPS prolongs survival of endotoxin-sensitized mice and protects rabbits from a lethal ST258 challenge. In vitro studies show that the biologic activity of A1102 includes complement- and Fc-independent LPS neutralization that requires divalent binding, and enhancement of human serum bactericidal killing and complement-dependent macrophage (RAW267.4 cell) uptake of ST258.

Interestingly, the in vitro activity of A1102 does not provide a singular correlate of how it mediates protection in vivo. For instance, although it exhibits bactericidal and opsonic activity that require complement in vitro, A1102 and an aglycosylated IgG1 mutant derivative which cannot bind C1q were protective in cobra venom-treated and normal mice, respectively. Based on their findings that complement and C1q-mediated complement activation are dispensable for A1102 efficacy in vivo, the authors suggest that Fc-independent LPS neutralization may be the main mechanism by which A1102 mediates protection in vivo. However, the role that Fc receptors might play in A1102-mediated protection was not directly examined in this study. Of relevance to the question of how A1102 neutralizes LPS in vivo, mAb-mediated toxin neutralization can be Fc receptor dependent.¹⁵ In fact, although complement was unnecessary for the efficacy of a non-opsonic Streptococcus pneumoniae capsular polysaccharide-binding mAb in vivo, FcγRIIIA (and the mAb's Fc) was required for protection against sepsis and pneumonia in mice. However, for other antibodies to S. pneumoniae, complement was required to promote bacterial uptake by phagocytes depending on the amount of antibody present.^16,17 Antibodies are versatile molecules capable of multiple modes of action ranging from direct effects on the targeted microbe to toxin neutralization to enhancement of phagocytosis and immune modulation.¹⁸ Thus, experiments to evaluate the efficacy of whole A1102 versus its F(ab′)2 fragments in normal and Fc receptor-deficient mice could be used in the future to dissect the roles that Fc dependent and Fc independent activities play in A1102-mediated protection in vivo. Since studies of A1102 efficacy in mice feature a species mismatch between human IgG1 and mouse Fc receptors, use of mice expressing human Fc receptors could be considered.¹⁹

Fcγ receptors regulate immune responses upon binding antigen-IgG complexes via an interplay of activating and inhibitory receptors. Different IgG subclasses have different affinities for Fc receptors that in concert can elicit or limit inflammation.²⁰ Human IgG1, the isotype of A1102, binds activating Fc receptors more avidly than the inhibitory Fc receptor.²¹ Thus, the inhibitory Fc receptor might enhance A1102 efficacy by balancing or dampening inflammation. This was the case for an opsonic mAb to S. pneumoniae that required the inhibitory Fc receptor to mediate protection.¹⁷ On the other hand, if F(ab′)₂ fragments of A1102 neutralize LPS in vivo, this might limit cytokine-mediated inflammation triggered by LPS and/or Fcγ receptor activation. However, notably, even though F(ab′)₂ fragments of a S. pneumoniae capsule-specific mAb reduced nasopharyngeal colonization and prevented lung dissemination in mice,²² the whole mAb was required to protect against systemic infection and pneumonia and modulate inflammation in the nasopharynx.^17,22 Along the same lines, F(ab′)₂ fragments of polyclonal IgG provided optimal protection against colonization in another mouse model of S. pneumoniae colonization.²³ In these examples, antibodies that were effective against nasopharyngeal colonization exhibited the ability to agglutinate bacteria. A similar result was obtained with human post-pneumococcal vaccine samples.²⁴

K. pneumoniae possesses a polysaccharide capsule, an important virulence factor and mechanism of antigenic variation. Thus, another possible therapeutic modality for K. pneumoniae infections are capsule-binding mAbs. Consistent with this concept, K1 capsular polysaccharide (CPS)-specific mouse IgG1 mAbs were protective in murine models of K. pneumoniae sepsis and pulmonary infection.²⁵ These antibodies enhanced in vitro Fc receptor mediated phagocytosis of K. pneumoniae and modulated cytokine levels in vivo. However, a drawback of this approach is the heterogeneous nature of the K. pneumoniae CPS, which precludes development of mAbs that could be used irrespective of CPS type. Therefore, antibodies such as A1102 that target conserved antigens, e.g. the LPS O-antigen, may find a faster place in the immunotherapy armamentarium for K. pneumoniae. Nevertheless, use of a multitude of mAbs that target different determinants and work by different mechanisms may be required to protect against various K. pneumoniae strains and disease manifestations.

A major strength of mAb A1102 lies in its potency against experimental K. pneumoniae infection. In general, endotoxin-neutralizing antibodies have had relatively low affinity and historically were largely unsuccessful in the clinic.²⁶ In contrast, A1102 exhibits high affinity, neutralizes LPS better than polymyxin B in vitro, and doses as low as 3 μg for mice and 2 μg/kg for rabbits mediate protection in vivo. The ability to achieve protection with small doses could significantly lower the cost of treatment, highlighting the promising potential of A1102 for therapy of K. pneumoniae infections. However, one caveat is that the current study shows A1102 mediates protection when given before infection. Although high risk patients might be candidates for pre-emptive or prophylactic mAb therapy, further work is needed to establish the potency of A1102 as a therapeutic agent when infection has already occurred. Nonetheless, in the serum therapy era, sera that were protective in experimental models when given before infection were protective in patients presenting with symptoms, though sera had to be given early in the course of disease to be effective.²⁷ In this regard, the revolution in rapid diagnostics that is making pathogen identification possible within hours will enhance the feasibility of mAb-based therapy for K. pneumoniae and other pathogens.

Although ST258 is the most predominant multi-drug resistant K. pneumoniae sequence type, one question raised by this study is how effective A1102 will be against other sequence types. Nevertheless, the data in this study indicate that A1102 is a potent mAb that holds significant promise as a therapeutic agent for multi-drug resistant K. pneumoniae which could help stem the rising tide of antibiotic resistance.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Funding

Funding was provided by NIH grants R01AI123654 and R01AG045044 to LP.