Emerging avian H5N1 and swine H1N1 influenza serotypes are currently the subject of major international research endeavors. Past influenza pandemics have proven that in the absence of proper safeguards, new and highly pathogenic strains of influenza can be extremely deadly. With the rise in the global population and the advent of international travel and commerce, the repercussions of a modern pandemic would be devastating [i
]. Since its initial isolation in 1997 [ii
], there have been a reported 500 cases of H5N1 in humans that have resulted in 296 deaths [iii
]. The majority of these reported cases have resulted from avian to human transmission, but isolated cases of human-to-human transmission have been reported as well [iv
]. As a precaution, governments are stockpiling drugs in the event that a vaccine is not created, is not efficient, or is not able to be produced in a sufficient, global quantity [v
]. Unfortunately, as has been evident with the prescription of broad-spectrum antibiotics [vi
], a few cases of drug-resistant H5N1 strains have already been reported [vii
]. Moreover, the preventative culling of high-risk poultry populations is a common practice, and has led to the destruction of well over 240 million birds [viii
]. The recent global emergence of H1N1 swine influenza, now officially listed by the World Health Organization (WHO) as a pandemic [ix
] and anticipated to infect as many as 2 billion people over the next two years, highlights the continued ability of this organism to evolve and impact world events.
Responding to disease threats of this type is a dual-armed problem. First, surveillance of human and animal populations is essential in order to understand the extent of infection, and to monitor the success of containment or treatment efforts. Second, continued vaccine development and assessment of efficacy is essential as viral populations change. In both cases, a high-throughput assay to monitor immune responses is highly desirable to assess the presence of infection or response to a candidate vaccine. In the context of influenza, standard methods require high biosafety level facilities and cannot be readily implemented in a high-throughput fashion. All of these issues separately point to the need for the development of simple, field-deployable surveillance systems of viral exposure and immune status. The availability of such systems would have implications for improving human health, stabilizing global food supplies, upholding the ethical treatment of animals (by limiting culling), and for anticipating future zoogenic serotypes of influenza.
Immunological assays, intended for population or vaccination monitoring, ultimately require an analytical biomarker indicative of infection or resistance. Hemagglutinin (HA) is the influenza antigen responsible for mediating host cell recognition via surface sialic acid receptors [x
] and is the main antigenic protein on the surface of the influenza virus [xi
]. HA anchored in viral constructs [xii
]. or in recombinant form [xiv
], is the current focus of efforts towards developing effective vaccines. Monitoring an immunologic response to a candidate vaccine typically requires the use of functional assays, such as hemagglutinin inhibition (HAI) and viral microneutralization (MN). As the reference standards to ascertain antibody titers in subject antisera, these tests must be performed in centralized, high biosafety level (BSL-2+ or BSL-3) facilities due to the use of cultures containing proliferative viruses. They provide the principal set of data supporting or refuting the efficacy of a vaccination, but are extremely time- and cost-intensive. HAI and MN assays are also commonly employed as tools for infection surveillance. It is difficult to envision a method for supplanting these assays entirely in the context of vaccine development, as one cannot demonstrate protective immunity without employing live viruses. However, demonstrating the presence of antibodies to specific influenza antigens independent of their protective capability is of considerable importance, and currently both ELISA and Western blot methods are commonly employed for this purpose [xvii
]. A rapid and consistent preliminary assay able to be safely performed in the field or in standard BSL-2 laboratories would dramatically simplify surveillance and vaccine development efforts, allowing rapid profiling of samples. Where necessary, the presence of neutralizing antibodies could be subsequently confirmed for strong-responder samples by HAI or MN. Proteome profiling via protein microarrays has proven useful in many studies focused on understanding basic biochemical processes [xviii
], but, more germane to the research reported herein, microarrays have also been used to discover antigenic proteins and monitor immunological responses to them [xxi
]. Thus, antigen arrays would seem to be ideal for the development of influenza surveillance and immunity screening tools. Unfortunately, most current microarray technologies rely on labeling schemes, and are too unwieldy for field use.
Over the past several years, we have been engaged in the development and characterization of Arrayed Imaging Reflectometry (AIR) [xxiv
]. AIR is an optical biosensor allowing direct observation of target binding-induced perturbation of an antireflective coating on the surface of a silicon substrate. Briefly, the antireflective condition arises when s-polarized light of a specific wavelength and angle is incident upon a thin layer of silicon dioxide, appended with capture molecules, of a particular thickness. The resulting surface is thus highly sensitive to local deviations in the thickness of the interfering film: a film thickening due to specific capture of a target molecule, and the ensuing destruction of the local destructive interference condition, gives rise to signal generation in the form of reflected light. In this manner, multiple probe/target interactions may be rapidly and simultaneously monitored due to the spatial separation in an array without any requirement for secondary antibodies or labeling. As such, AIR appeared to us to be an ideal platform for the development of a rapid influenza immunity screening tool, and we therefore describe here the preparation and evaluation of AIR hemagglutinin isoform arrays for this purpose.
To evaluate the effectiveness of the array, we examined antisera previously obtained from a blinded pool of trial subjects as part of a trial of an inactivated subvirion H5N1 avian influenza vaccine [xxvi
]. These antisera were analyzed to determine immunogenic responses, distinguish placebo subjects, and quantify antigen cross reactivity over a panel of HAs. AIR data were then compared to previously acquired ELISA and Western Blot information. We also report the extension of this methodology and the AIR technique to a microarray format.