Complementary Assays Helping to Overcome Challenges for Identifying Neuraminidase Inhibitors

Martina Richter; Lilia Schumann; Elisabeth Walther; Anja Hoffmann; Heike Braun; Ulrike Grienke; Judith M Rollinger; Susanne von Grafenstein; Klaus R Liedl; Johannes Kirchmair; Peter Wutzler; Andreas Sauerbrei; Michaela Schmidtke


Future Virology. 2015;10(2):77-88. 

In This Article

Abstract and Introduction


Aims In this study, we analyze the challenges involved in detecting novel neuraminidase inhibitors (NAIs) and offer strategies to overcome them with complementary bioassays.

Materials & Methods We investigated the inhibitory activities of NAIs (oseltamivir, zanamivir, DANA, katsumadain A and remazol) as well as non-NAIs (amantadine, nucleozin and rifampicin) on influenzaviral and bacterial (Streptococcus pneumoniae, Clostridium perfringens and Vibrio cholerae) neuraminidases (NAs) with chemiluminescence (CL)- and fluorescence (FL)-based assays. Furthermore, hemagglutination-based NA inhibition assays were established.

Results Our study shows three types of signal interference affecting the readout of biochemical assays: self-FL (katsumadain A and remazol), FL quenching (rifampicin) and CL quenching (rifampicin, remazol, nucleozin and katsumadain A). These challenges were overcome by hemagglutination-based assays.

Conclusion The latter allow a robust performance in discriminating NAIs and non-NAIs.


Influenza A viruses express two surface glycoproteins, hemagglutinin and neuraminidase (NA), which function in a highly balanced manner.[1] The binding of viral hemagglutinin to sialic acid (functioning as receptor on the cell surface) mediates the attachment of the virus to host cells, while influenza NA allows progeny virus particles to elute from infected cells by enzymatic cleavage of terminal sialic acid residues.[2] In this way, viral NA prevents self-aggregation of progeny virions and facilitates their spread across the respiratory tract. NA therefore represents a target of influenza therapeutics.

Because of the widespread resistance of influenza viruses to ion channel blockers, current antiviral therapy options are limited to NA inhibitors (NAIs). These include oseltamivir and zanamivir, both of which are licensed worldwide since 1999, and peramivir and laninamivir, which have been approved in Japan, China and South Korea in 2010.[3,4] Resistance to NAIs is observed relatively rarely because of the essential catalytic function of these proteins. The emergence and global spread of oseltamivir-resistant H1N1 influenza viruses with H275Y substitution in the season of 2007/2008[5,6] however demonstrated that resistant strains could emerge suddenly. Therefore, the search for novel NAIs is ongoing. Encouraged by the observation that oseltamivir-resistant strains maintain susceptibility to zanamivir, several research groups are investigating structural modifications of oseltamivir[7,8] to overcome loss of activity induced by mutation,[4] in particular H275Y. Other groups are searching for novel scaffolds for NAI acting against oseltamivir- and zanamivir-resistant influenza viruses.[3,9]

Also bacteria, such as Streptococcus pneumoniae, one of the most common causative agent of flu-associated secondary bacterial pneumonia,[10] express three different NAs as virulence factors.[11,12] Pneumococcal NAs were shown to play an important role in the lethal synergism of influenza viruses and pneumococci.[13] According to x-ray data, the conformation of the cleavage site of pneumococcal NA resembles that of the viral enzyme.[14] This renders bacterial NA a pharmaceutically relevant target.

Assays most commonly used to monitor the NAI susceptibility of influenza A viruses are based on the detection of chemiluminescence (CL) or fluorescence (FL) signals.[4,15–18] The CL-based NA-inhibition assay (henceforth referred to as CL assay) uses NA-Star® (1,2-dioxetane derivative of sialic acid; Applied Biosystems, Darmstadt, Germany) as chemiluminescent substrate. The FL-based NA inhibition assay (henceforth referred to as FL assay) is based on 2′-(4-methylumbelliferyl)-α-D-N-acetylneuraminic acid (MUNANA) as fluorogenic substrate. Both assays are standardized and available in kit format. In addition, different protocols for in house FL assay exist. The CL assay is more sensitive than its FL counterpart.[4,19] Studies on the specificity of these assays revealed good correlation between the 50% inhibitory concentration (IC50) values of oseltamivir and zanamivir for NAI susceptible influenza viruses.[4,18,19] However, the FL assay showed the largest difference between IC50 values measured for the H1N1 wild-type and the oseltamivir-resistant virus variants with H275Y substitution,[4] suggesting that this assay is more sensitive in detecting these drug-resistant viruses. Opposing trends with regard to assay sensitivity were identified for the zanamivir-resistant H3N2 influenza viruses carrying the D151G substitution.[20] These findings imply that resistance-conferring substitutions of amino acids do have an impact on the sensitivity of both assays. According to the manufacturer of the CL and FL assays, both tests are also suitable for the assessment of novel potential inhibitors of viral and bacterial NA. Using these assays, several novel NAIs (synthetic molecules and natural products) have been identified and characterized.[7–9,21–23] Recently, the reliability of these assays was questioned, specifically for the testing of flavonoids, the most prominent scaffolds found among various chemical classes of NA inhibiting natural products.[24] Some of them were reported as quenching the FL signal of the fluorogenic cleavage product of MUNANA, 4-methylumbelliferone (4-MU), leading to false-positive results.[25] For these flavonoids also quenching of the CL signal was reported.[24] It was suggested that the quenching effects measured at flavonoid concentrations of 142 and/or 714 µM were mediated by the phenyl-benzpyran scaffold.

Miniaturized assays using fetuin as substrate provide a further possibility to measure anti-NA activity. These assays are typically used to determine the titer of anti-NA antibodies.[26,27] They were also evaluated for their suitability to detect NAI resistance.[28] Fetuin-based assays are colorimetric assays. In contrast to antibodies, oseltamivir and zanamivir, many natural compounds are colored and hence hold the risk of color-based assay interference.

In searching for novel inhibitors of viral and bacterial NAs, we tested more than 500 small molecules for their inhibitory activity on various NAs. This selection comprised synthetic molecules from virtual screening and natural products selected by empirical knowledge (data not shown). As a result, we identified several compounds showing strong inhibitory activity in the CL assay but poor or no measurable activity in the FL assay. In this work, we present a representative collection of compounds that we investigated in detail regarding their behavior in these biochemical assays: known NAIs (oseltamivir, zanamivir and N-acetyl-2,3-dehydro-2-deoxyneuraminic acid [DANA]); novel NAIs (katsumadain A,[21] remazol[9]); and non-NAIs (rifampicin, nucleozin, amantadine). Our aim was to evaluate the significance of the assays for a diverse set of compounds and rationalize the underlying biochemical processes. We also elaborated alternative assays working under conditions similar to the physiological environment (pH, substrate and salt concentration) that reveal more reliabl0e results. In particular, we focused on measuring the inhibition: of viral NA activity in elution of hemagglutination (HA); and of bacterial NAs lectin-based HA (both henceforth referred to as HA assays).