Da

Ta

b

ARRAA

KHHHHHCCA

1

tancfapGfttca

j((

6

n

0d

Journal of Virological Methods 173 (2011) 182–188

Contents lists available at ScienceDirect

Journal of Virological Methods

journa l homepage: www.e lsev ier .com/ locate / jv i romet

evelopment of a high-throughput human rhinovirus infectivity cell-basedssay for identifying antiviral compounds

im Phillipsa,∗, Lesley Jenkinsona,1, Christopher McCraeb,2, Bob Thonga, John Unitta

Bioscience, AstraZeneca R&D Charnwood, Loughborough, Bakewell Road, Leicestershire LE11 5RH, United KingdomAstraZeneca R&D Mölndal, Pepparedsleden 1, SE-431 83 Mölndal, Sweden

rticle history:eceived 29 October 2010eceived in revised form 25 January 2011ccepted 1 February 2011vailable online 15 February 2011

eywords:

a b s t r a c t

Asthma and chronic obstructive pulmonary disease exacerbations are associated with human rhinovirus(HRV) lung infections for which there are no current effective antiviral therapies. To date, HRV infectiv-ity of cells in vitro has been measured by a variety of biochemical and immunological methods. Thispaper describes the development of a high-throughput HRV infectivity assay using HeLa OHIO cellsand a chemiluminescent-based ATP cell viability system, CellTiter-Glo from Promega, to measure HRV-induced cytopathic effect (CPE). This CellTiter-Glo assay was validated with standard antiviral agents andemployed to screen AstraZeneca compounds for potential antiviral activity. Compound potency values in

uman rhinovirusRV serotypesigh-throughput screenTSeLa OHIOytopathic effect

this assay correlated well with the quantitative RT-PCR assay measuring HRV infectivity and replicationin human primary airway epithelial cells. In order to improve pan-HRV screening capability, compoundpotency was also measured in the CellTiter-Glo assay with a combination of 3 different HRV serotypes.This HRV serotype combination assay could be used to identify quickly compounds with desirable broadspectrum antiviral activity.

ellTiter-Glontiviral agents

. Introduction

HRV is a member of the picornavirus family and an impor-ant causative agent of the common cold. Although HRV infectionsre mostly mild and self-limiting they represent a significant eco-omic burden, especially in loss of working hours to society andommerce (Rohde, 2009). This global financial burden is increasedurther by recent findings that HRV is a common pathogen associ-ted with acute exacerbations in asthma and chronic obstructiveulmonary disease (van Rijt et al., 2005; Seemungal et al., 2000;ern, 2009; Leung et al., 2010; Jackson and Johnston, 2010). There-

ore the economic and commercial benefit of an effective antiviral

herapy to treat these clinical conditions is enormous. Althoughhere are many current therapies for relieving the symptoms of theommon cold (Simasek and Blandino, 2007), there are no approvednti-HRV agents on the market.

∗ Corresponding author. Tel.: +44 01509 645926; fax: +44 01509 645506.E-mail addresses: tim.phillips@astrazeneca.com (T. Phillips),

enkinsonl@MedImmune.com (L. Jenkinson), christopher.mccrae@astrazeneca.comC. McCrae), bob.thong@astrazeneca.com (B. Thong), john.unitt@astrazeneca.comJ. Unitt).

1 Present address: MedImmune, Milstein Building, Granta Park, Cambridge CB21GH, United Kingdom.2 Present address: AstraZeneca R&D Mölndal, Pepparedsleden 1, SE-431 83 Möl-dal, Sweden.

166-0934/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.jviromet.2011.02.002

© 2011 Elsevier B.V. All rights reserved.

In a recent publication by De Palma et al. (2008), they reviewedpotential mechanisms for inhibiting HRV infectivity and replica-tion. These mechanisms include (i) blocking viral attachment, entryand uncoating (e.g. viral capsid proteins, intercellular cell adhe-sion molecule-1 receptor (ICAM-1R) and low density lipoproteinreceptors (LDL-R)), (ii) protein processing and replication (e.g. 3Cand 2A proteases, and polymerases), and (iii) assembly and release(e.g. RNA encapsidation). There are known compounds that inhibitselectively each of these processes in the virus lifecycle.

Of interest are several structurally related small molecule com-pounds that inhibit HRV attachment to host cells. These compoundsinclude pirodavir, pleconaril and BTA-798 (Hayden et al., 2003;Barnard et al., 2004; De Palma et al., 2008; Rohde, 2009) that workby binding to the hydrophobic pocket within the virus protein 1(VP-1), part of the virus capsid. Integration of the small moleculeinhibitor into the hydrophobic pocket prevents HRV attachmentand entry into the host cell either via the host cell proteins ICAM-1R or LDL-R. It has now been recognised that the major groupHRV serotypes employ ICAM-1R and the minor group use LDL-Rfor entry into the host cells. Although these anti-HRV compoundsshow promising preclinical and/or preliminary clinical results by

reducing lung viral load following infection, they face further chal-lenges (e.g. DMPK, broad spectrum activity, resistance) to achievethe acceptable clinical efficacy required by the regulatory bodies.There are dedicated efforts by some pharmaceutical companies (e.g.Merck, Janssen, Biota) to find novel anti-HRV agents and this paper

logical Methods 173 (2011) 182–188 183

ditatp

ee2S2vocratriiencavsasa

L1wgotbttse

2

2

scC

2

t(WsEe

C(f

Table 1Properties of HRV serotypes used in the CellTitre-Glo assay.

Human rhinovirus Source HRV-receptor

HRV-1b ATCC VR-481 Minor (LDL-R)HRV-2 ATCC VR-482 Minor (LDL-R)HRV-4 ATCC VR-484 Major (ICAM-1R)HRV-7 ATCC VR-1117 Major (ICAM-1R)HRV-9 ATCC VR-489 Major (ICAM-1R)HRV-10 ATCC VR-490 Major (ICAM-1R)HRV-14 ATCC VR-284 Major (ICAM-1R)HRV-16 ATCC VR-283 Major (ICAM-1R)HRV-22 ATCC VR-1132 Major (ICAM-1R)HRV-24 ATCC VR-1134 Major (ICAM-1R)HRV-29 ATCC VR-1139 Minor (LDL-R)HRV-41 ATCC VR-339 Major (ICAM-1R)HRV-70 ATCC VR-1180 Major (ICAM-1R)

T. Phillips et al. / Journal of Viro

escribes the undertakings here at AstraZeneca. The key objectives to identify novel compounds with better oral DMPK propertieshat inhibit the entry and replication of HRV in the host cell, and solleviate viral-driven exacerbations in asthma and chronic obstruc-ive pulmonary disease, leading to an improved quality of life foratients.

There are many biochemical and immunological methodsmployed to screen compounds for antiviral activity (Last-Barneyt al., 1991; Andries et al., 1992; Smee et al., 2002; Barnard et al.,004; Campbell et al., 2007; Noah et al., 2007; Chu and Yang, 2007;everson et al., 2008; Mondal et al., 2009; Li et al., 2009; Yu et al.,009; Iro et al., 2009). A common method relies on prevention ofirus mediated cell death and is quantified by the measurementf cell viability/cell death post-viral infection. Usually assays forell viability or death may involve a variety of colorimetric, fluo-ometric or other detection systems. A 384-well high-throughputssay using HeLa OHIO cells, which are known to be susceptibleo cell death following exposure to HRV has been developed andeported in this manuscript. This measures HeLa OHIO cell viabil-ty post-infection with HRV and the ability of test compounds tonhibit HRV mediated cell death. Cell viability is assessed at thend of the assay using CellTiter-Glo. CellTiter-Glo determines theumber of viable cells by measuring ATP, a marker for metaboli-ally active cells, using a luciferase based reaction which produceschemiluminescent readout. The CellTiter-Glo cell assay has beenalidated with standard antiviral agents and then used to screen aelection of AstraZeneca compounds. Results from the CellTiter-Glossay agreed with data generated from a physiologically relevantystem, i.e. measuring HRV infectivity and replication in humanirway epithelial cells by quantitative RT-PCR.

To date, there are over 101 HRV serotypes (Uncapher et al., 1991;edford et al., 2004; Piralla et al., 2009) grouped into major (ICAM-R) and minor (LDL-R) subtypes. It is important to identify agentsith the broadest possible anti-HRV serotype inhibition, as no sin-

le HRV serotype is uniquely associated with asthma or chronicbstructive pulmonary disease exacerbations. However, to test allhese HRV serotypes individually in the CellTiter-Glo assay woulde laborious and costly on a routine basis. Initially, a selection ofhree HRV serotypes (representing two major and one minor sub-ypes) was used in combination in the assay. Preliminary findingsuggest that broad spectrum antiviral activity could be identifiedfficiently by this combination method.

. Materials and methods

.1. Chemicals and buffers

All reagents were purchased from Sigma, UK, unless otherwisetated. CellTiter-Glo was purchased from Promega, UK. AstraZenecaompounds and standard antiviral agents were synthesized by thehemistry Department, AstraZeneca R&D Charnwood, UK.

.2. Cell culture

HeLa OHIO cells (Flow Laboratories, Irvine, Scotland) were cul-ured in Eagle’s MEM supplemented with 10% (v/v) foetal calf serumFCS), 1% (v/v) non-essential amino acids and l-glutamine (2 mM).

hen infecting HeLa OHIO cells with HRV to generate serotypetocks (see Section 2.3), cells were cultured in pH indicator freeagles’s MEM (Gibco 51200) supplemented with 1% (v/v) non-

ssential amino acids and l-glutamine (2 mM) only.

Cryopreserved HeLa OHIO cells were routinely used in theellTiter-Glo assay. Cells (5 × 106 cells/ml) were slowly frozen1 ◦C/min) to −80 ◦C in a freezing container (Nalgene 5100 Cryo 1 ◦Creezing container) and stored for up to 12 months at −80 ◦C, or for

HRV-74 ATCC VR-1184 Major (ICAM-1R)HRV-86 ATCC VR-1196 Major (ICAM-1R)HRV-92 ATCC VR-1293 Major (ICAM-1R)

up to 2 years in a liquid nitrogen cryovessel. Cell freezing mediumcontained 80% (v/v) FCS, 10% (v/v) cell culture sterile filtered DMSOand 10% (v/v) Eagle’s MEM.

Human primary airway epithelial cells were isolated from twohuman lung donors and cultured in cell growth medium (Lonzabronchial epithelial basal medium [BEBM, Lonza CC-3171]) andsupplement kit (BEGM Singlequots, Lonza CC-4175).

2.3. Generating HRV serotype stocks

In these experiments, the following major and minor HRVserotypes were used (see Table 1).

Viral stocks were generated by infecting monolayer cultures ofHeLa OHIO cells until cytopathic effects were fully developed (usu-ally 72 h–1 week). Cells and supernatants were then harvested.Cells were disrupted by freezing and thawing (−20 ◦C overnight),cell debris was pelleted by centrifugation and the resulting virus-containing supernatants were stored at −80 ◦C. These unpurifiedsupernatants were used directly in the CellTiter-Glo assay for allserotypes apart from HRV-16, which underwent further purifica-tion before use.

HRV-16 was purified from the cell culture supernatant bypolyethyleneglycol (PEG) precipitation (7% (w/v) PEG 6000, 0.5 MNaCl). HRV-16 virus was resuspended from the PEG pellet inPBS and centrifuged to remove insoluble matter. The HRV-16preparation was passed through a 0.2 �m syringe filter and thenbuffer-exchanged into PBS using an Amicon Ultra (100.000 NMCO)centrifugal filtration device. The resulting purified HRV-16 prepa-ration was dispensed into aliquots and stored at −80 ◦C.

CPE of the viral stocks was assessed by measuring HeLa OHIO cellviability using CellTiter-Glo. A dilution of virus stock that caused90% CPE (the titre used was serotype and batch dependent, with themultiplicity of infection (MOI) ranging from 0.23 to 10) was selectedfor use in the screening assay. Further details of the CellTiter-Gloassay are included below.

2.4. CellTiter-Glo assay using a single HRV serotype

The CellTiter-Glo assay measures the ability of test compoundsto inhibit HRV-induced CPE in HeLa OHIO cells. Cryopreserved HeLaOHIO cells were rapidly thawed in a 37 ◦C water bath, transferredinto 35 ml warmed assay medium (Eagle’s MEM supplemented

with 1% (v/v) FCS and l-glutamine (2 mM)), and centrifuged at300 × g for 5 min. The resulting cell pellet was suspended in assaymedium. Then 5000 cells per well in a volume of 30 �l weredispensed into a Greiner Bio-one 384-well white, solid bottommicro-titre assay plate.

1 logica

aedccw

9wt(ptpattE

uf

swsmaf(fi(cfl

2

usc2

2a

(Awpirot3w(pa

Qt

•

ral agents. Some examples of assay characteristics are shown inTable 2. The assay typically produces a signal-to-noise ratio of 3-fold or greater and exhibits more than acceptable %CV values (<10)and Z′ factors (>0.5) for screening.

Table 2Typical assay characteristics for HRV serotypes 1b, 2, 4, 16, 22, 29, 41, 70, 74 and 92in the CellTitre-Glo assay.

Assay type Positivecontrola (%CV)

Negativecontrolb (%CV)

Z′-factor N

HRV-1b CPE 10.0 8.8 0.6 16HRV-2 CPE 11.4 6.4 0.7 16HRV-4 CPE 8.1 5.7 0.7 16HRV-16 CPE 5.6 8.2 0.6 16HRV-22 CPE 5.5 5.0 0.7 16HRV-29 CPE 7.1 3.1 0.8 16HRV-41 CPE 8.7 5.8 0.6 16HRV-70 CPE 4.8 9.7 0.7 16HRV-74 CPE 3.9 5.9 0.8 16HRV-92 CPE 5.5 5.0 0.7 16

84 T. Phillips et al. / Journal of Viro

Test compounds were dissolved in dimethyl sulphoxide (DMSO)nd were serially diluted in DMSO using half log dilutions to gen-rate 10-point concentration response curves. These were furtheriluted in assay medium and then 5 �l transferred to the assay plateontaining HeLa OHIO cells, such that the top concentration of eachompound was usually 30 �M and the final DMSO concentrationas 0.3% (v/v).

For each HRV serotype test, a virus concentration that caused0% CPE in HeLa OHIO cell death was used. Each HRV serotypeas diluted to a 90% CPE concentration and 5 �l was added to

he assay plate. Assay plates were incubated at 33 ◦C at in 95%/5%v/v) air/CO2 and 95% relative humidity for 48 h. Thereafter, thelates were removed from the incubator and allowed to equilibrateo room temperature for 30 min. CellTiter-Glo reagents were pre-ared according to the manufacturer’s instructions and 5 �l weredded to each well. Assay plates were agitated in a plate mixer andhe chemiluminescence reaction allowed to progress for 10 min inhe dark. Assay plates were read for chemiluminescence using annvision plate reader.

For all HRV serotypes the known anti-viral agent pirodavir wassed as the positive control compound to protect HeLa OHIO cellsrom HRV induced CPE.

Compound induced HeLa OHIO cell toxicity (at 48 h) was mea-ured using an identical format to the CellTiter-Glo assay (as above)ith the exception that HeLa cells were not infected with HRV

erotypes. Compound toxicity was further confirmed in a humanonocytic leukaemia cell line THP-1 (ATC TIB202.F-11838) assay

lso in the absence of virus. The THP-1 cell toxicity assay was per-ormed in a Corning 96-well plate containing 90 �l THP-1 cells40,000 cells/well) and 10 �l compound or vehicle (1% v:v DMSOnal assay concentration). After 24 h at 37 ◦C, 11 �l resazurin45 �M final assay concentration) was added to all wells and theonversion of resazurin to resorufin in the live cells was measureduorometrically at Ex560 nm and Em590 nm.

.5. CellTiter-Glo assay using multiple HRV serotypes

A combination of serotypes HRV-1b, HRV-16 and HRV-70 wassed in the CellTiter-Glo assay. In this assay format, individual viruserotypes were mixed together in assay medium, such that the finaloncentration of the virus used was the same as that used in Section.4. All other assay conditions were identical.

.6. RT-PCR assay for HRV-1b, HRV-14 and HRV-16 infectivitynd replication

Human airway epithelial cells were cultured as indicated abovesee Section 2.2). Cells at passage 4 were used in the infection assay.pproximately 30,000 cells per well (in 100 �l culture medium)ere plated out in a Corning flat bottomed 96-well polystyrenelate and incubated overnight at 37 ◦C/5% (v/v) CO2 in a humidified

ncubator. Compounds were added to give a final concentrationange of 10 nM to 30 �M, as previously described. Virus stocksf HRV-1b, HRV-14 and HRV-16 were diluted in culture mediumo give a MOI of 0.04. The plates were then incubated for 24 h at3 ◦C 95%/5% (v/v) air/CO2 in a humidified incubator. Supernatantsere aspirated from the wells. Aliquots of 100 �l of RLT lysis buffer

Qiagen) were added to the cells (including T = 2 h controls). RNAreparation was then performed using the RNeasy 96 kit (Qiagen),ccording to the manufacturer’s protocol.

Taqman was performed using a one-step RT-PCR kit (Qiagen

uantitect Probe One step RT-PCR kit), according to the manufac-

urer’s protocol.Primers and probe sequences:

HRV-1b Forward: GCAACTCTCCAGGTTGTCTAAG

l Methods 173 (2011) 182–188

• HRV-1b Reverse: TGCGGGTAACGATATCAGTTGT• HRV-1b Probe: AGCACTTCTGTTTCCCCGGTTGACGT• HRV-14 Forward: TTCCCTCCACTAGTTTGGTCGAT• HRV-14 Reverse: AAGGGCGTCCCAGCATAAG• HRV-14 Probe: CCTAGCCTGCGTGGCGGCC• HRV-16 Forward: AGGATGTGTTGGAGAAAGGCATAC• HRV-16 Reverse: GTTATGGTTGAGTCGCCTCTTGTAAT• HRV-16 Probe: AAGTGTAGAAGCTTGTGGATACTCTGATAGA

The RT-PCR was carried out on a Stratagene MX3000P ther-mocycler, using the following conditions: 1 cycle of 30 min/50 ◦Cand 15 min/95 ◦C; 40 cycles of 15 s/94 ◦C and 60 s/60 ◦C, fluores-cence reading taken at end of each cycle. The standard curvewas plotted using the MX3000P software, and the software per-formed RNA quantity calculations. Viral RNA measurements takenat 2 h post-infection were subtracted from all other viral RNAdata, to control for viral RNA present in the initial inoculum.Viral RNA measurements were then normalised to total RNAlevels measured in each well. This was carried out using theRibogreen assay (Invitrogen R11490 Quant-iT Ribogreen Assaykit), according to the manufacturer’s protocol. A cell toxicityassay from Roche (cat no. 05015944001) was used to assesscompound induced toxicity in the human airway epithelialcells.

2.7. Data analysis

The pIC50 value (equivalent to −log10[compound IC50 value]M) of each test compound was determined using a 4-parameterlogistic equation in a non-linear curve fitting routine (Baud, 1993).The Z′ is defined as 1 − (3 × STDEV Pos controls + 3 × STDEV Negcontrols)/[mean Pos controls − Mean Neg controls], where 1 is theoptimal condition (Zhang et al., 1999).

3. Results

3.1. Characteristics of HRV serotypes in the CellTiter-Glo assay

In the CellTiter-Glo assay, a range of different major and minorHRV serotypes were shown to be effective in killing HeLa OHIOcells. This HRV-induced CPE could be prevented by the use of antivi-

Z′-factor is described in Section 2.7.a Positive control (5 �M pirodavir), with the number of replicates per experiment

(N).b Negative control (0.3% DMSO in assay buffer), with the number of replicates per

experiment (N).

T. Phillips et al. / Journal of Virological Methods 173 (2011) 182–188 185

NNN

O

OO

Pirodavir

N

NOF

FF

O O N

Pleconaril

N

NN N

O

3a

as1aiiti6

FHst

1E-41E-51E-61E-71E-81E-9

-20

0

20

40

60

80

100

120 Compound 1

Compound 2

Compound 3

% I

nh

ibitio

n

[Compound] (M)

OO BTA-798

Fig. 1. Structures of standard anti-HRV agents.

.2. HRV serotypes CellTiter-Glo assay profiles of standardntiviral agents and AstraZeneca compounds

The CellTiter-Glo assay was used to profile three standardntiviral agents, pirodavir, pleconaril and BTA-798 (see Fig. 1 fortructures). In sixteen different HRV serotypes (1b, 2, 4, 7, 9,0, 14, 16, 22, 24, 29, 41, 70, 74, 86 and 92), all three antiviralgents were able to fully protect HeLa Ohio cells from virus-

nduced cell death in a concentration-dependent manner. Fig. 2llustrates the potency and concentration inhibition curves of thehree standard antiviral agents for HRV-16. The potency ordern this assay is pirodavir (pIC50 7.2 ± 0.4, n = 94) > BTA-798 (pIC50.8 ± 0.2, n = 10) > pleconaril (pIC50 6.0 ± 0.2, n = 34).

1E-41E-51E-61E-71E-81E-9

-20

0

20

40

60

80

100

120

Pleconaril

Pirodavir

BTA-798

% Inhib

itio

n

[Compound] (M)

ig. 2. Typical concentration inhibition curves of standard antiviral agents in theRV-16 CellTiter-Glo assay (note: the highest concentration of pleconaril (red

quare) has been excluded from the curve fit) (For interpretation of the referenceso color in this figure legend, the reader is referred to the web version of the article.).

Fig. 3. Typical concentration inhibition curves of three AstraZeneca compounds inthe HRV-16 CellTiter-Glo assay (note: the highest concentration of Compound 3 (redtriangle) has been excluded from the curve fit) (For interpretation of the referencesto color in this figure legend, the reader is referred to the web version of the article.).

A selection of AstraZeneca molecules was screened in theCellTiter-Glo assay and the profiles of 3 interesting compounds(Compound 1 pIC50 7.1 ± 0.2, n = 5; Compound 2 pIC50 6.7 ± 0.2,n = 4; Compound 3 pIC50 7.0, n = 2) versus HRV-16 are shown inFig. 3.

Interestingly, both pleconaril and Compound 3 displayed a bell-shaped inhibition profile, with <100% inhibition seen with themaximal concentration of antiviral agent used. In addition to com-pounds 1–3, other AstraZeneca molecules were identified withequivalent activity to the standards used.

3.3. Determination of broad spectrum antiviral activity in theCellTiter-Glo assay using a combination of 3 HRV serotypes

As an extension to the individual HRV CellTiter-Glo assay,an alternative assay format was investigated that used multi-ple serotypes in combination (up to 3 serotypes in one well) toinfect HeLa OHIO cells. This assay would allow the identificationof compounds capable of inhibiting multiple serotypes in a sin-gle assay and mimic more the disease situation in multiple HRVserotypes infecting the same cells. For a compound to demonstratean inhibitory effect under such conditions it would be required toshow activity against all serotypes. However, a compound with lackof inhibition against a single serotype would potentially allow theuninhibited serotype to cause HeLa OHIO cell death and thus appearinactive in the assay. Using these experimental conditions, threestandard antiviral agents and four AstraZeneca small moleculeswere examined against a combination of 3 HRV serotypes (i.e. HRV-1b, HRV-16 and HRV-70).

As shown in Fig. 4, pleconaril has a similar potency againstHRV-1b (pIC50 6.6 ± 0.4, n = 34), HRV-16 (pIC50 6.0 ± 0.2, n = 34)and HRV-70 (pIC50 6.2 ± 0.2, n = 10), when tested individually or incombination (pIC50 6.0 ± 0.2, n = 3). This finding suggests that viralserotype–serotype interaction is minimal and that the serotypesbehave the same when tested alone or in combination. This profilewas also observed for pirodavir and BTA-798. Interestingly, simi-lar compound inhibition profiles were seen with 5 combined HRVserotypes in the assay (1b, 2, 4, 16 and 70) (data not shown).

Four AstraZeneca molecules were selected for testing in the

HRV serotype combination assay. One compound (Compound 5)was inactive against all three individual serotypes, two compounds(Compounds 4 and 6) had equal potency against 2 out of 3 serotypes(i.e. inactive at HRV-1b or HRV-70, respectively), and the final

186 T. Phillips et al. / Journal of Virologica

1E-41E-51E-61E-71E-81E-9

-20

0

20

40

60

80

100

120

140 Pleconaril vs HRV-1b

Pleconaril vs HRV-16

Pleconaril vs HRV-70

Pleconaril vs all 3 HRVs

% In

hib

itio

n

[Compound] (M)

Fig. 4. The effect of pleconaril against single and combined HRV serotypes in theCellTiter-Glo assay. This is a representative plot from at least three independentexperiments (note: the highest concentrations of pleconaril have been excludedfrom the curve fit). (For interpretation of the references to color in this figure legend,the reader is referred to the web version of the article.)

Table 3Compound potencies (pIC50) versus HRV-1b, 16 and 70 in the CellTiter-Glo assay.

Compound HRV-1b HRV-16 HRV-70

Compound 1 7.1 ± 0.2 (5) 7.0 ± 0.3 (5) 5.9 ± 0.2 (3)Compound 4 NAa (3) 6.9 ± 0.2 (3) 6.4 ± 0.2 (3)Compound 5 NAa (3) NAa (3) NAa (3)

o

ci

HttauH

Ftisr

Compound 6 6.0 ± 0.1 (3) 6.3 ± 0.1(3) NAa (3)

a NA, not active (at 10 �M). Arithmetic mean pIC50 value ± SD from the numberf separate determinations in parenthesis.

ompound (Compound 1) had similar potency against all threendividual serotypes (Table 3).

Compound 4 was inactive against HRV-1b and active againstRV-16 and 70 in individual serotype assays. When tested against

hese three serotypes in combination, Compound 4 remained inac-ive (see Fig. 5) despite the fact it had good inhibitory activitygainst HRV-16 and HRV-70. It is likely that HeLa OHIO cell deathnder these conditions was caused entirely by the uninhibitedRV-1b serotype. This observation was also seen for the other

1E-51E-61E-71E-81E-9

-20

0

20

40

60

80

100

120 Compound 4 vs HRV-1b

Compound 4 vs HRV-16

Compound 4 vs HRV-70

Compound 4 vs all 3 HRVs

% In

hib

itio

n

[Compound] (M)

ig. 5. The effects of Compound 4 against single and combined HRV serotypes inhe CellTiter-Glo assay. This is a representative plot from three independent exper-ments for a single serotype and one independent experiment for the combinederotypes. (For interpretation of the references to color in this figure legend, theeader is referred to the web version of the article.)

l Methods 173 (2011) 182–188

AstraZeneca compound that was inactive against HRV-70 (Com-pound 6, data not shown). As expected, the compound that waseither active (Compound 1) or inactive (Compound 5) against allthree individual HRV serotypes was also equally active or inactiverespectively, when tested against a combination of serotypes (datanot shown).

3.4. HRV profiles of AstraZeneca compounds in the RT-PCR assayusing primary airway epithelial cells

The ability of HRV to infect human primary airway epithelialcells was demonstrated with quantitative RT-PCR which measuresviral RNA production. Compounds 1–3 were tested against HRV-1b,HRV-14 and HRV-16 infection of human primary airway epithe-lial cells using two different donors’ cells (Table 4). The potencyof these three AstraZeneca compounds agreed well with the val-ues obtained from the CellTiter-Glo assay. Under these conditions,there was no obvious toxicity of the three compounds (WST-1 kitassay by Roche).

4. Discussion

The principal events involved in any viral infection of host cellsare attachment, absorption, uncoating, nucleic acid/protein synthe-sis, assembly and release. Degenerative changes in host cell viabilitydue to viral infection are known collectively as the cytopathic effect(CPE), which can be observed after 24–48 h post-infection. Theextent and nature of these cellular changes depend on the virusserotype, type of host cells and multiplicity of infection (MOI) used.It is well recognized that for HRVs, there are two main evolution-ary groups that have evolved to use either the ICAM-1R (major) orLDL-R (minor) for host cell entry (Uncapher et al., 1991; Hofer et al.,1994; Neumann et al., 2003).

Normally cell-based assays for HRV infectivity are low-throughput (e.g. 96-well or lower well format densities) thatrequire virus plaque determination or a crystal violet readoutneeding multiple cell fixative steps and handling of corrosivechemicals (Smee et al., 2002; Noah et al., 2007; Campbell et al.,2007; Li et al., 2009). Four different commercial kits were evalu-ated (Toxilight [Lonza]; Tox-8 [Sigma]; MultitoxFluor [Promega];CellTiter-Glo [Promega]) for measuring cell toxicity/viability usingPirodavir and BTA-798, and compared to crystal violet (data notshown). Crystal violet was not the preferred method because itwas inefficient and laborious, involving many serial steps (i.e. fix-ative and wash steps) before the assay plate was read. The aimwas to develop a simpler and quicker method with the samefidelity, but ideally with only a single addition step for the detec-tion system and measuring live HeLa OHIO cells rather than deadcells. Based on these assay requirements, the CellTiter-Glo assaywas chosen as the most suitable to develop for high-throughputscreening.

The CellTiter-Glo assay described here is a 384-well high-throughput homogeneous assay using cryopreserved HeLa OHIOcells, a cell line that is highly susceptible to HRV infection, result-ing in cell death. One advantage of this assay is that the readoutis very stable with a half-life generally >5 h, allowing greater timeand flexibility for batch processing of multiple assay plates. Thehomogeneous format also reduces the number of plate handlingsteps and has higher sensitivity compared with other colorimetricand fluorometric assays tested. However, any compound that inter-

feres directly with chemiluminescence or cell toxicity would resultin a false positive. Such compounds were filtered-out during thescreening process with an appropriate counter-screen assay. Thisassay was identical to the CellTiter-Glo assay with the exceptionthat HeLa OHIO cells were not infected with HRV serotypes.

T. Phillips et al. / Journal of Virological Methods 173 (2011) 182–188 187

Table 4Comparison of compound pIC50 values in the human airway epithelial cells and CellTiter-Glo assays.

HRV-serotype Compound AEC donor 1 (pIC50) AEC donor 2 (pIC50) RT-PCR pIC50 CellTiter-Glo pIC50

HRV-14 Compound 1 6.6 6.2 6.4 6.3Compound 2 NDa 6.1 6.1 6.1Compound 3 6.6 6.6 6.6 6.9

HRV-16 Compound 1 7.0 6.6 6.8 7.1Compound 2 NDa 7.1 7.1 6.7Compound 3 7.0 7.3 7.2 7.0

HRV-1b Compound 1 6.6 6.9 6.8 7.15

tions.

fwsaasn2ft(ci

icmtsDv

sphsairb

tlpwppTap

tdasetrsfi

Compound 2 6.0

a ND, not determined. Arithmetic mean pIC50 value from two separate determina

The CellTiter-Glo assay was robust, reproducible and suitableor screening of both major and minor HRV serotypes. Pleconarilas the only agent that showed activity against all the 17 serotypes

uggesting that the CellTiter-Glo assay will identify broad spectrumntiviral agents. The HRV CPE potency data for the three standardntiviral agents were similar to that reported in the literature, inpite of the differences in CPE detection methods and host cell phe-otype (Andries et al., 1992; Barnard et al., 2004; Ledford et al.,004; Ryan et al., 2005). The bell-shaped inhibition curve obtainedor pleconaril (see Figs. 2 and 4) is unlikely to be a result of celloxicity, as the compound was inactive up to 30 �M in HeLa OHIO48 h) and THP-1 (24 h) cell toxicity assays. Similarly, AstraZenecaompounds that showed a bell-shaped inhibition curve were alsonactive in the THP-1 or HeLa OHIO toxicity assay.

The results support the use of CellTiter-Glo for measuring HRV-nduced cell death of HeLa OHIO cells and potencies of antiviralompounds. This is the first report of the use of CellTiter-Glo toeasure specifically HRV-induced cell death following HRV infec-

ion in HeLa OHIO cell. Previously, this chemiluminescent detectionystem has been used in influenza virus H3N2 infection of Madinarby canine kidney cells (Noah et al., 2007) and in blue tongueirus infection of a BSR cell line (Li et al., 2009).

HRV CPE inhibition profiles for AstraZeneca compounds and thetandard antiviral agents were similar, in that all displayed fullrotection of HeLa OHIO cells from virus-induced CPE. The assayas successfully identified novel AstraZeneca compounds that haveimilar pIC50 values and HRV serotype activities to the standardntiviral agents. This data also suggests that the CellTiter-Glo assays capable of identifying potent anti-HRV compounds. Furthermore,esults from the CellTiter-Glo assay correlated well with an ICAM-1inding FRET assay to live HRV-16 (Newton, in preparation).

HeLa OHIO cells are useful model cells to measure viral infec-ivity, but are far removed from the human diseased target cell, i.e.ung epithelial cells. Therefore it was prudent to confirm compoundotency in human primary airway epithelial cells. RT-PCR detectionas used to measure HRV infection and replication within theserimary lung epithelial cells. All of the three AstraZeneca com-ounds tested had potencies similar to the CellTiter-Glo assay (seeable 4), providing further confidence that the assay will identifyctive compounds in the CellTiter-Glo assay as well as in the humanrimary host cells.

The presence of multiple HRV serotypes in the infected host isypically seen in exacerbations of chronic obstructive pulmonaryisease or asthma. To simulate more closely this clinical conditionnd improve screening capability, a combination of multiple HRVerotypes was used in the CellTiter-Glo assay. Interestingly, the

xperiments consistently identified compounds with broad spec-rum antiviral activity. This method was very cost effective as iteduced the total use of HRV virus, compounds and reagents, andhortened the time for iterative screening. Although only up tove different serotypes were used, it will be interesting to test the

.9 6.0 6.1

AEC – human airway epithelial cells.

maximal serotype capacity that can be used in the combinationassay.

Lastly, this high-throughput CellTiter-Glo assay for measuringHRV infectivity in HeLa OHIO cells has proved useful for identifyingand progressing novel anti-HRV agents in early drug discovery.

Acknowledgements

Thanks to Professor Paul Corris, Freeman Hospital, Newcastleupon Tyne, for supply of human chronic obstructive pulmonarydisease transplant tissue. We are grateful to Paul Willis, Sarah King,James Crawford, Philip Newton, Paul Sharpe, Derek Ogg, AudreySoars, Gary Allenby, Paul Harper and Andy Walkland for their usefuldiscussions and constant support.

References

Andries, K., Dewindt, B., Snoeks, J., Willebrords, R., Van Eemeren, K., Stokbroekx,R., Janssen, P.A.J., 1992. In vitro activity of pirodavir (R77975), a substi-tuted phenoxy-pyridazinamide with broad-spectrum antipicornaviral activity.Antimicrob. Agents Chemother. 36, 100–107.

Barnard, D.L., Hubbard, V.D., Smee, D.F., Sidwell, R.W., Watson, K.G.W., Tucker, S.P.T.,Reece, P.A.R., 2004. In vitro activity of expanded-spectrum pyridazinyl oximeethers related to Pirodavir: novel caspid-binding inhibitors with potent antipi-cornavirus activity. Antimicrob. Agents Chemother. 48, 1766–1772.

Baud, M., 1993. Data analysis, mathematical modeling. In: Masseheff, R.F., et al.(Eds.), Methods of Immunological Analysis, vol. 1: Fundamentals. VCH Publish-ers Inc., New York.

Campbell, C.E., Laane, M.M., Haugarvoll, E., Giaever, I., 2007. Monitoring viral-induced cell death using electric cell-substrate impedance sensing. Biosens.Bioelectron. 23, 536–542.

Chu, J.J.H., Yang, P.L., 2007. C-Src protein kinase inhibitors block assembly and mat-uration of dengue virus. Proc. Natl. Acad. Sci. USA 104, 3520–3525.

De Palma, A.M., Vliegen, I., De Clercq, E., Neyts, J., 2008. Selective inhibitors ofpiconarvirus replication. Med. Res. Rev. 28, 823–884.

Gern, J.E., 2009. Rhinovirus and the initiation of asthma. Curr. Opin. Allergy Clin.Immunol. 9, 73–78.

Hayden, F.G., Herrington, D.T., Coats, T.L., Kim, K., Cooper, E.C., Villano, S.A., Liu, S.,Hudson, S., Pevear, D.C., Collett, M., McKinlay, M., 2003. Efficacy and safety oforal pleconaril for treatment of colds due to picornaviruses in adults: resultsof 2 double-blind, randomized, placebo-controlled trials. Clin. Infect. Dis. 36,1523–1532.

Hofer, F., Gruenberger, M., Kowalski, H., Machat, H., Huettinger, M., Kuechler, E.,Blaas, D., 1994. Members of the low density lipoprotein receptor family mediatecell entry of a minor-group common cold virus. Proc. Natl. Acad. Sci. USA 91,1839–1842.

Iro, M., Witteveldt, J., Augus, A.G.N., Woerz, I., Kaul, A., Bartenschlager, R., Patel, A.H.,2009. A reporter cell line for rapid and sensitive evaluation of hepatitis C virusinfectivity and replication. Antiviral Res. 83, 148–155.

Jackson, D.J., Johnston, S.L., 2010. The role of viruses in acute exacerbations of asthma.J. Allergy Clin. Immunol. 125, 1178–1187.

Last-Barney, K., Marlin, S.D., McNally, E.J., Cahill, C., Jeanfavre, D., Faanes, R.B., Mer-luzzi, V.J., 1991. Detection of major group rhinoviruses by soluble intercellularadhesion molecule-1 (sICAM-1). J. Virol. Methods 35, 255–264.

Ledford, R.M., Patel, N.R., Demenczuk, T.M., Watanyar, A., Herbertz, T., Collet, M.S.,

Pevear, D.C., 2004. VP1 sequencing of all human rhinovirus serotypes: insightinto genus phylogeny and susceptibility to antiviral capsid-binding compounds.J. Virol. 78, 3663–3674.

Leung, T.F., To, M.Y., Yeung, A.C.M., Wong, Y.S., Wong, G.W.K., Chan, P.K.S., 2010.Multiplex molecular detection of respiratory pathogens in children with asthmaexacerbation. Chest 137, 348–354.

1 logica

L

M

N

N

P

R

R

S

88 T. Phillips et al. / Journal of Viro

i, Q., Maddox, C., Rasmussen, L., Hobrath, J.V., White, L.E., 2009. Assay develop-ment and high-throughput anti-viral drug screening against Bluetongue virus.Antiviral Res. 83, 267–273.

ondal, R., Koev, G., Pilot-Matias, T., He, Y., Ng, T., Kati, W., Molla, A., 2009. Develop-ment of a cell-based assay for high-throughput screening of inhibitors againstHCV genotypes 1a and 1b in a single well. Antiviral Res. 82, 82–88.

eumann, E., Moser, R., Snyers, L., Blaas, D., Hewat, E.A., 2003. A cellular receptor ofhuman rhyinovirus Type 2, the very low density lipoprotein receptor, binds totwo neighboring proteins of viral capsid. J. Virol. 77, 8504–8511.

oah, J.W., Severson, W., Noah, D.L., Rasmussen, L., White, E.L., Jonsson, C.B., 2007.A cell-based luminescence assay is effective for high-throughput screening ofpotential influenza antivirals. Antiviral Res. 73, 50–59.

iralla, A., Rovida, F., Campanini, G., Rognoni, V., Marchi, A., Locatelli, F., Gerna,G., 2009. Clinical severity and molecular typing of human rhinovirus C strainsduring a fall outbreak affecting hospitalized patients. J. Clin. Virol. 45, 311–317.

ohde, G., 2009. Drug targets in rhinoviral infections. Infect. Disord.-Drug Targets 9,126–132.

yan, J., Tucker, S.P., Luttick, A., Hamilton, S., Nearn, R.H., 2005. A new oral rhi-novirus inhibitor BTA798. In: 18th Int. Conf. Antivir. Res. , Barcelona, April 11–14,Abstract LB-11.

eemungal, T.A.R., Harper-Owen, R., Bhowmik, A., Jeffries, D.L., Wedzicha, J.A., 2000.Detection of rhinovirus in induced sputum at exacerbation of chronic obstruc-tive pulmonary disease. Eur. Respir. J. 16, 677–683.

l Methods 173 (2011) 182–188

Severson, W., Mcdowell, M., Ananthan, S., Chung, D.-H., Rasmussen, L., Sosa, M.I.,White, E.L., Noah, J., Jonssson, C.B., 2008. High-throughput screening of a100,000-compound library for inhibitors of influenza A virus (H3N2). J. Biomol.Screen. 13, 879–887.

Simasek, M., Blandino, D.A., 2007. Treatment of the common cold. Am. Fam. Physi-cian 75, 515–522.

Smee, D.F., Morrison, A.C., Barnard, D.L., Sidwell, R.W., 2002. Comparison of colori-metric, fluorimetric and visual methods for determining anti-influenza (H1N1and H3N2) virus activities and toxicities of compounds. J. Virol. Methods 106,71–79.

Uncapher, C.R., Dewitt, C.M., Colonno, R.J., 1991. The major and minor group recep-tor families contain all but one human rhinovirus serotype. Virology 180,814–817.

van Rijt, L.S., van Kessel, C.H.G., Boogaard, I., Lambrecht, B.N., 2005. Respiratory viralinfections and asthma pathogenesis: a critical role for dendritic cells. J. Clin.Virol. 34, 161–169.

Yu, X., Sainz Jr., B., Uprichard, S.L., 2009. Development of a cell-based Hepatitis

C virus infection fluorescent resonance energy transfer assay for high-throughput antiviral compound screening. Antimicrob. Agents Chemother. 53,4311–4319.

Zhang, J.-H., Chung, T.D.Y., Oldenburg, K.R., 1999. A simple statistical parameter foruse in evaluation and validation of high throughput screening assays. J. Biomol.Screen. 4, 67–73.