Abstract Book of EAVLD2012 - eavld congress 2012

S1 - O - 07
SUBTYPING OF SWINE INFLUENZA VIRUSES BY MULTIPLEX REAL-TIME PCR
Kees van Maanen 1 , Ingrid Wiggers 1 , Chris Schouten 3 , Tom Duinhof 1 , Paolo Cordioli 2 , Remco Dijkman 1
1
GD-Animal Health Service, Diagnostics,Research and Epidemiology,Deventer , the Netherlands
2 Istituto Zooprofilattico Esperimentale Brescia, the Netherlands
3 DAP Aadal, Erp, the Netherlands
Swine Influenza Virus; real-time PCR; subtyping; lung lavage samples; diagnosis
Introduction
Swine influenza is an acute respiratory disease of swine caused
by type A influenza viruses. In Europe, swine influenza is
considered one of the most important primary pathogens of swine
respiratory disease and infection is primarily with H1N1, H1N2
and H3N2 influenza A viruses. Avian-like H1N1 and human-like
H3N2 swine influenza viruses (SIV) have been considered
widespread among pigs in Western Europe since the 1980s, and
a novel H1N2 reassortant with a human-like H1 emerged in the
mid 1990s. These viruses have remained endemic in European
pig populations but significant differences in the circulation of
these strains occur at a regional level across Europe (1,2).
For diagnosis of swine influenza infections serological and
virological tests can be used. For detection of subtype-specific
antibodies HI-tests are commonly used. However, crossreactions
can occur, and the interpretation can be difficult
depending on the vaccination and infection history of the herd.
For diagnosis of recent infections paired serum samples are
required collected with a 2-3 week interval.
For rapid detection and discrimination of the different swine
influenza subtypes several PCRs have been described. However,
these methods were either conventional PCRs, targetted only the
Haemagglutinin gene, or the coverage for European SIV strains
was too low according to our BLAST analysis results.
Therefore we decided to develop an in-house subtype-specific
SIV real-time PCR. In the course of the project a prototype realtime
PCR for detection and differentiation of SIV (Life
Technologies) became available and was also evaluated.
Materials & methods
In this study two RNA extraction methods (High Pure nucleic acid
kit, Roche Applied Science, and MagMax AM1836, Life
Technologies) were compared in combination with either the inhouse
designed primer/probe sets + AgPath-ID multiplex onestep
RT-PCR kit (Life Technologies) or a prototype SIV
Screening qRT-PCR kit (Life Technologies). All tests were run on
an ABI7500 fast platform.
Primers and probes were designed in silico using GenBank SIV
sequences of the last 15 years. Primers and probes should have
a coverage of at least 95%, and were used in working dilutions of
10 and 5 µM, respectively. Primers and probes were designed to
enable all sorts of combinations (singleplex, multiplex H+N,
multiplex N and multiplex H in parallel). Test results were
compared with the results of an influenza A real-time PCR
targetting the M gene that had been validated and implemented
several years ago.
Detection limits and efficiency were determined for decimal
dilution series of H1N1, H3N2, and H1N2 reference viruses, as
compared with the influenza A matrix real-time PCR. Subtypespecificity
was evaluated by testing 22 different SIV strains
provided by the IZSLER institute, Brescia, Italy. Diagnostic
performance for lung lavage samples was evaluated by testing
influenza A PCR positive pooled lung lavage samples (n=17) of
4, 7 and 10 week old piglets. For some pools the lung lavage
samples that contributed to the pool were also tested individually
(n=25) in order to compare the reproducibility of subtyping results
and the presence of dual infections.
Results
For the haemagglutinin genes of H1N1, H1N2, and H3N2 viruses
81, 36, and 61 entries were aligned for the development of
primer/probe sets. For the neuraminidase genes of the same
subtypes 67, 28, and 25 entries were aligned for the development
of primer/probe sets.
Both for the in-house PCRs and for the prototype SIV Screening
qRT-PCR kit RNA extraction with the MagMax AM1836 kit
yielded 10-100 times lower detection limits than the High Pure
nucleic acid kit. Using the MagMax AM1836 kit for RNA
extraction, the detection limits of the in-house PCRs and the
prototype commercial kit were comparable for the neuraminidase
genes, whereas detection limits showed more variation for the
haemagglutin genes. Overall, the detection limits of the subtypespecific
primer/probe sets were equivalent to 10 times higher
than the detection limit of the influenza A real-time PCR targeting
the matrix gene.
Subtype-specificity of the in-house PCR and the commercial kit
were investigated by blind testing of a panel of Swine influenza
virus isolates (H1N1 n=10; H1N2 n=4; H3N8 n=8) generously
provided by the IZSLER institute, Brescia, Italy. The in-house
PCRs failed to subtype several strains and also one probe
appeared to be labile. In contrast the prototype commercial kit
subtyped 21/22 strains correctly (for one sample a dual infection
was reported, whereas this sample only contained H3N2). All
strains were detected with the influenza A matrix PCR with Ctvalues
ranging from 20-33.
A first diagnostic evaluation of the commercial kit was performed
by testing influenza A positive (results from another laboratory)
pooled lung lavage samples (n=17) from piglets of 4-10 weeks of
age. All pools were also tested in our in-house influenza A PCR
and results were confirmed. In the subtype-specific PCR 15/17
pools yielded a single subtype (H1N2: n=9; H1N1: n=5; H3N2:
n=1), whereas 2 pools yielded 2 or 3 subtypes. Retesting of the
individual lung lavage samples that contributed to the latter pools
confirmed the presence of dual infections in the majority of these
piglets. Retesting of the individual lung lavage samples of three
pools that were subtyped unequivocally yielded exactly the same
subtype in all individual samples.
Discussion & conclusion
Influenza viruses are highly variable RNA viruses that show rapid
evolution of especially the haemagglutinin and neuraminidase
genes. Therefore, development of reliable PCR protocols for the
subtyping of these viruses is not easy. Despite the effort we
spent on selection of primers and probes, the results of our inhouse
PCR were disappointing and inferior when compared to a
prototype commercial real-time PCR kit.
The latter kit showed a good performance when combined with
RNA extraction on a MagMax AM1836 platform. Detection limits
were comparable to tenfold higher than the influenza A matrix
PCR, 95% of well-defined SIV strains were subtyped correctly,
and the test also demonstrated a good diagnostic performance
for subtyping of influenza A positive lung lavage samples.
Although some practices in the Netherlands are very experienced
in collecting lung lavage samples, the method is cumbersome;
collecting other types of samples like nasal swabs or chewing
ropes is much less invasive laborious. However, nasal and oral
virus excretion is often short-lived. In the next study we will
evaluate the diagnostic performance of real-time influenza A PCR
and subtype-specific PCRs for nasal swabs and chewing ropes in
fattening pigs with acute respiratory disease.
Acknowledgements
We thank Life Technologies for generously providing the
prototype SIV Screening qRT-PCR kit for this study.
References
1. Van Reeth, K, Brown, IH, Dürrwald, R, Foni, E, Labarque, G, Lenihan,
P, Maldonado, J,,Markowska-Daniel, I, Pensaert, M, Pospisil, Z, Koch, G
(2008). Seroprevalence of H1N1, H3N2 and H1N2 influenza viruses in pigs
in seven European countries in 2002-2003. Influenza Other Respi Viruses,
2(3), 99-105.
2: Brown, IH (2012). History and Epidemiology of Swine Influenza in
Europe. Curr Top Microbiol Immunol Jan 11. [Epub ahead of print].