Highly heterogeneous temperature sensitivity of 2009 pandemic influenza A ( H 1 N 1 ) viral isolates , northern France

I Pelletier1,2, D Rousset3,4,5,6, V Enouf3,4,5,6, GROG7, F Colbère-Garapin1,2 S van der Werf3,4,5,6, N Naffakh (nadia.naffakh@pasteur.fr)3,4,5 1. Institut Pasteur, Unité de Biologie des Virus Entériques, Département de Virologie, Paris, France 2. INSERM U994 (French National Institute of Health and Medical Research) Paris, France 3. Institut Pasteur, Unité de Génétique Moléculaire des virus à ARN, Département de Virologie, Paris, France 4. CNRS URA3015 (French National Centre for Scientific Research), Paris, France 5. Université Paris Diderot, Sorbonne Paris Cité, Unité de Génétique Moléculaire des virus à ARN, Paris, France 6. Institut Pasteur, Centre National de Référence des virus influenzae (Région Nord), Paris, France 7. Groupes Régionaux d’Observation de la Grippe, Open Rome, Paris, France


Introduction
A novel influenza A(H1N1) virus emerged in April 2009 [1][2][3] and rapidly spread all over the world.In France, the first cases were identified in early May 2009.The 2009 pandemic A(H1N1) virus presented a unique combination of genomic segments that had not been reported previously [4].The segments coding for the neuraminidase (NA) and the matrix (M) proteins of the virus were related to the Eurasian lineage of swine influenza A(H1N1) viruses, whereas the six remaining gene segments were related to triple swine-human-avian influenza A(H1N1) reassortants that have been isolated from humans in contact with pigs in North America since 1998 [5,6].Although the properties of isolates of the 2009 pandemic influenza A(H1N1) virus have already been largely examined in vitro and in vivo (for a review, see [7]), sensitivity to elevated temperature has not been characterised precisely.Temperature sensitivity is an important viral phenotypic marker, as it may be involved in host species restriction, tissue specificity and/or virulence [8][9][10][11].In humans and pigs, influenza A viruses initially replicate in the upper respiratory tract at temperatures close to 33 °C and 37 °C, respectively, whereas in aquatic birds, influenza A viruses with low pathogenicity preferentially replicate in the intestinal tract at a temperature close to 40 °C [12][13][14].The sensitivity of avian influenza A viruses to low temperature (33 °C) has been clearly demonstrated [15,16].In contrast, no reduction in viral multiplication at 33 °C was observed for the swine viruses, and it has been proposed that temperature sensitivity might represent a specific, host-dependent signature of influenza A viruses [17].Depending on the optimal temperature for viral multiplication, fever in infected patients may either limit or facilitate viral multiplication and consequently the administration of anti-pyretic drugs may or not be beneficial.Treatment of ferrets infected with influenza virus with sodium salicylate (an anti-pyretic) resulted in increased viral loads in nasal washes [18].
In order to characterise and compare the temperature sensitivity of both pandemic influenza A(H1N1) viral isolates and seasonal viruses isolated in northern France in 2007/08 and 2008/09 before the emergence of the pandemic virus, we developed a test to compare viral multiplication at 34 °C, 37 °C and 40 °C.

Virus samples and reference isolates
We included 23 isolates of 2009 pandemic influenza A(H1N1) virus and 18 seasonal influenza A(H1N1) viral isolates in our study.The pandemic isolates were collected in northern France between weeks 39 and 51 ( as a control to assess the reproducibility of our experimental conditions.A further 12 seasonal influenza A(H1N1) viral isolates from 2007/08, either susceptible or resistant to oseltamivir, and five seasonal influenza A(H1N1) viral isolates from 2008/09, all resistant to oseltamivir, were also included.These seasonal viral isolates were chosen at the beginning and peak of the influenza seasons in northern France, as for the pandemic isolates, and, for the 2007/08 seasonal isolates, we also took into account the co-circulation of viruses sensitive or naturally resistant to oseltamivir.
Among the 23 pandemic influenza A(H1N1) viral isolates included in our study, we defined two distinct groups of viruses according to the disease severity of the patients (Table1).Information about the existence of underlying conditions prone to increase disease severity was noted when available (Table 1).Severe influenza cases were those who were hospitalised in an intensive care unit or died as a result of their infection.Patients with mild disease were matched as much as possible by the week and geographical area of collection.Swine/Hong Kong/1578/2003) [20], respectively, were tested in parallel.

Preparation and analysis of viral isolates
In order to produce suitable viral stocks, all isolates were amplified by two serial passages at a multiplicity of infection of 10 -3 plaque forming units per cell at 35 °C in MDCK cells in serum-free minimal essential medium (MEM) containing 1µg/ml trypsin treated with L-(tosylamido-2-phenyl) ethyl chloromethyl ketone (TPCK).We assumed that the pandemic viral isolates would be able to grow efficiently at 35 °C (as the temperature of the human upper respiratory tract is about 33 °C) and we therefore chose to amplify the virus at 35 °C rather than 37°C in order to avoid the preselection of variants that grow preferentially at high temperature.Viral stocks were clarified and aliquots for single use were kept frozen at -80 °C.

Temperature-sensitivity assays
Confluent three-day-old cultures of MDCK cells in 96-well plates, prepared in MEM containing 5% foetal calf serum and 50 µg/ml gentamycin, were washed twice with serum-free MEM before infection.Serumfree MEM (170 µl/well) containing trypsin-TPCK (1µg/ ml) and gentamycin (50 µg/ml) were added to cultures.Ten-fold dilutions of each virus sample in MEM (30 µl/ well, 10 wells/dilution, 3 plates/virus sample) were added to cells.Plates were sealed with an adhesive membrane and covered with lids and incubated at 34 °C, 37 °C and 40 °C.Incubators were used for incubation at 34 °C and 37 °C, whereas a water bath was used to incubate plates at exactly 40 °C.Cytopathic effects were observed under the microscope three days after infection and virus titres as 50% tissue culture infectious doses (TCID 50 ) per mL were determined as previously described by Reed and Muench [23].The reproductive capacity at the high, potentially restrictive temperature of 40 °C (RCT 40 value) is the difference, in log values, between the titres at 40 °C and at 37 °C for each viral isolate.Similarly, the reproductive capacity at 34 °C (RCT 34 value) is the difference in log values between the viral titres at 34 °C and at 37 °C.Both RCT values are expressed as the mean ± standard deviation (SD) from at least three independent experiments.
On average, pandemic viral isolates were about three-fold less sensitive at 40 °C than the pool of the 2007/08 and 2008/09 seasonal viruses (RCT 40 values of -2.55±0.82 and -3.06±0.46,respectively; p<0.05,Student's t-test) and showed a significantly higher variability in temperature sensitivity (variance ratio: 3.18; p<0.025,Fisher's exact test).No statistically significant differences were seen in RCT 40 values regardless of whether the pandemic viral isolates had been isolated from severe cases with or without underlying condition (n=15) or from mild cases (n=8) (Table 2).Interestingly, two human isolates representative of the North American triple reassortant influenza A(H1N1) viruses (A/Illinois/09/2007 and A/Ohio/02/2007) grew similarly at 40 °C and 37 °C (Table 2).Their growth was thus clearly more resistant to high temperature than that of the pandemic viral isolates.The Hong Kong TRIG swine influenza A(H1N2) and Eurasian swine influenza A(H1N1) viruses included in our study showed an intermediate phenotype between the triple reassortant and pandemic viruses (Table 2

and Figure).
A D222G substitution in the receptor binding site of HA was seen in two of the viral isolates included in our study (isolates 20097101 and 20097105).This substitution has been detected sporadically, with some degree of correlation between the presence of the substitution and the severity of the disease [24][25][26][27].Isolates 20097101 and 20097105 showed RCT 40 values of -2.25±0.57and -1.33±0.12,respectively (data not shown).The 20097101 virus was isolated from the brain of a young patient who died after infection and showed a G residue at position 222 of the HA (Table 1).The viruses detected in the initial brain specimen showed a D at this position, but probably contained a low, undetectable fraction of viruses of the HA-222G genotype upon amplification in MDCK cells.
The 20097097 virus isolated from the lung of the same patient showed a D residue at position 222 of the HA (Table 1).No statistically significant difference in temperature sensitivity was observed between the 20097101 and 20097097 isolates.
One of the pandemic viral isolates (20097214) included in our study had the H275Y substitution in the NA (Table 1) that is associated with oseltamivir resistance [28,29] and was characterised by a marked sensitivity to high temperature, with an RCT 40 value of -3.90±0.57.However, the two panels of oseltamivir-resistant and -sensitive seasonal isolates from 2007/08 showed no statistically significant difference in temperature sensitivity (Table 2 and Figure).Overall, our results suggest that neither the D222G substitution in the HA nor the H275Y substitution in the NA have a major impact on the viral sensitivity to high temperature.
The NA and M gene sequences of the 23 pandemic viral isolates included in our study were determined: the Global Initiative on Sharing Avian Influenza Data (GISAID) accession numbers are shown in Table 3.The NA and M1 amino acid sequences of the 23 pandemic viral isolates included in our study were were aligned with the corresponding sequences of the swine and triple reassortant viruses.The pandemic virus-derived sequences showed very few variations: their NA and M1 sequence shared about 91% and 94% identity with the respective Eurasian swine virus-derived sequences and 81% and 88% identity with the respective triple reassortant virus-derived sequences.

Discussion and conclusion
A panel of seasonal and pandemic influenza A(H1N1) viral isolates from northern France in 2007/08 to 2008/09 grew with similar efficiency at 34 °C and 37 °C, suggesting that these viruses are well adapted to the physiological temperatures of the upper and lower respiratory tract.In contrast, they replicated less efficiently at 40 °C than at 37 °C.As compared with seasonal isolates, the pandemic viral isolates showed a marked heterogeneity in temperature sensitivity as indicated by a significantly higher variability in the corresponding RCT 40 values.This heterogeneity probably reflects ongoing evolution and genetic diversification of the virus since its introduction in the human population in April 2009.
The sensitivity to high temperature of isolates of the pandemic virus from severe cases of influenza was not statistically significantly different from that of isolates from mild cases, but the numbers were small.These results suggest that there was little or no correlation between temperature sensitivity of pandemic viruses and clinical severity.However, this finding should be confirmed by analysing a larger panel of viruses, given   the strong heterogeneity in temperature sensitivity, the possible bias due to the fact that the severity of the disease in up to 25% of severe cases during the pandemic was due to bacterial secondary infections rather than the characteristics of the pandemic virus [30,31] and the fact that host factors, such as underlying conditions identified as risk factors, seem to have contributed substantially to the clinical course of severe cases with 2009 pandemic influenza A(H1N1) [32,33].
The Eurasian swine influenza A(H1N1) virus, a Hong Kong TRIG swine influenza A(H1N2) virus and two A(H1N1) triple reassortant viruses included in our study showed a lower sensitivity to elevated temperature (40 °C) than the pandemic and seasonal viral isolates on average, in agreement with the fact that the normal body temperature of pigs varies between 38.5 °C and 39.2 °C [34].All the pandemic viral isolates included in our study replicated less efficiently at 40 °C than did the triple reassortant viruses although their genomic segments, except for the NA and matrix (M) segments, are phylogenetically related to the triple reassortants.
No specific sequence signature was observed for the viruses that showed the highest RCT 40 (data not shown).Overall, our observations suggest that the sensitivity to high temperature of the pandemic viral isolates is determined by complex gene constellation and/or mutation effects.
In conclusion, our small dataset shows that the pandemic viruses that circulated in northern France in 2009 were more heterogeneous with respect to their ability to grow at high temperature (40 °C) than the seasonal viruses that circulated there in 2007/08 and 2008/09.They point to the impact of viral temperature sensitivity on the genetic evolution and diversification of the pandemic virus during the first year after its introduction into the human population and they call for a close monitoring of this phenotypic marker related to host and tissue tropism during the coming years.
c Sequence information refers to the viruses isolated after one passage in Madin-Darby canine kidney (MDCK) cells.d Viruses 20097097 and 20097101 were isolated from the lung and brain, respectively, of the same patient.e The sequence of the virus present in the original specimen was also determined, and a D was found at residue 222 of the haemagglutinin.

Figure
Figure Reproductive capacity of 2009 pandemic influenza A(H1N1) viral isolates (n=23) and 2007/08 and 2008/09 seasonal influenza A(H1N1) viral isolates (n=18) at 34 °C and 40 °C, relative to 37 °C, northern France 24 September to 16 December) in 2009; pandemic activity in this area started at week 42 in 2009, peaked at week 49 and ended at week 2 in 2010 [19].The isolates for the 2007/08 season were collected in northern France between week 44 (29 October) in 2007 and week 3 (14 January) in 2008 and those for the 2008/09 season between week 45 (3 November) in 2008 and week 4 (19 January) in 2009.One of the 2007/08 seasonal influenza A(H1N1) viral isolates, (A/Paris/1149/2008), was included in most experiments (12/13, due to a technical problem in one)