Results
Over the course of the 2019-2020 influenza season, an increasing number
of influenza A(H1N1)pdm09 viruses exhibited decreased reactivity towards
ferret antiserum against the vaccine virus A/Brisbane/02/2018 (Figure
1A), indicating an antigenic profile distinct from the vaccine virus.
For these viruses, an average 6 log2 reduction in
reactivity compared to the vaccine virus was observed (Figure 1 A).
These viruses emerged in December 2019 and continued to circulate until
March 2020, when influenza activity in Germany declined due to the
lockdown and safety measures in response to the SARS-CoV-2 pandemic
(Figure 2). In a total, of 631 A(H1N1)pdm09 viruses detected in
2019-2020 the antigenic profiles of 417 viruses were determined and HI
assay results indicated good vaccine coverage for influenza viruses of
other subtypes (A(H3N2) and B) circulating in Germany (Figure 1B-D).
164 A(H1N1)pdm09 influenza viruses (mainly whole genomes) were sequenced
by NGS in 2019-2020. The HA gene analysis of A(H1N1)pdm09 viruses
circulating between week 42/2019 and 5/2020 revealed that the majority
of viruses (69 out of 71 = 97%) belonged to the 6B.1A5A clade,
represented by A/Norway/3433/2018 and characterized by the
clade-specific substitutions N129D and T185I in HA1. The remaining two
viruses fell into the 6B1.A7 clade, displaying the K302T substitution in
HA1 and the I77M, N169S, E179D substitutions in HA2. Among the
6B1.A5A-viruses, five variants were identified, which carried additional
substitutions, namely D187A, Q189E in HA1 (51%, cluster 1); R205K in
HA1 (3%, cluster 2); N156K, A195E in HA1 (7%, cluster 3); K130N,
N156K, L161I, V250A in HA1, E179D in HA2 and L8M in the signal peptide
partially with K209M substitution in HA1 (20%, cluster 4); and V193A in
HA2 (15%, cluster 5; Figure 3A, Table 1). HA gene analysis demonstrated
that all viruses with reduced reactivity towards antiserum against the
vaccine virus A/Brisbane/02/2018 carried the N156K substitution, which
is located in the Sa antigenic site on the hemagglutinin. Interestingly,
the N156K substitution was detected in two different virus clusters (3
and 4) and also in one virus of cluster 1 (A/MVP/2/2020). Although these
cluster 1, 3 and 4 viruses had a different genetic background, all
N156K-viruses showed reduced reactivity in the hemagglutination
inhibition test (Figure 1A). For a 6B.1A5A virus (A/THR/125/2019, near
cluster 4) with polymorphisms at amino acid position 156 (N156N/K) and
at other positions (L8L/M in signal peptide, K130K/N, V250V/A in HA1,
E179E/D, V193V/A in HA2), an only slight HI-titer reduction was observed
(≤4-fold, Figure 3A).
All test viruses for which the genetic group or subgroup were known,
with collection
dates after 2018-08-31, fell into clade 6B.1 which is defined by the
amino acid
substitutions S84N, S162N (introducing a potential N-linked
glycosylation site) and
I216T in HA1 with the viruses clustering into a
genetic subclade designated as 6B.1A
and defined by the HA1 amino acid substitutions S74R ,S164T (which alters the
glycosylation motif at residues 162 to 164 ) and I295V .
The NA gene analysis showed a phylogeny similar to that based on the HA
gene (Fig. 3B). The majority of viruses clustered with 6B.1A5A-viruses
(97%), carrying the Q51K, T74S, I389K, T452I substitutions. The two
6B.1A7-viruses (3%) displayed the M314I and F322L substitutions in NA.
Consistent with HA gene phylogeny, five clusters could be discerned:
Cluster 1 (S52N, partially also I40T, T72N, 49%), cluster 3 (I396M,
4%) and cluster 4 (M19T, Y66F, N222K, partially also T72I, 23%)
exhibited cluster-specific substitutions, whereas cluster 2 (3%) and 5
(17%) did not. Of interest, two viruses clustered differently in the NA
gene than in the HA gene: A/BWB/11/2020 (HA cluster 1) clustered with
viruses of cluster 4 in the NA gene and A/SAS/3/2020 (HA cluster 3) did
not belong to any of the five cluster in the NA gene (Figure 3B).
HA and NA sequences were examined using the
https://flusurver.bii.a-star.edu.sg/ online tool to assess for
substitutions which are located close to common drug, host receptor or
antibody binding sites; or which promote the loss or creation of a
glycosylation motif (relative to the closest reference
A/Brisbane/02/2018). One virus (A/NRW/142/2019) featured the S247N
substitution, which is described to be associated with mild drug
resistance (14). This virus belonged to the HA-N156K variant (cluster 4)
and displayed HI titers that were >4 fold reduced. Three
viruses carrying the HA-N156K substitution, belonging to clusters 1
(A/MVP/2/2020), 3 (A/MVP/12/2019) and 4 (A/NRW/142/2019), were further
analyzed, using flusurver to deduce the localization of the identified
substitutions in a structural model, as shown in Figure 4A-F, which
indicates substitutions relative to the closest reference
A/Brisbane/02/2018.
Next, cross neutralization was assessed via HI assay, including
specifically generated ferret antisera against A(H1N1)pdm09 HA-N156K .
These experiments revealed clear differences between A(H1N1)pdm09
HA-N156K and other A(H1N1)pdm09 viruses (Table 2). Although viruses of
both groups are still recognized by antisera against either group,
antigenic reactivity between these variants is reduced 3-7
log2 fold.
To gain first insight into a potential clinical impact of the HA-N156K
mutation, patient data were evaluated, assessing specifically whether
(i) these viruses were preferentially detected in certain patient groups
(age group, patients with chronic disease, patients with pneumonia,
vaccinated patients) and / or (ii) patients were prescribed antiviral
agents (as a proxy of disease severity) and / or (iii) time to physician
consultation after symptoms onset was short (as another proxy of disease
severity) (Supplemental Table 2). Patients infected with viruses
carrying the HA-H156K mutation were more frequently treated with
neuraminidase inhibitors than those infected with other A(H1N1)pdm09
viruses. Moreover, the time delay between symptom onset and physician
consultation that was observed for these patients was shorter. No other
differences were noted between patients infected with A(H1N1)pdm09
HA-N156K viruses and those infected with other A(H1N1)pdm09 viruses.