﻿JOURNAL OF VIROLOGY, JUlY 1991, p. 3704-3714 Vol. 65, No. 7
0022-538X/91/073704-11$02.00/0
Copyright © 1991, American Society for Microbiology
Evolution of Influenza A Virus Nucleoprotein Genes: Implications
for the Origins of HlNi Human and Classical Swine Viruses
OWEN T. GORMAN,1t WILLIAM J. BEAN,' YOSHIHIRO KAWAOKA,l ISABELLA DONATELLI,2
YUANJI GUO,3 AND ROBERT G. WEBSTER'*
Department of Virology and Molecular Biology, St. Jude Children's Research Hospital, 332 North Lauderdale,
P.O. Box 318, Memphis, Tennessee 38101-03181; Dipartimento di Virologia, Istituto Superiore di Sanitd,
00161 Rome, Italy2; and National Influenza Centre, Institute of Virology, Chinese Academy of
Preventative Medicine, Beijing 100052, People's Republic of China3
Received 7 January 1991/Accepted 1 April 1991
A phylogenetic analysis of 52 published and 37 new nucleoprotein (NP) gene sequences addressed the
evolution and origin of human and swine influenza A viruses. HiN1 human and classical swine viruses (i.e.,
those related to Swine/Iowa/15/30) share a single common ancestor, which was estimated to have occurred in
1912 to 1913. From this common ancestor, human and classical swine virus NP genes have evolved at similar
rates that are higher than in avian virus NP genes (3.31 to 3.41 versus 1.90 nucleotide changes per year). At
the protein level, human virus NPs have evolved twice as fast as classical swine virus NPs (0.66 versus 0.34
amino acid change per year). Despite evidence of frequent interspecies transmission of human and classical
swine viruses, our analysis indicates that these viruses have evolved independently since well before the first
isolates in the early 1930s. Although our analysis cannot reveal the original host, the ancestor virus was
avianlike, showing only five amino acid differences from the root of the avian virus NP lineage. The common
pattern of relationship and origin for the NP and other genes of HlNl human and classical swine viruses
suggests that the common ancestor was an avian virus and not a reassortant derived from previous human or
swine influenza A viruses. The new avianlike HlNl swine viruses in Europe may provide a model for the
evolution of newly introduced avian viruses into the swine host reservoir. The NPs of these viruses are evolving
more rapidly than those of human or classical swine viruses (4.50 nucleotide changes and 0.74 amino acid
change per year), and when these rates are applied to pre-1930s human and classical swine virus NPs, the
predicted date of a common ancestor is 1918 rather than 1912 to 1913. Thus, our NP phylogeny is consistent
with historical records and the proposal that a short time before 1918, a new HlNl avianlike virus entered
human or swine hosts (0. T. Gorman, R. 0. Donis, Y. Kawaoka, and R. G. Webster, J. Virol. 64:4893-4902,
1990). This virus provided the ancestors of all known human influenza A virus genes, except for HA, NA, and
PB1, which have since been reassorted from avian viruses. We propose that during 1918 a virulent strain ofthis
new avianlike virus caused a severe human influenza pandemic and that the pandemic virus was introduced
into North American swine populations, constituting the origin of classical swine virus.
The influenza virus pandemic of 1918 claimed more than
20 million lives (9). Resolving the origin of this virulent virus
has been a major focus of virological research for more than
60 years. Antigenic and seroarcheological studies since the
1930s (see, e.g., references 2, 11, 12, 17, 18, and 27) have
indicated that the early human and classical swine influenza
A HlNl viruses (i.e., those related to Swine/Iowa/15/30)
were very similar. Swine are susceptible to human influenza
viruses and vice versa (19, 20), and avianlike HlNl and
H3N2 viruses have been isolated from pigs (23, 38, 44).
Scholtissek et al. (39) proposed that swine may provide an
efficient "mixing vessel" for the introduction of reassortant
viruses into the human population. This hypothesis may
explain the appearance of new pandemic viruses in the
human population (40).
The nucleoprotein (NP) gene has been chosen for evolu-
tionary studies of influenza A viruses (see, e.g., references
13 and 15) because its purported role as a determinant ofhost
range (39, 45, 47) predicts that NP gene evolution should be
host specific. Thus, the NP gene may serve as a model for
*
Corresponding author.
t Present address: U.S. Fish and Wildlife Service, Flagstaff, AZ
86001.
host-specific evolution of influenza viruses. Using RNA
hybrization techniques, Bean (3) showed that NP genes fall
into five host-specific groups. This finding was confirmed by
showing that NP genes have evolved into five major host-
specific lineages that correspond to Bean's RNA hybrization
groups (15). The inclusion ofboth human and classical swine
virus NPs into one lineage suggests that they share a
common ancestor. The above-mentioned antigenic studies of
HlNl human and classical swine viruses predict the same
close relationship for surface proteins, hemagglutinin (HA)
and neuraminidase (NA). Evolutionary analyses of other
internal protein genes of influenza A viruses also show a
close relationship between human and classical swine vi-
ruses (M and NS [31], PB1 [22], PA [35], PB2 [16]), and these
viruses appear to be evolving away from avianlike ancestors
(15, 16). This common pattern of evolution and origin for all
virus genes suggests that the common ancestor was an avian
virus and not a reassortant derived from previous human or
swine influenza A viruses (16).
To date, evolutionary analyses of influenza A virus genes
include sequences for only a few classical swine viruses, and
human virus gene sequences are heavily biased toward
recent virus isolates. Clearly, the lack of representative
sequence data for swine viruses precludes an evolutionary
analysis of classical swine viruses, and the bias in human
3704
EVOLUTION OF HUMAN AND SWINE HlNl VIRUSES 3705
virus sequence data weakens the accuracy of published
evolutionary analyses. For example, Gammelin et al. (13)
and Gorman et al. (15) show NP genes of recent classical
swine virus isolates as diverging before the common ances-
tor of human and the earliest classical swine viruses. We
believe that this portrayal of two separate lineages of clas-
sical swine virus NPs is an artifactual result of not having
representative classical swine virus NPs between 1931 and
1977. These deficiencies hamper our understanding of the
origin of human influenza viruses and the potential role of
swine hosts in that origin.
Our aim has been to determine the evolutionary relation-
ships of human and swine influenza viruses. Specifically, we
have addressed the following questions. (i) Do human and
classical swine viruses share a single common ancestor? If
so, what was the host ofthis common ancestor and what was
the date of divergence into human and swine virus lineages?
(ii) How does virus evolution differ in the two host species?
Have HlNl human and classical swine viruses evolved
independently; i.e., is there any evidence for a common gene
pool for human and swine viruses? (iii) What can the
evolution of new avianlike viruses in swine reveal about the
early evolution of HlNl human and classical swine viruses?
(iv) What is the relationship between the origin of HlNl
human and classical swine viruses and the 1918 influenza
pandemic? To address these questions, we have included 37
additional sequences that address the deficiencies in previ-
ous analyses. The present expanded evolutionary analysis of
89 NP genes is intended to complement our previous analy-
sis of 41 NP genes (15).
MATERIALS AND METHODS
We selected 37 influenza A virus isolates (16 swine, 7
human, 13 avian, and 1 equine) from which to clone and
sequence NP genes (Table 1). Many of the early swine virus
isolates (SWOH35, SW29-37, SWJMS42, SWIA46, SW41-
49, SWMAY54, and SWWIS57) were kindly provided by
H. F. Maassab and were from the Francis Historical Influ-
enza Collection at the University of Michigan. Remaining
isolates were selected to complement 24 NP genes that we
previously sequenced (15) and 28 others from literature and
databank sources (Table 1).
Molecular cloning, sequencing, and sequence analysis
were performed as previously described (15, 16). Briefly,
cDNAs of viral NP RNAs were ligated into plasmid vectors
and were then transfected into competent Escherichia coli.
For each virus strain, two to five NP clones were sequenced.
Phylogenetic analysis of sequence data was performed with
PAUP (Phylogenetic Analysis Using Parsimony) software,
version 2.4 (David Swofford, Illinois Natural History Sur-
vey, Champaign, Ill.).
RESULTS
Evolutionary tree of NP gene nucleotide sequences. A
phylogenetic analysis of 89 influenza A virus NP gene
sequences is presented as an evolutionary tree rooted to an
aligned influenza B virus NP (Fig. 1A). The general topology
of the tree is similar to that of our previous analysis (15),
except that now the human and classical swine virus NPs are
shown sharing a single common ancestor (upper star in Fig.
1A) and EQPR56 is slightly closer to other NP lineages. The
human lineage contains three examples of humanlike viruses
isolated from swine: SWCAM35, SWHK76, SWDAN83. It
is notable that SWCAM35 is not a classical swine virus and
is most closely related to WS33, the first human virus isolate.
The isolates SWHK76 and SWDAN83 represent humanlike
swine viruses from China. Within the classical swine virus
NP lineage are two examples of swinelike viruses isolated
from humans, NJ76 and WIS88. These two viruses are
independently derived and are closely related to contempo-
rary strains isolated from pigs (SWTN77 and SWIA88,
respectively). This pattern emphasizes the susceptibility of
swine to human viruses and vice versa, but these interspe-
cies transmissions appear to represent evolutionary dead
ends since there is no evidence that they circulate exten-
sively or leave descendant viruses.
The classical swine NP lineage contains a subgroup of
isolates that have not evolved significantly for 19 years
relative to the original SWIA30 isolate (SWJMS42, SWIA46,
and SW41-49). The lack of accompanying data for these
three isolates precludes any explanation for this pattern.
However, the remaining isolates fall in the expected chrono-
logical order within the lineage. The only Asian isolate repre-
sented in the HlNl classic swine virus lineage, SWHK82,
appears to be a reassortant containing avian H3N2 surface
proteins (23). The classical swine virus NP of this isolate is
related to North American strains in the mid-1970s. This
agrees with the appearance of classical swine viruses anti-
genicly similar to North American strains in the late 1970s in
Southeast Asia (43) and Japan (34, 51).
A distinct sublineage of Italian swine virus isolates
(SWIT76, SWIT79, SWIT41-81, and SWIT47-81) supports
the reported introduction of classical swine viruses into Italy
in 1976 (32). However, antigenic analyses of Italian swine
virus isolates (12a) indicates that classical swine viruses
disappeared from Italy after 1986 and have been replaced by
avianlike HlNl swine viruses. These new avianlike swine
viruses first appeared in Europe in 1979 (38), but antigenic
analyses suggest that they began to circulate in Italy in 1985
(12a). The timing of the appearance of the new avianlike
HlNl viruses in Italy (1985) and the disappearance of
classical swine viruses (1986) suggests that the new swine
virus in some way hastened the extinction of classical swine
viruses in Italy.
Within the avian NP lineage, only one Asian swine isolate
is represented (SWHK2-82) and is closely related to a H3N2
Asian duck isolate (DKHK75). Kida et al. (23) provide
additional examples of H3N2 avianlike viruses isolated from
pigs in southern China. Antigenic studies of swine virus
isolates from southern China suggest that pigs are commonly
infected with H3N2 human and avian viruses (43). However,
there are no data to indicate whether these H3N2 avian
viruses circulate and evolve in the Asian swine host reser-
voir. In contrast, the new avianlike European swine viruses
form a distinct sublineage within the avian lineage
(SWGER81, SWNED85, and SWIT89 [Fig. 1A]). The for-
mation of a swine-specific lineage indicates that these new
swine viruses are circulating in the swine host reservoir and
have begun to evolve separately from their avian ancestor.
Other examples of avianlike swine viruses isolated from
mammals include those isolated from whales (WHALEM84
and WHALEP76), seals (SEAL80), and mink (MINKSW84)
(Fig. 1). In every case these isolates appear to be indepen-
dently derived from avian viruses. The inability to isolate
descendant strains indicates that transmissions of avian
viruses to marine mammals represent evolutionary dead
ends (25).
Evolutionary tree of amino acid sequences. To evaluate the
effect of gene evolution on NP proteins, nucleotide se-
quences were translated and a phylogenetic tree of amino
VOL. 65, 1991
3706 GORMAN ET AL.
TABLE 1. Influenza virus strains used in phylogenetic analyses
Strain Abbreviation GenBnk Reference
Equine strains
A/Equine//Prague/1/56 (H7N7) EQPR56 M63748 This report
A/Equine/Miami/63 (H3N8) EQMI63 M22575 14
A/Equine/London/1416/73 (H7N7) EQLON73 M30750 15
A/Equine/Kentucky/2/86 (H3N8) EQKY86 M30751 15
A/Equine/Tennessee/5/86 (H3N8) EQTN86 M30758 15
Human strains
B/Lee/40 BLEE40 K01395 4
A/Wilson-Smith/33 (HlNl) WS33 M30746 15
A/Puerto Rico/8/34 (HlNl) PR834 NAa 50
A/Puerto Rico/8/34 (HlNl) (Cambridge) PR8CAM34 J02147 48
A/Hickox/40 (HlNl) HICKOX40 M63749 This report
A/Ft. Monmouth/47 (HlNl) FM47 M63750 This report
A/Ft. Warren/50 (HlNl) WARREN50 NA 1
A/England/19/55 (HlNl) ENG55 M63751 This report
A/Singapore/1/57 (H2N2) SING57 M63752 This report
A/Ann Arbor/6/60 (H2N2) AARBOR60 M23976 8
A/Victoria/5/68 (H2N2) VIC68 M63753 This report
A/NT/60/68 (H3N2) NT60-68 J02137 21
A/Udom/307/72 (H3N2) UDORN72 M14922 5
A/USSR/77 (HlNl) USSR77 NA 1
A/Brazil/78 (HlNl) BRAZIL78 NA 1
A/Texas/1/77 (H3N2) TEXAS77 NA 1
A/California/10/78 (HlNl) CALIF78 NA 1
A/Hong Kong/5/83 (H3N2) HK83 M22577 14
Human strains, swinelike viruses
A/New Jersey/8/76 (HlNl) (Ft. Dix) NJ76 M63754 This report
A/Wisconsin/3523/88 (H1N1) WIS88 M63755 This report
Swine strains, classical viruses
A/Swine/Iowa/15/30 (HlNl) SWIA30 M30747 15
A/Swine/1976/31 (HlNl) SW31 M22578 14
A/Swine/Ohio/23/35 (HlNl) SWOH35 M63756 This report
A/Swine/29/37 (HlNl) SW29-37 M63757 This report
A/Swine/Jamesburg/42 (HlNl) SWJMS42 M63758 This report
A/Swine/Iowa/46 (HlNl) SWIA46 M63759 This report
A/Swine/41/49 (HlNl) SW41-49 M63760 This report
A/Swine/May/54 (HlNl) SWMAY54 M63761 This report
A/Swine/Wisconsin/1/57 (HlNl) SWWIS57 M63762 This report
A/Swine/Wisconsin/1/61 (HlNl) SWWIS61 M63763 This report
A/Swine/Italy/437/76 (HlNl) SWIT76 M63764 This report
A/Swine/Tennessee/24/77 (HlNl) SWTN77 M30748 15
A/Swine/Italy/2/79 (HlNl) SWIT79 M63765 This report
A/Swine/Italy/141/81 (HlNl) SWIT141-81 M63766 This report
A/Swine/Italy/147/81 (HlNl) SWIT147-81 NA 13
A/Swine/Ontario/2/81 (HlNl) SWONT81 M63767 This report
A/Swine/Hong Kong/127/82 (H3N2) SWHK82 M22570 14
A/Swine/Iowa/17672/88 (HlNl) SWIA8R M63768 This report
Swine strains, humanlike viruses
A/Swine/Cambridge/1/35 (HlNl) SWCAM35 M63769 This report
A/Swine/Hong Kong/6/76 (H3N2) SWHK76 M22571 14
A/Swine/Dandong/9/83 (H3N2) SWDAN83 M63770 This report
Swine strains, avianlike viruses
A/Swine/Hong Kong/126/82 (H3N2) SWHK2-82 M63771 This report
A/Swine/Germany/2/81 (HlNl) SWGER81 M22579 14
A/Swine/Netherlands/12/85 (HlNl) SWNED85 M30749 15
A/Swine/Italy/839/89 (HlNl) SWIT89 M63772 This report
Miscellaneous mammalian strains, avianlike viruses
A/Whale/Maine/328/84 (H13N2) WHALEM84 M30759 15
A/Seal/Massachusetts/1/80 (H7N7) SEAL80 M27518 25
A/Whale/Pacific Ocean/76 (H1N3) WHALEP76 M27517 25
A/Mink/Sweden/84 (H1ON4) MINKSW84 M24454 36
Continued on following page
J. VIROL.
EVOLUTION OF HUMAN AND SWINE HlNl VIRUSES 3707
TABLE 1-Continued
Strain Abbreviation GenBank Reference
accession no.
H13 gull virus strains
A/Gull/Maryland/704/77 (H13N6) GULMD77 M30754 15
A/Gull/Maryland/1824/78 (H13N9) GULMD78 M30755 15
A/Gull/Maryland/181579 (H13N6) GULMD79 M30756 15
A/Gull/Minnesota/945/80 (H13N6) GULMN80 M30757 15
A/Gull/Massachusetts/26/80 (H13N6) GULMA80 M30752 15
A/Gull/Astrakhan/227/84 (H13N6)b GULAST84 M30753 15
Avian strains
A/Duck/New Zealand/31/6 (H4N6) DKNZ76 M30760 15
A/Grey Teal/Australia/279 (H4N4) GTAUS79 M30761 15
A/Duck/Manitoball/53 (H1ON7) DKMAN53 M63773 This report
AfTurkey/Ontarion732/66 (H5N9) TYONT66 M63774 This report
A/Duck/Pennsylvania/1/69 (H6N1) DKPEN69 M63775 This report
A/Duck/Memphis/928M4 (H3N8) DKMEM74 M63776 This report
A/Gull/Maryland/5/77 (H11N9) GULM5-77 M63777 This report
A/Mallard/NY/6750/78 (H2N2) MLRDNY78 M14921 5
AfTurkey/Minnesota/833/80 (H4N2) TYMN80 M30769 15
A/Turkey/Minnesota/1661/81 (HlNl) TYMN81 M63778 This report
A/Chicken/Pennsylvania/1/83 (H5N2) CKPENN83 M30768 15
A/Ruddy Turnstone/NJ/47/85 (H4N6) RTNJ85 M30766 15
A/FPV/Dobson/"Dutch"/27 (H7N7) FPV27 M63779 This report
A/FPV/Rostock/34 (Giessen) (H7N1) FPV34 M24456 26
A/Chicken/Germany/"N"/49 (H1ON7) CKGER49 M24453 36
A/Duck/Czechoslovakia/56 (H4N6) DKCZ56 M30762 15
A/Duck/England/1/56 (H11N6) DKENG56 M63780 This report
A/Duck/Ukraine/2/60 (H11N8) DKUK60 M30763 15
A/Tern/South Africa/61 (H5N3) TERNSA61 M30767 15
A/Duck/England/1/62 (H4N6) DKENG62 M63781 This report
A/Shearwater/Australia/72 (H6N5) SHWTR72 M27298 13
A/Parrot/Ulster/73 (H7N1) PARU73 M22344 46
A/Duck/Hong Kong/7/75 (H3N2) DKHK75 M22573 14
A/Duck/Bavaria/2/77 (HlNl) DKBAV77 M22574 14
A/Budgerigar/Hokkaido/lM7 (H4N6) BUDHOK77 M30765 15
A/Duck/Beijing/1/78 (H3N6) DKBEI78 M63782 This report
A/Duck/Australia/749/80 (HlNl) DKAUS80 M63783 This report
A/Teal/Iceland/29/80 (H7N7) TEALIC80 M63784 This report
A/Mallard/Astrakhan/244/82 (H14N6)b MAST82 M30764 15
A/Mallard/Astrakhan/263/82 (H14N5)b MAST2-82 M63785 This report
a NA, not applicable.
b Astrakhan locality is synonymous with Gurjev.
acid sequences was constructed based on the branching
topology of the nucleotide tree (Fig. 1B). This approach
permits direct comparison of the corresponding branches in
nucleotide and amino acid trees for differences in genetic
versus protein evolution (15). In essence, the amino acid tree
represents a phylogeny based on nonsilent changes; the
collapse of internal branches in the avian lineage indicates
that nearly all the homologous internal branches in the
nucleotide tree are composed of silent changes. A compari-
son of human and classical swine virus sublineages reveals
that at the nucleotide level they have evolved roughly the
same amount from their common ancestor (i.e., the two
branches are similar in length [Fig. 1A]). However, at the
amino acid level the classic swine virus NPs are evolving
much more slowly (Fig. 1B), which suggests that swine
viruses are subjected to less selective pressure than are
human viruses.
The avian lineage is characterized by evolutionary stasis
of NPs (Fig. 1B) as noted previously (15). SWHK2-82, an
Asian avianlike swine virus isolate, is very similar to an
Asian duck virus isolate, DKHK75 (three amino acid differ-
ences [Fig. 1B]). In contrast to other NPs of the avian
lineage, those from avianlike European swine virus isolates
(SWGER81, SWNED85, and SWIT89) form a distinct sub-
lineage that shows a progressive accumulation of amino acid
changes consistent with dates of isolation.
Comparison of amino acid sequences. To detect patterns of
derived (synapomorphic) amino acid changes for each lin-
eage, we compared NP sequences with baseline sequences
that are closest to the root of the phylogenetic tree (a table of
aligned sequences is available upon request). The DKBAV77,
MAST82, and MAST2-82 isolates are one amino acid change
from the root node of the avian lineage, which represents the
hypothetical ancestor for all avian NPs (Fig. 1B, lower star).
This analysis permits identification of the specific amino acid
changes that have occurred in the evolution of host-specific
NPs and permits reconstruction of sequences for hypothet-
ical ancestral NPs (Fig. 2). Two possible sequences for
hypothetical ancestral avian NPs (Fig. 2, AVIAN 1 and
AVIAN 2) differ at amino acid positions 105 and 450. These
amino acid differences distinguish the common ancestors for
H13 gull and North American avian from Old World avian
groups. The common ancestor of human and classical swine
NP lineages (Fig. 2, HUM-SWINE) is characterized by five
VOL. 65, 1991
3708 GORMAN ET AL.
EOPR56((H7N7)
EOM163 (H3N
()43
6O00ON73(H7N7)
*EOKY86 (H3N) X
EOTN86(H3N48)
SWCAM35 (H11)
4WS33 (HIM)
PR834 (H14N)
1824 PR8CAM34 (H14)
I
I I I I r HlCK~~~WOX40 (H1Nl)
18928 F47 (H")
I8815 (HIM)
U18811SSR77 (H1 )
1BRAZL78 (H1N1)
| 1 19471 _ ~~~~~~ENG55(H1N1)
5S4457 (H2N2)
I
I I Ll I r ~~~~~~~~~VIC68
(M2N2)
1868| 14788NT60-68 (H3N2)
|S12 | W 13 SW476 (H3N2)
112
*U8 1(50OR1472 (H3N2)
18 70TEXAS77 (H3N2)
CALF78 (H1N1)
WDAN683((H3N2)
8683 (H3N2)
AARBOR60 (H2N2)
1SWLA30 (HI)
4SW31 (H11)
| | SI46 (H1lNl)
1SW4642 (HI)
1825 8S41-49 (H1N1)
140SWOH35 (HIM)
5128-37
(HI) SWIT141-81 (H )
L
-4~ ~ ~ ~ r --601(411
.CSW14(T17-81 (HWI)
]
I1
7 9 ( 1)
1873 rNJ76 (H8)
146SWW82 (H3N2)v
1 4ST7t (HINI)
1sww,s 514SWONT81 (H11)
61 (H") WIS88(H1M)
141SWN8 (HI)
GLXAST84 (MIMS)
,GLLM77 (H13NG)
IGLMD78 (H13N9)
IGLMD79 (H13N6)
GLLWOO (HU3NG)
WHALEM84 (H13N2)
| - GULM~~~~~~ASO
(Hl3N6)
|DKMAN53 (H11N7)
TYONT66 (HSN9)
GULM5-77 (H149)
I I I r ~~~DKPEN69 (H6Ni)
DKMEM74 (H3N8)
.MLRDNY78 (H2N2)
RTNJ85 (H4N6)
TYMN80(H4N2)
TYWM48 (HN1)
_(PPNN4835(H5N2)
IEALIM (H7N7)
IDKNZ76 (H4N6)
|
-G
T A U S~~~79 (H4N4)
_FPV4I71
1688 * CRGER4940l~A~ (14104N7)
DKCZ
56(H46)
DJI
KENG56 (HtN6)
1DKUK80 (H1 8)
188 2(14DKENG62(H4N1)
4TERNSAG161 (13)
LlpWHALEP76 (HIN3)
188,9 18PARU73 (H7N1)
06KB6077 (HIM)
1841 61
BUDHOK77 (H4N6)
6DKW75(H3N2)
108 | rl -SWHK2-82 (H3N2)
06KAU580(H414)
Nudeob0@ ChtrQ m | KE78 (H3N6)
8n5782 (H141N)
84MAS2-82(141488)
7641(588(H7N7)
LMINKSW84 (H10N4)
51WTR72((H6N5)
SGER81 (HlNl)
1619 5146NEW5 (HIM)
1-4SW 89 (H1N1)
II
B
III
'Cl iv
*
EOPR56 (H7N7)
EOMt63 (H3N8)
NO773 (H17N17)
EOTN86 (H3N8)
S 5CAM3S (H1N1)
WS33 (H1N4)
PR834 (HtNl)
-PR8CAM34 (HN41)
0HICKOX40 (HINM)
FM47 (HNl)
WARRENSO (HlN)
BRAZL78
1EN455 (HINI)
_*lit - S4NG57 (H2N2)
_V1C68 (H2N2)
NT60-66 )H )2(
_ _r1SWHK76 (H3N2)
SW)A30(HtN1) _
1SW31 (HNl)
SLSWIA46 (HlNl) _
SWJMS42(HIN1)
_SW41-49 (HtNI) AAF
SWOH35 (H1141)
SW29-37 (H144)
SWMAY54 (H1N1)
SWWE57 (H144)
SWIT141-81 (HIM)
SWIT147-81 (HNt)
4SWIT79 (HIM)
SWIT76 (H1N1)
4NJ76(H1N1)
SWHK82 (H3N2)
_ 1*
STN77 (14141)
WN8148(H14N1)
H1)
LSWWIS561 (HIN1)
5LAST84 (HW3N6)
GUMD77 (H13N6)
G
qLMD78 (H13N9)
GULMD79 (H13N6)
G_-G
ULM 80 (H13N6)
_GULMA80 (H13N6)
WHALEM84 (H13N2)
r 6KMAN53(H1ON7)
- TYONT66((HSN9)
- G8LM5-77(H119)
D
DKPEN69 (H6N1)
DKMEM74 (H3N8)
*0MRDNY78 (H2N2)
17RTNJ85 (H4N6)
4TYMN80((H4N2)
CKPENN83 (H5N2)
SEALBO (H7N7)
1TYMN81(H1N1)
DKNZ76 (H4N6)
- GTAUS79 (H4N4)
- FPV?77(H7N7)
_ FPV34 (H7Nt)
CKGER49 (H1ON7)
DKCZ56 (H4N6)
6DKEN1G56 (HllN6)
0KUK60 (H11N8)
D^KENG62(H4N6)
ITERNSA61 (H5N3)
16WHALEP76 (H1N3)
PARU73 (H7Nt)
D*6KB677
(H1N1)
0
BuGRH1077 (H4N6)
DK8675 (H3N2)
1SW4K2-82 (H3N2)
DKBE178 (H3N6)
MAST82 (H14N6)
M*AST2-82 (H41N5)
DKAUS80 (HINI)
TEALIC80 (H7N7)
MINSW84 (HtON4)
SHWTR72(H6N5)
1-S45ER81(HN1)
1SWNED85 (HIN1)
1SWIT89 (H1NI)
-UDORN72 (H3N2)
TEXA577 (H3N2)
CALF78 (HN1I)
SW0AN83 (H3N2)
HK83 (H3N42)
RBOR60 (H2N2)
*w
Amino Ac
Chnes
FIG. 1. (A) Phylogenetic tree of 89 influenza A virus NP gene nucleotide sequences rooted to the NP of B/Lee/40 (37 sequences from this
study plus 52 published sequences). Sequences were analyzed with PAUP software, which uses a maximum parsimony algorithm to find the
shortest trees (MULPARS, SWAP=GLOBAL, and HOLD= 10 options). The arrow indicates the direction of the B/Lee/40 NP from the root
node. Horizontal distance is proportional to the minimum number of nucleotide differences to join nodes and NP sequences. Vertical lines
are for spacing branches and labels. Roman numerals indicate Bean's (3) RNA NP hybridization groups: I, Equine/Prague/56; II, recent
equine; III, human and classical swine; IV, H13 gull; V, avian. Animal symbols indicate host specificities of the lineages. Dates for
hypothetical ancestor nodes were derived by dividing branch distance by evolutionary rate estimates (see Fig. 4). Stars indicate the
hypothetical ancestors for the human-swine NP lineage (upper) and avian NP lineage (lower). (B) Phylogenetic tree of influenza A virus NP
amino acid sequences. The amino acid phylogeny conforms to the topology of the nucleotide tree (panel A). Sequences represented in these
trees are listed in Table 1.
amino acid changes relative to either of the ancestral avian
NPs (positions 33, 136, 351, 425, and 473). Two of these
changes (positions 351 and 473) occur as unique convergent
characters in some avian NPs (GULM5-77 and FPV34 [see
Fig. 3 in reference 15]). This suggests that the three changes
at positions 33, 136, and 425 are unique in the evolution of
the human-swine NPs. Furthermore, the small number of
unique amino acid changes emphasizes the avian character
of the human-swine NP ancestor.
Specific amino acid changes that are synapomorphic for
post-1930 human and classical swine NPs can be identified.
For the classical swine NP lineage these include five amino
acid changes at positions 100, 105, 289, 350, and 447 (Fig. 2,
C-SWINE). For the human NP lineage, 13 changes occur at
positions 16, 31, 61, 100, 127, 253, 283, 313, 357, 375, 408,
421, and 472 (Fig. 2, HUMAN). Recent human and classical
swine virus NPs (HK83 and SWIA88) show an accumulation
of 34 and 14 additional amino acid changes, respectively.
A
J. VIROL.
I
11
9 III
Iv
l Iv
4---
EVOLUTION OF HUMAN AND SWINE HlNl VIRUSES 3709
1.10. 20. 30. 40. 50. 60. 70. 80. 90.100.110.120.
SWIA88 D I V V
C-SWINE I I I
HK83 D K ID L R R K V R V
HUMAN D K I L V
HUM-SWINE I
AVIAN 1 MASQGTKRSYEQMETGGERQNATEIRASVGRMVGGIGRFYIQMCTELKLSDYEGRLIQNSITIERMVLSAFDERRNKYLEEHPSAGKDPKKTGGPIYRRRDGKWVRELILYDKEEIRRIWRQANN
AVIAN 2 M
AE-SWINE G KK
..130. 140. 150. 160. 170. 180..190. 200. 210. 220. 230. 240. 250
SWIA88 I IA
C-SWINE M
HK83 D R M T I K S
HUMAN D M
HUM-SWINE M
AVIAN 1 GEDATAGLTHLMIWHSNLNDATYQRTRALVRTGMDPRMCSLMQGSTLPRRSGMGMVKGVGTMVMELIRMIKRGINDRNFWRGENGRRTRIAYERMCNILKGKFQTAAQRAMMDQVRESRNPGN
AVIAN 2
AE-SWINE
....... 260. 270. 280. 290. 300. 310. 320. 330. 340. 350. 360. 370.
SWIA88 H K KK K V A
C-SWINE H KK
HK83 S P S K K Y LL K S K DA E
HUMAN F P Y K K E
HUM-SWINE K
AVIAN 1 AEIEDLIFLARSALILRGSVAHKSCLPACVYGLAVASGYDFEREGYSLVGIDPFRLLQNSQVFSLIRPNENPAHKSQLVWMACHSAAFEDLRVSSFIRGTRWPRGQLSTRGVQIASNENMETMD
AVIAN 2
AE-SWINE V K
..380 . 390 . 400 . 410 . 420 . 430 . 440 .450 .460 . 470 .480 .490 .498
SWIA88 K V S N V K L
C-SWINE V I S
HK83 A KS A G K E R
HUMAN I D V AS
HUM-SWINE V S
AVIAN 1 SSTLELRSRYWAIRTRSGGNTNQQRASAGQISVQPTFSVQRNLPFERATIMAAFTGNTEGRTSDMRTEI IRMMENARPEDVSFQGRGVFELSDEKATNPIVPSFDMSNEGSYFFGDNAEEYDN
AVIAN 2 S
AE-SWINE K
FIG. 2. Predicted amino acid sequences of hypothetical ancestor NPs. The ancestor sequences show the synapomorphic characters that
define each lineage relative to the avian ancestral root NP (lower star in Fig. 1B) and are derived from analysis of amino acid sequences
grouped by phylogenetic relationship (Fig. 1; see text). AVIAN 1 and AVIAN 2 represent one of two possible sequences for the avian
ancestral root NP and differ by two amino acids (positions 105 and 450). Amino acid differences at these positions distinguish H13 gull and
North American avian (AVIAN 1) from Old World avian (AVIAN 2) groups. AVIAN 2 is equivalent to baseline Old World avian NP
sequences DKBAV77, MAST82, and MAST2-82 (Fig. 1B). Other ancestor NP sequences shown are human-swine lineage (HUM-SWINE;
indicated by the upper star in Fig. 1), ancestor of classical swine NPs (C-SWINE; 1925 node in Fig. 1), a recent classical swine virus NP
(SWIA88), the ancestor of human NPs (HUMAN; 1924 node in Fig. 1), and a recent human virus NP (HK83). The sequence shown for the
European avianlike swine virus NPs (AE-SWINE) represents a consensus sequence of characters that are shared in at least two isolates. The
only character that is shared by all AE-SWINE isolates and distinguishes them from the ancestral avian baseline is a valine at position 284.
Recent HlNl avianlike swine viruses can be compared for
parallel patterns of evolution with early classical swine virus
NPs. The common ancestor of the new HlNl avianlike
swine NP lineage shows two changes relative to the avian
baseline at positions 31 and 284; by 1985, four additional
changes had appeared, at positions 98, 99, 351, and 384 (Fig.
2, AE-SWINE); and the 1989 isolate SWIT89 shows four
more changes at positions 49, 323, 377, and 497 (not shown).
With the exception of the lysine at position 351, none of
these amino acids are found in classical swine or human
virus NPs. The lysine at position 351 appears to be an
example of convergence with human and classical swine
NPs.
Evolutionary rate analysis. A regression of branch dis-
tances of NP isolates from common ancestor nodes permits
estimation of evolutionary rates for the human lineage and,
for the first time, the classical swine virus lineage (Fig. 3). At
the nucleotide level, human and swine virus NP genes are
evolving at nearly identical rates (3.31 to 3.41 nucleotide
changes per year [Fig. 3]), but at the protein level the swine
virus NPs are evolving at half the rate of human virus NPs
(0.66 versus 0.34 amino acid change per year [Fig. 3]). The
per-nucleotide annual rate of change is 2.12 x 10-3 to 2.18 x
10-3 change per year for swine and human virus NPs,
respectively. These values are higher than our previous
estimate of 1.62 x 10-3 change per year for the human NP
lineage (15), which was based on fewer human virus NPs (6
versus 20 isolates), but are very close to the 2.2 x 10-3
change per year estimated for human virus NPs (1) and are
similar to the estimate of 2.0 x 10-3 change per year for the
NS gene (6). Our estimate of evolutionary rate for Old World
avian NPs (1.90 changes per gene per year, 1.21 x 10-3
change per nucleotide per year [Fig. 4]) is lower than
previously reported (2.17 changes per gene per year; 1.39 x
10-3 change per nucleotide per year) (15). The new estimate
may be regarded as more accurate since it is based on a
larger sample of avian NPs (22 versus 13 NPs).
NPs of reemergent HlNl human viruses (USSR77 and
BRAZIL78) share a common ancestor with the WARREN50
isolate as previously shown (1) or predicted (29, 41). Assum-
ing a 1950 origin, we treated these reemergent isolates as
appearing in 1951 (USSR77) and 1952 (BRAZIL78). The
close position ofthese isolates to the regression line supports
the interpretation that they had not been in circulation for 27
years.
Dates for common-ancestor nodes can be derived from
estimates of evolutionary rates for each lineage. Indepen-
dent estimates from the human and classical swine virus
VOL. 65, 1991
3710 GORMAN ET AL.
NP Nucleotide Evolutionary Rate NP Nucleotide Evolutionary Rate
25
20
15
0
0 OW avian b-1.90
iO
S0
i0o
SO °
10
o ~
1910 1920 1930 1940 1950 1960 1970 1980 1990
year of isolation
NP Amino Acid Evolutionary Rate
b
r
a
n
c
h
d
i
s
t
a
n
c
e
b
r
a
n
c
h
d
s
t
a
n
c
e
1910 1920 1930 1940 1950 1960 1970 1980 1990
year of isolation
FIG. 3. Evolutionary rates for NP genes and proteins for human
and swine virus isolates. The evolutionary rate is estimated by
regression of the year of isolation against the branch distance from
the common ancestor node of the nucleotide and amino acid
phylogenetic trees (Fig. 1). Regression statistic b provides rate
estimates. Shown are regressions for NPs of human viruses, classi-
cal swine viruses, and European avianlike swine viruses. The
reemergent HlNl human virus isolates (USSR77 and BRAZIL78,
indicated by +) are not included in the regression but are shown for
reference (they are treated as appearing 27 years earlier). Three
classical swine virus isolates (SWIA46, SWJMS42, and SW41-49,
indicated by x) that show no evolution are not included in the
regression but are shown for reference. Nucleotide regressions have
been extrapolated beyond the first isolates and show estimated dates
of origin for human (1912), classical swine (1913), and European
avianlike swine (1979) virus lineages. Similar treatment of amino
acid regressions yields earlier estimates for human (1900) and
classical swine (1905) virus lineages (not shown). Dotted lines show
average estimated evolutionary rates of human and classical swine
NPs if 1918 is assumed to be the date of origin.
lineages agree on 1912 to 1913 as the date of the common
ancestor (Fig. 1A and 3). This common ancestor is relatively
close to the hypothetical ancestor of the avian lineages (five
amino acid changes) and is well within the range of distances
for all avian NPs (Fig. 1B). The date for the hypothetical
v11
1910
NP Amino Acid Evolutionary Rate
1920 1930 1940 1950 1960 1970 1980 1990
year of isolation
FIG. 4. Evolutionary rates for NP genes and proteins for avian
virus isolates. The evolutionary rate is estimated by regression of
the year of isolation against the branch distance from the common
ancestor node of the nucleotide and amino acid phylogenetic trees
(Fig. 1). Regression statistic b provides rate estimates. Extrapola-
tion of the nucleotide regression estimates 1904 as the date of the
ancestor of Old World avian virus NPs related to FPV27.
ancestor of the Old World avian lineage is estimated at 1904
(Fig. 1A).
Evolutionary rate analysis of the avianlike swine virus
NPs indicates that they are evolving at a higher rate than
classical swine virus NPs (Fig. 3): 4.5 versus 3.31 changes
per year at the nucleotide level, and 0.74 versus 0.34 change
per year at the amino acid level. We estimate 1979 as the
date for the common ancestor of these avianlike swine virus
NPs, which matches the date when the first avianlike swine
virus was isolated in Europe (38). The new lineage shows
relatively rapid, divergent evolution away from an avianlike
ancestor within the avian lineage (Fig. 1).
DISCUSSION
Common ancestry for human and classical swine viruses.
Our evolutionary analysis of influenza A virus NP nucleotide
sequences shows that human and classical swine viruses
30
b
r
a
n
c
h
d
i
t
a
n
c
e
b
r
a
n
c
h
d
i
t
a
n
c
e
14-
1
0 OW avian b--0.01
12-
10 K0
8
6 )0
4 - 0 0 (X> 0 0
2~~~ ~ O
0O
2 )0
n K
J. VIROL.
EVOLUTION OF HUMAN AND SWINE HlNl VIRUSES 3711
(i.e., those related to Swine/Iowa/15/30) share a single com-
mon ancestor and that this ancestor is estimated to have
emerged in 1912 to 1913. Moreover, our analysis shows that
NP genes in human and swine virus lineages have evolved at
similar rates that are higher than for avian virus NP genes
(Fig. 1A, 3, and 4). It is not possible from this analysis to
determine the host in which the virus first appeared. If there
had been some asymmetry in the lineages, e.g., early swine
virus NPs forming a sister group to NPs of later human and
classical swine viruses, this would have provided some
evidence that pigs were the original host. Instead, the two
nucleotide lineages are highly symmetrical and estimate
virtually the same date for a common ancestor.
Gammelin et al. (13) have proposed that the common
ancestor for avian and human virus NPs may have existed as
far back as 1837. In their analysis, human virus NP amino
acid evolutionary rates over the past 50 years were extrap-
olated back 150 years to the common ancestor with the avian
virus NP, DKBAV77. Such lengthy extrapolations may be
misleading (15, 16). The disparity in our estimates arises
because (i) our estimates are based on nucleotide phyloge-
nies and (ii) we attempt to estimate only the date of the
immediate common ancestor for human and classical swine
virus NPs and then evaluate the relationship of that ancestor
to those of avian virus NPs. We believe that more accurate
rates of evolution and dates of common ancestors can be
estimated from nucleotide phylogenies. This is because
protein evolution is dependent on nucleotide evolution and
nucleotide sequences contain much more information to
resolve ambiguous relationships, particularly when a high
proportion of silent mutations occurs (e.g., in avian virus
lineages). Moreover, because selection acts directly on the
proteins (the phenotype) whereas the genetic code is degen-
erate, amino acid evolutionary rates may be less constant
over time, particularly after a virus is introduced into a new
host. During this early adaptation phase, selection pressure
on proteins and their evolutionary rates are expected to be
relatively high. Thus, linear extrapolation of long-term
amino acid evolutionary rates is expected to provide earlier
estimates of dates of common ancestors. For example, on
the basis ofamino acid evolutionary rates in our analysis, the
common ancestor of human and classical swine virus NPs is
estimated to have appeared in 1900 or 1905, 7 to 13 years
earlier than estimates based on nucleotide evolutionary rates
(Fig. 3).
The time frame given by Gammelin et al. (13) for NP
evolution suggests that the virus may have been in humans
for 75 years before our estimated divergence of human and
classical swine viruses in 1912 to 1913. Our analysis does not
support a long history of NP evolution in mammalian hosts
prior to 1912. The internal branch that connects the human
and swine NP amino acid lineages to the avian root (between
the upper and lower stars in Fig. 1B) represents the changes
in the common ancestor that are shared by human and swine
virus NP lineages. The closeness ofthis common ancestor to
the avian root (five amino acid changes, only three of which
are not shared with any avian NP [Fig. 2]) suggests that this
ancestor NP had recently been acquired from an avian virus.
Comparison of the homologous internal branches in nucleo-
tide and amino acid trees (Fig. 1) shows that the bulk of the
nucleotide changes are silent coding changes (coding-to-
noncoding ratio, 1:10.6). In this respect, the human-swine
NP common ancestor is avianlike; i.e., evolutionary stasis of
the protein limits coding changes at the nucleotide level (15,
16). This parallel suggests that the majority of the nucleotide
changes in this common ancestor gene were inherited from
an avian ancestor.
Differences in protein evolution between human and swine
virus NPs are evident in the earliest virus isolates and
represent host-specific signatures of virus evolution. After
divergence of human and swine NP lineages, the ratio of
coding to noncoding changes in the human virus NP in-
creased and remained stable at 1:4.17 (1933 to 1983 mean;
ratio ofamino acid to nucleotide evolutionary rate regression
slopes - 1 [Fig. 3]), which indicates a shift in gene evolution
relative to the avianlike common ancestor (Fig. 1). Unlike
evolutionary stasis among avian NPs, the human virus NP
has undergone relatively rapid, divergent protein evolution
which has not abated over the 50 or more years human
viruses have been isolated. In comparison, evolution in the
swine lineage remains more avianlike; i.e., the swine virus
NP protein has evolved more slowly away from its avian
ancestor, as indicated by a smaller ratio of coding to non-
coding changes (1930 to 1988 mean, 1:8.74 [Fig. 4]).
Divergent evolution in human and classical swine viruses.
The common ancestor for human and classical swine virus
NPs marks the point where the evolution of the two viruses
diverged. That divergence reflects a split in the virus gene
pool into human and swine host reservoirs and shows that
the viruses have evolved independently in each reservoir. It
is apparent that regular interspecies transmissions of human
and swine viruses (see, e.g., references 7, 10, 19, and 37) and
examples from this report (i.e., NJ76, WIS88, SWCAM35,
SWHK76, and SWDAN83) have not affected virus evolution
in the two host species. Thus, the continued divergent
evolution of the viruses in the two host reservoirs suggests
that there are as yet unidentified factors that allow only
certain viruses to persist, circulate, and evolve in each host
species. The closeness of the human-swine common ances-
tor to avian NP proteins, the divergence and rapid parallel
evolution of NP genes in human and swine viruses, and the
different signatures of host-specific protein evolution in
human and swine viruses suggest that the avianlike ancestor
could have circulated for only a short time in one of the two
host reservoirs before entering the other.
Evolutionary divergence in human and classical swine
viruses is probably related to differences in immune protec-
tion as it is related to host population age structure. Human
populations are characterized by older individuals (>10
years) with extensive immunological experience, whereas in
swine populations young, immunologically naive individuals
predominate. As a result of these differences, a newly
established virus would be subjected to strong immune
selection pressure in human populations which would result
in rapid evolution of virus antigens; the reverse would be
expected for a virus in swine populations. Following estab-
lishment of the same new virus in human and swine popu-
lations, only a small fraction ofhumans would continue to be
susceptible to the swine virus strains, and such cases would
not lead to epidemics. On the other hand, pigs could con-
tinue to be infected with human virus strains, but it is
apparent that human-to-swine virus transmissions in North
America over the past 60 years have not resulted in new
persistent swine virus strains or lineages (Fig. 1). This
evolutionary model is appropriate for the HA and NA
surface proteins of influenza viruses because they are the
principal targets of neutralizing antibodies (28). However,
because the evolution of the internal NP protein is concor-
dant with this model, it is possible that selection pressure via
T-cell immune response has affected the evolution of NP
genes. In contrast to virus protein evolution, host demo-
VOL. 65, 1991
3712 GORMAN ET AL.
If
swine
1 9 18
swine
A swi
human
1918
human
swine
1918
Si
B
avian
human
,S
IfWS33
human
swine?
avian
FIG. 5. Hypothetical phylogenies for early human and classical swine viruses adapted from Fig. 1. Names of earliest isolates identify
lineages for classical swine (SWIA30) and human (WS33) viruses. All phylogenies show the common ancestor as being recently derived from
an avian virus. (A) Divergence of human and classical swine viruses from a 1912 to 1913 common ancestor. To account for historical records,
coincident pandemics and epizootics had to arise independently from virulent strains in 1918, which is not likely. Also, this phylogeny requires
a new human virus to circulate and evolve for 6 years (1912 to 1918) without causing pandemics. (B) Divergence of multiple lineages of swine
viruses from a 1912 to 1913 common ancestor. Of the two surviving lineages, the one that gives rise to classical swine viruses also caused the
human influenza pandemic of 1918, but present-day human viruses are shown as being derived from a sister lineage of swine viruses. This
model requires multiple strains of swine viruses to circulate and evolve independently for 6 years or more after 1912. This pattern of early
evolution has not yet been demonstrated for recent avianlike swine viruses. The proposed post-1918 origin of human viruses from a sister
swine virus strain would require higher pre-1933 evolutionary rates for human virus genes than is predicted by assuming a 1918 origin (Fig.
3). (C) Human and swine viruses diverging from a common ancestor in 1918. In this model the classical swine viruses are derived from the
1918 human pandemic virus. The pre-1918 host is unknown but could have been swine. This phylogeny requires higher evolutionary rates for
human and classical swine virus over the period from 1918 to the early 1930s. However, these estimated rates are similar to those for the new
avian-like HlNl swine viruses.
graphic differences appear to have had little impact on the
evolution of NP genes; the similar nucleotide evolutionary
rates observed in the two host populations suggest that the
viruses may have similar replication (and mutation) rates.
An important characteristic of classical swine virus is that
it has circulated and evolved in North American swine herds
for more than 60 years. However, when classical swine
viruses have been introduced into Europe or Asia, they
usually have not persisted for long periods. The reasons for
this pattern are not understood. The association of classical
swine viruses with North America suggests a unique regional
ecology which is probably related to differences in swine
husbandry practices in North America, Europe, and Asia.
A 1918 origin for HlNl human and classical swine viruses?
The estimated date of 1912 to 1913 for the common ancestor
of human and swine viruses (Fig. SA) does not agree with
results of seroarcheological studies that suggest that 1918
marked the appearance of a new HlNl virus in the human
population (see, e.g., reference 27), nor with 1918 historical
records of a severe pandemic of human influenza and epi-
zootics of swine virus in North America (9). An alternative
explanation might be that a number of virus lineages may
have diverged around 1912 to 1913 and the only surviving
lineages are those that now represent swine and human hosts
(e.g., Fig. 5B). A major assumption in estimating the date of
the common ancestor is that evolutionary rates within a
lineage remain constant from its origin. If 1918 is accepted as
the date for the divergence ofthe human and swine virus NP
lineage (Fig. 5C), evolutionary rates from 1918 to the early
1930s were significantly higher than after that period. Our
evolutionary analysis suggests that human and classical
swine viruses originated from a common event and that the
ancestor virus had only a short prior existence in human or
swine host reservoirs.
Model for evolution of early human and swine viruses.
Evolution of recent avianlike HlNl swine viruses in Europe
may serve as a model for evolution of early human and
classical swine viruses. The NP genes of these avianlike
viruses have been evolving in the European swine host
reservoir for some 10 years (1979 to 1989) at a higher rate
than is observed for human or classical swine virus NPs
(4.50 versus 3.41 and 3.31 nucleotide changes per year,
respectively [Fig. 3]). Higher nucleotide evolutionary rates
may be typical for an avian virus recently introduced into the
swine host reservoir. If the evolutionary rate of the new
avianlike swine virus NPs is applied to the human and
classical swine virus lineages for the pre-1930s period, the
estimated date of the common ancestor is 1918 to 1919.
Conversely, if 1918 is assumed as the date of the common
ancestor, the predicted initial evolutionary rate is 4.7 nucle-
otide changes per year for human NPs (between 1918 and
1933) and 4.75 changes per year for classical swine NPs
(between 1918 and 1930). With this modification, the pre-
dicted amino acid evolutionary rate for 1918 to 1930 classical
swine virus NPs becomes more similar to that of the recent
avianlike swine virus NPs (0.9 versus 0.74 changes per year)
than to that of the pre-1930s human NPs (1.35 changes per
year). The much higher predicted evolutionary rate for early
human virus NP proteins than for classical swine viruses
mirrors the pattern for post-1930s human and classical swine
viruses (Fig. 3). The parallel between pre-1930 classical
swine virus NPs and the new avianlike swine virus NPs
suggests that the SWIA30 isolate is nearly as avianlike as the
recent SWIT89 isolate; from the root of the avian lineage,
SWIA30 has diverged by 14 amino acid changes in 12 to 17
years and SWIT89 has accumulated 11 changes in 10 years.
The unusually close agreement in the predicted evolution of
these human and swine virus NPs makes a post-1912 date of
origin for human and classical swine viruses plausible.
An unusual aspect of evolution in the new avianlike swine
virus NPs is that the rate of amino acid evolution is twice as
high as that in classical swine viruses while nucleotide
J. VIROL.
EVOLUTION OF HUMAN AND SWINE HlNl VIRUSES 3713
evolution is only 30 to 40% higher (Fig. 3). The dispropor-
tionally higher rate of NP protein evolution in the new
avianlike swine viruses compared with the classical swine
viruses suggests that they are under stronger selective pres-
sure, which, as discussed above, is expected for new vi-
ruses. The same situation would apply to pre-1930 classical
swine viruses if their date of origin is assumed to be 1918
(Fig. 3).
A proposal consistent with our evolutionary analysis and
historical records would be that an HlNl avianlike virus
entered human or swine hosts a short time before 1918. In
1918 a virulent strain of this virus caused a severe human
influenza pandemic, and, as suggested by Crosby (9), the
human pandemic virus was subsequently introduced into
North American swine populations, giving rise to the clas-
sical swine viruses (Fig. SC). The suggestion that classical
swine virus was derived from the human pandemic virus of
1918 does not imply that swine did not play a role in the
appearance of this new virus in the human population. It is
entirely possible that somewhere in the world, swine may
have served as an intermediate host prior to the emergence
of the 1918 pandemic virus. Clearly, such an event would be
highly transitory and therefore impossible to detect in an
evolutionary analysis. The potential of avian viruses to
enter, circulate, and evolve in the swine host reservoir is
demonstrated by the new avianlike swine viruses in Europe.
The possibility that the 1918 virus was not a reassortant
virus but an entirely novel avianlike virus may be a contrib-
uting factor to its virulence, predicted rapid evolution, and
entrance into and persistence in the swine host reservoir. A
comparison of mortality figures shows that the reassortant
pandemic viruses of 1957 and 1968 were apparently milder
than the highly virulent 1918 virus (9). There is evidence that
wide dissemination of human virus during the 1968 pandemic
resulted in an increased incidence of human viruses isolated
from pigs (see, e.g., references 24, 30, and 49). Unlike the
1918 pandemic virus, there is as yet no evidence that the
pandemic viruses of 1957 or 1968 resulted in new swine-
specific strains that have persisted to the present day. Also,
there is no evidence that the pandemics of 1957 and 1968 had
any detectable effect on the evolution of NPs of human or
classical swine viruses (Fig. 1 and 4). Therefore, although
new strains of swine viruses may be expected to appear
following human pandemics, few would be expected to
persist and evolve in the swine host reservoir.
The mode of interspecies transmission of influenza viruses
or genes from the avian host reservoir to the human popu-
lation remains unresolved. The 1957 and 1968 pandemic
viruses originated from China, and there is anecdotal evi-
dence that 19th century pandemic viruses also originated
from China (28). Humanlike and avianlike viruses are regu-
larly isolated from pigs in China, Taiwan, and Southeast
Asia (24, 33, 42, 44), but persistent, widely circulating,
swine-specific strains comparable to classical swine viruses
in North America or the new avianlike swine viruses in
Europe have not been found. Common Oriental agricultural
practices place humans, swine, and domestic ducks in close
association and enhance the likelihood of interspecies trans-
mission of influenza viruses. Further understanding of the
origin of human pandemic viruses will require detailed
knowledge of the ecology and evolution of human, swine,
and avian influenza viruses in a variety of situations around
the world where interspecific exchanges are likely.
ACKNOWLEDGMENTS
We thank Raphael Onwuzuruigbo and Evelyn Stigger for techni-
cal assistance, Clayton Naeve and the SJCRH Molecular Resource
Center for preparation of oligonucleotides, and Patricia Eddy and
the SJCRH Molecular Biology Computer Facility for computer
support.
This work was supported by U.S. Public Health research grant
AI-29680, National Institute of Allergy and Infectious Diseases
grant AI-29599 and AI-08831, Cancer Center Support (CORE) grant
CA-21765, National Research Service Award 5T32-CA09346 to
O.T.G., and American Lebanese Syrian Associated Charities.
REFERENCES
1. Altmuller, A., W. M. Fitch, and C. Scholtissek. 1989. Biological
and genetic evolution of the nucleoprotein gene of human
influenza A viruses. J. Gen. Virol. 70:2111-2119.
2. Andrewes, C. H., P. P. Laidlaw, and W. Smith. 1935. Influenza:
observations on the recovery of virus from man and antibody
content of human sera. Br. J. Exp. Med. 17:579-581.
3. Bean, W. J. 1984. Correlation of influenza A virus nucleoprotein
genes with host species. Virology 133:438-442.
4. Briedis, D. J., and M. Tobin. 1984. Influenza B genome:
complete nucleotide sequence of the influenza B/Lee/40 virus
genome RNA segment 5 encoding the nucleoprotein and com-
parison with the B/Singapore/222/79 nucleoprotein. Virology
133:448-455.
5. Buckler-White, A. J., and B. R. Murphy. 1986. Nucleotide
sequence analysis of the nucleoprotein gene of an avian and a
human virus strain identifies two classes of nucleoproteins.
Virology 155:345-355.
6. Buonagurio, D. A., S. Nakada, J. D. Parvin, M. Krystal, P.
Palese, and W. M. Fitch. 1986. Evolution of human influenza A
viruses over 50 years: rapid and uniform rate of change in the
NS gene. Science 232:980-982.
7. Chambers, T. M., V. S. Hinshaw, Y. Kawaoka, B. C. Easterday,
and R. G. Webster. Influenza viral infection of swine in the
United States 1988-1989. Arch. Virol., in press.
8. Cox, N. J., F. Kitame, A. P. Kendal, H. F. Maassab, and C.
Naeve. 1989. Identification of sequence changes in the cold-
adapted live attenuated influenza vaccine strain A/Ann Arbor/
6/60 (H2N2). Virology 167:554-567.
9. Crosby, A. W. 1989. America's forgotten pandemic, the influ-
enza of 1918. Cambridge University Press, New York.
10. Dasco, C. C., R. B. Couch, H. R. Six, J. F. Young, J. M.
Quarles, and J. A. Kasel. 1984. Sporadic occurrence of zoonotic
swine influenza virus infections. J. Clin. Microbiol. 20:833-835.
11. Davenport, F. M., A. V. Hennessy, J. Drescher, J. Mulder, and
T. Francis. 1964. Further observations on the relevance of
serologic recapitulation of human infection with influenza vi-
ruses. J. Exp. Med. 120:1087-1097.
12. Davenport, F. M., A. V. Hennessy, and T. Francis. 1953.
Epidemiologic and immunologic significance of age distribution
of antibody to antigenic variants of influenza virus. J. Exp. Med.
98:641-656.
12a.Donatelli, I. Unpublished data.
13. Gammelin, M., A. Altmuller, U. Reinhardt, J. Mandler, V. R.
Harley, P. J. Hudson, W. M. Fitch, and C. Scholtissek. 1990.
Phylogenetic analysis of nucleoproteins suggests that human
influenza A viruses emerged from a 19th-century avian ances-
tor. Mol. Biol. Evol. 7:194-200.
14. Gammelin, M., J. Mandler, and C. Scholtissek. 1989. Two
subtypes of nucleoproteins (NP) of influenza A viruses. Virol-
ogy 170:71-80.
15. Gorman, 0. T., W. J. Bean, Y. Kawaoka, and R. G. Webster.
1990. Evolution of the nucleoprotein gene of influenza A virus.
J. Virol. 64:1487-1497.
16. Gorman, 0. T., R. 0. Donis, Y. Kawaoka, and R. G. Webster.
1990. Evolution of influenza A virus PB2 genes: implications for
evolution of the ribonucleoprotein complex and origin of human
influenza A virus. J. Virol. 64:4893-4902.
17. Hennessy, A. V., and F. M. Davenport. 1958. Epidemiologic
implications of the distribution by age of antibody response to
experimental influenza virus vaccines. J. Immunol. 80:114-121.
VOL. 65, 1991
3714 GORMAN ET AL.
18. Hennessy, A. V., F. M. Davenport, and T. Francis. 1955. Studies
of antibodies to strains of influenza virus in persons of different
ages in sera collected in a postepidemic period. J. Immunol.
75:401-405.
19. Hinshaw, V. S., W. J. Bean, R. G. Webster, and B. C. Easterday.
1978. The prevalence of influenza viruses in swine and the
antigenic and genetic relatedness of influenza viruses from man
and swine. Virology 84:51-62.
20. Hinshaw, V. S., and R. G. Webster. 1982. The natural history of
influenza A viruses, p. 79-104. In A. S. Beare (ed.), Basic and
applied influenza research. CRC Press, Boca Raton, Fla.
21. Huddleston, J. A., and G. G. Brownlee. 1982. The sequence of
the nucleoprotein gene of human influenza A virus, strain
A/NT/60/68. Nucleic Acids Res. 10:1029-1038.
22. Kawaoka, Y., S. Krauss, and R. G. Webster. 1989. Avian to
human transmission of the PB1 gene of influenza A viruses in
the 1957 and 1968 pandemics. J. Virol. 63:4603-4608.
23. Kida, H., K. F. Shortridge, and R. G. Webster. 1988. Origin of
the hemagglutinin gene of H3N2 influenza viruses from pigs in
China. Virology 162:160-166.
24. Kundin, W. D. 1970. Hong Kong A-2 influenza virus infection
among swine during a human epidemic in Taiwan. Nature
(London) 228:857.
25. Mandler, J., 0. T. Gorman, S. Ludwig, E. Schroeder, W. M.
Fitch, R. G. Webster, and C. Scholtissek. 1990. Derivation of the
nucleoproteins (NP) of influenza A viruses isolated from marine
mammals. Virology 176:255-261.
26. Mandler, J., and C. Scholtissek. 1989. Localisation of the
temperature-sensitive defect in the nucleoprotein of an influenza
A/FPV/Rostock/34 virus. Virus Res. 12:113-122.
27. Mulder, J., and N. Masurel. 1958. Pre-epidemic antibody
against 1957 strain of Asiatic influenza in serum of older people
living in the Netherlands. Lancet i:810-814.
28. Murphy, B. R., and R. G. Webster. 1985. Influenza viruses, p.
1179-1239. In B. N. Fields (ed.), Virology. Raven Press, New
York.
29. Nakajima, K., U. Desselberger, and P. Palese. 1978. Recent
human influenza viruses are closely related genetically to strains
isolated in 1950. Nature (London) 274:334-339.
30. Nakajima, K., S. Nakajima, K. F. Shortridge, and A. P. Kendal.
1982. Further genetic evidence for maintenance of early Hong
Kong-like influenza A(H3N2) strains in swine until 1976. Virol-
ogy 116:562-572.
31. Nakajima, K., E. Nobusawa, and S. Nakajima. 1984. Genetic
relatedness between A/swine/Iowa/15/30 (HlNl) and human
influenza viruses. Virology 139:194-198.
32. Nardelli, L., S. Pascucci, G. L. Gualandi, and P. Loda. 1978.
Outbreaks of classical swine influenza in Italy in 1976. Zen-
tralbl. Veterinaermed. Reihe B 25:853-857.
33. Nerome, K., M. Ishida, A. Oya, C. Kanai, and K. Suwicha. 1982.
Isolation of an influenza HlNl virus from a pig. Virology
117:485-489.
34. Nerome, K., M. Ishida, A. Oya, and K. Oda. 1982. The possible
origin of HlNl (HswlNl) virus in the swine population ofJapan
and antigenic analysis of the isolates. J. Gen. Virol. 62:171-175.
35. Okazaki, K., Y. Kawaoka, and R. G. Webster. 1989. Evolution-
ary pathways of the PA genes of influenza A viruses. Virology
172:601-608.
36. Reinhardt, U., and C. Scholtissek. 1988. Comparison of the
nucleoprotein genes of a chicken and a mink influenza A H10
virus. Arch. Virol. 103:139-145.
37. Schnurrenberger, P. R., G. T. Woods, and R. J. Martin. 1970.
Serologic evidence of human infection with swine influenza
virus. Am. Rev. Respir. Dis. 102:356-361.
38. Scholtissek, C., H. Burger, P. A. Bachmann, and C. Hannoun.
1983. Genetic relatedness of hemagglutinins of the Hi subtype
of influenza A viruses isolated from swine and birds. Virology
129:521-523.
39. Scholtissek, C., H. Burger, 0. Kistner, and K. F. Shortridge.
1985. The nucleoprotein as a possible factor in determining host
specificity of influenza H3N2 viruses. Virology 147:287-294.
40. Scholtissek, C., and E. Naylor. 1988. Fish farming and influenza
pandemics. Nature (London) 331:215.
41. Scholtissek, C., V. Von Hoyningen, and R. Rott. 1978. Genetic
relatedness between the new 1977 epidemic strain (HlNl) of
influenza and human influenza strains isolated between 1947 and
1957 (HlNl). Virology 89:613-617.
42. Shortridge, K. F., A. Cherry, and A. P. Kendal. 1979. Further
studies of the antigenic properties of H3N2 strains of influenza
A isolated from swine in Southeast Asia. J. Gen. Virol. 44:251-
254.
43. Shortridge, K. F., and R. G. Webster. 1979. Geographical
distribution of swine (HswlNl) and Hong Kong (H3N2) influ-
enza variants in pigs in Southeast Asia. Intervirology 11:9-11.
44. Shortridge, K. F., R. G. Webster, W. K. Butterfield, and C. H.
Campbell. 1977. Persistence of Hong Kong influenza virus
variants in pigs. Science 196:1454-1455.
45. Snyder, M. H., A. J. Buckler-White, W. T. London, E. L.
Tierney, and B. R. Murphy. 1987. The avian influenza virus
nucleoprotein gene and a specific constellation of avian and
human virus polymerase genes each specify attenuation of
avian-human influenza A/pintail/79 reassortant viruses for mon-
keys. J. Virol. 61:2857-2863.
46. Steuler, H., B. Schroder, H. Burger, and C. Scholtissek. 1985.
Sequence of the nucleoprotein gene of influenza A/parrot/
Ulster/73. Virus Res. 3:35-40.
47. Tian, S. F., A. J. Buckler-White, W. J. London, L. J. Recle,
R. M. Chanock, and B. R. Murphy. 1985. Nucleoprotein and
membrane protein genes are associated with restriction of
replication of influenza A/mallard/NY/78 virus and its reassor-
tants in squirrel monkey respiratory tract. Virus Res. 1:625-630.
48. Van Rompuy, L., W. Min Jou, D. Huylebroeck, R. Devos, and
W. Fiers. 1981. Complete nucleotide sequence of the nucleopro-
tein gene from the human influenza strain A/PR/8/34 (HON1).
Eur. J. Biochem. 116:645.
49. Webster, R. G., and W. G. Laver. 1975. Antigenic variation of
influenza viruses, p. 269-314. In E. D. Kilbourne (ed.), The
influenza viruses and influenza. Academic Press, Inc., New
York.
50. Winter, G., and S. Fields. 1981. The structure of the gene
encoding the nucleoprotein of human influenza virus A/PR/8/34.
Virology 114:423-428.
51. Yamane, N., J. Arikawa, T. Odagiri, and N. Ishida. 1979.
Annual examination of influenza virus infections among pigs in
Miyagi Prefecture, Japan, 1979. Acta Virol. 23:240-248.
J. VIROL.
