nature outlook influenza 2011

INFLUENZA8 December 2011 / Vol 480 / Issue No. 7376

C O N T E N T S

Editorial Herb Brody, Michelle Grayson, Tony Scully

Art & Design Wes Fernandes, Alisdair MacDonald, Andrea Duffy

Production Karl Smart, Emilia Orviss, Leonora Dawson-Bowling

Sponsorship Patrick Murphy, Reya Silao, Yvette Smith, Gerard Preston

Marketing Elena Woodstock, Hannah Phipps

Project Manager Helen Anthony, Claudia Deasy

Art Director Kelly Buckheit Krause

Magazine Editor Tim Appenzeller

Editor-in-Chief Philip Campbell

COVER ART: NIK SPENCER

Amidst the series of tragic events that was the first half of the twentieth century, the 1918 influenza pandemic stands out for its swift lethality. Over the course

of about 18 months at the close of the First World War, an estimated 50 million people died from this viral infection.

Since the last flu pandemic in 2009, public health officials have started to stand guard against another catastrophe (page S11). Yet echoes of this outbreak have reverberated through the decades and the disease, in its various guises, kills at least a quarter of a million people each year worldwide. The virus is adept at adaptation, leaving scientists chasing a moving target. The virus’s surface proteins mutate rapidly and can combine into dozens of variants (page S2); the antibodies we produce to fight one year’s flu strain can’t stop the others.

Influenza evades the effects of drugs developed against it. First-generation antivirals are now almost useless, and newer classes of drugs — the neuraminidase inhibitors oseltamivir (Tamiflu) and zanamivir (Relenza) — are beset by resistance(page S9). And, owing to a number of factors, including genetics, individual responses to the virus vary widely; researchers are starting to get a handle on why some people infected with influenza become very sick or die while others are unscathed (page S14).

Fears of widespread contagion have grown because the virus seems to be making a habit of jumping from its original bird hosts into domesticated mammals and birds (page S4). Perhaps the great hope lies in flu prevention. Researchers are reporting progress in developing a vaccine that could be effective against a broad range of flu strains (page S6), a feat that might mark a turning point against this killer.

We are pleased to acknowledge the financial support of Sanofi Pasteur, Crucell and Baxter and we also acknowledge the unrestricted grant from F.Hoffmann-La Roche Ltd. As always, Nature retains sole responsibility for all editorial content.Herb BrodySupplements Editor, Nature Outlook.

S2 EPIDEMIOLOGY Racing against the flu

Life isn’t easy living with one of humanities biggest killers

S4 Q&A The flu catcher

Richard Webby’s keeps track of the various ecological niches of influenza circulating in animal populations

S6 PREVENTION Vaccine for all seasons

Targeting the more stable parts of influenza’s surface proteins is showing some promise, and might lead to a universal vaccine

S9 DRUGS Lines of defence

In search of other antiviral treatments to prepare for the next pandemic

S11 PUBLIC HEALTH Life lessons

Without learning from influenza’s tragic history we are doomed to repeat it

S14 MORBIDITY A personal response Some people get sicker than most, and

others die. What’s behind these drastic differences?

8 D E C E M B E R 2 0 1 1 | V O L 4 8 0 | N A T U R E | S 1

OUTLOOK

Nature Outlooks are sponsored supplements that aim to stimulate interest and debate around a subject of interest to the sponsor, while satisfying the editorial values of Nature and our readers’ expecta-tions. The boundaries of sponsor involvement are clearly delineated in the Nature Outlook Editorial guidelines available at http://www.nature.com/advertising/resources/pdf/outlook_guidelines.pdf

CITING THE OUTLOOKCite as a supplement to Nature, for example, Nature Vol XXX, No. XXXX Suppl, Sxx–Sxx (2011). To cite previously published articles from the collection, please use the original citation, which can be found at the start of each article.

VISIT THE OUTLOOK ONLINEThe Nature Outlook Influenza supplement can be found at http://www.nature.com/nature/outlook/Influenza/

All featured articles will be freely available for 6 months.

SUBSCRIPTIONS AND CUSTOMER SERVICESFor UK/Europe (excluding Japan):Nature Publishing Group, Subscriptions, Brunel Road, Basingstoke, Hants, RG21 6XS, UK. Tel: +44 (0) 1256 329242. Subscriptions and customer services for Americas – including Canada, Latin America and the Caribbean: Nature Publishing Group, 75 Varick St, 9th floor, New York, NY 10013-1917, USA. Tel: +1 866 363 7860 (US/Canada) or +1 212 726 9223 (outside US/Canada). Japan/China/Korea:Nature Publishing Group — Asia-Pacific, Chiyoda Building 5-6th Floor, 2-37 Ichigaya Tamachi, Shinjuku-ku, Tokyo, 162-0843, Japan. Tel: +81 3 3267 8751.

CUSTOMER [email protected] Copyright © 2011 Nature Publishing Group

COLLECTIONS17 Live attenuated influenza virus

vaccines by computer-aided rational designSteffen Mueller et al.

S21 Long-term evolution and transmission dynamics of swine influenza A virusDhanasekaran Vijaykrishna et al.

S25 Dissolving polymer microneedle patches for influenza vaccinationSean Sullivan et al.

S 2 | N A T U R E | V O L 4 8 0 | 8 D E C E M B E R 2 0 1 1

B Y D U N C A N G R A H A M - R O W E

Influenza kills more than 250,000 people each year, worldwide. And with new strains ever emerging, there is always the threat of

a repeat of the Spanish flu pandemic of 1918, which claimed more than 50 million lives. Indeed, despite vaccines and antiviral drugs that offer some protection to the vulnerable, nearly a century after that catastrophe we still seem no closer to being able to predict and ultimately stave off such a virulent outbreak.

One reason this has proved so difficult is the speed at which this acute viral infection can spread, transmitted by coughs and sneezes. “It’s a very stable respiratory virus that survives

well on surfaces and people’s hands,” says flu specialist Anthony Mounts with the Global Influenza Programme at the World Health Organization in Geneva, Switzerland.

Moreover, the flu virus changes all the time, mutating at the genetic level, which alters the surface level — in the hemagglutinin and neu-raminidase antigens that coat it (and provide the familiar H-N designations). These proteins are what our antibodies attach to; by continu-

ally changing, these anti-gens enable the virus to infect people who have developed immunity to other strains. It is because of its ability to

rapidly mutate that the virus is able to develop resistance to previously effective drugs, such as oseltamivir (marketed as Tamiflu by Swit-zerland-based Roche) and amantadine.

On average between 5–15% of the global population are infected each year. Most at risk are children younger than 2 years and adults 65 years or older, as well as people of any age with asthma, diabetes or chronic heart dis-ease. But no one is entirely free from risk. “This was particularly notable in this last pandemic,” says Mounts, referring to the 2009 outbreak, in which as many as half of those contracting the disease had fully func-tioning immune systems and were perfectly healthy beforehand.

E P I D E M I O L O G Y

Racing against the fluInfluenza mutates fast and spreads easily, earning a place among humanity’s biggest killers.

H5N12003

H1N12009

>1,000500–1,000100–50050–10010–501–100

THE FAST OR THE FURIOUSTo exact the greatest toll, in�uenza needs to be both fast-spreading and highly lethal. Unlike the 1918 ‘Spanish �u’, the two recent outbreaks — H5N1 ‘bird �u’ and H1N1 ‘swine �u’ — only had one attribute each.

CASES ON THE RISEThe mortality rate has risen with the number of cases worldwide

CASES ON THE RISE

1960195519501918 201019701965 1985 1990 1995 2000 2005

CANADA921 CASES1 DEATH

AZERBAIJAN8 CASES5 DEATHSIRAQ

3 CASES2 DEATHS

EGYPT152 CASES52 DEATHS

CAMBODIA18 CASES16 DEATHS

LAOS2 CASES2 DEATHS

VIETNAM119 CASES59 DEATHS

THAILAND25 CASES17 DEATHS

MYANMAR1 CASES0 DEATHS

PAKISTAN3 CASES1 DEATHS

INDONESIA182 CASES150 DEATHS

CHINA40 CASES26 DEATHS

BANGLADESH3 CASES0 DEATHS

DJIBOUTI1 CASE0 DEATHS

USA6,764 CASES10 DEATHS

CHILE86 CASES0 DEATHS

COSTA RICA33 CASES1 DEATH

MEXICO4,541 CASES83 DEATHS

UK137 CASES0 DEATHS

AUSTRALIA39 CASES0 DEATHS

SPAIN138 CASES0 DEATHS

INDIA1 CASE0 DEATHS

JAPAN360 CASES0 DEATHS


BRAZIL9 CASES0 DEATHS

GERMANY17 CASES0 DEATHS2004–2006 H5N1 Cases Deaths

1890

The total number of deaths caused by the H5N1 in�uenza virus up to November 2011.


335Con�rmed deaths in 214 countries caused by the H1N1 virus up to August 2010.

Con�rmed deaths in 214 countries caused by the H1N1 virus up to August 2010.

18,449

The ‘Spanish Flu’ caused by the H1N1 in�uenza virus, kills approximately 50 million people.

Several new strains of in�uenza emerge throughout the twentieth century, circulating without turning pandemic.

1977H1N1 outbreak of Russian �u in China, similar to strain in circulation prior to 1957, spreads around the world disproportionately infecting people under the age of 23.50M H5N1

1997H7N71995

H7N3H10N7

2004H3N21968

H7N71959

H2N21957

H1N11918

H5N1H9N2

H7N7H3N2

H7N22003 1890

RUSSIAFirst recorded in�uenza pandemic. 1 million die worldwide.

1999: FDA approves the neuraminidase inhibitor oseltamivir.

1997: First case of direct trans-mission from birds to humans.

2009: Cases of Tami�u resistant H1N1 reported.

2010: August,the WHO announces end of H1N1 pandemic.

PANDEMICThe origin of H1N1 has been the subject of many studies — some suggest it may have gone directly from birds to humans, while others suggest it may have involved incubation in an intermediate host, such as the pig.

H9N21998

H1N12009

60% approximate proportion of fatalities for immunocompetent people infected with H5N1.

0.03%approximate proportion of fatalities for immunocompetent people infected with H1N1.

71–167BILLION

The estimated cost (US$) of seasonal �u to the US

economy each year

ASIAThe second in�uenza pandemic strikes in Asia, killing 100,000 people worldwide.

HONG KONGFlu pandemic kills 700,000 people worldwide. Virus still circulates today.

ASIAH5N1, �rst detected in humans in 1997, evolved in Hong Kong chickens, remains largely an avian virus but known to kill people in close contact with birds (see map, left).

MEXICOFirst pandemic of in�uenza in 40 years. Labs in the US con�rm the re-emergence of the dreaded H1N1 strain, the virus behind the 1918 pandemic.

THE SLOW BUT DEADLY SPREAD OF H5N1Between 2003 and 2011, H5N1, commonly known as bird �u, spreads across Asia but remains largely an avian virus.

THE SLOW BUT DEADLY SPREAD OF H5N1

H1N1, SNAPSHOT OF A RAPID PANDEMICIn May 2009, 48 countries reported 13,398 cases of in�uenza H1N1, the �rst in�uenza pandemic in four decades.H1N1, SNAPSHOT OF A RAPID PANDEMIC

300

250

200

150

100

50

02004

Peo

ple

2005 2006

People worldwide die each year from seasonal in�uenza infection.

250,000+2003–2004 Outbreak of the avian in�uenza H7N7 in Holland, among mainly poultry workers.Panama strain of H3N2 mutates into Fijian strain of H3N2.

NATURE.COMKeep up to date with the latest news and research on influenza: go.nature.com/azxjrn

INFLUENZAOUTLOOK


The virulence of an outbreak is a combina-tion of the viral lethality and infectiousness. These, in turn, depend on which of the three varieties of human flu virus is the pathogen: type A, type B or type C. Type A is by far the most virulent; with 16 known subtypes, it mutates into countless different genetic and antigenic strains at a rate far faster than type B or type C. This ability to mutate quickly allows type A flu to spread faster. All the flu strains discussed in this Nature Outlook are influenza type A.

Some of the most virulent strains encoun-tered are those that have managed to pass from other species to humans. The avian H5N1 flu virus had a very high lethality, kill-ing 56% of those who were infected. However it was also not terribly infectious: between 2003 and 2006 there were only 263 cases of H5N1, mitigating its overall impact. In con-trast, the H1N1 swine flu virus of 2009–2010 had a case fatality rate of only around 0.03% but, as it can be passed easily from human to human, it was responsible for more than 13,000 cases in a single month.

INFLUENZA MENAGERIEFlu has been found in a variety of mammals, from horses, cats, dogs, and pigs, to seals, fer-rets, camels and even whales. But it is mainly through wild aquatic birds that the virus is able to jump species — most likely because the virus originated in these avian species. Although it is next to impossible to predict where and when an inter-species jump is likely to take place, or which of these will result in a particularly virulent human strain, clues are beginning to emerge.

By analysing the genetic and antigenic varia-tions of 13,000 strains of the H3N2 virus, a group led by infectious-disease experts Colin Russell and Derek Smith at the University of Cambridge in the UK were able to track the evolution of the virus as it moved around the world between 2002 and 2007. They found that, at least for the common H3N2 virus, strains tend to originate in East and Southeast Asia, and mutate their way around the world until they end up in what Smith calls an “evolutionary graveyard” in South America — a continent that, in terms of flu at least, he says, is therefore “the safest place to be

because you get the biggest warning.”The conditions in Asia are ripe for new influ-

enza strains to emerge, says Russell. For one thing, the virus is able to exist in circulation almost perpetually, giving it more of a chance to mutate. Because tropical countries have no winter, seasonal influenza epidemics typically occur during the rainy season, which can hap-pen at different times of the year in neighbour-ing places like Bangkok and Kuala Lumpar, says Russell. This means that within East and South-east Asia there will be an influenza epidemic happening somewhere at any point in a year. “Given the travel and trade between cities and countries in the regions, influenza seems to be able to spread readily from place to place,” says Russell. “We often talk about it as being simi-lar to runners passing a baton in a relay race, as viruses move from epidemic to epidemic.”

It’s a race that researchers, public health authorities, and the public continue — with mixed success — to try to disrupt. ■

Duncan Graham-Rowe is a science writer in Brighton, UK.

H5N12003

H1N12009

>1,000500–1,000100–50050–10010–501–100

THE FAST OR THE FURIOUSTo exact the greatest toll, in�uenza needs to be both fast-spreading and highly lethal. Unlike the 1918 ‘Spanish �u’, the two recent outbreaks — H5N1 ‘bird �u’ and H1N1 ‘swine �u’ — only had one attribute each.

CASES ON THE RISEThe mortality rate has risen with the number of cases worldwide

CASES ON THE RISE

1960195519501918 201019701965 1985 1990 1995 2000 2005

CANADA921 CASES1 DEATH

AZERBAIJAN8 CASES5 DEATHSIRAQ

3 CASES2 DEATHS

EGYPT152 CASES52 DEATHS

CAMBODIA18 CASES16 DEATHS

LAOS2 CASES2 DEATHS

VIETNAM119 CASES59 DEATHS

THAILAND25 CASES17 DEATHS

MYANMAR1 CASES0 DEATHS

PAKISTAN3 CASES1 DEATHS

INDONESIA182 CASES150 DEATHS


BANGLADESH3 CASES0 DEATHS

DJIBOUTI1 CASE0 DEATHS

USA6,764 CASES10 DEATHS

CHILE86 CASES0 DEATHS

COSTA RICA33 CASES1 DEATH

MEXICO4,541 CASES83 DEATHS

UK137 CASES0 DEATHS

AUSTRALIA39 CASES0 DEATHS

SPAIN138 CASES0 DEATHS

INDIA1 CASE0 DEATHS

JAPAN360 CASES0 DEATHS


BRAZIL9 CASES0 DEATHS

GERMANY17 CASES0 DEATHS2004–2006 H5N1 Cases Deaths

1890



335Con�rmed deaths in 214 countries caused by the H1N1 virus up to August 2010.

Con�rmed deaths in 214 countries caused by the H1N1 virus up to August 2010.

18,449

The ‘Spanish Flu’ caused by the H1N1 in�uenza virus, kills approximately 50 million people.

Several new strains of in�uenza emerge throughout the twentieth century, circulating without turning pandemic.

1977H1N1 outbreak of Russian �u in China, similar to strain in circulation prior to 1957, spreads around the world disproportionately infecting people under the age of 23.50M H5N1

1997H7N71995

H7N3H10N7

2004H3N21968

H7N71959

H2N21957

H1N11918

H5N1H9N2

H7N7H3N2

H7N22003 1890

RUSSIAFirst recorded in�uenza pandemic. 1 million die worldwide.

1999: FDA approves the neuraminidase inhibitor oseltamivir.

1997: First case of direct trans-mission from birds to humans.

2009: Cases of Tami�u resistant H1N1 reported.

2010: August,the WHO announces end of H1N1 pandemic.

PANDEMICThe origin of H1N1 has been the subject of many studies — some suggest it may have gone directly from birds to humans, while others suggest it may have involved incubation in an intermediate host, such as the pig.

H9N21998

H1N12009

60% approximate proportion of fatalities for immunocompetent people infected with H5N1.

0.03%approximate proportion of fatalities for immunocompetent people infected with H1N1.

71–167BILLION

The estimated cost (US$) of seasonal �u to the US

economy each year

ASIAThe second in�uenza pandemic strikes in Asia, killing 100,000 people worldwide.

HONG KONGFlu pandemic kills 700,000 people worldwide. Virus still circulates today.

ASIAH5N1, �rst detected in humans in 1997, evolved in Hong Kong chickens, remains largely an avian virus but known to kill people in close contact with birds (see map, left).

MEXICOFirst pandemic of in�uenza in 40 years. Labs in the US con�rm the re-emergence of the dreaded H1N1 strain, the virus behind the 1918 pandemic.

THE SLOW BUT DEADLY SPREAD OF H5N1Between 2003 and 2011, H5N1, commonly known as bird �u, spreads across Asia but remains largely an avian virus.

THE SLOW BUT DEADLY SPREAD OF H5N1

H1N1, SNAPSHOT OF A RAPID PANDEMICIn May 2009, 48 countries reported 13,398 cases of in�uenza H1N1, the �rst in�uenza pandemic in four decades.H1N1, SNAPSHOT OF A RAPID PANDEMIC

300

250

200

150

100

50

02004

Peo

ple

2005 2006

People worldwide die each year from seasonal in�uenza infection.

250,000+2003–2004 Outbreak of the avian in�uenza H7N7 in Holland, among mainly poultry workers.Panama strain of H3N2 mutates into Fijian strain of H3N2.

SO

UR

CES

: WH

O/G

IP A

ND

FAU

CI,

A. S

. MIL

BA

NK

MEM

OR

IAL

LEC

TUR

E 1–2

2 (2006).

OUTLOOKINFLUENZA


INFLUENZAOUTLOOK

What is the basic ecology of influenza? All influenza type A viruses likely originated in wild aquatic birds, and they’ve moved out to other hosts from there. The next layer is the intermediate hosts. Domestic animal species — mainly poultry and pigs — most often transmit the viruses to humans. Once the viruses have found their way from their natural wild hosts into domestic species, they behave differently. Something about these domestic reservoirs seems to increase the virus’s ability to infect people.

What are the biggest unanswered questions when it comes to these interspecies jumps?What, all of a sudden, makes an animal virus a human virus? What specific virologic factors, what host factors, what environmental factors, align to allow that to happen? We’re starting to

gain greater under-standing of the types of viruses circulating in animal populations, but we don’t have a firm grasp on which of them we should be most concerned about.

The wildlife component is probably still one of the biggest missing pieces and it’s one of the hardest to study. At least with the domesticated species you don’t have to chase them around too long to catch them.

Still, we have learned a lot in the past decade. You pick up an old textbook and it’ll say if a virus is to jump from the aquatic bird reservoirs to humans, it probably has to go through an inter-mediate mammalian host, like a pig. The H5N1 strain of the influenza A virus— the bird flu that emerged in Asia in the late 1990s — taught us that viruses can also use domestic poultry as the intermediate host. I can’t think of many viruses that can pass directly between birds and humans besides influenza — certainly none that have the same capacity to cause human disease.

Did bird flu teach us anything else?H5N1 resulted in a lot more resources being put into understanding the animal reservoirs of influ-enza. Before bird flu emerged, what informa-tion we had typically came from fairly confined geographical regions. But in the last ten years we’ve increased that knowledge a lot, and now we understand more about the virus in different parts of the world. Of course, there are still many gaps to fill, but we’re certainly much better off.

Do wild birds experience flu like people do?Not really. Flu-infected wild birds don’t usually appear to be sick. It’s a fairly stable host–micro-organism relationship, most likely evolved over millions of years, and the virus doesn’t cause many symptoms in the birds. When influenza viruses first enter a new host species, they might cause a lot of disease but mortality drops over time. The virus is now simply using wild birds as a means to propagate and maintain itself. Highly

Q & A

The flu catcher Richard Webby studies the ecology of influenza, trying to better understand how certain strains of influenza can leap across the species divide from animals to people. Nature Outlook sat down with him to learn more about his research.

“If there’s anything that keeps many of us up at night, it’s the H5 virus.”

ST. J

UD

E P

HO

TOG

RA

PH

Y


OUTLOOKINFLUENZA

pathogenic forms of avian flu typically arise through evolution in domestic poultry species, and when they transmit back to wild birds, sometimes they’ve become more virulent or severe and do cause mortality in that bird spe-cies. But that’s the exception rather than the rule.

How do wild birds transmit flu amongst themselves?That’s still unclear. One research angle is try-ing to understand hotspots of virus activity. The Delaware Bay region on the US east coast, which my colleagues and I have been studying, is a prime example. Shorebirds stopping in on their northward migration in May to feed on the eggs of spawning horseshoe crabs seem to have a very high prevalence of influenza. There’s something special about that environment — probably that there are huge numbers of birds, many of them naive for influenza, coming together in one location at one time — and the virus spreads very rapidly. A few studies look-ing at those same bird populations further up or down their migratory path have shown much lower influenza burdens. Flu hotspots also appear in Australia, and we have some ongoing studies in Alberta, Canada. But they probably occur in many other parts of the world as well, where waterfowl gather to feed or breed.

How many types of influenza viruses circulate in wildlife?The influenza virus has two major proteins on its surface: hemagglutinin (H) and neuramini-dase (N). Within the aquatic bird populations there are 16 serologically and genetically dis-tinguishable hemagglutinins and 9 different neuraminidases. This is where the flu subtypes get their names, like H5N1. Most — but not all — of the 144 possible combinations have been found circulating in wild birds. When we start to move away from that natural reservoir into domestic animals and humans, the diversity of viruses diminishes.

Are other animals besides wild birds and domesticated poultry and pigs part of the influenza ecosystem?In terms of wildlife, it’s not just birds. We also have good serologic evidence that wild pigs can be infected by flu, and there are very likely other wild hosts that we know nothing about at all. When it comes to domesticated species, horses tend to harbour a horse-specific lineage of influenza virus, though that virus recently transferred to dogs. And the H5N1 bird flu virus and the 2009 H1N1 pandemic virus, which was widely called swine flu, have both jumped into dogs and cats. Exactly what role those animals play in terms of the ecology of

flu is unclear, and I don’t think we have a good handle on what the true prevalence of the virus is within their populations. But certainly they can be

hosts, they can get infected from other species, and they can probably transfer the virus within their own species as well.

Why does the influenza virus seem to be so active in jumping between animals and people right now?That’s the question everyone is asking. Certainly the much larger demand on protein from the global population and rearing animals in larger numbers in smaller areas play a role in the evolution of some of these viruses. And these practices are also bringing domestic species into more contact with wild species.

What questions are you trying to answer with your current research?We’re focused on understanding how viruses behave over time within a given animal popu-lation, whether that be wild bird populations, poultry markets or swine herds. We are going back to some locations repeatedly to under-stand how the viruses themselves and their prevalences change over time. Then, if we do detect interspecies transmissions, we study

these viruses in animal models to understand the virologic basis for the differences in epidemiology that we see on the ground. For instance, we’re undertaking a fairly large study with col-leagues at the National Research Center in

Cairo to study people regularly in contact with poultry to understand the real prevalence of the H5N1 bird flu virus jumping into humans. Typically, we only detect the virus in people with severe disease symptoms, since they’re the ones that come into hospitals. But that’s very likely just the tip of the iceberg — and we have absolutely no idea how big the iceberg is. We’ve also been going back to some US swine farms to study circulating viruses — how they come into these populations, how long they stay, and how they leave.

It’s very practical research. For example, here at this WHO meeting we’re using detection data from over the past year to make sure that we have the right diagnostic capacities to detect the most current strains, and ultimately have the right vaccines to protect the public.

Nearly 3 years on, what’s the lesson from the 2009 H1N1 swine flu pandemic?The most important lingering question is finding out what was special about the H1N1 pandemic virus that allowed it to become estab-lished in humans. There’s a lot of work studying the differences between it and the viruses that were circulating in pigs before the pandemic that never successfully jumped to humans. Had somebody identified the pandemic virus in pigs before the outbreak, nobody would have been

jumping up and down. It had none of the hall-marks that would have had us scuttling to make vaccines against it — no functional PB1-F2 pro-tein, which contributes to a flu virus’s lethality, for instance, and no novel gene segments. That’s obviously not good. We hope this research will help us predict which viruses we need to be most concerned about.

The H5N1 bird flu virus so far hasn’t shown much ability to pass from person to person. What’s the likelihood that the worst-case bird flu pandemic scenario will happen?That’s a tough one. The threat of H5N1 becom-ing a pandemic is likely lower than for any H1, H2 or H3 influenza virus because H5 isn’t as contagious in humans. But if there is a virus we do not want in humans, it is the H5 because it’s so deadly. One of the hallmarks of highly patho-genic influenza viruses is that they accumulate additional amino acids in the hemagglutinin protein, which are thought to allow the virus to generate a systemic infection rather than just a respiratory tract infection. Only H5 and H7 viruses are known to do this naturally.

H5N1 has had 10 years to try, which tells us that it’s not one, two, or even three or four changes that are required for it to become a human pathogen — it’s probably more. On the other hand, there are examples of viruses taking several decades to jump species, despite prolonged contact. So we now have a little bit of confidence that H5N1 can’t become transmis-sible easily, but unfortunately I think it’s still a possibility. If there’s anything that keeps many of us up at night, it’s the H5 virus.

Are there other influenza strains that researchers have their eye on as well?There are a handful of viruses that likely pose the most threat. H5 being one because we know it infects humans and when it does, it causes severe disease. H7 poses a similar concern. Then there are the H2s, which have the capacity to be a successful human pathogen. There was an H2 pandemic in 1957, and the virus disappeared in 1968, so there’s a large percentage of the popula-tion that has no immunity to that virus now.

What do you find most interesting about influenza?It still amazes me how little we know about these viruses. There’s not a whole lot to them—they’re fairly simple viruses — but we really have no good feel at all for what allows them to jump species. It’s an exciting field to be in. ■

Richard Webby, a virologist at St Jude Children’s Research Hospital in Memphis, Tennessee, directs the World Health Organization’s collaborating center for studies on the ecology of influenza in lower animals and birds.

Interview by Rebecca Kessler, a freelance science journalist in Providence, Rhode Island.

“The wildlife component is probably still one of the biggest missing pieces and it’s one of the hardest to study.”

NATURE.COMScientific American looks at where flu is lurking: go.nature.com/lqkgwr


B Y J A N A S C H L Ü T T E R

While the eyes of flu researchers were focused on Asian bird markets in 2009, a pandemic began to emerge

at the opposite end of the world in Mexico and California: H1N1 influenza, a version of the 1918 flu virus circulating in pigs for nearly a cen-tury, had suddenly leaped back into the human population, which now lacked herd immunity. The virus took the world by surprise.

It took months to develop a vaccine matched to H1N1— too long to thwart the pandemic, which peaked in March 2009 and then again in early November, 2009. “We were fortunate that the virus was not really a nasty bastard and did not kill so many people,” says virologist Robert Webster, whose laboratory at St Jude Children’s Research Hospital in Memphis, Tennessee, tracks flu viruses and guides the development of flu vaccines to stop them. “If we had an H5N1

virus that was transmitting from human to human: God help us! We would have run out of antivirals almost overnight .”

The 2009 outbreak reignited the hunt for a universal vaccine. In contrast to the seasonal flu jab or infection, the spreading pandemic virus, because it was so unlike its predecessors, elicited more cross-reactive, broadly neutral-izing antibodies than usual — and now sci-entists had the technology to find them. One patient even managed to make an antibody that could inactivate all 16 subtypes of the influ-enza type A virus — the so-called FI6 antibody. (Type A influenza is the most virulent kind, and is responsible for virtually all major human

flu outbreaks.)It was, however, a rare

find. Antonio Lanza-vecchia, an immunolo-gist at the Institute for Research in Biomedicine

in Bellinzona, Switzerland, and his colleagues had to screen 104,000 white blood cells (B cells) collected from eight donors until they found one cell that produced the FI6 antibody. FI6 binds to hemagglutinins representative of all the 16 sub-types of type A influenza, including the H1N1 swine flu and avian influenza H5N1, as well as to the more common H3N2 strains. In theory, this FI6 antibody provides a blueprint for design-ing a one-size-fits-all shot. “People have been searching for it forever,” says Webster. “Now, it looks as if the Holy Grail for flu really might be achievable.”

STEADY TARGETThe flu virus can skillfully play a chess-like game with the human immune system. Its surface is packed with varieties of two main proteins: neuraminidase and hemagglutinin. Hemagglu-tinin in particular plays two vital roles: its stem contains the machinery that allows the virus

NATURE.COMFor some of the latest research on influenza vaccines: go.nature.com/tezyqI

P R E V E N T I O N

Vaccine for all seasonsAs researchers map the stable parts of the proteins that stud the surface of influenza, the decades-long quest for a universal flu vaccine is showing signs of progress.

DU

NG

VO

TR

UN

G/L

OO

K A

T S

CIE

NC

ES/S

CIE

NC

E P

HO

TO L

IBR

AR

Y

Scientists developing a human vaccine against the avian influenza.

INFLUENZAOUTLOOK


to fuse with host cells; while flashy loops on its globular head act as decoys. To an antibody, each virus appears to be covered with some-thing resembling a dense tropical forest. The freely moving chains of amino acids through the treetops offer alluring binding sites. But these chains mutate and alter their shapes rap-idly. The antibody that the host’s immune system tailors to fit into the branches of one virus will be limited to that virus, and not work against any new influenza.

Because existing flu vaccines mimic a natural infection, they fall into the same trap: even if our body manages to make a few broadly neutral-izing antibodies they are vastly outnumbered by strain-specific ones that bind to the globu-

lar head — and most people have to get a flu shot each year to regain immunity.

To sidestep the trap, the immune system would need to refocus its attack, perhaps onto the bases of the trees: a highly stable region of the hemagglutinin

stem that tends to stay the same even after the virus mutates. FI6 and most other cross-reactive antibodies push their way in and go for these hard-to-find targets. “This is the first time that an antibody can neutralize every single influ-enza virus,” says Rino Rappuoli, head of vaccines

research at Swiss pharmaceutical company Novartis. But while calling FI6 a “milestone”, Rappuoli is cautious about seeing it as the key to making a universal vaccine. “That’s tough,” he says. “We don’t know how to do it yet, but it is a dream that can come true .”

ROOT OF THE PROBLEMThere are two ways to use broadly reactive antibodies: as a passive immunization to treat severely ill patients who did not respond to anti-virals, or as a template to develop a prophylactic vaccine for everyone. The latter is the more diffi-cult task — it requires sophisticated engineering of the antibody to derive a protein that not only fits the antibody’s binding site but also manages to coax the human immune system into produc-ing enough ammunition to prevent infection. Many researchers favour starting with the first option. “It’s effective, it’s extremely broad, and it’s human — there should be no problem injecting antibodies into humans,” says Lanzavecchia. “To be frank, our first goal is to use the antibody as a therapeutic.”

Gary Nabel, director of the Vaccine Research Center at the National Institute of Allergy and Infectious Diseases (NIAID) in Bethesda, Maryland, agrees on the difficulty in making a universal preventative flu vaccine. “We’ve been doing similar work with HIV,” he says. Creating a prophylactic vaccine “sounds simple, but it takes time.” The challenge, then, is in creating a vac-cine that will induce this specific yet unnatural

response. Even if one presents a specific site of a protein to the immune system, there are many ways in which antibodies can bind to it. “That’s clearly an important problem to solve for any infectious disease, and flu can be our poster child,” says Nabel. “When we understand the basics behind this biology we will be in the driv-er’s seat to design a lot of vaccines that could be very, very promising.”

While eliciting ‘anti-stem’ antibodies in humans is difficult, it is not impossible. Nabel and his team were the first to apply a so-called prime-boost approach in humans to achieve this goal. This technique entails giving a combina-tion of two vaccines administered in sequence to induce the strongest immune response pos-sible against a broad array of influenza strains, including H5N1, H5N2 and H9N2. The prim-ing shot was a DNA vaccine: small, circular pieces of bacterial DNA genetically engineered to code for a specific protein that targets an avian flu hemagglutinin. Then, 24-weeks later, volunteers were given a booster vaccine made of whole inactivated H5N1 virus. This regimen enhanced the immune response against avian flu. In addition, three individuals were able to make broadly neutralizing antibodies directed at the hemagglutinin stem, a result that Nabel says is a “proof of concept” for a universal influ-enza vaccine.

The mice in microbiologist Peter Palese’s lab at Mount Sinai School of Medicine in New York are already protected against various flu strains

“When we understand the basics behind this biology we will be in the driver’s seat to design a lot of vaccines.”

TARGETING A SHAPE-SHIFTING VIRUSA universal �u vaccine needs to zero in on the parts of the virus that do not change as the virus evolves.

Nucleoprotein (RNA)

Lipid envelope

Protein shell

HemagglutininNeuraminidase

M1

M2

M2

Conserved stem region (yellow)

Receptor binding pocket (green)

Two parts of the hemagglutinin protein remain unchanged as the virus mutates and are thus good targets for vaccines: a conserved region on the stem and more recently discovered the receptor binding pocket on the head.

OUTLOOKINFLUENZA


by experimental, prophylactic flu vaccines. Palese has shown that two truncated versions of hemagglutinin, one with its globular head chopped off and the other only partly expressed, work better than full-length hemagglutinin, but they’re not perfect: they offer good protection against H3 strains of influenza virus, but they are weaker against H1. Moreover, Palese echoes the familiar caution from anyone working with animal models: “It’s promising,” he says, “but mice are not men”.

Nonetheless, scientists and vaccine com-panies around the world are hard at work. Advances in high-throughput technology allow them to screen individual B cells from healthy volunteers or patients and search for rare anti-bodies. In 2008, a group based at the Dutch company Crucell was the first to find broadly neutralizing antibodies against influenza in humans. Apart from exploring therapeutic approaches, they are also trying to vaccinate rodents with conserved viral structures.

Normally, the immune system overlooks the virus’s stem (because it’s a target that’s hard to see and to get to) and produces instead many antibodies that target the viral protein’s head. Most broadly neutralizing antibodies, on the other hand, bind to conserved domains on the stem. Crucell is trying to recreate these conserved areas and expose them to the immune system to induce the production of broadly neutralizing antibodies. It’s crucial to use the right binding domain, because other-wise the antibody might be just as strain-specific as anything that binds to the head. “We know that we can rebuild the stem in a way that it looks like the binding domain,” explains Cru-cell immunologist Katarina Radosevic. But the question is how to rebuild these structures in a stable form so that they can induce an immune response in animals. “It’s a complex task,” Rado-sevic says. “You can’t just chop off a part of a molecule and expect the rest to fold and func-tion the same way .”

IN THE POCKETRecently, another target for a universal vaccine was discovered, one that is conserved but does not hide away in the protein folds. At the crown of hemagglutinin’s globular head is the recep-tor-binding pocket, which allows the virus to infect host cells. With minor modifications, this pocket has the same structure on all the 16 sub-types of hemagglutinin. But the pocket is tiny — much smaller than a typical antibody. Nobody expected it to be a good vaccine target, and yet this appears to be the case.

To get snapshots of how the immune sys-tem reacts to the seasonal vaccine, research-ers from Duke University in Durham, North Carolina, screened antibodies from volunteers several times after vaccination. One of these antibodies, dubbed CH65, seemed to be a little more effective than others. The Duke team sent the information to Stephen Harrison, a struc-tural biologist at Harvard Medical School and

Children’s Hospital in Boston, Massachusetts, for analysis. “We were surprised to find that it docks to the receptor-binding pocket,” says Har-rison. “About half of the contacts are with amino acid residues in the pocket itself. The degree and intimacy of mimicry is striking.”

What’s more, point mutations in the area sur-rounding the pocket did not hinder the antibody from docking. When the researchers tested the antibody against 36 strains of H1 that had cir-culated within the last three decades, it blocked 30 of them. “Over 30 years of evolution of the virus, there were hardly any mutations that had a strong effect on binding,” Harrison says. Another possibility, he adds, is that since 1977, the anti-body arose so rarely that the virus did not have to escape its attack.

Harrison’s research interest is not so much the creation of a universal flu vaccine as in advanc-ing the basic science of affinity maturation: When exposed to the same antigens time and again, the immune system produces antibodies with slight mutations that will bind much more effectively to the invader. But why do some small changes make such a difference to the immune

response — and how can one induce them?

While some scien-tists are busy mapping the conserved regions on the hemagglutinin molecule, others are still investigating exist-ing vaccine targets that had for many years

been the only ones available. These include the ion channel M2, the matrix protein 1 (M1) and the nucleoprotein that the virus needs to keep its genome stable. “We have known about these conserved proteins for at least 20 years,” says Rappuoli. While acknowledging that “theoreti-cally” one might prove to be a good target for a universal vaccine, he says that “there hasn’t been much progress” on these fronts.

Many companies are still pursuing these tar-gets. VaxInnate, a biotech company in Cran-bury, New Jersey, has engineered a hybrid molecule that consists of four copies of M2e — part of the ion channel M2 that sits on the virus surface — fused into the bacterial pro-tein flagellin. VaxInnate recently reported that its vaccine, called Vax102, safely produces an immune response in humans that should be protective against all strains of influenza A. DynaVax, a biotech company in Berkeley, California, has fused M2e and nucleoprotein to create a vaccine candidate called N895; the goal is to encourage antibodies against M2e as well as T cells against the nucleoprotein. Acambis, the UK biotech company based in Cambridge, now owned by Sanofi Pasteur of Lyon, France, evaluated its own M2e vaccine in a phase I study in 2008 and found that it was immuno-genic against influenza A, and well-tolerated. The company also conducted another study that it says showed the vaccine protecting about

70% of ferrets against bird flu. Further development has stagnated. Sanofi

Pasteur is hesitant about moving its M2e vac-cine to phase II trials. “M2e alone is unlikely to be better then the current seasonal vaccines,” says Jeffrey Almond, Sanofi’s vice president for discovery research and external R&D. “So we stepped back a little and thought about what we could add to M2e. There is no reason why a uni-versal vaccine shouldn’t have multiple compo-nents.” Meanwhile, Merck, based in New Jersey, has halted its own trials.

A DIFFERENT TACKAnother path toward a universal vaccine involves not trying to elicit antibodies, but rather boosting the body’s T-cell response to infection. While antibodies prevent the virus from infecting host cells, T cells help to clear the virus from the body by killing flu-infected cells. That’s the approach being pursued by vaccin- ologist Sarah Gilbert at the Jenner Institute — a vaccine research organization based at Oxford University in the UK. Gilbert uses an attenu-ated poxvirus called MVA, which presents the flu nucleoprotein and M1 to the immune system. As a result, Gilbert reports, “we saw a very large peak in T-cell response — everybody improved.” This T-cell-mediated immunity might provide another line of defense against flu, possibly in combination with a protein vac-cine that targets the stem of the hemagglutinin. Gilbert invited 11 vaccinated volunteers and 11 non-vaccinated volunteers to a quarantine facil-ity. To test their vaccine, the researchers dripped H3N2 into the noses of their subjects and moni-tored their flu symptoms as well as their T-cell response.

Combining Gilbert’s vaccine with the sea-sonal flu shot yields a preparation that boosts not only T cells but also the antibody response. Following vaccination with the seasonal flu shot alone, young people respond well by producing antibodies, whereas the same vaccine is less effective in older people. “Using both vaccines together might be extremely useful for vaccinating the elderly,” Gilbert says. Usually, the seasonal flu jabs don’t work as well for older people, because their immune response diminishes quickly, and it becomes more and more difficult to make new antibodies as we age. “We boost what people already have,” Gilbert explains. “We are not trying to prime new responses.”

Even a potent universal protein vaccine modelled after the perfect antibody would not change the fact that different age groups react differently to a flu jab. Thus a once-in-a-lifetime shot that protect against all strains might never be achieved for everyone: “The question is what universal means,” says Rappuoli. “Should it cover all the pandemic strains and the seasonal strains? That’s almost impossible. But it’s OK to have a dream to move forward.” ■

Jana Schlütter is a science writer in Berlin.

“We boost what people already have.We are not trying to prime new responses.”

INFLUENZAOUTLOOK


B Y R O X A N N E P A L M E R

The influenza virus has adapted to render the first generation of flu antivirals impotent, and there are signs of resistance

to newer remedies. In late August 2011, the International Society for Infectious Diseases, based in Boston, Massachusetts, reported that 25 influenza patients in eastern Australia were infected with a strain of the notorious swine flu — influenza A (H1N1) — resistant to the widely used drug oseltamivir (Tamiflu), made by Roche, headquartered in Basel, Switzerland. The Australian case is the largest cohort of oseltamivir-resistant flu yet reported as scientists and pharmaceutical companies explore ways to outsmart the virus.

The good news is that resistance to antiviral drugs is not widespread in influenza strains for the upcoming 2011–2012 season — at least not

yet. “Of the circulating strains of influenza in humans, we do not see resistance to oseltami-vir,” says Charles Penn, an antiviral drug expert with the global influenza program of the World Health Organization (WHO) in Geneva, Switzerland. Although resistance to oseltami-vir was reported in as many as 1% of H1N1 flu samples collected during the winter of 2010–2011, says Penn, that level doesn’t pose enough of a risk to warrant the WHO to change its treat-ment recommendations.

That doesn’t mean that drug resistance is not a problem. Oseltamivir has successfully treated millions of patients since 1999, but mutations that confer resistance were described as early as 1998 — and given the ever-changing nature of flu, it can be hard to predict the trajectory of oseltamivir resistance. “In a span of two to three years we’ve seen a blossoming of resistance. At any time, the whole game can change,” says

Zachary Taylor, an infectious disease fellow at the Kaiser Permanente Fontana Medical Center in Sacramento, California. In part to safeguard against the possibility of such game-changing developments, drug developers are slowly filling the pipeline with alternative therapies (see ‘Drugs to treat influenza infection’). Each drug come with side effects, which make them only worthwhile for those whom the flu could be potentially lethal — the elderly and the immu-nocompromised.

Given the wily history of the influenza virus, any sudden appearance of drug resistance is certain to concern public health officials. The first antiviral drugs to combat the disease — the adamantanes, which target the M2 channel protein to block virus entry into host cells — are now essentially useless. The US Centers for Dis-ease Control and Prevention (CDC) found that 100 % of seasonal H3N2 flu in the 2009–2010 season and 99.8% of 2009 pandemic H1N1 flu were resistant to adamantanes.

Oseltamivir belongs to a class of drugs called neuraminidase inhibitors. These agents block the active site of a viral protein called neurami-nidase (N), thereby arresting the influenza virus’ ability to leave the host cell after it proliferates. The most common way for the influenza virus to evade oseltamivir is via the H275Y muta-tion (also known as H274Y) of neuraminidase, which replaces a single histidine amino acid with a tyrosine. This alteration interferes with the drug’s ability to bind to the protein — a prob-lem acknowledged by the maker of oseltamivir. “There remains a medical need and room for additional treatment options, especially for the management of severe infections and for improved pandemic preparedness,” says Klaus Klumpp, Roche’s top virologist. Klumpp says the Roche is supporting research into new therapies targeting viral replication as well as other mecha-nisms, but notes that these efforts are preclinical.

Fortunately, viruses with the H275Y mutation are still susceptible to a different neuraminidase inhibitor: zanamivir (marketed by UK-based GlaxoSmithKline (GSK) as Relenza). Zanami-vir, the first neuraminidase inhibitor discovered, is generally administered by oral inhalation; its side effects include dizziness and nose irritation.

The WHO recommends using zanamivir to treat patients afflicted with oseltamivir-resistant flu strains, and it’s already a viable alternative to oseltamivir. Zanamivir was actually the first neuraminidase inhibitor on the market, but its cousin oseltamivir was approved shortly after-wards and captured a larger market share. In January 2011, GSK began a double-blind study comparing intravenous zanamivir with oral oseltamivir. The study, which has enrolled 462 adolescents and adults, is due to be completed in September 2013.

Whi le zanamivir remains an arrow in the quiver to be used in case of an oseltamivir-resistant strain, another drug

D R U G S

Lines of defenceAntiviral treatments are a critical component of an effective healthcare response to influenza, but drug resistance to the treatment-of-choice has public health officials searching for other options.

TAKING ON A TOUGH VIRUSFlu drugs tend to stop working after the virus mutates enough to become resistant to them, and the arms race continues apace.

Adomantanestarget surfaceproteins

Fludase targetshost cell receptorsto prevent initialviral attachment

Neuraminidaseinhibitors stopbudding virusesfrom escaping

Nucleoprotein (RNA)

Lipid envelope

Capsid

NeuraminidaseHost cell

Flu virus

Neuraminidase

Hermagglutinin

Receptor with sialic acid

Fludase targetshost cell receptorsto prevent initialviral attachment

Neuraminidase inhibitors block neuraminidase, preventing the virus from leaving the host cell.

Host cell

In�uenzavirus

Neuraminidase

Hemagglutinin

Receptor

Adamantanes targetthe proteins that block viral entry into host cells; the virus is almost completely resistant tothese drugs.

mpletely resistant tothese drugs.

NATURE.COMFor some of the latest research into drugs for influenza: go.nature.com/wkjvpp

OUTLOOKINFLUENZA

S 1 0 | N A T U R E | V O L 4 8 0 | 8 D E C E M B E R 2 0 1 1

targeting viral neuraminidase is already avail-able in Asia and might be useful for patients who can’t tolerate the other neuraminidase inhibitors. This third neuraminidase inhibitor — peramivir, developed by BioCryst Pharmaceuticals of Dur-ham, North Carolina — is available in Japan (as Rapiacta) and South Korea (as Peramiflu) but is still undergoing clinical trials in the United States. Its known side effects are similar to those of oseltamivir — diarrhea, nausea and vomit-ing. As an intravenous drug, peramivir can be administered to very sick or hospitalized patients who wouldn’t be able to swallow oseltamivir or inhale zanamivir. In October 2009, the US Food and Drug Administration issued an emergency use authorization (EUA) for peramivir for peo-ple unable to take oseltamivir or zanamivir. The peramivir EUA expired in June 2010, and although the approval process has been fast-tracked in the United States, general approval of the drug is still several years away; the phase III clinical trial is not expected to be completed until May 2013.

THE VIRTUE OF VARIETYOther pharmaceutical companies, cognizant of fears of resistance to neuraminidase inhibitors, are exploring a variety of ways to attack flu. One drug making its way along the pipeline is DAS 181, an oral antiviral drug developed by NexBio in San Diego, California, under the trade name Fludase, which is now in phase II clinical trials. The drug is a fusion protein, which is created by combining genetic sequences that encode two or more other proteins into a single protein. It aims to prevent infection by inactivating viral receptors on cells in a patient’s respiratory tract, making it harder for the virus to latch onto host cells. A study by NexBio published in 2009 reports that Fludase was effective against 11 oseltamivir-resistant strains of the H1N1 virus — including a couple of variations that showed signs of reduced sensitivity to zanamivir. In Sep-tember 2011, NexBio presented results from the phase II trial at the Interscience Conference on

Antimicrobial Agents and Chemotherapy in Chicago, Illinois, that showed Fludase had cut the level of virus in patients’ blood after just the second day of treatment, much as zanamivir and oseltamivir do. However, viral load has not been found to correlate with clinical severity of a flu patient’s illness, and its use as a surrogate end-point is controversial.

Any single flu drug has drawbacks. A mix-ture of treatments might be a better strategy to combat the virus — and in some cases, overcome resistance. Adamas Pharmaceuticals, based in Emeryville, California, for example, has been developing a three-drug combination treat-ment strategy of oseltamivir, amantadine — an adamantane drug thought to be obsolete — and ribavirin (another antiviral drug commonly used to treat hepatitis C infection). In cell cul-tures, the three drugs worked better than any pair, and were even able to inhibit viral activity in strains of influenza resistant to oseltamivir.

Nobody knows for sure how the three drugs work together. Adamas researcher Jack Nguyen speculates that each drug can take out the viruses that slipped past the others — a one-two-three punch combo. “After each stage you have fewer and fewer viruses making it through the cycle,” Nguyen says. It may also be possible that oseltamivir’s binding to neuraminidase causes other proteins on the virus’ surface to change shape and become more susceptible to the other drugs, Nguyen adds.

In April 2010, Adamas released the results of a pilot study of its three-drug combination in seven patients with weakened immune sys-tems. Five of the six patients that received the triple treatment responded by day 10 of therapy. The lone patient that was given only oseltamivir did not respond after 20 days. “This pilot study was an important first step in validating that the combination of three antivirals can provide a virologic and clinical benefit to patients at risk for complications of influenza,” says Janet Englund, a clinical investigator who led the trial at Seattle Children’s Hospital in Washington. The

US National Institute of Allergy and Infectious Diseases in Bethesda, Maryland is recruiting 720 people for a phase II trial testing the efficacy of Adamas’ triple combination therapy versus oseltamivir in patients with severe medical conditions, such as heart or lung disease, which make them more susceptible to serious compli-cations from a bout of the flu.

More options are being explored. Recently researchers at the Scripps Research Institute in La Jolla, California, infected mice with a fatal dose of influenza, then gave some of the animals compounds that inhibited their ability to pro-duce cytokines — cell signalling molecules that summon T cells and other immune system effec-tors to the site of infection. Mice that received these cytokine blockers plus the antiviral drug oseltamivir had a much higher survival rate (96%) than the other experimental groups that received only the cytokine-blocking compound (82% survival); oseltamivir alone (50%); or no treatment at all (21%). This suggests that much of the damage caused by flu is not due to the virus itself, but to an overenthusiastic response from a person’s immune system. ‘Cytokine storms’, trig-gered by an overproduction of immune signal-ing molecules, can cause significant damage to tissue. Indeed, the phenomenon is thought to be responsible for a large proportion of deaths during the 1918 influenza pandemic.

Still, targeting the patient instead of the virus has its risks. “When you start tweak-ing the immune system, you have to wonder about whether you’re under correcting or over-correcting the immune response,” explains Dean Blumberg, head of paediatric infectious diseases at the University of California, Davis.

Overcorrect, and you exacerbate the prob-lem of cytokine storms; under correct, and you run the risk of leaving the body defenceless against pathogens besides influenza.

Blumberg says some patients shrug at the idea of using antiviral

drugs at all, figuring that once contracted, the disease will simply run its course. In most cases, it will. In fact, neuraminidase inhibitors usually only shorten the duration of illness by about one day, and that’s if the drug is taken within 48 hours of the first sign of flu symptoms. Yet Blumberg says that personally, “as a doctor and as a parent” he usually recommends a course of antivirals, to cut down on the length of illness.

With no signs of an avian or swine flu pan-demic looming, the need for a wide spectrum of antiviral treatments might seem over cautious. But when warring against viruses, carefully laid preparations made in peacetime may serve well in the thick of the fight. ■

Roxanne Palmer is a science writer in Brooklyn, New York.

DRUGS TO TREAT INFLUENZA INFECTION

Clinical name Type of drug Brand name and manufacturer

Status

Oseltamivir (oral) Neuraminidase inhibitor

Tamiflu (Roche) Commercially available in most countries

Zanamivir (inhaled) Neuraminidase inhibitor

Relenza (GSK) Commercially available in most countries

Peramivir (intravenous) Neuraminidase inhibitor

Rapiacta, in Japan); Peramiflu in South Korea (BioCryst)

Available in Japan and South Korea; in Phase III testing in U.S.

DAS181 (inhaled) Fusion protein Fludase (NexBio) Phase II testing in U.S.

ADS-8902 (adamantine, ribavarin, oseltamivir) (oral)

Triple combination N/A (Adamas) Phase II testing in U.S.

Amantadine (oral) Adamantane Symmetrel (Endo Pharmaceuticals)

Commercially available; not recommended for influenza due to resistance

Rimantidine (oral) Adamantane Flumadine (Forest Pharmaceuticals)

Commercially available; not recommended for influenza due to resistance

“In a span of two to three years we’ve seen a blossoming of resistance. At any time, the whole game can change.”

INFLUENZAOUTLOOK

8 D E C E M B E R 2 0 1 1 | V O L 4 8 0 | N A T U R E | S 1 1

B Y L A U R A V A R G A S P A R A D A

Adela Gutiérrez, a door-to-door census-taker for the Mexican government, started feeling unwell with fever, headache and

a sore throat, in early April 2009. Gutiérrez, a 39-year-old mother of three, lived in the southern state of Oaxaca. By the time she arrived at the state hospital, she was unable to breathe and was diag-nosed with severe atypical pneumonia that might have been complicated by her diabetes. Within four days, Gutiérrez was dead.

Lab tests at the Oaxaca state hospital could not confirm the cause of death, so throat swabs were sent to the National Microbiology Labora-tory in Winnipeg, Canada, and to the US Cent-ers for Disease Control and Prevention (CDC) in Atlanta, Georgia. Ten days later, Cruz would become one of the first confirmed fatalities of a new influenza A (H1N1) virus, which originated in pigs. By then, nearly half of all the people who were in contact with Cruz in the hospital had developed respiratory symptoms and one preg-nant nurse had fallen ill. Two months later, the World Health Organization (WHO) declared a global health emergency.

Early observations showed that this strain of flu disproportionately afflicted young people. A study of patients in Mexico found that the

median age of fatalities was 39 years (ref. 1). Epi-demiologists feared this outbreak could resem-ble the infamous 1918 influenza pandemic, also an H1N1 strain, which killed an estimated 50 million people (around 3% of the then global population) — half of whom were healthy adults between the ages of 20 years and 40 years old. The US President’s Council of Advisors on Science and Technology calculated a possible scenario of 30,000–90,000 deaths in the United States alone. Fortunately, the transmission rate and virulence of the virus could not produce the pandemic that was initially dreaded.

GAINING MORE EXPERIENCENevertheless, the 2009 H1N1 pandemic proved to be a thorough test of the 2005 International Health Regulations (IHR), designed to be the world’s first line of defence during public-health emergencies. (IHR is a legal agreement that dates back to 1969 and is binding by 194 party states, including all WHO members.) The 2009 out-break was the first major assessment of the new IHR and raised critical questions about why it was so difficult to determine the severity of the threat, how effective the preparations were, and whether the crisis was properly managed.

In late 2009, director-general Margaret Chan of the WHO advised a review of the

pandemic to learn from any lessons and eval-uate whether the IHR fulfilled its purpose. The WHO assigned this task to a newly cre-ated IHR review committee, comprised of 25 international experts from diverse scientific fields with experience in public health. The committee presented its final report in May 2011, during the 64th World Health Assembly in Geneva, Switzerland.

The emergence of human infections of avian influenza A (H5N1) in 1997 and the corona- virus causing severe acute respiratory syndrome (SARS) in 2002 expedited international prepara-tions for a pandemic. The IHR was updated in 2005 to respond to these new threats and entered into force worldwide in 2007, and are expected to be fully operational by 2012.

A major point of concern raised during the outbreaks of bird flu and SARS was the perfor-mance of disease warning systems in place. To assure a two-way channel of communication between the WHO and the party states, the IHR required each country to establish, by 2012, what it calls a focal point — offices that liaise with the WHO at all times and deploy resources for dis-ease surveillance, early warning systems and the response to a disease outbreak.

The IHR review committee reported the IHR was still “not yet fully operational” world-wide. “There are still many states that have not developed the plans and infrastructure speci-fied in the IHR,” says José Ignacio Santos, who heads the infectious disease unit of the school of medicine at the National Autonomous Uni-versity of Mexico in Mexico City and a member of the IHR review committee. Although three quarters of the 194 party states had plans to co- ordinate national efforts in the case of a pan-demic influenza outbreak, that statistic didn’t tell the whole story. Only a dozen of the 128 countries that answered a WHO question-naire had taken the steps required to put such a plan in practice, including: enacting legislation; allocating funding; putting enough people in place for detection and alert operations; and establishing procedures for surveillance, event detection, risk assessment and information provision2. In addition, many National IHR focal points are unable to communicate infor-mation about public health emergencies to the WHO in a timely fashion.

According to the IHR review committee “the IHR played a central role in the global response to the pandemic.” In Mexico, the IHR probably saved lives thanks to the progress made towards establishing plans and stocking antiviral medica-tions. But what would happen if a new virus were to appear in countries that are not yet prepared, such as some in Africa and Asia?

The WHO did not escape blame in the IHR review committee’s assessment. For instance, the WHO was inconsistent in its defi-nition of ‘pandemic’. Some WHO documents described pandemics in terms of the number of deaths and illness caused by the disease, whereas the latest WHO definition is based on the degree

P U B L I C H E A LT H

Life lessonsThe 2009 pandemic arrived suddenly and lethally, exposing our plans to reality. Are we now better prepared?

THE

IMAG

E W

OR

KS

/ T

OP

FOTO

OUTLOOKINFLUENZA


of geographical spread. The WHO also failed to come up with consistent measures to assess the severity of an outbreak. The review committee stressed the need for the WHO to strengthen its capacity to mount a sustained response, and to improve its communication policies — in particular by issuing timely guidance in the organization’s six official languages and hold-ing routine press conferences. The review com-mittee’s report concluded with a stark warning: “The world is ill-prepared to respond to any similarly global, sustained and threatening public-health emergency.”

SHARPENING PREDICTIONSDespite recent efforts to improve preparedness for a global influenza pandemic, there are still many things we don’t understand about the virus. “Influenza is the name of one disease but is caused by a family of viruses,” says Sylvie Briand, head of the global influenza programme at the WHO in Geneva, Switzerland. “We don’t know very much about how these viruses circulate at the global level.” It is still an enigma as to when, and to what extent, influenza viruses change their genetic make-up, and how they spread around the globe.

A reminder of how influenza can still surprise us came on 2 September, 2011, when a new swine influ-enza virus A (H3N2) was isolated from two children in the United States — one in Indiana and one in Pennsylvania3. Human H3 and H2 influenza strains were transmitted to pigs in the 1990s. Those viruses evolved in swine and now differ from the seasonal influenza that circulates in humans. People born before the 1990s “still have antibodies to the swine viruses because their immune system was exposed to the original viruses,” says Nancy Cox, director of the influenza division at the CDC. “But young kids are very susceptible.” The CDC has identified a vaccine candidate against this virus — just in case, says Cox.

In an ominous report published in 2011, researchers studied H2N2 viruses that have not circulated in human populations for several decades, but are still common infections of birds and swine4. The study found that people under 50 years of age have little or no immunity to this strain of the virus, reigniting fear that an H2N2 pandemic could, like H1N1, jump from animals to humans.

The 2009 swine flu was a reassortant virus that crossed from pigs to humans (a reassortant virus contains genetic material from two or more related viruses). Cox believes that this exchange between human and porcine viruses is more common than had been previously thought: “This is probably going on in a lot of other coun-tries. In Switzerland they have detected swine infections in humans,” Cox says. In a retrospec-tive study published in 2007, researchers found evidence of 50 apparent cases of influenza that were transmitted from pigs to humans. Most of the cases were reported in the United States — probably, Cox says, because the US has a very good surveillance system, but there were also six cases in the Czech Republic, four in the Nether-lands, three in Russia, three in Switzerland, and one case each in Canada and Hong Kong5.

Prior to 2009, three global networks were in place for the early detection and/or surveillance of influenza: the 2005 IHR, the WHO’s Global Influenza Surveillance Network (GISN) and sys-tematic event detection (that monitors informa-tion published not only formally in journals but also in newspapers and online forums). In March 2009, just weeks before the swine flu outbreak, more than half of WHO member states (104 of 193 countries) had no or very limited seasonal influenza surveillance capacity, according to a 2011 article co-authored by Briand6. (Both the United States and Mexico had good surveillance

systems, which is why the new virus was identi-fied in those countries early on.) The same report, echoing some of the themes articulated by the IHR review committee report, identified other deficiencies that were revealed during the H1N1 pandemic. Among them were a lack of standards for reporting illness, risk factors and mortality. As a result, different countries used different crite-ria to define influenza-like illness and collected information on different risk factors. This incon-sistency makes it difficult to do the sort of com-parisons that would help public health officials assess how well they are doing relative to other places in the world. After the swine flu pandemic, the WHO worked to better define standards and develop a web-based platform called FLuID (Flu Informed Decisions) for direct reporting of epi-demiological data, such as intensity of influenza transmission, number of hospitalizations, and risk factors for severe disease. FLuID is similar to the WHO’s FluNet surveillance system, which collects information concerning the viral type and subtype.

Scientists are confident that a good surveillance system, along with the GISN monitoring antibod-ies against the virus, will help detect new viruses before they become a threat to global health. And if this is the case, says Briand, “we have a protocol for rapid containment [if] we detect an outbreak early enough to contain it at its source.”

NEW DIRECTIONSOne of the main challenges during an outbreak is to assess the severity of the pandemic disease. Even with today’s technology, there are a few critical days of uncertainty after the outbreak is identified, but before enough epidemiological and clinical data are gathered to allow public health officials to plot the most effective response.

“In the 2009 pandemic, some decisions had

THE NEXT PANDEMIC

1900 1910 201020001920 1930

1957 1957

1968

2009

1918

1940 1950 1960 1970 1980 1990

Experts recommend the following actions:

• Strengthen surveillance and response in much of the world, especially in under-resourced countries.

• Define pandemic, develop measures to estimate severity.

• Standardize reporting of influenza data.

• Focus on new vaccine production technologies, new antivirals, and extending the shelf lives of medications.

• Vaccine distribution worldwide needs to be revised and strengthened.

1918The Spanish flu (H1N1) strikes killing an estimated 50 million people.

Camp Funston, Kansas, site of the first recorded cases of Spanish flu (left), nurse draws water for flu patients (top).

Noted flu researchers. D. E. Rogers and E. D. Kilbourne test for the influenza virus (top), a New York hospital at the height of the epidemic (right).

A sign in Iowa casts flu as a foreign invader (left).

A pig sty is disinfected in Indonesia (right).

20091968700,000 people die from H3N2 infection (descended from H2N2); originates in Hong Kong.

1957H2N2 outbreak that began in Asia kills 100,000 people.

The return of H1N1 — swine flu — surfaces in Mexico and spreads rapidly — including to healthy people, but with a low fatality rate.

FRO

M L

EFT:

NAT

ION

AL

MU

SEU

M O

F H

EALT

H A

ND

MED

ICIN

E/SC

IEN

CE

PH

OTO

LIB

RAR

Y; N

ATIO

NAL

AR

CH

IVES

/SC

IEN

CE

PH

OTO

LIB

RAR

Y; T

IME,

BET

TMAN

N/

CO

RB

IS &

LIF

E PIC

TUR

ES/G

ETTY

IMAG

ES, B

ETTM

AN

N/C

OR

BIS

, BET

TMAN

N/C

OR

BIS

, CD

C/S

CIE

NC

E PH

OTO

LIB

RAR

Y, K

PA/Z

UM

A/R

EX F

EATU

RES

.

INFLUENZAOUTLOOK


to be taken before we had a clear picture,” Briand says. Drawing on that experience, she says, health officials are trying to improve their ability to assess an outbreak at its earliest stages. “If we isolate the virus,” she explains, “we can determine how dif-ferent it is from the current circulating viruses and if it is susceptible to available antivirals.” Testing blood samples for antibodies to influenza viruses will allow researchers to determine how many people have been exposed to the virus and how prevalence fluctuates over time.

The absence of important data can be over-come. During the 2009 pandemic, determining the fatality rate was sometimes difficult because information about the number of persons infected and the number of deaths was incomplete. Exacer-bating the problem, the WHO, as well as national authorities, overwhelmed public health officers with requests for specific data, such as the number of cases. In countries with limited epidemiological and laboratory infrastructure, the need to deliver so many statistics led to an unfortunate diversion of personnel away from patient care. And in coun-tries during the early stages of the pandemic, this information proved to be less useful for assessing severity than rates of hospitalization and complica-tions, and actual number of deaths.

As it turned out, however, other parameters proved to be good proxies to gauge the severity of disease. Measures included the proportion of cases that required hospitalization for treatment or that required intensive care and mechanical ventilation. Other useful data were the proportion of previously healthy individuals without under lying risk factors that developed severe disease. And as always, the severity of a pandemic will depend on conditions inside a country, such as access to health services, general health of the population, and various social and behavioral factors.

As the influenza virus continues to circulate

in human and animal populations across the globe, it is crucial to identify changes in the viral genome — especially those mutations that con-fer resistance to antiviral drugs. In 2009, Briand says, “we were fortunate because the virus was sensitive to neuraminidase inhibitors that we had in our stockpile.” Neuraminidase is a viral sur-face protein that allows the virus to spread from infected cells to other cells (see ‘Lines of defence’, page S9). But any strategic stockpiling of anti-flu drugs needs to account for the limited shelf lives of the medications. At the WHO, public health officers are looking for technical solutions to prolong the usefulness of the antivirals to at least two or three more years beyond the roughly 5 year shelf lives of today’s medications, as well as “encouraging the development of new antivirals”, says IHR review committee member Santos.

The best weapon against flu is still the vac-cine. Once the H1N1 2009 virus was identified, the WHO needed just one month to select the pandemic virus and develop the seed strains for vaccine development. But that’s only the first step. “It takes at least six months to produce significant quantities of a vaccine,” Klaus Stöhr of Novartis claimed in 2010 (ref. 7). Stöhr is vice president of influenza strategy at Novartis Vaccines and Diagnostics in Cambridge, Massachusetts. One vexing question is how much vaccine to produce. By the time vaccines against the 2009 H1N1 virus were made available, the production was enough to treat only 10% of the population worldwide. Santos also highlights pharmaceutical deficits. “The technology to produce influenza vaccine has not changed much in the past 60 years,” he says.

Even if vaccines are produced in abundance, there’s no guarantee that the medication will get to the people who need it. “Vaccine distribution worldwide is the other challenge,” says Santos. Access to vaccines, he says, can be improved by

legal agreements between manufacturers and countries. Other paths to more equitable access to vaccines include differential pricing so that poorer countries can get the vaccines cheaper — already in practice, but says Santos, needs to be made more universal — as well as vaccine donations by manufacturers to the WHO. At the national level, there is the need to assure the maintenance of the cold chain — the continuous sequence of properly refrigerated vessels to transport the vaccines from factory to the patient — and to plan for local vac-cine distribution and administration. “One way to facilitate both vaccine and antiviral distribution,” Santos says, “could be to regionalize the effort, giv-ing each WHO regional office the responsibility to secure the resources needed for the area.”

No one can predict with precision when and where the next global outbreak will arise. But, in view of what has been learned from recent influenza pandemics, it is possible to define the most likely scenarios and to prepare for them. “It is like your personal life,” says Briand. “You can-not predict anything but there are scenarios that are more likely than others.” The lessons learned from the latest pandemic “will certainly drive our preparedness to different directions now.”

One key lesson is the need to be flexible in our response to unexpected conditions and to be able to respond in times of uncertainty. We must, in other words, be like the flu virus itself. ■

Laura Vargas Parada is a freelance science writer in Mexico City.

1. Domíguez-Cherit, G. et al. JAMA 302, 1880–1887 (2009).2. http://apps.who.int/gb/ebwha/pdf_files/WHA64/

A64_10-en.pdf. 3. Nalluswami, K. et al. MMWR 60, 1213–1215 (2011).4. Nabel, G. J. et al. Nature 471, 157–158 (2011).5. Myers, K. P. et al. Clin. Infect. Dis. 44, 1084–1088 (2007).6. Briand, S. et al. Public Health 125, 247–256 (2011).7. Stohr, K. Nature 465, 161 (2010).

THE NEXT PANDEMIC

1900 1910 201020001920 1930

1957 1957

1968

2009

1918

1940 1950 1960 1970 1980 1990

Experts recommend the following actions:

• Strengthen surveillance and response in much of the world, especially in under-resourced countries.

• Define pandemic, develop measures to estimate severity.

• Standardize reporting of influenza data.

• Focus on new vaccine production technologies, new antivirals, and extending the shelf lives of medications.

• Vaccine distribution worldwide needs to be revised and strengthened.

1918The Spanish flu (H1N1) strikes killing an estimated 50 million people.

Camp Funston, Kansas, site of the first recorded cases of Spanish flu (left), nurse draws water for flu patients (top).

Noted flu researchers. D. E. Rogers and E. D. Kilbourne test for the influenza virus (top), a New York hospital at the height of the epidemic (right).

A sign in Iowa casts flu as a foreign invader (left).

A pig sty is disinfected in Indonesia (right).

20091968700,000 people die from H3N2 infection (descended from H2N2); originates in Hong Kong.

1957H2N2 outbreak that began in Asia kills 100,000 people.

The return of H1N1 — swine flu — surfaces in Mexico and spreads rapidly — including to healthy people, but with a low fatality rate.

OUTLOOKINFLUENZA


Your spouse had the flu last week and could barely get out of bed for five full days — she was feverish and her

aching body exhausted. Then your four-year-old became sick and you had to rush her to the hospital when her fever spiked. Now that you’ve come down with the virus, you’re tired, a bit achy, but that only lasts for two or three days. How is it that the flu virus can have such a variable impact on people?

In order for the flu to pose a danger to any individual, two major factors must align: the host person has to be a clean slate — someone who hasn’t been exposed to an influenza virus related to the one circulating; and the virus itself has to be especially potent. The range and severity of symptoms the host experiences depend on the answers to a few other questions, such as whether the person’s immune system is compromised, and whether some genetic vulnerability exists.

The first question to delve into when asking about why a particular person gets very sick is: what makes a particular strain of flu more likely to take hold in a human host? The answer lies in the basic structure of the patho-gen. The influenza virus always consists of two types of major proteins: one hemagglutinin (H) and one neuraminidase (N). There are at least 16 types of neuraminidase and 9 of hema gglutinin, in principle, capable of forming more than 100 strains of flu virus. In reality, however, the flu viruses that commonly infect humans have been limited to only three types of hemagglutinin and two types of neuraminidase1, giving six possible flu strains. Each season, there’s a chance that, through mutations and recombinations, a novel virus will emerge — one that few, or no, humans have encountered before. Rolling this viral game of dice leads to deadly consequences. 1957 saw the emergence of the H2N2 strain. This represented a major structural shift for the virus; for the previous 40 years, all the flu cases had been caused by H1N1 viruses. This new virus killed 70,000 people in the United States alone2.

A more recent example of a novel virus was the H5N1 swine flu that circulated in 2009. Unlike seasonal flus, which are most severe in the elderly and those with other health problems, this virus killed a large number of young people — about a quarter of the deaths were in people younger than 24 years (many of them without pre-existing illness), and more than 60% were in people aged 25–64 years — again, many of them otherwise healthy. The reason for this reversal of for-tunes was that the virus was similar to ones that had circulated many years ago. Many elderly people, therefore, had some immunity to it, but their children and grandchildren had never been exposed to any influenza like this one and so lacked any resistance to it3.

Another issue that sets flu apart from some other infectious diseases is that antibodies against influenza don’t have a very long shelf

M O R B I D I T Y

A personal responseSome people get horribly sick from the flu, and even die. Others just rest for a few days. What’s behind this fateful variation? B Y C H R I S T I N E J U N G E

BR

AN

D N

EW IM

AG

E/ G

ETTY

IMAG

ES

INFLUENZAOUTLOOK


life. In certain groups of people, there are addi-tional concerns: after peaking, antibody levels decline faster in older people than in younger people. In people with compromised immune systems, such as someone with HIV/AIDS or someone taking certain cancer drugs, it’s double trouble — antibody levels don’t rise as high after exposure to the virus in the first place, nor do they last as long. “It’s difficult to say whether any individual is still protected after a year or not, which is why we recommend getting the flu shot each year,” says Lisa Grohskopf an infectious disease specialist who works in the influenza division of the Centers for Disease Control and Prevention (CDC) in Atlanta, Georgia. She adds that protection can wane from the beginning of one flu season to its end, since the season can stretch for seven or eight months.

Infants or young children who haven’t come into contact with the flu virus before — either by getting infected or from vaccination — don’t have antibodies specific to the viruses circulating in a given year, putting them at risk for getting the flu.

For reasons that aren’t always clear, accord-ing to microbiologist Peter Palese at the Mount Sinai School of Medicine in New York, people with certain chronic illnesses also have a harder time resisting flu or fighting it once they do get it. Flu is more dangerous for people with heart disease, diabetes, asthma, chronic obstructive pulmonary disease (COPD) and morbid obe-sity. Palese explains that some of those diseases, such as asthma and COPD, cause persistent lung problems, so any illness that causes fluid to accumulate in the lungs, as flu does, would be more problematic for these people. For the obese, he says, breathing can be difficult, so the act of clearing mucus from the lungs might also be difficult, and the virus would linger in the lungs longer. Grohskopf adds that various health

problems can create a vicious cycle. For instance, she says, “diabetes affects the immune system and so may affect the body’s ability to fight influ-enza; in turn, being ill with the flu can cause dif-ficulty with keeping blood sugar under control,” an outcome that can complicate diabetes.

Flu’s effects can be amplified not only by ill-ness but by our genes. Indeed, just as the genetics of the flu virus itself (which determines which H

and N type it is) can make it stronger or weaker, the genetics of the person who encounters the influ-enza virus can make that person more or less likely to become ill. One thing that has become clear in recent years is that single nucleotide

polymorphisms in one of a number of genes governing immunity affects a person’s suscep-tibility to infectious diseases, and how sick he or she becomes from those infections. A 2008 study by Ute Vollmer-Conna, a geneticist at the University of New South Wales in Australia, who studies the genetic response to infection, advanced that idea. Vollmer-Conna’s research looked at 300 people infected with one of three viruses: Epstein-Barr, Coxiella burnetii (which causes Q fever) and the Ross River virus4. Her conclusion: those who had the most severe symptoms, and those who were ill for the longest, had polymorphisms that signifi-cantly influenced pro- and anti-inflammatory cytokine production. Vollmer-Conna explains that these genetic predispositions intensify the inflammatory response to an acute infection, as well as prolonging recovery time. Although

the scope of her study did not include influenza, she hypothesizes that “the same principles apply across a much broader range of common infec-tions, including flu”.

Another, more recent study looked at the changes in gene expression caused by the flu virus specifically. Researchers exposed 17 healthy subjects to flu virus type H3N2. Nine came down with flu symptoms and eight did not. When the scientists looked at blood from each subject, they noticed that different genes were up-regulated, depending on whether the person fell ill or not. “Individuals who became symptomatic had up-regulation of genes involved in the inflammatory response,” says Geoffrey Ginsburg, director of the center for genomic medicine at the Duke Institute for Genome Sciences and Policy in Durham, North Carolina, and one of the authors of the study. “The exuberance of the inflammatory response is likely at the root of why we have symptoms of infection,” he says. In those who did not become ill, he explains, the genes that were up-regulated were those involved in protein synthesis and oxidative stress. “Asymptomatic subjects may have mechanisms at work that prevent active infection and therefore reduce the inflammatory response.”

Aside from more research into the human genes that control the immune response to flu, there are plenty of other lines of investigation that would help explain why different people respond to flu so differently. Vollmer-Conna argues for more research to tease out which genes and biological processes are responsible for recovery after an infection. This knowledge, she writes, could let doctors identify those at risk for severe illness following infection with com-mon diseases, and therefore set them up with special prevention and treatment programmes.

In a 2004 research paper, Palese posed sev-eral questions, including what makes it possible for some influenza viruses to hop from one ani-mal species to another; questions that he hoped researchers would have answered by now. He concedes that seven years later, the questions on his list are still pertinent. One in particular that still remains an important area of research, says Palese, are the viral genes that determine transmission versus the host or environmental factors that promote its spread.

Clearly researchers have made some pro-gress, but as the poet Robert Frost once wrote, there are miles to go before we sleep. Frost was writing about stopping in the woods on a snowy evening in winter — which, of course, is flu season. ■

Christine Junge is a science writer in North Easton, Massachusetts. 1. Palese, P. Nature Medicine 10, S82–S87 (2004). 2. Dolin, R. How to understand your risk and protect your health (Harvard Health Publications, 2009). 3. Vollmer-Conna, U. et al. Clinical Infect. Dis. 47, 1418–1425 (2008). 4. Huang, Y. et al. PLoS Genet. 7, e1002234 (2011).

“The exuberance of the inflammatory response is likely at the root of why we have symptoms of infection.”

Up close and personal: H5N1 (green) in cell culture.

CD

C/S

CIE

NC

E P

HO

TO L

IBR

AR

YOUTLOOKINFLUENZA

LETTERdoi:10.1038/nature10004

Long-term evolution and transmission dynamics ofswine influenza A virusDhanasekaran Vijaykrishna1,2,3, Gavin J. D. Smith1,2,3, Oliver G. Pybus4, Huachen Zhu1,2, Samir Bhatt4, Leo L. M. Poon1,Steven Riley5, Justin Bahl1,2,3, Siu K. Ma1, Chung L. Cheung1, Ranawaka A. P. M. Perera1, Honglin Chen1,2, Kennedy F. Shortridge1,2,Richard J. Webby6, Robert G. Webster1,6, Yi Guan1,2 & J. S. Malik Peiris1,7

Swine influenza A viruses (SwIV) cause significant economic lossesin animal husbandry as well as instances of human disease1 andoccasionally give rise to human pandemics2, including that causedby the H1N1/2009 virus3,4. The lack of systematic and longitudinalinfluenza surveillance in pigs has hampered attempts to recon-struct the origins of this pandemic4. Most existing swine data werederived from opportunistic samples collected from diseased pigs indisparate geographical regions, not from prospective studies indefined locations, hence the evolutionary and transmissiondynamics of SwIV are poorly understood. Here we quantify theepidemiological, genetic and antigenic dynamics of SwIV in HongKong using a data set of more than 650 SwIV isolates and morethan 800 swine sera from 12 years of systematic surveillance in thisregion, supplemented with data stretching back 34 years. Inter-continental virus movement has led to reassortment and lineagereplacement, creating an antigenically and genetically diverse viruspopulation whose dynamics are quantitatively different from thosepreviously observed for human influenza viruses. Our findingsindicate that increased antigenic drift is associated with reassort-ment events and offer insights into the emergence of influenzaviruses with epidemic potential in swine and humans.

All major SwIV lineages of North American or European origin—classical swine (CS), European or Eurasian avian-like swine (EA) andtriple-reassortant swine (TRIG) (see Supplementary Information foran overview)—co-circulate in southern China4,5, and both human(H3N2) and avian (H5N1 and H9N2) viruses have been isolated fromswine in the region6–9. To address the critical lack of structured swineinfluenza data, we undertook virological and serological analysis ofHong Kong SwIV surveillance samples. Most (80–95%) of the swineslaughtered in Hong Kong originate from provinces in mainlandChina (Supplementary Fig. 1 and Supplementary Table 1), the regionwith the world’s largest swine population10–12.

We isolated and subtyped 573 H1N1 and H1N2, 97 H3 and 2 H9N2viruses from fortnightly sampling of swine slaughtered between May1998 and January 2010 (Fig. 1a, b). We found no H5N1 viruses. FromAugust 1998 to December 2002, the isolates were mostly CS H1N1viruses. EA H1N1 viruses were detected only from 2001 onwards andTRIG H1N2 from 2002 onwards. During 2002–05, viruses classified asCS, EA, TRIG and H3N2 co-circulated and fluctuated in relative pre-valence (Fig. 1b). After 2005, EA H1N1 viruses became dominant andH3N2 viruses disappeared, although CS H1N2 and TRIG H1N2viruses continued to be isolated sporadically (Fig. 1b). All threeSwIV H1 lineages (CS, EA and TRIG) have co-circulated withH1N1/2009 after the introduction of the latter virus into pigs5.

Comprehensive phylogenetic analyses of the genes encoding surfaceantigens haemagglutinin (HA) and neuraminidase (NA) in all H1N1and H1N2 isolates (including 93 H1N1 viruses isolated during 1976–79

and 1993–94) confirmed that most isolates belonged to the CS or EAlineages (Fig. 2 and Supplementary Fig. 2a–c). All pre-1998 viruses wereCS except two ‘pure avian’ viruses from 1993 (ref. 13); much greater HAdiversity was observed after 1998 (Fig. 2a). Notably, our CS isolates donot form a single monophyletic group; rather, they are interspersed withNorth American CS viruses, indicating multiple introductions of CS intothe study area. In contrast, EA viruses are monophyletic, indicating asingle introduction. All Hong Kong TRIG viruses form a single group

1State Key Laboratory of Emerging Infectious Diseases & Department of Microbiology, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong, China.2International Institute of Infection and Immunity, Shantou University Medical College, Shantou, Guangdong, China. 3Laboratory of Virus Evolution, Program in Emerging Infectious Diseases, Duke-NUSGraduate Medical School, 8 College Rd, 169857, Singapore. 4Department of Zoology, University of Oxford, South Parks Road, Oxford OX1 3PS, UK. 5Department of Community Medicine and School ofPublic Health, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Pokfulam, Hong Kong, China. 6Virology Division, Department of Infectious Diseases, St Jude Children’s Research Hospital,Memphis, Tennessee 38015, USA. 7HKU-Pasteur Research Centre, The University of Hong Kong, Pokfulam, Hong Kong Special Administrative Region, China.

Rela

tive f

req

uency o

f S

wIV

0

20

40

60

80

100

1998 2000 2002 2004 2006 2008 2010

Year

0

2

4

6

8

10

12

14

2000 2002 2004

Year

2006 2008 20101998

Perc

enta

ge o

f sam

ple

s

testing

po

sitiv

e f

or

H1 o

r H

3

a

Human H1N1 (n = 2)

H1N1/2009 (n = 10)

CS (n = 321)

EA (n = 188)

TRIG (n = 52)

H3N2 (n = 97)

b

0

100

200

300

400

500

600

Pigs sampled

Pig

s s

am

ple

d

Figure 1 | Prevalence and relative frequency of swine influenza H1 and H3subtypes. a, b, Percentage prevalence (a) and year-averaged relative frequency(b) of the major HA variants of SwIV. Colour codes and numbers of isolates (n)of H1 and H3 subtype viruses detected from swine in Hong Kong between 1998and 2010 are shown. The viruses detected include CS, EA, TRIG, H1N1/2009and human seasonal H1N1 and H3N2 viruses. The orange line indicates thenumber of pigs sampled during the surveillance period.

2 6 M A Y 2 0 1 1 | V O L 4 7 3 | N A T U R E | 5 1 9

Macmillan Publishers Limited. All rights reserved©2011

www.nature.com/doifinder/10.1038/nature10004

except for isolate Sw/HK/78/2003, indicating that this virus was intro-duced from North America separately. The Hong Kong TRIG virusesdiverge from the North American TRIG and H1N1/2009 HA lineagessoon after the emergence of TRIG H1N2 viruses in North America andthus constitute a third distinct TRIG HA gene lineage (Fig. 2a)14.

Molecular clock phylogenies of 221 whole genomes (33.2% of iso-lates) revealed that CS viruses isolated from 1976 to 1994 containedonly CS genome segments: no reassortment was detected during thisperiod (Fig. 3a and Supplementary Fig. 3). However, several reassortantSwIV were detected between 1998 and 2010. In addition to the CS, EA,TRIG and H1N1/2009 viruses, we detected 14 ‘genotypes’ generated byreassortment between circulating swine and human/avian lineages(Supplementary Figs 3 and 4). Most of the newly identified reassortantswere observed only transiently and usually contained genome segmentsfrom viruses that were dominant at that time (Supplementary Fig. 4).Excepting the CS H1N2 virus, which acquired the human H3N2neuraminidase gene repeatedly, we detected no preferential directionof horizontal gene transfer among SwIV strains.

Three of the 14 reassortant genotypes were isolated repeatedly(Supplementary Fig. 4); specifically, (1) CS H1N2 viruses, (2) novelH1N2 reassortants and (3) Sw/HK/72/2007-like (EA-reassortant)strains, which have acquired an NS gene from TRIG viruses and whichbelong to a divergent EA lineage (Fig. 2). Since their initial detection in2007, EA-reassortant viruses have become the dominant EA lineage,constituting 12.5% of all EA viruses in 2007, 15.4% in 2008 and 41.4%in 2009. Because most reassortant ‘genotypes’ were isolated only once,we hypothesize that few are adapted for continuous circulation(although we cannot exclude stochastic demographic effects or samplingbias as alternative explanations). SwIV diversity in our population isprobably increased by pig movements: breeding pigs constitute the bulkof live pigs imported into China and data indicate that imports haveincreased since 1990 (refs 11,12).

For all genome segments, molecular clock phylogenies exhibitedlong branches leading to several reassortant lineages (Fig. 3). This

was also observed for the H1N1/2009 virus4 and indicates a long periodof unsampled diversity. Upon first detection, these reassortant lineagestend to be more closely related to viruses circulating in our populationfive to eleven years previously, rather than to co-circulating strains(that is, they do not arise from the contemporaneous part of the phylo-genetic ‘backbone’).

We found extensive antigenic crossreaction among CS, TRIG andH1N1/2009 viruses (Supplementary Table 2 and Supplementary Fig.5). Ferret antisera to these viruses also crossreacted with early EAviruses (2001–03) but reacted poorly with more recent EA-reassortantstrains (Fig. 4). Interestingly, the six novel EA-reassortant virusestested (sampled between 2007 and 2009) crossreacted weakly withall ferret antisera used, including the antiserum to Sw/HK/NS29/2009. This group thus represents an antigenically distinct SwIVlineage. Excepting the earliest EA reassortant (Sw/HK/72/2007), allremaining EA reassortants reacted well against antisera raised to theEA-reassortant virus Sw/HK/1559/2009 (Fig. 4), indicating progres-sive antigenic change of EA-reassortant viruses during our study.

The earliest of the above-mentioned EA reassortants (Sw/HK/72/2007) had acquired two amino acid changes in HA antigenic sites andlater EA reassortants (for example, Sw/HK/1559/2008 and Sw/HK/1532/2009) had a further five changes at antigenic sites (Supplemen-tary Figs 3a and 6). These findings support the hypothesis that EA-reassortant viruses have antigenically drifted away from crossreactingantibodies arising from CS, TRIG and early-EA virus infection. Theobservation that antigenic change occurred in the reassortant EA viruslineage rather than in the parental lineage indicates that reassortmentmay facilitate the generation of SwIV antigenic diversity.

Although SwIV isolation rates declined after EA viruses became pre-dominant (Fig. 1), serological data indicate that overall SwIV seropre-valence has not declined (Supplementary Tables 3 and 4). To testwhether EA viruses have a competitive advantage over CS strains, weintranasally infected five-week-old, previously influenza-naive pigs withSwIV representative of the lineages isolated here (Methods and

EA

CS

TRIG

H1N1/2009 swine

Hong Kong surveillance

Swine (seasonal influenza)

Avian

Human

GenBank

Swine

0.07

a c

0.04

b

0.03

Avia

nH

um

an

H1

N1

Cla

ssic

sw

ine

Avia

n

Avia

nH

um

an

/sw

ine

Hu

man

H1

N1

Cla

ssic

sw

ine

Eu

rop

ean

‘avia

n-l

ike’

sw

ine

Eu

rop

ean

‘avia

n-l

ike’

sw

ine

H1 N1 N2

AA

Figure 2 | Genetic relationships of swine influenza A viruses for genesencoding surface proteins. a–c, Haemagglutinin H1 (a), neuraminidase N1(b), neuraminidase N2 (c). Representative avian, swine and human influenza Aviruses obtained from GenBank are represented by black, dark grey and lightgrey, respectively. Colour codes of H1N1 and H1N2 subtype viruses detected

from swine during 1977–2010 in Hong Kong are shown in the key. Arrow Aindicates the antigenically divergent EA-reassortant viruses discussed in thetext (for example, Sw/HK/72/2007, Fig. 4). Scale bars represent substitutionsper site. Fully detailed phylogenies including sequence names are provided inSupplementary Fig. 2a–c.

RESEARCH LETTER

5 2 0 | N A T U R E | V O L 4 7 3 | 2 6 M A Y 2 0 1 1


Supplementary Fig. 7). EA viruses showed the highest and most pro-longed virus shedding, closely followed by TRIG viruses; CS virusesshowed lower peak viral titres. Thus, the replicative advantage of EAviruses, together with the low prevalence of crossreactive antibodies toEA in swine (15% in 2000, 26% in 2004; Supplementary Table 4), mayhelp to explain the replacement of other SwIV lineages with EA viruses.

We tracked the evolution in our EA viruses of amino acids previ-ously associated with adaptation of avian influenza to other species15–17.Purported avian residues were maintained at most of these sites(Supplementary Table 5) despite the circulation of these viruses inswine for more than 30 years18,19. However, the PDZ- (post-synapticdensity protein, Drosophila disc large tumor suppressor and zonulaoccludens-1 protein) ligand at the 39 end of EA virus non-structural(NS) 1 genes showed significant host-specific evolution: early EuropeanEA viruses had the avian ESEV motif, with a change to GSEV/GPEV

motifs observed in several hosts. By 1999, most viruses sampled had theGPKV motif previously described from pigs16. CS and TRIG virusesthat contributed the NS gene to H1N1/2009 have a truncated NS gene,as do the antigenically variant Sw/HK/72/2007-like viruses. The role ofthe truncated NS gene in inter-species transmission clearly meritsfurther study. Furthermore, a modest but significant (P , 0.01) changein selection pressure was observed between European EA viruses iso-lated shortly after cross-species transmission (non-synonymous tosynonymous (dN/dS) substitution rate ratio of 0.24; 95% confidenceinterval 5 0.22–0.27) and those isolated later (dN/dS 5 0.17; 95% con-fidence interval 5 0.14–0.20), consistent with the hypothesis that host-specific selection increased viral adaptation after the introduction ofEA viruses into swine (Supplementary Table 6).

Our unique longitudinal study reveals a genetically and antigenicallydynamic SwIV population within a single region and provides a baselinefor future studies of the virus elsewhere. The epidemiology and evolu-tion of SwIV seem to be strongly shaped by gene flow among continentsand species, facilitating the reassortment of diverse lineages and occa-sionally resulting in antigenic change. Although we confirm that theH1N1/2009 virus was not generated within our study’s catchment, theprocesses of lineage emergence, importation, reassortment and replace-ment described here are probably representative of the H1N1/2009source population. We show that reassortments between EA andTRIG viruses do occur, generating reassortants that establish themselvesas stable lineages in swine. SwIV reassortants containing H1N1/2009-like genome segments have also been transiently detected5,20.

Despite clear evidence of inter-continental SwIV movement, geneflow is not so frequent that the global SwIV population acts as a singlegene pool (as observed for human influenza A21,22); instead a higherdiversity of mammalian-adapted viruses in global swine populations issupported. Crucially, the co-circulation of multiple SwIV lineages facili-tates the production of new genomic combinations. The evolutionaryconsequences of increased SwIV movement are hard to predict butrequire consideration given an increasingly globalized future.

a

b

Segment1 2 3 4 5 6 7 8

CS

EA

TRIG

H1N1/2009

Avian

Human H3N2

1974 1978 1982 1986 1990 1994 1998 2002 2006 2010

Year

Figure 3 | Phylogenies and divergence times of the haemagglutinin genes ofclassical swine and European avian-like SwIV. a, CS; b, EA. Coloured boxesadjacent to tips show the lineage classification of each gene segment of SwIVisolated in this study. Arrows indicate the long branches that lead to newlydetected reassortant SwIV. Purple node bars represent 95% credible intervals oflineage divergence times. A fully detailed HA phylogeny including sequencenames is shown in Supplementary Fig. 3a.

4167/

1999

1304/

2003

1110/

2006

Cal/4/

2009

NS29/

2009

1559/

2008

Test antigens

CS H1N1 CS H1N2 TRIG H1N2 H1N1/09 EA H1N1 EA* H1N1

Sw/HK/4167/1999 CS H1N1 1:20,480 1:320 1:10,240 1:2,560 1:2,560 <1:10

Sw/HK/1304/2003 CS H1N2 1:1,280 1:2,560 1:640 1:80 1:40 <1:10

Sw/HK/1110/2006 TRIG H1N2 1:40,960 1:1,280 1:10,240 1:640 1:5,120 <1:10

Cal/04/2009 H1N1/2009 1:640 1:640 1:2,560 1:1,280 160 <1:10

Sw/HK/NS29/2009 EA H1N1 1:640 1:160 1:1,280 1:80 1:10,240 <1:10

Sw/HK/1559/2008 EA* H1N1 <1:10 <1:10 <1:10 <1:10 1:40 1:5,120

Sw/HK/8512/2001 EA H1N1 1:10,240 1:1,280 1:5,120 1:2,560 1:10,240 1:320

Sw/HK/72/2007 EA* H1N1 <1:10 <1:10 <1:10 <1:10 1:40 <1:10

Sw/HK/247/2009 EA* H1N1 <1:10 <1:10 <1:10 <1:10 1:40 1:2,560

Sw/HK/NS613/2009 EA* H1N1 <1:10 <1:10 <1:10 <1:10 1:10 1:2,560

Sw/HK/2481/2009 EA* H1N1 <1:10 <1:10 <1:10 <1:10 1:20 1:2,560

Sw/HK/NS186/2009 EA* H1N1 <1:10 <1:10 <1:10 <1:10 1:40 1:2,560

Ferret antisera

Sw/HK/1669/2002 EA H1N1 1:5,120 1:1,280 1:2,560 1:1,280 1:10,240 1:160

Sw/HK/NS129/2003 EA H1N1 1:5,120 1:1,280 1:2,560 1:1,280 1:10,240 1:160

Sw/HK/1716/2006 EA H1N1 1:2,560 1:640 1:1,280 1:640 1:2,560 <1:10

Sw/HK/NS952/2008 EA H1N1 1:2,560 1:640 1:640 1:320 1:2,560 <1:10

Figure 4 | Antigenic characterization of SwIV measured by haemagglutinininhibition assays. Titres are shaded according to their respective major SwIVHA lineages (see Figs 1–3); low (1:20, 1:40) and non-reactive titres (,1:10) areshaded in lighter colours. Underlined values represent homologous antibodytitres. EA-reassortant viruses (indicated by asterisks) showed poorcrossreactivity against antisera raised towards CS, TRIG, H1N1/2009 and late(2006–2009) ‘pure’ EA viruses. Excepting the earliest EA-reassortant virus (Sw/HK/72/2007), all remaining EA-reassortants reacted well against antiseraraised towards the EA-reassortant virus Sw/HK/1559/2009, indicatingprogressive antigenic change of this novel reassortant.

LETTER RESEARCH

2 6 M A Y 2 0 1 1 | V O L 4 7 3 | N A T U R E | 5 2 1


Our study reveals a frequent generation of new reassortants but thesurvival and persistence of only a few, a process we term ‘recombinantchatter’23. Our data also indicate that reassortment and antigenic changeare linked. This phenomenon was described in North America CSviruses after the events that generated the TRIG viruses24; it has alsobeen observed in human influenza21. After reassortment, evolution inHA antigenic domains may arise for several reasons: (1) because ofherd-immune selection pressure; (2) because those residues are underweak selective constraint; or (3) to compensate for fitness costs of muta-tions accruing elsewhere in the genome. The role of reassortment indriving genome-wide evolution requires detailed investigation.

We found that the quantitative dynamics of SwIV genomic diversityand lineage turnover (Supplementary Fig. 8) are slower, less periodic andless predictable than the repeated annual replacements typically seen forhuman influenza A. The reasons for SwIV lineage change are unclear:previously, selection arising from herd immunity was considered lessimportant for pigs than for humans because the short lifespan of farmedswine (,150 days) lowers the chance of re-infection, reducing the cross-protection that probably drives antigenic drift. Furthermore, maternallyacquired swine immunity does not seem to interrupt SwIV infection ortransmission, despite masking clinical illness25. Our surveillance data,animal infection experiments and serological data show that one reasonfor lineage change may be a competitive advantage of EA over CS andTRIG viruses.

The hypothesis that pigs are important in pandemic emergence, asfacilitators of reassortment among influenza viruses26, has regainedfavour after the emergence of H1N1/2009. This virus represented asubtype already endemic in humans, implying that other H1 and H3viruses prevalent in swine are credible pandemic candidates, especiallywhen corresponding immunity in humans is absent. Indeed, we foundthat there are other swine viruses (for example, Sw/HK/NS29/09) towhich humans lack herd immunity (Supplementary Table 7). Hence,future assessment of zoonotic potential must combine the evaluation ofcrossreactive immunity in humans, the assessment of transmissibilityin animal models and ongoing surveillance of SwIV genetic diversity.The H1N1/2009 virus has already infected swine and reassorted withother SwIV, indicating that circulating SwIV will continue to acquirenovel non-SwIV genes5 (notably, avian viruses such as H9N2 andH5N1 are occasionally detected in swine in Asia6,8,9,27). Avian-to-swineand swine-to-human host adaptation of influenza viruses are bothpoorly understood in comparison to avian-to-human adaptation andare a priority for future research.

METHODS SUMMARYSystematic swine influenza A virus surveillance was initiated in May 1998 at a centralslaughterhouse in Hong Kong. During 1999–2007, about 15–20% of the pigs werefarmed locally in Hong Kong and the remainder were imported from several pro-vinces in China; however, since 2008 the proportion of locally produced pigs fell to5% (see Supporting Information). About 128 nasal and tracheal swabs were collectedtwice monthly from August 1998 to April 2009; since May 2009, sample numberswere doubled (Fig. 1a). Genes encoding surface proteins (HA and NA) weresequenced for all 573 H1N1 viruses isolated from 1998 to 2010 and for 93 swineH1 viruses from our repository, isolated during the periods 1976–1978 and 1993–1994. Full genome sequencing was carried out for 221 representative viruses. Toestimate the genetic diversity and the level of gene reassortment, phylogenetic treeswere constructed for each genomic segment independently (Supplementary Fig. 2).On the basis of the phylogenetic relationships of each gene segment, major swinevirus lineages circulating in Hong Kong were identified and a more detailed Bayesianphylogenetic analysis for each lineage was conducted, thereby estimating rates ofviral evolution and dates of divergence (Supplementary Fig. 3).

Full Methods and any associated references are available in the online version ofthe paper at www.nature.com/nature.

Received 22 September 2010; accepted 17 March 2011.

1. Shinde, V. et al. Triple-reassortant swine influenza A (H1) in humans in the UnitedStates, 2005–2009. N. Engl. J. Med. 360, 2616–2625 (2009).

2. Smith, G. J. D. et al. Dating the emergence of pandemic influenza A viruses. Proc.Natl Acad. Sci. USA 106, 11709–11712 (2009).

3. Garten, R. J. et al. Antigenic and genetic characteristics of swine-origin 2009A(H1N1) influenza viruses circulating in humans. Science 325, 197–201 (2009).

4. Smith, G. J. D. et al. Origins and evolutionary genomics of the 2009 swine-originH1N1 influenza A epidemic. Nature 459, 1122–1125 (2009).

5. Vijaykrishna, D. et al. Reassortment of pandemic H1N1 viruses in swine. Science328, 1529 (2010).

6. Peiris, J. S. M. et al. Cocirculation of avian H9N2 and contemporary ‘‘human’’H3N2 influenza A viruses in pigs in southeastern China: potential for geneticreassortment? J. Virol. 75, 9679–9686 (2001).

7. Yu, H. et al. Genetic evolution of swine influenza A (H3N2) viruses in China from1970–2006. J. Virol. 46, 1067–1075 (2008).

8. Cong, Y. L. et al. Antigenic and genetic characterization of H9N2 swine influenzaviruses in China. J. Gen. Virol. 88, 2035–2041 (2007).

9. Li, H. Y. et al. Isolation and characterization of H5N1 and H9N2 influenza virusesfrom pigs in China. Chinese J. Vet. Prev. Med. 26, 1–6 (2004).

10. Clements, A. C. A., Pfeiffer, D. U., Otte, M. J., Morteo, K. & Chen, L. A global livestockproduction and health atlas (GLiPHA) for interactive presentation, integration andanalysis of livestock data. Prev. Vet. Med. 56, 19–32 (2002).

11. Zhang, J. & Beckman, C. People’s Republic of China: Agricultural situation:Livestock and Products 2008. (USDA Foreign Agriculture Service, 2008).

12. Wang, R. China – pork powerhouse of the world. Advances Pork Prod. 17, 33–46(2006).

13. Guan, Y. et al. Emergence of avian H1N1 influenza viruses in pigs in China. J. Virol.70, 8041–8046 (1996).

14. Lorusso, A. et al. Genetic and antigenic characterization of H1 influenza virusesfrom United States swine from 2008. J. Gen. Virol. 92, 919–930 (2011).

15. Finkelstein, D. B. et al. Persistent host markers in pandemic and H5N1 influenzaviruses. J. Virol. 81, 10292–10299 (2007).

16. Obenauer, J. C. et al. Large-scale sequence analysis of avian influenza isolates.Science 311, 1576–1580 (2006).

17. Taubenberger, J. K. et al. Characterization of the 1918 influenza virus polymerasegenes. Nature 437, 889–893 (2005).

18. Pensaert, M., Ottis, K., Vanderputte, J., Kaplan, M. M. & Buchmann, P. A. Evidencefor the natural transmission of influenza A virus from wild ducks to swine and itspotential for man. Bull. World Health Organ. 59, 75–78 (1981).

19. Dunham, E. et al. Different evolutionary trajectories of European avian-like andclassical swine H1N1 influenza A viruses. J. Virol. 83, 5485–5494 (2009).

20. Moreno, A. et al. Novel H1N2 swine influenza reassortant strain in pigs derivedfrom the pandemic H1N1/2009 virus. Vet. Microbiol. 149, 472–477 (2011).

21. Rambaut, A. et al. The genomicandepidemiological dynamics of human influenzaA virus. Nature 453, 615–619 (2008).

22. Russell, C. A. et al. The global circulation of seasonal influenza A (H3N2) viruses.Science 320, 340–346 (2008).

23. Wolfe, N. D., Dunavan, C. P. & Diamond, J. Origins of major human infectiousdiseases. Nature 447, 279–283 (2007).

24. Webby, R. J. et al. Evolution of swine H3N2 influenza viruses in the United States.J. Virol. 74, 8243–8251 (2000).

25. Loeffen, W. L. A., Heinen, P. P., Bianchi, A. T. J., Hunneman, W. A. & Verheijden, J. H.M. Effect of maternally derived antibodies on the clinical signs and immuneresponse in pigs after primary and secondary infection with an influenza H1N1virus. Vet. Immunol. Immunopathol. 92, 23–35 (2003).

26. Scholtissek, C., Hinshaw, V. S. & Olsen, C. W. Influenza in pigs and their role as theintermediate host. In Textbook of Influenza (eds Nicholson, K. G., Webster R. G. &Hay A. J.) 137–145 (Blackwell Scientific, 1998).

27. Nidom, C. A. et al. Influenza A (H5N1) viruses from pigs, Indonesia. Emerg. Infect.Dis. 16, 1515–1523 (2010).

Supplementary Information is linked to the online version of the paper atwww.nature.com/nature.

Acknowledgements This research was supported in part by the National Institute ofAllergy and InfectiousDiseases (NIAID) contractHHSN26600700005Candthe AreaofExcellence Scheme of the University Grants Commission (grant AoE/M-12/06) of theHong Kong SAR Government. We acknowledge the Food and Environmental HygieneDepartment of Hong Kong for facilitating the study. We acknowledge support from TheRoyal Society of London (O.G.P.), UK COSI (S.B.), NIAID (G.J.D.S.), the Agency forScience, Technology and Research and the Ministry of Health, Singapore (D.V., G.J.D.Sand J.B.). We thank C. Y. H. Leung for producing some of the ferret antisera used in thisstudy.

Author Contributions J.S.M.P. and Y.G. conceived the study, conducted surveillance,performed analyses and co-wrote the paper. D.V., G.J.D.S. and O.G.P. conceived thestudy, performed analyses, co-wrote the paper and contributed equally to this work.H.Z., S.B., L.L.M.P., S.R., J.B., R.A.P.M.P. and H.C. performed analyses, S.K.M. conductedsurveillance, C.L.C. conducted sequencing, K.F.S. and R.G.W. initiated surveillance in1976 and provided viruses and R.J.W. provided viruses and reagents. All authorscommented on and edited the paper.

Author Information Sequences generated in this study have been deposited withGenBank under the accession numbers CY084470–CY085121, CY085301–CY086876 and CY087041–CY087142. Reprints and permissions information isavailable at www.nature.com/reprints. The authors declare no competing financialinterests. Readers are welcome to comment on the online version of this article atwww.nature.com/nature. Correspondence and requests for materials should beaddressed to Y.G. ([email protected]) or J.S.M.P. ([email protected]).

RESEARCH LETTER

5 2 2 | N A T U R E | V O L 4 7 3 | 2 6 M A Y 2 0 1 1


www.nature.com/nature


www.nature.com/reprints


mailto:[email protected]

mailto:[email protected]

METHODSSurveillance. Systematic influenza surveillance was conducted from May 1998until January 2010 in swine at an abattoir in Hong Kong, where tracheal or nasalswabs were collected fortnightly from slaughtered swine. During 1999–2007,about 15–20% of the pigs were farmed locally in Hong Kong and the remainderwere imported from several provinces in China; however, since 2008 the propor-tion of locally produced pigs fell to 5% (see Supporting Information). Serumsamples were collected from 50 slaughtered pigs each month. Routine virologicalsurveys were also conducted in Hong Kong in 1977–79 and 1993–94. Swab mate-rials were inoculated into nine-to-ten-day-old embryonated chicken eggs andMadin Darby canine kidney (MDCK) cells; virus isolates were identified andsubtyped by haemagglutination inhibition assays as previously described24.Virus isolation and sequencing. Viral RNA extraction, complementary DNAsynthesis, PCR and sequencing were carried out as described5,28. Viral RNA wasextracted directly from infected allantoic fluid or cell culture using the QIAampviral RNA minikit (Qiagen). cDNA was synthesized by reverse transcription; geneamplification by PCR was performed using specific primers for each gene segment.PCR products were purified with the QIAquick PCR purification kit (Qiagen) andsequenced using synthetic oligonucleotides. Reactions were performed using theBig Dye-Terminator v3.1 Cycle Sequencing Reaction Kit on an ABI PRISM 3700DNA Analyser (Applied Biosystems) following the manufacturer’s instructions.All sequences were assembled and edited with Lasergene version 6.1 (DNASTAR).The HA and NA genes were sequenced for all viruses collected in this study and fullgenome sequencing was conducted for representative viruses, selected on the basisof HA and NA gene diversity and including representative viruses sampled on eachpositive sampling occasion. All novel reassortants detected on the basis of fullgenome sequencing were subjected to plaque cloning and full genome sequencing(of at least six randomly selected clones per virus) to confirm that the reassortantwas not an artefact of mixed infection.Antigenic analyses. The antigenic characteristics of SwIV were compared using ahaemagglutination inhibition assay with ferret antisera raised against representativeinfluenza A viruses. Ferret antisera raised against Sw/HK/4167/1999 (CS H1N1), Sw/HK/1110/2006 (TRIG H1N2), Sw/HK/NS29/2009 (EA H1N1) and A/California/4/2009 were produced at the Department of Infectious Diseases at St Jude Children’sResearch Hospital, Memphis, Tennessee and the Department of Microbiology, TheUniversity of Hong Kong. The haemagglutination inhibition assay started at 1:40dilutions for ferret antisera.

To detect antibody prevalence towards major SwIV lineages, we used the hae-magglutination inhibition assay with five representative viruses including theantigenically divergent Sw/HK/72/2007-like EA viruses. This allowed us to quantifychanges in seroprevalence in serum collected from swine during 2000, 2004, 2009and 2010.Experimental infection of pigs. To characterize in vivo replicative behaviour ofviruses from the major SwIV lineages, we experimentally infected local domestichybrid (Putian white and Nianbian variant) pigs (Sus scrofa domesticus) obtainedfrom a commercial herd and confirmed to be sero-negative and free of influenza virusby HI assays and virus isolation in MDCK cells. Pigs were infected with representativestrains belonging to the CS (Sw/HK/4167/1999, Sw/HK/1304/2003), TRIG (Sw/HK/

1110/2006), EA (Sw/HK/NS29/2009) and novel EA-reassortant (Sw/HK/72/2007)lineages isolated in this study. Two five-week-old pigs (one male and one female)were intranasally infected with 1 ml of Eagle’s minimal essential medium (MEM)containing 106 50% tissue culture infectious doses (TCID50) of a virus strain. Nasalswabs were collected for 14 d after inoculation from each piglet and placed in 0.6 ml ofvirus transport medium. Virus shedding in the nasal swabs of pigs was calculated inMDCK by the 50% end-point method29 and was expressed as TCID50 ml21 of swab.Animal experiments were carried out in biosafety level three containment facilities at20–21 uC and 76.5 6 2.1% relative humidity. Experiments were approved by theShantou University Medical College and conducted in compliance with universityguidelines on animal ethics and welfare.Molecular evolution and adaptation. Global dN/dS rate ratios for each HongKong swine lineage and the haemagglutinin gene of European EA viruses wereestimated using the codon-based single likelihood ancestor counting method30.To determine whether selection was acting differentially on major lineages, thedN/dS rate ratio estimate for a lineage was enforced to other co-circulatinglineages. A likelihood ratio test was conducted to evaluate whether this fit wassignificantly worse than unconstrained analysis (and vice versa), with a criticalP value of 0.01. This test was repeated using the upper and lower limits of theconfidence interval.Phylogenetic analyses. Phylogenetic trees were inferred using the neighbour-joining method, using genetic distances calculated by maximum likelihood underthe Hasegawa, Kishino and Yano (HKY) model with gamma-distributed among-site rate variation (HKY1C). The parameters of this model were estimated usingmaximum likelihood on an initial tree. Temporal phylogenies and rates of evolu-tion were inferred using a ‘relaxed molecular clock’ model that allows evolutionaryrates to vary among lineages in a Bayesian Markov chain Monte Carlo (MCMC)framework31 This was used to sample phylogenies and dates of divergence whileconstraining each sequence to its known date of sampling. A model comprisingcodon-position-specific HKY1C substitution models was used. For all analysesemploying Bayesian MCMC sampling, a chain length of at least 50 million stepswas used with a 10% ‘burn-in’ removed. At least two independent runs of eachchain were performed and compared to ensure adequate sampling. To estimatechanges in genetic diversity during our sampling period we used a coalescent-basedflexible demographic model32 to the above MCMC approach. An estimate of therelative genetic diversity (Net, where Ne is the effective population size and t is thegeneration time) is obtained by integrating uncertainty across the tree topologies.

28. Poon, L. L. M. et al. Rapid detection of reassortment of pandemic influenza H1N1.Clin. Chem. 56, 1340–1344 (2010).

29. Reed, L. J. & Muench, H. A. Simple method of estimating fifty percent endpoints.Am. J. Hyg. 27, 493–497 (1938).

30. Kosakovsky Pond, S. L. & Frost, S. D. W. Not so different after all: a comparison ofmethods for detecting amino acid sites under selection. Mol. Biol. Evol. 22,1208–1222 (2005).

31. Drummond, A. J., Ho, S. Y., Phillips, M. J. & Rambaut, A. Relaxed phylogenetics anddating with confidence. PLoS Biol. 4, e88 (2006).

32. Drummond, A. J., Rambaut, A., Shapiro, B. & Pybus, O. G. Bayesian coalescentinference of past population dynamics from molecular sequences. Mol. Biol. Evol.22, 1185–1192 (2005).

LETTER RESEARCH


nature medicine VOLUME 16 | NUMBER 8 | AUGUST 2010 915

t e c h n i c a l r e p o r t s

Influenza prophylaxis would benefit from a vaccination method enabling simplified logistics and improved immunogenicity without the dangers posed by hypodermic needles. Here we introduce dissolving microneedle patches for influenza vaccination using a simple patch-based system that targets delivery to skin’s antigen-presenting cells. Microneedles were fabricated using a biocompatible polymer encapsulating inactivated influenza virus vaccine for insertion and dissolution in the skin within minutes. Microneedle vaccination generated robust antibody and cellular immune responses in mice that provided complete protection against lethal challenge. Compared to conventional intramuscular injection, microneedle vaccination resulted in more efficient lung virus clearance and enhanced cellular recall responses after challenge. These results suggest that dissolving microneedle patches can provide a new technology for simpler and safer vaccination with improved immunogenicity that could facilitate increased vaccination coverage.

The effectiveness of influenza vaccination is limited by the quality and breadth of the immune response and the time required for vaccine delivery1. Traditional intramuscular (i.m.) injection requires hypodermic needles that can cause needle phobia and generate biohazardous waste. An advantageous immunization scenario would involve transdermal delivery of the vaccine with a device that promises increased vaccine immunogenicity, enhanced patient compliance via simple selfadministration and mass immunization, and elimination of hypodermic needles and their associated biohazardous waste.

This study presents dissolving microneedle patches to increase vaccine immunogenicity by targeting antigen delivery to skin. Microneedles are micronscale structures that painlessly pierce into the skin to administer vaccines in a minimally invasive and targeted manner2. The skin is a highly active immune organ containing a large population of resident antigenpresenting cells3. Human clinical studies have shown evidence for dose sparing of intradermal influenza vaccination compared to i.m. immunization, although some other studies have not4–7. Intradermal influenza vaccinations at full dose

(15 μg hemagglutinin antigen per strain) and reduced dose (9 μg hemagglutinin per strain) have recently been licensed for human use in some countries (for example, Intanza and IDflu, Sanofi Pasteur). Widespread use of intradermal immunization has been limited by traditional intradermal injections that use the Mantoux technique, which requires specifically trained personnel and is often unreliable8. Needlefree transdermal patches have been reported, but the skin’s outer layer (stratum corneum) must be disrupted for delivery of large vaccine molecules9. In contrast, microneedles are designed to reliably administer antigen at a specific skin depth that maximizes interaction with resident antigenpresenting cells.

Previous studies show that nondissolving metal and silicon microneedle patches can be painless10 and can effectively administer vaccine in animals11,12, including the influenza vaccine13–15. Watersoluble microneedles have been shown to encapsulate bioactive molecules and deliver their cargo into skin16–19, but vaccination using this approach has not been studied before.

In this study, we compare standard i.m. immunization to vaccination with polymer microneedles that dissolve within minutes and completely resorb in the skin, resulting in no biohazardous sharps. We show that a single vaccine dose with dissolving microneedles induces protective immune responses superior to those obtained with i.m. injection at the same dose, including increased lung viral clearance. Dissolving microneedles also offer additional benefits, both to the individuals vaccinated and in regard to logistics, including small storage and disposal size, inexpensive fabrication and ease of use to enable selfadministration at home.

RESULTSDesign and fabrication of dissolving polymer microneedlesWe designed the polymer material, microneedle geometry and device fabrication process to encapsulate influenza virus while preserving its antigenicity, to insert into skin without mechanical failure and to rapidly dissolve into safe dissolution products. The resulting microneedles measured 650 μm tall with sharp tips tapering to a 10μm radius of curvature (Fig. 1a) and were assembled into a multineedle array (Fig. 1b) that encapsulated 3 μg of inactivated influenza virus vaccine per patch.

Dissolving polymer microneedle patches for influenza vaccinationSean P Sullivan1,4, Dimitrios G Koutsonanos2,4, Maria del Pilar Martin2, Jeong Woo Lee3, Vladimir Zarnitsyn3, Seong-O Choi3, Niren Murthy1, Richard W Compans2, Ioanna Skountzou2 & Mark R Prausnitz1,3

1Wallace H. Coulter Department of Biomedical Engineering at Emory University and Georgia Tech, Georgia Institute of Technology, Atlanta, Georgia, USA. 2Department of Microbiology & Immunology and Emory Vaccine Center, Emory University School of Medicine, Atlanta, Georgia, USA. 3School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta, Georgia, USA. 4These authors contributed equally to this work. Correspondence should be addressed to M.R.P. ([email protected]) or I.S. ([email protected]).

Received 30 July 2009; accepted 23 April 2010; published online 18 July 2010; doi:10.1038/nm.2182

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

http://www.nature.com/naturemedicine/

http://www.nature.com/doifinder/10.1038/nm.2182


916 VOLUME 16 | NUMBER 8 | AUGUST 2010 nature medicine

We fabricated the microneedles by roomtemperature (23 °C) photopolymerization of a liquid monomer (vinyl pyrrolidone) within a microneedle mold to form polyvinylpyrrolidone (PVP) microneedles that encapsulate the lyophilized vaccine. This process avoids the need for organic solvents or elevated temperatures that can damage vaccine or other biomolecule stability. We chose PVP as the structural material for the polymer microneedles used in this study because it is biocompatible, mechanically strong and highly water soluble20.

Insertion and dissolution of microneedles in skinThe resulting microneedles were able to be inserted into porcine skin with gentle force applied by the thumb (Fig. 1c). We determined the fracture force of the microneedles to be 0.13 ± 0.03 N per needle, which provides a twofold margin of safety over the force (0.058 N per needle) required for insertion into skin using microneedles of this geometry, according to previous measurements21. Upon insertion into porcine cadaver skin, microneedles penetrated to a depth of approximately 200 μm and deposited their encapsulated payload largely within the epidermis (Fig. 1d,e). This localization is likely to be similar in human skin, which has comparable thickness to porcine skin22.

To characterize the kinetics of dissolution in skin, we inserted microneedles into porcine skin and monitored them over time. Significant dissolution occurred within 1 min, and after 5 min the microneedles were 89 ± 3% (by mass) dissolved (Fig. 2a). Given the similarity of porcine and human skin, we expect that microneedle dissolution in human skin could also be complete within just a few minutes. Because we used mouse skin for the in vivo vaccination experiments described below, we also measured the dissolution kinetics of dissolving microneedles encapsulating the viral antigen in mice. In this scenario, microneedle dissolution was slower but nonetheless increased with time (P < 0.05), depositing 34 ± 17%, 63 ± 10% and

83 ± 6% of the polymer in the skin after 5, 10 and 15 min, respectively, and leaving almost no residue on the skin surface (Fig. 2b).

Antigen stabilityTo assess the stability of the inactivated influenza vaccine in dissolving microneedles, we identified two steps during the fabrication of PVP microneedles that might cause damage: the initial lyophilization of vaccine and the subsequent encapsulation within microneedles during polymerization.

To analyze the individual effects of lyophilization and PVP, we administered inactivated influenza virus i.m. in mice as the original vaccine solution, after lyophilization, as the original vaccine solution mixed with PVP and after lyophilization and encapsulation within PVP microneedles. Compared to naive mice, all four vaccinated groups showed elevated influenzaspecific IgG titers and hemagglutination

250 µm

a

1 mm

c

2 mm

b

200 µm

d

200 µm

e

Figure 1 Dissolving polymer microneedle patches. (a) Side view of dissolving polymer microneedles. (b) Relative height of an array of microneedles next to a US nickel coin. (c) En face view of porcine cadaver skin after insertion and removal of microneedles, showing delivery of the encapsulated compound (sulforhodamine). (d) Fluorescence micrograph of pig skin histological section after insertion of dissolving microneedles ex vivo. (e) Brightfield micrograph of the same skin section with H&E staining.

1200 min

200 µm

200 µm

200 µm

1 min

5 min

5 m

in

inser

tions

Per

cent

age

ofsu

lforh

odam

ine

dose

10 m

in

inser

tions

15 m

in

inser

tions

100

80

60

40

20

4,000 128

0

a b

c3,000

2,000

1,000

N NLyo.

Lyo.

Lyo.

+ P

VP

Lyo.

+ P

VP

Unpro

c. +

PVP

Unpro

c. +

PVP

Unpro

c.

Unpro

c.0

64 HA

I titers

IgG

(ng

ml–1

)

32

16

8

On skin surfaceIn skin In needle

Figure 2 Delivery to skin using microneedles. (a) Polymer microneedle dissolution in pig skin ex vivo. Top, before insertion; middle, remaining polymer 1 min after insertion in skin; bottom, remaining polymer 5 min after insertion in skin. (b) Dissolving microneedle delivery efficiency to mice in vivo. Sulforhodamine was encapsulated within microneedles and administered to mice (n = 5 for each time point). The delivery efficiencies for the three time points were statistically different from one another (Student’s t test, P < 0.05). (c) Effect of PVP and lyophilization on vaccine immunogenicity. Mice (n = 3) were immunized i.m. with 20 μg inactivated influenza virus (A/PR/8/34) that was either lyophilized or in solution with or without PVP added. Serum IgG antibody titers and HAI were measured 14 d after immunization. Unproc., unprocessed inactivated influenza virus in PBS; Lyo., lyophilized inactivated influenza virus redissolved in PBS; Unproc. + PVP, unprocessed inactivated influenza virus in PBS mixed with PVP; Lyo. + PVP, lyophilized inactivated influenza virus encapsulated in PVP; N, naïve mice. Error bars represent s.d. from three to five independent experiments.

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.



inhibition (HAI) titers (Fig. 2c, P < 0.01). Among the four vaccinated groups, there was no significant effect of vaccine processing or formulation on IgG or HAI titers (P > 0.05).

Humoral immune responsesThe efficacy of skin immunization with dissolving microneedles was determined in BALB/c mice that received a single dose of 6 μg of whole, encapsulated, inactivated influenza virus. The microneedle patches were applied on the caudal dorsal area of skin for approximately 15 min, which was sufficient to dissolve the microneedles and deliver at least 80% of the antigen into skin. We compared induction of humoral immune responses after i.m. immunization, which is the standard influenza vaccination method, with those generated using dissolving microneedles at the same vaccine dose (Fig. 3a–d). We collected blood on days 14 and 28 after immunization to determine the concentrations of influenzaspecific antibodies. Mice immunized with microneedles showed slightly lower influenzaspecific IgG titers than the i.m. group by day 14 (Fig. 3a, P < 0.0009). Titers were at similar levels for both i.m. and microneedle groups at day 28 (P = 0.9).

We also determined the concentrations of influenzaspecific isotypes, IgG1 and IgG2a, at 14 and 28 d after immunization. At day 14, microneedleimmunized mice had more pronounced IgG1 titers than the i.m. group (Fig. 3b, P = 0.03), whereas the i.m.immunized mice showed significantly stronger IgG2a responses than the microneedle group (Fig. 3c, P = 0.0006). At day 28 there were no significant differences in the isotype levels between the groups. This indicates that the i.m. group had T helper type 1 (TH1)biased responses early after immunization (IgG1/IgG2a ratio = 0.2), but levels of these isotypes were similar after 1 month (IgG1/IgG2a = 0.9). In contrast, the microneedle group showed a slight predominance of IgG1 production over time (IgG1/IgG2a in the range of 1.35 to 1.53) (Fig. 3b,c).

HAI activity is generally used as the serological measure for functional antibodies associated with protection. We observed high HAI titers after one immunization (Fig. 3d). HAI titers detected in the microneedle group were similar to each other on days 14 and 28 and to i.m. group titers too (Fig. 3d), demonstrating that a single microneedle immunization induced high levels of functional antibodies.

Protection against lethal viral challengeTo determine whether microneedle immunization can confer protective immunity, we challenged the immunized groups with five times

the halfmaximal lethal dose (LD50) of mouseadapted PR8 influenza virus 30 d after vaccination. All immunized animals survived challenge (Fig. 3e) and lost <5% body weight (Fig. 3f), showing that vaccine delivery with dissolving microneedles provided protection equal to the i.m. group. In contrast, the unimmunized group did not survive beyond 6 d after challenge (Fig. 3f).

We then investigated the ability of challenged mice to clear influenza virus from the lung 90 d after vaccination to assess longevity and efficiency of recall responses. On day 4 after challenge, the i.m.immunized mice showed a decrease in lung viral titers of a factor of 1 × 103 compared to unimmunized infected mice, whereas microneedleimmunized mice showed a marked decrease in lung viral titers of a factor of 1 × 106 (Fig. 4a). As the challenge of the mice took place three months after vaccination, these findings indicate that microneedle immunization induced more robust recall responses than i.m. vaccination, as shown by more efficient virus clearance.

Recall immune responsesTo evaluate the induction of local immune responses, we measured influenzaspecific IgG and IgA titers in lungs of challenged mice 90 d after immunization. We found that soluble IgA titers were modestly increased in vaccinated groups and were similar among microneedle and i.m. groups (Fig. 4b). Lung IgG titers were also similar in microneedle and i.m.immunized mice, including IgG1 and IgG2a isotype profiles (Fig. 4c). Systemically, we observed that challenged mice had serum influenzaspecific IgG titers similar to those observed 28 d after immunization, with no significant differences among immunized groups (Fig. 4d). Serum HAI titers were also similar in all immunized challenged groups, consistent with total antibody levels (Fig. 4e). Although we noted an increase in IgG1 titers after infection in vaccinated mice, microneedleimmunized mice had a higher IgG1/IgG2a ratio than the i.m. group, as observed in prechallenge samples (Fig. 4f). Thus, changes in antibody levels were consistent with protective responses in immunized mice. Overall, these data demonstrate that microneedle vaccination induces similar antibody recall responses compared to i.m. vaccination.

Antibodysecreting cells (ASCs) are partly responsible for recall immune responses that confer protection against influenza infection. We examined mice challenged 90 d after immunization for influenza IgG ASCs in spleen and lungs on day 4 after infection. In spleen, ASC numbers were elevated in both the microneedle and

Per

cent

age

surv

ival

Per

cent

age

ofB

W lo

ss

1101009080706050403020

0 1 2 3 4 5 6 7Time after challenge (d) Time after challenge (d)

8 9 10 11 12 13 14

100

1101051009590858075706560

Inf.IMMN

Inf.IMMN

1 2 3 4 5 6 7 8 9 10 11 12 13 14

e f

Ser

um lg

G2a

(ng

ml–1

)

Day 14 Day 28

8,0006,0004,0002,000

50

25

0

P = 0.0006

c N IM MN

HA

I tite

rs(G

MT

+ 9

5% C

l)

Day 14 Day 28

4,096

1,024

256

64

16

4

d N IM MN20,00015,00010,000

Ser

um lg

G (

ng m

l–1)

5,000

500

250

0Day 14

N IM

P = 0.0009

MN

Day 28

a8,0006,0004,0002,000

50

25

0

Ser

um lg

G1

(ng

ml–1

)

P = 0.03P = 0.002

P = 0.0055NIMMN

Day 14 Day 28

b

Figure 3 Microneedle immunization studies. (a) Serum influenza-specific IgG titers 14 and 28 d after immunization. Mice (n = 12) were immunized i.m. with inactivated influenza virus (A/PR/8/34) or via a microneedle patch encapsulating the same amount of virus. (b–d) IgG1 titers (b), IgG2a titers (c) and HAI titers (d) on days 14 and 28. (e) Survival rates of immunized and naive mice upon lethal challenge with five times the LD50 of homologous influenza virus. (f) Percentage of body weight changes upon lethal challenge. N, naïve group; i.m., intramuscularly immunized group; MN, microneedle-immunized group; Inf., unimmunized challenged group. Data shown are means ± s.e.m. HAI titers are depicted as geometric mean titers (GMT) with 95% confidence interval (CI).

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.



i.m. groups; despite a lack of noticeable differences between groups, the microneedle group was the only one showing significantly higher numbers of ASCs than naive or infected mice (Fig. 4g, P < 0.05). In lungs, we observed that the microneedle and i.m. groups had three to five times higher ASC numbers than unimmunized infected or naive mice (Fig. 4h). These results suggest that a skin vaccination route using dissolving microneedles induces sustained humoral immune responses in lungs at least as strong as responses induced by i.m. immunization.

Induction of systemic cytokine responsesWe next investigated induction of cellular immune responses systemically upon challenge 90 d after immunization. We re stimulated splenocytes isolated from challenged mice on day 4 with hemagglutinin major histocompatibility complex (MHC) class I– and hemagglutinin MHC class II–restricted peptides or inactivated influenza virus for 48 h and 72 h to determine the contribution of CD4+ and CD8+ T lymphocytes secreting interleukin4 (IL4) and interferonγ (IFNγ) (Fig. 5a,b). IL4 secretion was higher in the i.m. group in the presence of class I or class II peptides, although increases were more prominent with class I, suggesting increased CD8+ T cellderived response (Fig. 5a). In contrast, levels of IFNγ secreted by CD8+ or CD4+ cells were two to threefold higher in the microneedle group when compared to i.m.injected mice (Fig. 5b). Naive mice did not show any differences in cytokine levels from unimmunized infected mice (data not shown). Elevated IFNγ concentrations in microneedleimmunized mice suggest that microneedle immunization generates strong TH1

and effector responses, which are necessary to support cytotoxic activity, events that are crucial for viral clearance23.

Assessment of cellular immune responses in lungsTo assess cellular immune responses elicited in the mucosal compartment, we restimulated lung cell suspensions in vitro with inactivated A/PR/8/34 influenza virus and assessed the amounts of IL21, IFNγ, tumor necrosis factorα (TNFα), and IL12 p70. IL21 is a pleiotropic cytokine known to upregulate genes associated with innate immunity and TH1 responses24, as well as regulating B cell isotype

108

107

106

105

104

103

102

101

100

1,500

250

500500

10,500

20,500

lgG2a lgG2algG1 lgG1lgG

N.

MN

MNIM

IMInf.

Inf.

N MN MNIM IMInf. Inf.N N

Day 4 after challenge Day 4 after challenge

P = 0.05P = 0.03

0 0

100

100

75

25

50

200

300

lgG

4 after chall.2814

0

500

1,000

Inf Inf.

Log 10

PF

U p

er g

lung

tiss

ue

lgG

isot

ypes

(ng

ml–1

in B

ALF

)

Spl

een

lgG

AS

C p

er 1

06 cel

ls

Lung

lgG

AS

C p

er 1

06 cel

ls

HA

I tite

rs

(GM

T +

95%

CI)

lgG

isot

ypes

(ng

ml–1

in s

erum

)

Ser

um lg

G1

/ lgG

2a r

atio

BA

LF lg

A (A

450n

m)

IM

IM

Inf.MN

NIM

Inf.MN

IMMN

N

MN IM MN

0.20

0 0

3.01,024

512256128

32

1248

16

64 2.0

1.0

1.5

2.5

0.5

0

0.05

0.15P = 0.0116

0.10

Time (d)

a b c d

fe g h

Figure 4 Long-lived immune responses. Mice immunized with influenza virus by microneedle or i.m. route of delivery were challenged with live homologous virus 90 d after immunization; bronchoalveolar lavage fluid (BALF) and sera were collected at 4 d after challenge. (a) Lung virus titers determined by plaque assay. PFU, plaque-forming units. (b) Lung IgA titers determined by quantitative ELISA. (c) Lung IgG titers and their isotypes. (d) Serum IgG titers and their isotypes. (e) Serum HAI titers after infection, determined as geometric mean titers (GMT) with 95% confidence intervals (CI). (f) Serum IgG1/IgG2a ratio on days 14 and 28 after immunization and on day 4 after challenge. (g) Influenza-specific IgG ASCs from splenocytes re-stimulated with inactivated influenza virus. (h) Lung influenza-specific IgG ASCs. Data shown are means ± s.e.m.

1,000

Spl

een

IL-4

(pg

ml–1

)Lu

ng IL

-21

(pg

ml–1

)Lu

ng T

NF

-α (p

g m

l–1)

Lung

IFN

-γ (p

g m

l–1)

Spl

een

IFN

-γ (p

g m

l–1)

Lung

IL-1

2 p7

0 (p

g m

l–1)

0

20,000

10,000

5,000

25,000

15,000

800

600

400

200

200

360

0

20

406060

160

260

0

100

200

300

400

500

200

0

20

6040

400600800

1,0001,200

100

150

50

0Inf. Inf.IMMN IMMNN N

0HAII HAII

P = 0.027

P = 0.016P = 0.0211

P = 0.004P = 0.0001

P = 0.0005

P = 0.02

MNIMInf.

MNIMInf.

Day 4 after challenge Day 4 after challenge

Day 4 after challengeDay 4 after challenge

Inf. IM IMMN MNN Inf.NDay 4 after challenge Day 4 after challenge

HAI HAI

a b

c

e f

d

Figure 5 Cellular immune responses after challenge. Cellular immune responses were determined in splenocyte cultures and lung suspensions. (a) IL-4 concentrations, determined from 72-h splenocyte culture after isolation from immunized and unimmunized mice at 4 d after challenge and re-stimulation with hemagglutinin (HA) MHC class I and II influenza-specific peptides. (b) IFN-γ concentrations in 72-h splenocyte cultures. (c) Lung IL-21 concentrations. (d) Lung IFN-γ concentrations. (e) Lung TNF-α concentrations. (f) Lung IL-12 p70 concentrations. Data shown are means ± s.e.m.

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.



class switching25. It also augments IFNγ production in vitro when combined with other cytokines26. We found that IL21 amounts in lungs of i.m.vaccinated mice were significantly higher than in other groups (Fig. 5c, P = 0.0211), with IFNγ production correspondingly upregulated in the same group (Fig. 5d). Unimmunized infected mice showed the highest IFNγ and TNFα concentrations (Fig. 5d,e), consistent with stronger inflammatory reaction in mice not protected by vaccination. Notably, both i.m. (P < 0.0001) and microneedle (P < 0.0005) groups had significantly higher IL12 p70 production than naive or infected groups, which correlates with the high INFγ, which was more prominent in the i.m. group (Fig. 5f).

Expression of IFNγ, IL12 p70 and IL21 induced after polyclonal restimulation in lung was higher in the i.m. compared to microneedle group, which suggests stronger local TH1 response in the microneedle group upon challenge. In contrast, influenza virus–specific MHC class I– and class II–restricted T cell responses were increased in the spleen of microneedleimmunized groups, indicative of increased recall CD4+ and CD8+ T cell responses systemically. The higher virusspecific IFNγ production in the microneedleimmunized group may reflect enhanced generation and maintenance of memory T cells that are responsible for the increased virus clearance observed in lungs when compared to the i.m. group. Overall, these data demonstrate that microneedle immunization can generate a robust cellular and humoral immune response similar to that observed with the conventional i.m. route, and they suggest that microneedle immunization can establish a sustained and broader immune response.

Comparison of dissolving polymer and metal microneedlesAs a final set of experiments, we compared the dissolving polymer microneedles used in this study to coated metal microneedles described previously13–15 by vaccinating mice with each of these microneedle technologies and measuring humoral and cellular immune responses after two weeks (Supplementary Data). Humoral immune responses were similar (Supplementary Fig. 1), but cellular responses differed (Supplementary Figs. 2 and 3), most notably shown through increased IL4 and IFNγ production from inguinal lymph node cells in response to inactivated influenza virus stimulation in mice vaccinated with dissolving polymer microneedles compared to coated metal microneedles. This result suggests that dissolving microneedles not only offer advantages over i.m. injection but may also represent an improvement over coated metal microneedles.

DISCUSSIONThis study aimed to evaluate use of a simple patchbased vaccination method designed to overcome the limitations of hypodermic needle injection, both in terms of targeting skin antigenpresenting cells and avoiding hypodermic needles27,28. We therefore designed, fabricated and analyzed a novel dissolving microneedle patch for skin vaccination. Because microneedles dissolve in skin’s interstitial fluid, there is no sharps waste, which makes dissolving microneedles impossible to reuse and thereby eliminates the risks of biohazardous sharps.

This new approach incorporates vaccine in a lyophilized form within the structural polymer material of the microneedle, thereby avoiding the need for reconstitution before administration. These polymer microneedles dissolve in the skin within minutes and are safely eliminated by the body, as evidenced by the historical use of PVP as a plasma expander29. The use of needles measuring just hundreds of microns in length not only eliminates pain10 and enables simple delivery through a thin patch, but also inherently targets antigen to the abundant antigenpresenting cells of skin’s epidermis and dermis3.

This study demonstrates that influenza vaccine delivery with dissolving microneedles can induce robust humoral and cellular immune responses after a single immunization with a low antigen dose that confers protective immunity against lethal viral challenge. Immunologic responses to microneedle vaccination were similar to those achieved by i.m. injection by some measures and were stronger by others. Overall, microneedle immunization yielded enhanced recall cellular immune responses, increased numbers of antibodysecreting cells and, notably, more efficient viral clearance.

Although it is possible that dissolving microneedles have strong immunogenicity because of an adjuvant effect caused by PVP, we believe that this is unlikely, because i.m. injection of inactivated virus with PVP did not enhance immune response compared to vaccination without PVP. It is also possible that skin flora are drawn into the skin during microneedle insertion and thereby serve as an adjuvant. We think this is also unlikely, as we carefully cleaned the skin before microneedle insertion and because hypodermic needle insertion for i.m. injection could similarly draw in skin flora.

Thus, dissolving microneedle patches may provide not only practical advantages compared to hypodermic needles but also better protective immunity. Similar reports in human studies have shown that intradermal immunization can induce primary immune responses that are equivalent to or surpass i.m. delivery of seasonal influenza vaccine, with possible dosesparing effects4–7. Although this study did not assess dose sparing, the key immunologic difference between vaccine delivery through dissolving microneedles versus i.m. immunization is the 1,000fold more efficient lung virus clearance after microneedle vaccination, which is expected to correlate with reduced morbidity and mortality. Of note, we observed this difference upon challenge 3 months after immunization, suggesting that microneedle immunization induces more robust recall immune responses.

These results may be due to higher numbers of antibodysecreting cells found in spleen and lungs of microneedleimmunized mice as well as enhanced cellular memory responses in spleens, as shown by increased IFNγ secretion after in vitro restimulation. Cellular immune responses may promote rapid viral clearance from lung and thereby decrease morbidity, for example, via preexisting CD8+ T cell–mediated immunity directed at peptides from conserved internal proteins of the influenza A virus30. The enhanced production of serum IgG1 antibodies after microneedle vaccination may also reflect the role of humoral immune responses that assist in effective virus clearance. These differences are probably due to the route of immunization, although antigen formulation, slower release kinetics and other features of the dissolving microneedle delivery system may also have a role.

Immunization via skin may target innate dendritic cell populations directly through lymphatics from proximal draining lymph nodes and simultaneously by activating the rich dendritic cell network that resides in skin. It is well established that the innate immune system has a pivotal role in adaptive immune responses31, possibly accounting for the differences we observe between dissolving microneedle patches and i.m. vaccination32,33. The early virus clearance from lungs that we observed may be the result of enhanced involvement and mobilization of innate and adaptive cell populations that induce broader humoral and cellular immune responses.

Overall, these results show that dissolving microneedle patches offer an attractive approach to administer influenza vaccine with improved safety, immunogenicity and logistical operations that may enable an increased population coverage for influenza vaccination. The dissolving microneedle vaccine patch developed in this study

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.



also provides a new platform technology for simple administration of other vaccines and medicines to skin without the need for hypodermic needles.

METHODSMethods and any associated references are available in the online version of the paper at http://www.nature.com/naturemedicine/.

Note: Supplementary information is available on the Nature Medicine website.

ACKNOWLeDGMeNtSThis study was carried out at the Emory Vaccine Center and the Georgia Tech Center for Drug Design, Development and Delivery and Institute for Bioengineering and Biosciences. The work was supported in part by US National Institutes of Health grants R01EB006369 and U01AI084579 and contract HHSN266200700006C. S.P.S. was a trainee supported by a fellowship from the US Department of Education Graduate Assistance in Areas of National Need program. M.d.P.M. was a trainee supported by contract HHSN266200700006C from the US National Institutes of Health–National Institute of Allergy and Infectious Diseases.

AUtHOR CONtRIBUtIONSS.P.S., D.G.K., M.d.P.M. and I.S. carried out most experimental studies; J.W.L. and V.Z. prepared microneedles and helped generate the Supplementary Data; S.O.C. prepared the molds used to fabricate microneedles; S.P.S., D.G.K., I.S. and M.R.P. designed the study and its analysis; S.P.S., I.S. and M.R.P. wrote the manuscript; and N.M., R.W.C., I.S. and M.R.P. supervised the project.

COMPetING FINANCIAL INteReStSThe authors declare competing financial interests: details accompany the fulltext HTML version of the paper at http://www.nature.com/naturemedicine/.

Published online at http://www.nature.com/naturemedicine/. Reprints and permissions information is available online at http://npg.nature.com/reprintsandpermissions/.

1. Centers for Disease Control and Prevention. Influenza activity—United States and worldwide, 2007–08 season. MMWR Morb. Mortal. Wkly. Rep. 57, 692–697 (2008).

2. Prausnitz, M.R., Mikszta, J.A., Cormier, M. & Andrianov, A.K. Microneedle-based vaccines. Curr. Top. Microbiol. Immunol. 333, 369–393 (2009).

3. Glenn, G.M. & Kenney, R.T. Mass vaccination: solutions in the skin. Curr. Top. Microbiol. Immunol. 304, 247–268 (2006).

4. Belshe, R.B. et al. Serum antibody responses after intradermal vaccination against influenza. N. Engl. J. Med. 351, 2286–2294 (2004).

5. Holland, D. et al. Intradermal influenza vaccine administered using a new microinjection system produces superior immunogenicity in elderly adults: a randomized controlled trial. J. Infect. Dis. 198, 650–658 (2008).

6. Van Damme, P. et al. Safety and efficacy of a novel microneedle device for dose sparing intradermal influenza vaccination in healthy adults. Vaccine 27, 454–459 (2009).

7. Hickling, J. & Jones, R. Intradermal Delivery of Vaccines: A Review of the Literature and the Potential for Development for Use in Low- and Middle-Income Countries. (Program for Appropriate Technology in Health, Ferney Voltaire, France, 2009).

8. Flynn, P.M. et al. Influence of needle gauge in Mantoux skin testing. Chest 106, 1463–1465 (1994).

9. Prausnitz, M.R. & Langer, R. Transdermal drug delivery. Nat. Biotechnol. 26, 1261–1268 (2008).

10. Gill, H.S., Denson, D.D., Burris, B.A. & Prausnitz, M.R. Effect of microneedle design on pain in human volunteers. Clin. J. Pain 24, 585–594 (2008).

11. Mikszta, J.A. et al. Improved genetic immunization via micromechanical disruption of skin-barrier function and targeted epidermal delivery. Nat. Med. 8, 415–419 (2002).

12. Widera, G. et al. Effect of delivery parameters on immunization to ovalbumin following intracutaneous administration by a coated microneedle array patch system. Vaccine 24, 1653–1664 (2006).

13. Koutsonanos, D.G. et al. Transdermal influenza immunization with vaccine-coated microneedle arrays. PLoS One 4, e4773 (2009).

14. Zhu, Q. et al. Immunization by vaccine-coated microneedle arrays protects against lethal influenza virus challenge. Proc. Natl. Acad. Sci. USA 106, 7968–7973 (2009).

15. Kim, Y.C. et al. Enhanced memory responses to seasonal H1N1 influenza vaccination of the skin with the use of vaccine-coated microneedles. J. Infect. Dis. 201, 190–198 (2010).

16. Miyano, T. et al. Sugar micro needles as transdermic drug delivery system. Biomed. Microdevices 7, 185–188 (2005).

17. Ito, Y., Yoshimitsu, J., Shiroyama, K., Sugioka, N. & Takada, K. Self-dissolving microneedles for the percutaneous absorption of EPO in mice. J. Drug Target. 14, 255–261 (2006).

18. Lee, J.W., Park, J.H. & Prausnitz, M.R. Dissolving microneedles for transdermal drug delivery. Biomaterials 29, 2113–2124 (2008).

19. Sullivan, S.P., Murthy, N. & Prausnitz, M.R. Minimally invasive protein delivery with rapidly dissolving microneedles. Adv. Mater. 20, 933–938 (2008).

20. Robinson, B.V. PVP: A Critical Review of the Kinetics and Toxicology of Polyvinylpyrrolidone (Povidone). (Lewis Publishers, Chelsea, Michigan, 1990).

21. Park, J.-H., Allen, M.G. & Prausnitz, M.R. Biodegradable polymer microneedles: fabrication, mechanics and transdermal drug delivery. J. Control. Release 104, 51–66 (2005).

22. Bronaugh, R.L., Stewart, R.F. & Congdon, E.R. Methods for in vitro percutaneous absorption studies II. Animal models for human skin. Toxicol. Appl. Pharmacol. 62, 481–488 (1982).

23. McGill, J. & Legge, K.L. Cutting edge: contribution of lung-resident T cell proliferation to the overall magnitude of the antigen-specific CD8 T cell response in the lungs following murine influenza virus infection. J. Immunol. 183, 4177–4181 (2009).

24. Strengell, M., Sareneva, T., Foster, D., Julkunen, I. & Matikainen, S. IL-21 up-regulates the expression of genes associated with innate immunity and TH1 response. J. Immunol. 169, 3600–3605 (2002).

25. Ozaki, K. et al. A critical role for IL-21 in regulating immunoglobulin production. Science 298, 1630–1634 (2002).

26. Zeng, R. et al. Synergy of IL-21 and IL-15 in regulating CD8+ T cell expansion and function. J. Exp. Med. 201, 139–148 (2005).

27. Miller, M.A. & Pisani, E. The cost of unsafe injections. Bull. World Health Organ. 77, 808–811 (1999).

28. Mitragotri, S. Immunization without needles. Nat. Rev. Immunol. 5, 905–916 (2005).

29. Ravin, H.A., Seligman, A.M. & Fine, J. Polyvinyl pyrrolidone as a plasma expander; studies on its excretion, distribution and metabolism. N. Engl. J. Med. 247, 921–929 (1952).

30. Doherty, P.C. & Kelso, A. Toward a broadly protective influenza vaccine. J. Clin. Invest. 118, 3273–3275 (2008).

31. Katsikis, P.D., Schoenberger, S.P. & Pulendran, B. Probing the ‘labyrinth’ linking the innate and adaptive immune systems. Nat. Immunol. 8, 899–901 (2007).

32. Kupper, T.S. & Fuhlbrigge, R.C. Immune surveillance in the skin: mechanisms and clinical consequences. Nat. Rev. Immunol. 4, 211–222 (2004).

33. Miller, L.S. & Modlin, R.L. Toll-like receptors in the skin. Semin. Immunopathol. 29, 15–26 (2007).

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.



nature medicinedoi:10.1038/nm.2182

ONLINE METHODSCells and virus stocks. MadinDarby canine kidney cells (American Type Culture Collection CCL 34, American Type Culture Collection) were maintained in DMEM (Mediatech) containing 10% FBS (Hyclone, ThermoFisher Scientific). Influenza virus stocks (A/PR/8/34, H1N1) were prepared, purified and inactivated as previously described34. Inactivated influenza virus suspensions in PBS were lyophilized using settings based on a prior study35 and described in the Supplementary Methods. Hemagglutination activity was determined using chicken red blood cells (LAMPIRE Biological Laboratories) as previously described36. The mouseadapted A/PR/8/34 strain was obtained by eight serial passages in lungs of BALB/c mice. The LD50 was calculated by the ReedMuench formula37, and viral titer was determined by plaque assay34.

Polymer microneedle fabrication and encapsulation of influenza vaccine. Dissolving polymer microneedles were created via in situ polymerization of liquid monomer within a microneedle mold, as described previously19. Briefly, a microneedle master structure was created via a lensbased, lithographic microfabrication process. A reusable inverse mold was created by pouring polydimethylsiloxane (184 Dow Corning) over the master structure, allowing it to cure overnight, and carefully peeling the resulting mold off the master structure. We then applied 100 μl of vinylpyrrolidone monomer (99%, SigmaAldrich), freeradical initiator azobisisobutyronitrile (1.0 mol%) and inactivated influenza virus (6 mg ml−1) to the mold surface and administered vacuum (−101 kPa) for 1–2 min to pull the solution into the microneedle mold and form the microneedles. Then, a second mixture of 100 μl of vinylpyrrolidone monomer and azobisisobutyronitrile initiator (without vaccine) was applied to the surface of the mold to form the patch backing. Finally, the system was placed under an ultraviolet lamp (100 W, 300 nm, BLAK RAY) to initiate photopolymerization. After 30 min, the PVP microneedle patch was carefully removed from the mold and stored in a desiccator for up to 30 d.

Antigen stability study. Initial studies were conducted to test the stability of the processed antigen. Four different vaccine preparations were administered i.m., as described below, to assess the effect of microneedle fabrication processes on antigen stability in comparison with naïve mice. For the first two groups, 100 μg untreated inactivated influenza virus was resuspended either alone or in combination with 83 mg of PVP in 1.0 ml water. For the third group, 100 μg lyophilized inactivated influenza virus was resuspended in 1.0 ml water. For the fourth group, 100 μg lyophilized inactivated influenza virus was encapsulated in a microneedle patch containing 83 mg PVP, which was dissolved in 1.0 ml water. Two weeks after immunization, sera were collected and tested for influenzaspecific IgG titers, as described below.

Immunizations. Female BALB/c mice (Charles River Laboratory) (11 mice per group, 6–8 weeks old) received a single dose of vaccine by microneedle

or i.m. immunization. For microneedle delivery, 2 d before immunization the mice were anesthetized with a ketamine and xylazine cocktail, the dorsal caudal surface was prepared and hair was removed as previously described34. Microneedles were manually inserted into the caudal site of the dorsal surface of the skin, left in place for 15 min and then removed. Immunization with 6 μg of vaccine was accomplished by inserting two arrays of microneedles at the same time, each encapsulating 3 μg of vaccine. The vaccine dose is reported as the mass of virus protein, which was composed of ~30% hemagglutinin protein. I.m. immunization was carried out by injecting 6 μg of the vaccine suspended in 50 μl of PBS into the upper quadrant of the gluteal muscle. Mouse studies were approved by the Emory University Institutional Animal Care and Use Committee.

Challenge of mice with influenza virus. To determine survival rates and immune responses after challenge, six mice per group were challenged 1 month after immunization by intranasal instillation of 50 μl (180 PFU) of live mouseadapted A/PR/8/34 virus and monitored for 14 d. For a control group, we included six unimmunized challenged mice. A weight loss exceeding 25% was used as the experimental end point, at which mice were killed. The challenged mice were monitored daily for signs of morbidity (body weight changes, fever and hunched posture) and mortality.

Characterization of immune response. As described in the Supplementary Methods, blood was collected 14 and 28 d after immunization to determine humoral immune responses (total IgG, IgG isotypes and HAI titers). Four days after challenge, blood was collected to determine humoral immune responses; spleens were collected to assay antibodysecreting cells and cytokine expressions levels, and lungs were collected to determine lung virus titers, IgG and IgA titers, antibodysecreting cells and cytokine expression levels.

Statistical analyses. The statistical significance of observed differences was calculated by twotailed unpaired Student’s t test and oneway analysis of variance, including Bonferroni’s multiple comparison test. Values were considered significant for P ≤ 0.05. Unless otherwise stated, data were pooled from at least two independent experiments.

34. Skountzou, I., Quan, F.S., Jacob, J., Compans, R.W. & Kang, S.M. Transcutaneous immunization with inactivated influenza virus induces protective immune responses. Vaccine 24, 6110–6119 (2006).

35. Amorij, J.P. et al. Rational design of an influenza subunit vaccine powder with sugar glass technology: preventing conformational changes of haemagglutinin during freezing and freeze-drying. Vaccine 25, 6447–6457 (2007).

36. Compans, R.W. Hemagglutination-inhibition: rapid assay for neuraminic acid-containing viruses. J. Virol. 14, 1307–1309 (1974).

37. Reed, L.J. & Muench, H. A simple method of estimating fifty per cent endpoints. Am. J. Hyg. 27, 493–497 (1938).

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

nature outlook influenza 2011

Documents

flu prevention

flu pandemic

influenza pandemic

flu catcherrichard webbys

unscathed page s14

birds page s4

catastrophe page s11

years flu strain