|Home | About | Journals | Submit | Contact Us | Français|
Secondary bacterial infections contribute to morbidity and mortality from influenza. Vaccine effectiveness is typically assessed using prevention of influenza, not secondary infections, as an endpoint. We vaccinated mice with formalin-inactivated influenza virus vaccine preparations containing disparate HA and NA proteins and demonstrated an ability to induce the appropriate anti-HA and anti-NA immune profiles. Protection from both primary viral and secondary bacterial infection was demonstrated with vaccine-induced immunity directed toward either the HA or the NA. This finding suggests that immunity toward the NA component of the virion is desirable and should be considered in generation of influenza vaccines.
Induction of neutralizing immunity toward the HA component is the goal of annual influenza vaccine campaigns (13), and changes in this protein dictate annual reformulations of influenza vaccines (8, 15, 31). In split-virion and -subunit vaccine preparations, there are HA and NA constituents, but immunity toward the NA component is rarely measured (18). Without consistent detection of immunity generated toward the NA (27), we are left unaware of the impact that annual vaccination has on immunity toward the NA component within the population (29).
Secondary bacterial infections, the majority of which are associated with Streptococcus pneumoniae (21, 23, 25, 32), make a significant contribution to deaths during influenza epidemics (5, 28, 30) and pandemics (1, 19, 22), through a phenomenon known as excess mortality. Population-based studies have indicated that vaccination against influenza virus greatly reduces the incidence of S. pneumoniae (4, 24), which would implicate vaccination against influenza as a method for limiting secondary bacterial complications. Evidence that prevention of influenza virus NA activity can significantly limit the severity of secondary bacterial infections (20) led us to hypothesize that matching the NA component in annual vaccines will limit secondary bacterial complications. To test this hypothesis, 6- to 8-week-old female BALB/cJ mice (Jackson Laboratories, Bar Harbor, ME) were vaccinated with formalin-inactivated influenza vaccines created using an 8-plasmid reverse genetics system as described previously (10, 11, 14, 15).
Viruses used to create these vaccines were generated using HA and NA genes from influenza virus strains A/Hong Kong/1/68 (HK; GenBank accession numbers AF348176 and AF348184, respectively) and A/Sydney/5/97 (SY; AF180584 and AJ291403, respectively) in combinations where the HA and NA genes were a match (HK/HK and SY/SY) and where the HA and NA were mismatched (HK/SY and SY/HK). As detailed previously, the HA gene for HK HA differed from the GenBank sequence at position N153I (A458T) (14). The remaining six genes (PB1, PB2, PA, NP, M, and NS) in all viruses created were derived from A/Puerto Rico/8/34 (kindly provided by Erich Hoffmann and Robert G. Webster). Rescued viruses were propagated in 10-day-old embryonated chicken eggs, concentrated, and purified over sucrose gradients, and the HA content was quantitated as described previously (14, 15). Mice were vaccinated with 3 μg HA in a 100-μl volume delivered intramuscularly (i.m.) in the right rear quadriceps two or three times at 4-week intervals, using 2 mg/ml alum (Reheis, Berkeley Heights, NJ) as an adjuvant (12).
Three weeks after inoculation, sera were collected from isoflurane-anesthetized mice via the orbital plexus and treated with receptor-destroying enzyme (RDE) as described previously (14, 15), and anti-HA antibodies were quantitated using a hemagglutination inhibition (HI) assay (Fig. (Fig.1).1). Chicken red blood cells (Rockland Immunochemicals, Inc., Gilbertsville, PA) were used for HI assays, and titers are reported as the final serum dilution that demonstrates inhibition of hemagglutination. As predicted, HI titers detected using viruses expressing HK HA were significantly increased when sera obtained from HK HA-vaccinated mice were compared to sera collected from SY HA-vaccinated mice. Similarly, SY HA-expressing viruses demonstrated increased HI titers when sera from SY HA-vaccinated mice were included in the HI assay, compared to sera from HK HA-vaccinated mice. Mice vaccinated with alum alone did not develop detectable levels of anti-HA immunity. Interestingly, after two inoculations, mice that received inactivated influenza virus (IIV) preparations containing the HK NA demonstrated significantly reduced HI titers compared to IIV preparations that contained the SY NA (data not shown). Mice in these groups received a third inoculation with IIV to enhance the immunity toward the HA component of these vaccines, and the reported results for all immune assays are from sera collected 1 week prior to inoculation with virus in our challenge model.
To assess serum reactivity toward the neuraminidase component of the individual vaccine preparations, we used a fetuin-based NA assay to detect inhibition of NA activity (Fig. (Fig.2).2). Sera collected from individual mice were analyzed for their ability to inhibit neuraminidase activity using a thiobarbituric acid assay, with fetuin substrate, as described previously (2, 3, 9, 33). Briefly, 50 μl of heat-inactivated, non-RDE-treated sera were mixed with 50 μl of each virus and 100 μl fetuin (Sigma, St. Louis, MO) and incubated for 18 h at 37°C. The colorimetric product of the reaction of free sialic acid with thiobarbituric acid was measured at 549 nm using a Shimadzu Biospec Mini spectrophotometer (Shimadzu Biotech, Manchester, United Kingdom). Samples were standardized based on the optical density (OD) values for the NA activity of each virus alone, and data are represented as the percent inhibition of NA activity. The NA activities of the four viruses used in this study were not significantly inhibited by sera obtained from mice vaccinated with alum, while the NA activities of viruses expressing SY NA were significantly inhibited by sera collected from mice vaccinated with IIV preparations that contained SY NA. In addition, the NA activities of viruses expressing HK NA were significantly inhibited by sera collected from mice vaccinated with HK NA-containing IIV preparations, but not with sera collected from mice vaccinated with IIV preparations that contained SY NA. Interestingly, NA activity of the HK/HK virus was inhibited by sera collected from mice in the SY/HK IIV group to a level that was significantly greater than that demonstrated by sera collected from HK/HK IIV vaccine recipients. Data from our immune assays demonstrate that IIV preparations used as vaccines in this study adequately induced immunity toward the individual HA and NA components contained within the specific IIV preparations.
One of the viruses (HK/SY) was chosen as the challenge strain for this study. Characteristics of this virus have been described previously (26). Viral challenges for these studies were with 100 μl of a sublethal dose of virus (0.1 50% lethal dose [LD50]; 105.25 50% tissue culture infective dose [TCID50]), and mice were subsequently monitored for weight loss (Fig. (Fig.3A).3A). The group that received alum alone demonstrated a significant reduction in body weight compared to mice vaccinated with IIV containing either an HA component that matched that of the challenge virus (HK HA) or an NA component that matched that of the challenge virus (SY NA). Similarly, mice that were vaccinated with the SY/HK IIV, which contained HA and NA components that did not match those expressed by the challenge virus, demonstrated a significant reduction in body weight compared to all other vaccine groups.
At day 6 after viral challenge, lung viral titers were measured (Fig. (Fig.3B)3B) through propagation in MDCK cell monolayers, as described previously (15), with a lung homogenate dilution range of 10−2 to 10−7. These results demonstrated detectable titers in the lungs of mice vaccinated with alum (4/4), HK/HK (2/5), and SY/HK (5/5). The group that received the mismatched HA and NA (SY/HK) from the influenza virus used for challenge (HK/SY) demonstrated a significant increase in lung virus titers compared to mice in the groups vaccinated with HK/SY and SY/SY IIV preparations. Together with the weight loss data, these data demonstrate that vaccination with either a matching HA (HK HA) or NA (SY NA) component limited the ability of the HK/SY influenza virus to establish a primary infection. Surprisingly, the group that received a mismatched HA with a matched NA (SY/SY) demonstrated lower overall viral titers than the group that received a matched HA but a mismatched NA (HK/HK).
On day 7 after inoculation with 0.1 LD50 HK/SY reassortant virus, mice were inoculated with 5 × 104 CFU S. pneumoniae (strain D39) intranasally (i.n.; 100 μl) (20, 21). Mice were monitored daily for signs of morbidity (weight loss) and mortality (Fig. (Fig.4)4) as described previously (14). Mice that were inoculated with alum demonstrated significantly reduced survival (25%) compared to the groups vaccinated with HK/SY (83%), HK/HK (100%), and SY/SY (86%). Mice that received IIV containing SY/HK demonstrated only 50% survival, which was significantly lower than that observed for the groups that received the HK/SY, HK/HK, and SY/SY IIV preparations. These data demonstrate that vaccination with a matching HA (HK HA) or NA (SY NA) component within an IIV preparation can limit progression toward secondary bacterial complications. Importantly, the NA-matched but HA-mismatched vaccine preparation (SY/SY) was equally effective at limiting secondary bacterial complications as vaccine preparations composed of matching HA constituents. Thus, a vaccine (SY/HK) containing HA and NA components that do not match the challenge virus (HK/SY), while capable of inducing the predicted anti-HA and anti-NA immune profiles, cannot prevent infection with the mismatched challenge virus and the subsequent progression toward secondary bacterial complications.
Current influenza vaccine efforts focus on stimulation of immunity toward HA, even though anti-NA immunity can provide protection against viral challenge (16-18). Through vaccination with IIV, we were able to stimulate the predicted immunity toward the appropriate vaccine component (HA or NA), which allowed us to demonstrate that immunity toward the NA component does indeed affect primary viral outcomes (with regard to disease progression and viral titers) and that this immunity can aid in the prevention of secondary bacterial infections. Interestingly, our data demonstrate that matching NA immunity alone (SY/SY vaccine group) was actually better than matching HA immunity alone (HK/HK group) at preventing virus titers in the lungs, and while this result is not expected to translate to humans, it provides further evidence that NA immunity is valuable.
Since NA content in vaccines is not standardized (29), and measurement of vaccine-induced anti-NA immunity is not currently monitored (7, 18), we cannot appreciate the impact that immunity toward this vaccine component has on limiting secondary complications. Recent evidence has demonstrated that individuals with adequate immunity directed toward the N1 component contained within seasonal vaccine preparations demonstrate cross-reactive immunity toward the N1 associated with avian influenza virus (H5N1) (29). Together with the results presented here, this further supports the idea that immunity toward the NA should not be ignored in future vaccine preparations (including recombinant HA or HA-DNA approaches) (6, 13), especially in the context of emerging influenza pandemics where immunity toward NA may be our only method for reducing the severity of secondary bacterial infections.
We thank Robert G. Webster and Erich Hoffmann for providing A/PR/8/34 plasmids used to create viruses by reverse genetics. We also thank the Hartwell Center for sequence analyses and the animal resource center at St. Jude Children's Research Hospital. This work was supported by the American Lebanese Syrian Associated Charities (ALSAC).
All animal experiments were conducted in accordance with guidelines established by the Animal Care and Use Committee at St. Jude Children's Research Hospital.
Published ahead of print on 3 February 2010.