Age-associated changes in the impact of sex steroids on influenza vaccine responses in males and females

Abstract

Vaccine-induced immunity declines with age, which may differ between males and females. Using human sera collected before and 21 days after receipt of the monovalent A/Cal/09 H1N1 vaccine, we evaluated cytokine and antibody responses in adult (18-45 years) and aged (65+ years) individuals. After vaccination, adult females developed greater IL-6 and antibody responses than either adult males or aged females, with female antibody responses being positively associated with concentrations of estradiol. To test whether protection against influenza virus challenge was greater in females than males, we primed and boosted adult (8-10 weeks) and aged (68-70 weeks) male and female mice with an inactivated A/Cal/09 H1N1 vaccine or no vaccine and challenged with a drift variant A/Cal/09 virus. As compared with unvaccinated mice, vaccinated adult, but not aged, mice experienced less morbidity and better pulmonary viral clearance following challenge, regardless of sex. Vaccinated adult female mice developed antibody responses that were of greater quantity and quality and more protective than vaccinated adult males. Sex differences in vaccine efficacy diminished with age in mice. To determine the role of sex steroids in vaccine-induced immune responses, adult mice were gonadectomized and hormones (estradiol in females and testosterone in males) were replaced in subsets of animals before vaccination. Vaccine-induced antibody responses were increased in females by estradiol and decreased in males by testosterone. The benefit of elevated estradiol on antibody responses and protection against influenza in females is diminished with age in both mice and humans.

Keywords: Humoral immunity; Influenza virus.

Fig. 1

Adult females of reproductive ages have greater seroconversion to a monovalent A/Cal/09 H1N1 vaccine, which is associated with concentrations of estradiol. Adult (18–45 years) and aged (65+ years) males (dark or light blue, respectively) and females (dark or light pink, respectively) were vaccinated with an inactivated, monovalent 2009 H1N1 vaccine and serum was collected prior to and at day 42 (i.e., 21 days after boost) (a). Seroconversion of hemagglutination inhibition titers after vaccination (b) and the proportion of individuals with ≥4-fold increase in HAI antibody titers after vaccination (i.e., rate of seroconversion) (c) were measured. Seroconversion of neutralizing antibody titers after vaccination (d) and the rate of seroconversion after vaccination (e) was measured. Serum concentrations of estradiol (E2) and testosterone (T) were measured (f, h) and the correlation between hormone concentrations and neutralizing antibody seroconversion was quantified using a linear regression model (g, i). Data represent mean ± standard error of the mean and significant differences between groups are denoted by asterisks (*p < 0.05) based on one-way ANOVAs (b, d), Chi square analyses (c, e), or t-tests (f, h). n = 20 adult males, 30 adult females, 47 aged males, and 48 aged females

Fig. 2

Adult female mice have greater antibody responses to an inactivated 2009 H1N1 influenza vaccine than their male counterparts, which is mitigated in aged mice. Adult (8–10 weeks) and aged (68–70 weeks) male (dark or light blue, respectively) and female (dark or light pink, respectively) mice were vaccinated with an inactivated ma2009 H1N1 vaccine and plasma was collected on days 28 and 35 (i.e., 7 and 14 days post boost) (a). Hemagglutination inhibition (HAI) titers (b), neutralizing antibody titers (c), anti-2009 H1N1 IgG (d), anti-2009 H1N1 IgG1 (e), anti-2009 H1N1 IgG2c (f), the ratio of IgG1/IgG2 (g), and antibody avidity (h) responses were measured. Data represent mean ± standard error of the mean from two independent replications (n = 9–10/group) and significant differences between groups are denoted by asterisks (*p < 0.05) based on two-way ANOVAs (b–f) or one-way ANOVAs (g, h)

Fig. 3

Sex differences in influenza vaccine-induced protection is more pronounced in adult than aged mice. Adult (8–10 weeks) and aged (68–70 weeks) male (dark or light blue, respectively) and female (dark or light pink, respectively) mice were vaccinated with an inactivated ma2009 H1N1 vaccine and 6 weeks post vaccination challenged with an ma2009 drift variant virus (a). To evaluate morbidity, the percent change in body mass was measured daily for 14 days post challenge (n = 9–10/group) (b). Lung virus titers were measured on days 3 and 5 post challenge (n = 9–10/group) (c), and cumulative inflammation scores from H&E stained lung sections were quantified prior to, and on days 3 and 5 post challenge (d, e) (n = 4–5/group). The concentrations of proinflammatory cytokines were assessed prior to and 3 and 5 days post challenge, and the induction (i.e., fold change) from baseline of TNFα (e), IL-6 (f), CXCL1 (g), IFNγ (h), and IL-10 (i) are presented. Data represents means ± standard error of the mean from two independent replications, and significant differences between groups are denoted by asterisks (*p < 0.05) based on a repeat measures ANOVA (b) or two-way ANOVAs (d–i). Significant differences in morbidity (b) between adult males and females are denoted by asterisks (*), between adult and aged females by pound sign (#), and between adult and aged males by plus sign (+); p < 0.05 based on a repeat measures ANOVA

Fig. 4

Estradiol in female mice increases, whereas testosterone in male mice decreases, antibody titers following vaccination. Adult (8–10 weeks) and aged (68–70 weeks) male (closed, opened, or stippled blue) and female (closed, open, or stippled pink) mice were vaccinated with an inactivated ma2009 H1N1 vaccine and 6 weeks post vaccination plasma estradiol and testosterone levels were measured (a, b) and correlated with neutralizing antibody responses quantified by a linear regression model (c, d) (n = 10/group). Female and male mice were gonadectomized (gdx), implanted with either estradiol (females, gdx + E2), testosterone (males, gdx + T) or placebo capsules, and provided 1 week to recover from surgery prior to vaccination with an inactivated ma2009 H1N1 vaccine (e). Plasma anti-2009 H1N1 IgG (f) and neutralizing antibody (g) titers were measured in gonad intact, gdx, gdx + T, and gdx + E2 mice at 35 days post vaccination (n = 5/treatment group). Data represent means ± standard error of the mean from two independent replications and significant differences between groups are denoted by asterisks (*p < 0.05) based on t-tests (a, b), regression analyses (c, d), or one-way ANOVAs (f, g)