Broad influenza-specific CD8+ T-cell responses in humanized mice vaccinated with influenza virus vaccines

Abstract

The development of novel human vaccines would be greatly facilitated by the development of in vivo models that permit preclinical analysis of human immune responses. Here, we show that nonobese diabetic severe combined immunodeficiency (NOD/SCID) beta(2) microglobulin(-/-) mice, engrafted with human CD34+ hematopoietic progenitors and further reconstituted with T cells, can mount specific immune responses against influenza virus vaccines. Live attenuated trivalent influenza virus vaccine induces expansion of CD8+ T cells specific to influenza matrix protein (FluM1) and nonstructural protein 1 in blood, spleen, and lungs. On ex vivo exposure to influenza antigens, antigen-specific CD8+ T cells produce IFN-gamma and express cell-surface CD107a. FluM1-specific CD8+ T cells can be also expanded in mice vaccinated with inactivated trivalent influenza virus vaccine. Expansion of antigen-specific CD8+ T cells is dependent on reconstitution of the human myeloid compartment. Thus, this humanized mouse model permits preclinical testing of vaccines designed to induce cellular immunity, including those against influenza virus. Furthermore, this work sets the stage for systematic analysis of the in vivo functions of human DCs. This, in turn, will allow a new approach to the rational design and preclinical testing of vaccines that cannot be tested in human volunteers.

Figures

Figure 1

The expansion FluM1-specific CD8+ T cells in LAIV-vaccinated humanized mice. (A) LAIV vaccine was given intraperitoneally/intravenously; TT vaccine was used as the control. (B) At day 12 after vaccination, the frequency of CD8+ T cells binding FluM1-tetramer and HIVgag-tetramer in the blood was analyzed by flow cytometry. Analysis gates are set for high intensity tetramer staining. Representative data from an experiment with 3 mice vaccinated with LAIV or TT. (C) The frequency of FluM1-specific CD8+ T cells in the blood at day 12 after vaccination from 4 experiments with 4 different donors. Two-tailed nonparametric Mann-Whitney test. (D) The frequency of CD8+ T cells binding FluM1-tetramer and HIVgag-tetramer in the spleen and lungs at day 14 after vaccination. (E) The frequency of FluM1-tetramer+ CD8+ T cells in the blood correlates with their frequency in the spleen (Spearman correlation, P < .001) and (F) in the lung (Spearman correlation, P < .001).

Figure 2

The expansion of FluM1-specific CD8+ T cells in TIV-vaccinated humanized mice. Experiment as in Figure 1, except that mice received TIV vaccine. (A) Specific CD8+ T cells binding FluM1-tetramer and HIVgag-tetramer by flow cytometry. Representative mice from an experiment with 4 mice vaccinated with TIV. (B) The frequency of FluM1-specific CD8+ T cells in the blood at day 12 after vaccination from 5 experiments with 4 different donors. Two-tailed nonparametric Mann-Whitney test. (C) CD8+ T cells in the spleen and lungs at day 14 after vaccination.

Figure 3

Cross-presentation of FluM1 in humanized mice. Experimental protocols: (A) NOD/SCID β2m−/− mice were reconstituted with 20 × 106 sorted T cells and with 10 × 106 sorted B cells (purity > 99%) and vaccinated with TIV intraperitoneally/intravenously. (B) NOD/SCID β2m−/− mice were transplanted with CD34+ HPCs and reconstituted with 20 × 106 autologous sorted total T cells (purity > 99%) before vaccination. TIV was given intraperitoneally/intravenously. (C,D) The frequency of FluM1-specific CD8+ T cells in NOD/SCID β2m−/− mice (C) or (D) in humanized mice at day 12 after vaccination. (E) The frequency of FluM1-specific CD8+ T cells (mean ± SEM, n = 3 for each cohort at each time point) at different time points after vaccination. (F) The numbers of total human CD8+ T cells (mean ± SEM, n = 3 for each cohort at each time point) measured in the same volume of blood at different time points after vaccination.

Figure 4

The breadth and effector phenotype of elicited influenza-specific CD8+ T cells. Vaccination as indicated in figure panels. Single-cell suspensions from tissues harvested at day 14 after vaccination were stimulated for 8 hours with indicated peptides and antibodies against CD28 and CD49d. The frequency of IFN-γ–secreting CD8+ T cells by flow cytometry. (A-D) Spleen analysis. (A) Representative experiment. (B) IFN-γ–secreting CD8+ T cells (mean ± SEM, n = 3) after stimulation with indicated peptides. (C) Analysis after vaccination with TIV or TT. (D) The frequency of IFN-γ–secreting CD8+ T cells (mean, n = 4) specific to indicated peptides. (E) Blood: specific CD8+ T cells binding FluM1 and NS1 tetramer in humanized mice vaccinated with LAIV or TIV. (F) Blood and lungs of TIV or TT-vaccinated mice. IFN-γ and CD107a expression by CD8+ T cells in response to HIVgag or FluM1 peptides.

Figure 5

Intranasal vaccination with LAIV permits expansion of FluM1-specific CD8+ T cells. (A) Blood of mice vaccinated with LAIV intraperitoneally/intravenously (i.v./i.p.) as in Figure 1 or intranasally (i.n.). Frequency of CD8+ T cells binding FluM1 tetramer at day 12 after vaccination. (B) Flu-M1 specific CD8+ T cells in draining and control lymph node (LN) suspension pooled from 3 mice; blood, spleen, and lung from representative mouse.