NOD/scid IL-2Rg(null) mice: a preclinical model system to evaluate human dendritic cell-based vaccine strategies in vivo

Abstract

Background: To date very few systems have been described for preclinical investigations of human cellular therapeutics in vivo. However, the ability to carry out comparisons of new cellular vaccines in vivo would be of substantial interest for design of clinical studies. Here we describe a humanized mouse model to assess the efficacy of various human dendritic cell (DC) preparations. Two reconstitution regimes of NOD/scid IL2Rg(null) (NSG) mice with adult human peripheral blood mononuclear cells (PBMC) were evaluated for engraftment using 4-week and 9-week schedules. This led to selection of a simple and rapid protocol for engraftment and vaccine evaluation that encompassed 4 weeks.

Methods: NSG recipients of human PBMC were engrafted over 14 days and then vaccinated twice with autologous DC via intravenous injection. Three DC vaccine formulations were compared that varied generation time in vitro (3 days versus 7 days) and signals for maturation (with or without Toll-like receptor (TLR)3 and TLR7/8 agonists) using MART-1 as a surrogate antigen, by electroporating mature DC with in vitro transcribed RNA encoding full length protein. After two weekly vaccinations, the splenocyte populations containing human lymphocytes were recovered 7 days later and assessed for MART-1-specific immune responses using MHC-multimer-binding assays and functional assessment of specific killing of melanoma tumor cell lines.

Results: Human monocyte-derived DC generated in vitro in 3 days induced better MART-1-specific immune responses in the autologous donor T cells present in the humanized NSG mice. Moreover, consistent with our in vitro observations, vaccination using mature DC activated with TLR3 and TLR7/8 agonists resulted in enhanced immune responses in vivo. These findings led to a ranking of the DC vaccine effects in vivo that reflected the hierarchy previously found for these mature DC variations in vitro.

Conclusions: This humanized mouse model system enables comparisons among different DC vaccine types to be rapidly assessed in vivo. In addition, ex vivo analyses of human CD3+ T cells recovered from the spleens of these mice are also possible, including studies on lymphocyte subsets, Th1/Th2 polarization, presence of regulatory T cells and the impact of DC vaccination on their functions.

Figure 1

Efficacy of engraftment of human PBMC in NSG mice. (A) Schema of 9-wk engraftment protocol. Mice were irradiated 24 h prior to i.v. injection of 1 × 10⁶ human PBMC and spleens were isolated on day 63. (B) Schema of 4-wk engraftment protocol starting with i.v. injection of 10 × 10⁶ human PBMC on day 1. Isolation of spleens was performed on day 28. (C) Representative staining of CD3 vs. CD45 in blood withdrawals of individual mice engrafted with the 9-wk or the 4-wk protocol on week 2, 3 and 4. Italic numbers given beside the CD3/CD45-positive population indicate the percentage of CD62L-positive cells within the double-positive fraction.

Figure 2

Phenotypic and functional analysis of differently matured DC. (A) Percent of CD83-positive cells measured on 7-day, 4C-matured DC (n = 6) as well as 3-day, 4C- and 5C+R848-matured DC (n = 4). Percent of CD14-positive cells ranged between 0.9 and 5.7 percent in the tested populations. (B) Ratio of secreted IL-12(p70) to IL-10 assessed by standard ELISA of signal-3-assay supernatants. Analyzed DC were matured using the 7-day, 4C protocol (n = 7) or the 3-day protocol in combination with 4C or 5C+R848 mediated maturation (n = 4). (C) Ratio of the mean fluorescence intensity of CD80 to CD274 (B7-H1) assessed on 7-day, 4 C generated mDC (n = 3) as well as 3-day DC matured with 4C or 5C+R848 cocktails (n = 6). Significance was ascertained using a two-tailed Mann-Whitney U test and p-values were defined as following: ** = p < 0.005; * = p < 0.05; n.s. = p > 0.05.

Figure 3

Influence of vaccination on the engraftment efficacy of human PBMC. (A) Schema of 9-wk engraftment protocol in which vaccination was performed using 1 × 10⁶ mDC given on day 42 and day 49. Spleens were isolated on day 63. (B) Schema of 4-wk engraftment protocol using injection of 1 × 10⁶ mDC on day 14 and day 21. Spleens were isolated on day 28. (C) Representative examples for CD3 vs. SSC, CD4 vs. CD62L and CD8 vs. CD62L (from left to right) of flow cytometry analysis comparing engraftment efficacy of 9-wk (left of each block) and 4-wk (right of each block) schemes. (D) Statistical analysis of engraftment efficacy comparing 9-wk and 4-wk protocols and mice vaccinated using 7-day 4C-matured DC (circles), 3-day 4C-matured (squares) or 3-day 5C+R848-matured (triangles) DC; means are given in brackets. Statistical analyses were performed using a two-tailed Mann-Whitney test and p < 0.05 were considered significant. Engraftment experiments were performed at least 3 times with similar results. (E) Amounts of human IFN-γ in sera of vaccinated mice engrafted with the 9-wk or 4-wk protocol, respectively, were assessed using a standard ELISA. Depicted are means and SEM of 8 mice generated in two individual experiments. The open square and open circle represent levels in individual non-vaccinated mice.

Figure 4

In vivo priming using DC-based vaccination. (A) MART-1-specific multimer staining versus CD8 staining of in vivo-primed human T cells. Differently matured DC of an HLA-A2⁺ donor, electroporated with MART-1 ivt-RNA, were used for vaccination. Staining was performed one day after spleen isolation. (B) Killing capacity (% specific lysis) of human lymphocyte populations, shown in A, tested individually in a chromium-release assay 24 h after isolation (ex vivo). 2 × 10³ mel624.38 target cells were incubated with varying numbers of effector cells. Specific lysis of non-immunized mice is shown as open circles while filled circles represent specific lysis of lymphocytes from immunized mice. THP-1 cells (HLA-A2⁺, MART-1^-) cells were not recognized (data not shown). (C) Killing capacity (% specific lysis) human lymphocyte populations cultured in vitro after isolation from individual mice and tested separately in a chromium-release assay at day 7 after isolation. Responses of non-immunized mice are shown as open circles while filled circles represent responses of immunized mice. Individual mice shown in (A) are indicated with * and the analyzed lymphocyte populations correspond to the populations tested in B. THP-1 (HLA-A2⁺, MART-1^-) and K562 (HLA-A2^-, MART-1^-) cells were not recognized (data not shown). (D) Amount of secreted IFN-γ by human lymphocyte populations after stimulation with mel624.38 cells following in vitro culture for 7 days. Analysed cell populations correspond to those tested in B and C and were analysed on the same day as C. Values are given as means of four mice with SEM and PMA/I stimulation served as the positive control. (E) Specific lysis of melA375 (HLA-A2⁺, MART-1^-) and mel624.38 (HLA-A2⁺, MART-1⁺) melanoma cell lines as target cells. Shown are means (SEM) of four mice vaccinated with MART-1-expressing DC populations using the 4-wk protocol.

Figure 5

In vivo priming efficiency of MART-1-positive DC. (A) Specific lysis of mel624.38 (HLA-A2⁺, MART-1⁺) and THP-1 (HLA-A2⁺, MART-1^-) target cells by splenic-derived human lymphocytes isolated from mice vaccinated with various mDC types. Depicted are means of results from 16 mice total per group generated in 4 individual experiments and tested for significance using a two-tailed Mann-Whitney test (p values were considered as following * p < 0.05; ** p < 0.005; *** p < 0.0001) (B) Relative lysis of results shown in (A) adjusted for numbers of human CD8⁺ T cells. To account for individual experiments and mice, values were adjusted at an E:T of 20:1 as follows:% rel. specific lysis = % spec. lysis/(% CD8/100). Given are means with SEM (* p < 0.05; ** p < 0.005; *** p < 0.0001, assessed by a two-tailed Mann-Whitney test). (C) Percentage of responding mice after vaccination with MART-1-expressing mDC of an HLA-A2⁺ donor. Mice were considered as responders when human lymphocytes, isolated and cultured from spleens, showed a specific lysis of HLA-A2⁺, MART-1⁺ target cells (mel624.38) higher than 20% at an E:T of 40:1. Analyzed were 4 individual experiments with a total of 16 mice in each group.