Additional testing in A2.1 Tg mice reveals a more extensive repertoire overlap
We previously detected 14 different epitopes utilizing PBMC from A2.1-positive human volunteers vaccinated with Dryvax (Figure 1A) . Of these epitopes, only two were also independently identified in VACV-infected A2.1 Tg mice [5,13-16].
Based on these results, Terajima and Ennis hypothesized that the large
number of potentially immunogenic peptides encoded by a complex
pathogen such as VACV might explain the incomplete overlap observed . Herein, we addressed this issue in more detail.
Figure 1. Testing A2.1-restricted VACV T cell epitopes in A2.1 Tg mice using purified CD8+ T cells and individual peptides. (A) Splenic CD8+ cells were purified from VACV-infected A2.1 Tg mice and incubated with peptide-pulsed Jurkat-A2.1/Kb cells, after which number of IFN-γ-producing cells was enumerated in an ELISPOT assay. Shown are the average net SFC/106 CD8+ T cells in response to each peptide; black bars, epitopes identified by Oseroff et al.  in A2.1 positive donors and recognized in Tg mice upon re-screening; white bars, peptides identified by Pasquetto et al.  in Tg mice; grey bars, epitopes identified in both studies. (B) The specific T cell response to three different epitopes (B14R327–335, M1L374–383, and A14L51–59)
was assessed 7 days after VACV infection in A2.1 Tg mice. Shown is the
stimulation index for the epitopes tested either as individual peptides
(black bars) or in pools of ten peptides (white bars). (C) Shown is the
stimulation index for the epitope-specific responses by using either CD8+ T cells (black bars) or splenocytes (white bars). (D) Shown are the average net SFC/106 CD8+ T
cells in response to groups of epitopes recognized in Tg mice upon
re-screening (black bar) and identified in Tg mice in the original
study by Pasquetto et al. 
(white bar). The unpaired t test with Welch's correction was used to
determine if differences were significant. For all panels, average data
are shown from at least two independent experiments. For each
individual experiment samples are tested in triplicate. Error bars
First, A2.1 Tg mice were inoculated
with VACV and the reactivity against each of the 28 individual
A2.1-restricted epitopes was evaluated. Responses were detected for all
epitopes previously reported in HLA Tg mice (Figure 1A;
white and grey bars). In addition, responses were detected for 6
peptides that had been originally described in A2.1-positive humans [9,14,18], but not reported in studies utilizing Tg mice [5,15] (Figure 1A;
black bars). Altogether, responses directed against 8 of 14 (57%)
epitopes reported in humans were now also detected in A2.1 Tg mice.
The present epitope screen differs in two experimental settings from the original study by Pasquetto et al. .
First, individual peptides were used here instead of peptide pools.
Second, the initial epitope identification used splenocytes as effector
cells , whereas the present study utilized purified CD8+ T cells. Herein we examined the impact of these two experimental differences. When three representative epitopes (A14L51–59, B14R327–335, and M1L374–383) were tested either as individual peptides or in pools of 10 peptides, similar response magnitudes were observed (Figure 1B). We also tested the response to the same 3 epitopes using either purified CD8+ T cells or splenocytes, and found that the higher non-specific interferon (IFN)-γ production obtained when using splenocytes led to a reduced stimulation index, and thereby decreased sensitivity (Figure 1C).
We observed that the average responses against these epitopes not
initially identified in Tg mice were ~3 times lower than those against
the epitopes identified in the original screen (P = 0.0098; Figure 1D). Although the 3 dominant epitopes (A14L51–59, A17L81–90, and M1L347–383)
contribute to a significant portion of the higher response identified
in mice, exclusion of these 3 epitopes still results in a significant
difference in the magnitude of the response for the epitopes identified
in humans versus mouse (P = 0.0012). Taken together, these
results suggest that while 8 of 14 (57%) peptides reported in humans
can also be detected in A2.1 Tg mice, differences in the patterns of
immunodominance might render them less prominent and consequently make
their detection more difficult.
The mice utilized by Pasquetto et al. were A2.1 bxd mice .
To examine whether expression of endogenous mouse MHC molecules might
also contribute to the incomplete overlap, we tested the reactivity of
the 14 epitopes identified in humans using CD8+ T cells from VACV-infected HHD A2.1 mice, which do not express endogenous mouse MHC .
However, no additional epitopes were recognized in the HHD mice, and
the response magnitude and immunodominance pattern was similar to that
in A2.1 bxd mice (data not shown).
Testing of a broader panel of human vaccinees reveals a more extensive overlap
As described above, a total of 16 epitopes were initially identified in HLA Tg mice [5,13,15]. Of these, only two (13%) were also independently identified in human vaccinees (Figure 1A).
Most of the candidate peptides were tested in only 6 donors, with each
individual donor yielding a rather distinct pattern of epitope
recognition. While some epitopes were identified in multiple donors,
most of the epitopes were only detected in 1 out of the 6 donors
To test the hypothesis that the epitopes not initially detected in
humans were associated with lower and/or less frequent responses, we
examined the responses to the set of 16 A2.1-restricted epitopes
identified in Tg mice using PBMC from 8 different A2-positive
vaccinees. The samples tested included frozen cell aliquots from 5 of
the donors tested in the original study of Oseroff et al. ,
as well as PBMC from 3 additional A2-positive vaccinees. PBMC were
cultured in the presence of individual peptides, and the numbers of IFN-γ-producing
cells were enumerated in an enzyme-linked immunosorbent spot (ELISPOT)
assay. The same two peptides previously identified as positive in these
donors were also re-identified in the course of the present
experiments, and responses to an additional 5 peptides originally
identified in A2.1 Tg mice, but not identified in the human donor
screen of Oseroff et al. ,
were detected. Similar to what we observed in the mouse experiments,
responses to the newly detected epitopes were in general lower than to
epitopes identified in the initial screen (see Additional file 1).
Overall, responses directed against 7 of 16 (44%) peptides initially
identified in HLA Tg mice, were also detected in humans. Considering
the large donor-to-donor variation observed, it is likely that testing
of additional donors would further increase the overlap between the two
Additional file 1. Recognition of additional A2.1-restricted VACV T cell epitopes by human A2-positive VACV vaccinees. The data provided represent the net SFC/106 PBMCs of the human A2-positive VACV vaccinees in response to epitopes originally identified in A2.1 Tg mice.
Size: 77KB Download file