156 J. Hope and D. Werling
these cells have been associated with control of
chronic infections such as HIV, hepatitis C, leish-
maniasis and malaria in humans (Wilkinson
and Wilkinson, 2010), as well as PCV2 in pigs
(Koinig et al., 2015). A number of studies of
M. tuberculosis infection in humans indicate a
protective role for polyfunctional T cells but
these cells may also be associated with disease
progression (Sutherland et al., 2009; Wilkinson
et al., 2010; Geluk et al., 2012). It is clear that
the pattern of co-expression of multiple cyto-
kines by these polyfunctional T cells is important
for their potential protective effect. In humans
with active TB, T cells predominantly express
IFN-γ with TNF-α or TNF-α alone, whereas in
humans with latent TB, where it is assumed
infection is controlled, or in successfully treated
patients, higher frequencies of polyfunctional T
cells expressing IFN-γ, TNF-α and IL-2 can be
observed (Geluk et al., 2012). Similarly, bovine
polyfunctional CD4+ T cells which expressed
IFN-γ, IL-2 and TNF-α were shown to have an
effector phenotype (CD44hi CD62Llo CD45RO+)
and were associated with pathology rather than
protection (Whelan et al., 2011). In line with the
observation that more diverse cytokine profiles
reflect progression towards disease rather than
immunological control, Rhodes et al. (2014)
demonstrated that cattle producing both
antigen- specific IFN-γ and IL-2 were more likely
to present with visible pathology at post mortem
than those that expressed IFN-γ alone.
11.1.2 Tissue responses to infectionThe balance between protective immunity and
pathology is likely driven by the ability of myco-
bacteria to manipulate the CD4+ T-cell response
either directly or indirectly via antigen-
presenting cells. In tissues this may be evidenced
by slow, weak stimulation of antigen-specific T
cells enabling early growth and persistence of
mycobacteria within immune privileged sites
(granulomas). The delayed tissue response to
M. tuberculosis infection is well documented in
mice; the first antigen-specific T cells arrive in
lung-draining lymph nodes (mediastinal) after
~10 days (Reiley et al., 2008) and then in the
lungs several weeks following exposure (Reiley
et al., 2008). This delay contributes to the
inability of the host to clear the organism. While
the specific mechanisms are not yet well under-
stood, this delay likely reflects the time taken
for infected dendritic cells (DC) and recruited
macrophages (MO) to reach the lungs.
Both M. tuberculosis and M. bovis have been
shown to alter the functional capacity of DC and
MO, which likely alters their capacity to recruit
and activate T cells (Hope et al., 2004; Piercy
et al., 2007; Wolf et al., 2007). Using TCR trans-
genic mice, the initial recognition of mycobacte-
rial antigens was shown to occur in draining
lymph nodes rather than in the lung ( Chackerian
et al., 2002a; Reiley et al., 2008; Wolf et al.,
2008). In addition, it has been hypothesized that
poor stimulation/activation of T cells occurs in
the inflamed lung environment potentially
through limited availability of antigen (Bold
et al., 2011; Egen et al., 2011). This leads to poor
cytokine secretion and lack of engagement with
infected MO (Robinson et al., 2015). Whether
this occurs in cattle infected with M. bovis has
not been determined. However, given that the
kinetics of T-cell responses to infection are simi-
lar in cattle and humans (Waters et al., 2011),
and reflected by those observed in mice, it seems
likely that M. bovis also affects the timing and
development of antigen-specific CD4+ T-cell
responses in the bovine respiratory tract tissues.
However, in contrast to the murine environ-
ment, the bovine airway mucosal system is
already developed and contains significant num-
bers of DC at the time of birth (Hope and Werling,
unpublished data). This is an important consid-
eration for vaccine design and the definition of
immune correlates of protection. The response
of circulating T cells is relatively easily measured
in both humans and cattle but may not reflect
events occurring at local infection sites where
inflammation and low antigen availability may
be of importance. Therefore, consideration of
adjuvants, routes of immunization as well as
antigen/epitope availability is of crucial impor-
tance to protect the lung from infection.
Related to this is recent evidence that anti-
gens other than the immunodominant antigens
of M. tuberculosis and M. bovis (e.g. ESAT-6 and
CFP-10) may be recognized by specific CD4+ T
cells. It has been estimated that only ~5 to 20%
of murine lung T cells recognized ESAT-6 and/or
CFP-10 (Brandt et al., 1996; Winslow et al.,
2003; Wolf et al., 2008). Indeed, unbiased