182 R. Waters and M. Vordermeier
to antigen stimulation. Kasprowicz et al. (2011)
demonstrated the diagnostic potential of this
approach by evaluating monokine-induced by
IFN-γ (MIG) and IP-10 responses to ESAT-6/
CFP10 by leukocytes in whole blood samples
from humans in response to HIV, cytomegalovi-
rus, and TB infection. In regard to TB diagnosis,
the mRNA response correlated with the ex vivo
ELISPOT protein response and, importantly, the
assay could be performed with small volumes of
whole blood. With cattle, global gene-expression
studies have defined numerous candidate bio-
markers, particularly in the context of protective
versus non-protective responses to BCG vaccina-
tion (Bhuju et al., 2012). Of these, genes encod-
ing IL-22, IFN-γ, the zinc metallothionein MT-3,
IL-13 and CCL3 were strongly up-regulated in
response to PPD in vaccinated and protected
versus vaccinated and non-protected calves.
Functional analysis of the RNA-seq data indi-
cated that the most significantly modulated net-
work of expression was the cytokine-to-cytokine
receptor interaction pathway. These studies, as
well as other unpublished results in our labora-
tories, provide numerous candidate markers for
further validation by follow-up RT-qPCR and, if
available, protein assays (e.g. Th17-associated
genes [Waters et al., 2015a]).
Evaluation of gene expression in non-
stimulated blood cells may also be used to iden-
tify candidate biomarkers to discriminate active
versus latent TB and risk for recurrent disease
(Jacobsen et al., 2007; Maertzdorf et al., 2011;
Mihret et al., 2014). With this approach, TB can
also be distinguished from other pulmonary
diseases such as sarcoidosis, pneumonias and
lung cancers using transcriptional signatures of
blood leukocytes from humans (Bloom et al.,
2013). Most recently, Jenum et al. (2016) identi-
fied a host biomarker signature consisting of
BPI, CD3E, CD14, FPR1, IL4, TGFBR2, TIMP2
and TNFRSF1B that differentiated children with
active TB from asymptomatic siblings. Signa-
tures associated with a tendency towards active
disease consisted of FCGR1A, FPR1, MMP9,
RAB24, TNFRSF1A and TIMP2, whereas BLR1,
CD8A, IL7R and TGFBR2 were associated with a
decreased likelihood of TB-associated disease
in this population, thereby providing useful
information for the clinical management of
TB in children. This approach is particularly
appealing as it does not require an antigen-
stimulation phase. Thus, immune markers are
measured directly from blood leukocytes. Using
global gene expression, Zak et al. (2016) recently
identified a gene signature predictive of develop-
ment of active TB. The signature predicted dis-
ease development with a sensitivity/specificity
of 66.1%/80.6%, respectively. In addition to
mRNA, micro RNAs (miRNA) expressed in the
blood of infected patients have been explored for
the development of improved diagnostic tests
and to distinguish active versus latent TB as well
as to distinguish HIV co-infection from other
pulmonary diseases (Miotto et al., 2013). Using
a bovine microRNA microarray, Golby et al.
(2014) demonstrated that miR-155 is a poten-
tially useful biomarker of M. bovis infection as
well as a prognostic marker (in regard to protec-
tive BCG vaccination) to identify animals with
advanced pathology.
While biomarker discovery studies have
traditionally utilized transcriptomics, other
approaches such as evaluation of the proteome
and epigenome are also being developed to
define markers associated with clinical stages of
TB in humans (Esterhuyse et al., 2015). In a
large study evaluating serum protein markers in
HIV+ and HIV- patients with TB (active versus
latent) and other respiratory diseases, Achkar
et al. (2015) established that soluble CD14 and
SEPP1 were present in TB serum panels from
both HIV+ and HIV- patients. Several other
promising candidate biomarkers of TB infection
discovered in this study included: GP1BA
( pulmonary inflammation); SELL and LUM
( leukocyte homing); TNXB, COMP, PEPD and
QSOX1 (morphogenesis and extracellular matrix
remodeling); and APOC1 (lipid transport and
regulation). With cattle, Seth et al. (2009) and
Lamont et al. (2014) have identified multiple
host proteins in sera associated with M. bovis
infection of cattle; of which, vitamin D binding
protein had the greatest diagnostic potential.
Using metabolomics profiling, Lau et al. (2015)
identified four metabolites increased in plasma
from human TB patients as compared to plasma
from community-acquired pneumonia patients
and non-affected controls. Thus, emerging
technologies will likely prove useful for the dis-
covery of biomarkers of diagnostic relevance for
bovine TB.