196 S.I. Wanzala and S. Sreevatsan
yet to be agreed upon. In addition, varying
results may be due to differences in proteomic
techniques and their resolutions, study design,
case definitions and statistical analyses (Haas
et al., 2016). There are overlaps though, of
serum proteins that are differentially expressed
in active TB like CD14, S100A proteins, apolipo-
proteins, fibrinogen, orosomucoid and serum
amyloid A. The challenge with proteomics
research is that different investigators use
different selection criteria when assessing com-
mon protein signatures and the identified pro-
teins are not always evaluated for their
diagnostic potential (i.e. with receiver operator
curve analyses or decision trees); signatures
may not be cross- validated in independent data-
sets or evaluated with external datasets (Scaros
and Fisler, 2005).
A novel approach involves detecting circu-
lating mycobacterial peptides and/or lipids or
metabolites in the serum or plasma of infected
animals. Research in the Sreevatsan laboratory
has recently identified 16 M. bovis proteins
including, MB2515c (transcriptional regulator,
LuxR family), MB1895c (cell wall biosynthesis)
and MB1554c or Pks5 (polyketide synthetase 5)
in bovine TB-positive and -exposed cattle and
deer (Lamont et al., 2014a; Wanzala et al.,
2016). These proteins were first identified by gel-
free multi-dimensional isobaric tag for relative
and absolute quantitation (iTRAQ) proteomics
and subsequently validated using a well-
characterized cattle serum repository (Seth
et al., 2009; Lamont et al., 2014a). An indirect
ELISA using monoclonal antibodies synthesized
against these peptides was developed to detect
these biomarkers in serum and has been vali-
dated in bovine and primate TB (Sreevatsan,
Kaushal and Lamont, unpublished data). Given
that the current bovine TB diagnostics have a
‘one-size-fits-all’ testing method whereby dis-
ease prevalence status for a given region is not
considered, these pathogen-specific biomarkers
(Pks5, MB2515c and MB1895c) are unique in
that they take the disease prevalence status into
account and also detect TB.
Mechanisms of release of protein biomarkers
into circulation
Peptide biomarkers are the low molecular, less
abundant circulating proteome termed as
‘peptidome’ (Lai and Agnese, 2015). This pepti-
dome may consist of many types of diagnostic
information that may constitute the parent
protein, the peptide fragment, the quantity of
peptide or the nature of carrier protein to which
it is bound (Petricoin et al., 2006). According to
the peptidome hypothesis, many proteins and
peptides are shed into the local circulation from
the disease microenvironment. Apoptosis and
necrosis of cells are considered as the main
causes for release of proteins and peptides into
circulation from the disease microenvironment.
Mycobacterial lysis or its release into circulation
(mycobacteremia), as has been recently pro-
posed using a phage-based diagnostic test (Swift
et al., 2016), can also lead to release of bacterial
products into circulation. This is a phage-based
test which has recently been commercialized.
As a consequence, the blood peptidome could
contain ongoing recordings of the molecular
cascade of communication that takes place in
the tissue microenvironment (Petricoin et al.,
2006).
Researchers have explored all different
kinds of biological matrices from cell cultures
(lysates, supernatants) to clinical samples
(serum, plasma, cerebrospinal fluid, bronchoa-
veolar lavage and urine) for protein biomarker
discovery. Of these, serum has many attributes
that make it preferable over others for biomarker
discovery. Serum is readily available and has a
dynamic range of proteins. Serum continuously
perfuses through the tissues, and thus contains
proteins and peptides secreted/released from
cells and tissues from the disease microenviron-
ment. However, there are many challenges asso-
ciated with using serum for biomarker discovery.
Candidate biomarkers are expected to exist in a
very low concentration and are generally car-
ried with high-abundant blood proteins like
albumin, which exist in a billion-fold excess.
Moreover, serum constitutes 65–97% of high-
abundant proteins like albumin and immuno-
globulins that mask the biologically significant
variations among low-abundant serum proteins
and prevents their detection and identification in
proteomic studies (Govorukhina et al., 2003).
Thus, there is a need for the depletion of these
high-abundant proteins to enrich low-abundant
biomarkers for the proteomic analysis. If not
stored properly, serum protein may be depleted
due to repeated freeze–thaw cycles.