Biology Reference
In-Depth Information
a number of additional factors need to be considered to ensure that other relevant cri-
teria are fulfilled. These include the general requirements of targeted mass spectrom-
etry analyses, criteria specific to studies based on synthetic peptides (AQUA) or
criteria specific to the QconCAT approach. These factors are considered below.
Targeted mass spectrometry relies on the high specificity of precursor/fragment
ion pairs (transitions) generated in MS/MS experiments in triple quadrupole-based
instruments (
Lange
et al.
, 2008
). Adding to the mass filtering, these transitions are
measured in predetermined time windows during LC-MS runs to allow for multi-
plexed SRM analyses. To facilitate scheduled SRM assays, it is essential to know
the elution times of the peptides chosen for targeted analyses. Consequently,
if not already determined from previous studies, the column retention time of each
candidate peptide that is selected for SRM analysis has to be determined for
the various chromatography formats. With the two levels of mass selection
(i.e. Q1/Q3; precursor/fragment ion) and the appropriate time window established,
it is possible to achieve the required high specificity of co-eluting masses with very
high sensitivity and low background.
The technically challenging aspect of targeted proteomics workflows is to set up
validated SRM assays for peptides representing the specific proteins of interest
(
Kettenbach
et al.
, 2011
). Fragmentation patterns allowing selection of appropriate
peptides for a protein of interest are available either in public repositories (
Desiere
et al.
, 2005; Martens
et al.
, 2005
) or in existing MS/MS datasets. These sources yield
information on those peptides that are, in practice, good responders in an LC-MS/
MS experiment. For peptides that have been identified previously in such experi-
ments, it is important to recognize that larger peptides leading to high confidence
scores are not necessarily good candidates for SRM pairs. Here, the selection of
smaller peptides and peptides containing proline residues (incomplete fragmentation
patterns) will produce higher intensity fragment masses and thereby provide better
transitions. Problems may arise for peptides whose sequences are shared with homol-
ogous proteins or those sharing the same domains. Therefore, the chosen peptides
must be proteotypic, that is, they must uniquely identify the targeted protein
(
Mallick
et al.
, 2007
).
In addition to their unique identity, it needs to be recognized that peptides may
be identified on the basis of more than just their primary sequence. Peptides may be
post-translationally modified, for example, by the oxidation of methionine during
sample processing or even specifically in response to a native biological process.
Such modified peptides have to be handled with care and, consequently, are mostly
excluded from targeted analyses. This criterion is summed up by the term of a 'quan-
totypic peptide': peptides have to be chosen that are not only proteotypic but also
representative for the protein and its changes in abundance throughout the entire
analysis (
Brownridge
et al.
, 2011
).
Most studies that describe the workflow for targeted analyses are dependent on
trypsin for proteolytic digestion since, under optimum conditions, complete proteo-
lysis of a sample can be expected (
Barnidge
et al.
, 2004; Havlis and Shevchenko,
2004; Olsen
et al.
,2004
). Nevertheless, digestion efficiencies need to be checked
