Proteomics - Biomedical Informatics in Translational Research - page 97

do not affect the overall proteome coverage when based on the cysteinyl peptides,
but this is not the case with the phosphopeptide and glycopeptide affinity capture
procedures because proteins bearing these modifications in the proteome account
for only 30% and 65%, respectively.
Multidimensional separation improves the proteome coverage, but these
workflows are prone to variation. They also increase the analysis time and can
cause sample losses at each step. Moreover, this workflow requires every biological
replicate to be processed in the same labor-intensive manner and requires that
MS/MS be used to characterize the peptides. This approach is not feasible for a clin-
ical study, in which sample amounts are limited and protocols require a large num-
ber of samples to achieve statistical relevance. To circumvent these multiple
separation stages and to improve the peak capacities for 1D RPLC, high-pressure
LC systems have been built that utilize longer capillary columns packed with
smaller particles (<2 m) to achieve the high-resolution separation of peptides [45].

These setups improved the proteome coverage and the results obtained with 10-kpsi
RPLC-LCQ

μ

MS/MS were comparable to the multidimensional protein

identification technology (MudPIT) on LCQ [45].
To exploit the ability of the high-pressure (~ 20 kpsi) systems in extending the
dynamic range of detection for identifying the low-abundance proteins from clinical
samples, the proteomics group at Pacific Northwest National Laboratory (PNNL)
has developed a robust proteomics approach that utilizes routine multidimensional
LC/MS/MS to completely characterize a given sample system such as cells, tissue, or
body fluid and validates the peptide identifications from the accurate mass and
retention time recorded from a parallel multidimensional analysis on a high-perfor-
mance Fourier transform ion cyclotron resonance (FTICR) instrument. The data
from these two analyses will be matched to construct a database comprising accu-
rate mass and time (AMT) tags [46]. The creation of an AMT database obviates the
MS/MS analysis on subsequent, similar types of samples and increases the through-
put with the improved dynamic range. The potential of AMT mapping resides in the
ability of high-pressure LC systems to achieve reproducible 1D RPLC separations in
conjunction with highly accurate mass measurements using FTICR MS. This setup
demonstrated the identification of serum proteins ranging more than 6 orders of
magnitude in concentrations without depleting the highly abundant proteins. The
highly specialized instrumentation used to achieve such analytical capabilities is not
currently available in the market, is expensive to build, and needs highly skilled per-
sonnel to maintain. The recently introduced nanoAcquity UPLC (Waters) system
has demonstrated the high-resolution separations on peptide mixtures from a single
chromatographic run and has the ability to use capillary columns packed with
1.7- m particles (at >10 kpsi).

Other emerging technologies with the potential to provide robust, faster separa-
tions for proteomics applications are microfluidic chip technology [47, 48] and
gas-phase separations based on ion mobility mass spectrometry [49, 50]. With these
improvements in the instrumentation, robust platforms are being developed that
achieve high reproducibility and throughput with broad dynamic range and accu-
rate quantitation capabilities. These platforms may serve as diagnostic tools for the
analysis of clinically relevant samples.

μ