Biology Reference
In-Depth Information
quanti
cation of peptides and proteins spiked
into a human plasma matrix. Seven proteins
encompassed by eleven peptides were targeted
in a human plasma matrix and analyzed without
any sample fractionation. Five out of the seven
target proteins were nonhuman in origin and
there was therefore no interference on their quan-
titation from the sample matrix. The two human
target proteins were prostate-speci
how to minimize variation caused by sample
preparation and optimization of target peptide
release. The low fmol/ul limit of quantitation
(LOQ) achievable in phase I also established
that only biomarkers in plasma with sub to low
ug/ml or above can be reliably quanti
ed if the
current MRM-MS approach is applied without
any sample enrichment using current technology.
Kuzyk et al.
4
demonstrated quantitation of
45 endogenous proteins of moderate to high
abundance in plasma without sample enrich-
ment or fractionation in a single LC-MS run using
MRM-MS in combination with 45 concentration-
balanced stable isotope-labeled peptide stan-
dards according to their endogenous protein
level. This study demonstrated the high through-
put and multiplexing capability potential of
this approach for protein biomarkers verifi-
can igen
(PSA) and C-reactive protein (CRP). PSA is
a well-known prostate cancer biomarker present
at a concentration of less than 4 ng/ml in healthy
men.
20
CRP is an acute phase protein present at
sub mg/ml concentration in healthy population,
which has been implicated in both cardiovascular
diseases and cancers.
21,22
Heavy isotope-labeled
versions of the 11 peptide targets are used as
internal standards. Phases I and II of the study
demonstrated excellent inter- and intralaboratory
reproducibility and precision of MRM-MS assay
performed at multiple sites when prepared
samples were analyzed following standard oper-
ation procedure. The majority of interlaboratory
CVs were less than 15% in both phase I and
phase II. In phase III, all the sample handling
and preparation steps were performed onsite as
would be expected if the approach were applied
to a verify large number of biomarker candidates
at different clinical sites. The median percent
sample recovery of the peptides at the midcon-
centration point (46 fmol/ul) was 48.9% for phase
III (compared to 119.8% for phase I and 79.6% for
phase II). The median interlaboratory CVs of
eight of eleven peptides, however, were still less
than 25%. This invaluable study helps to establish
some baseline parameters when applying
targeted MRM-MS assay with isotope-labeled
internal standards to verify biomarkers in
a complex sample matrix. Overall, this study
systematically demonstrated that the MRM-MS
assay could be highly reproducible in a highly
complex sample matrix such as plasma even
across labs on different instrument platforms.
For a future large-scale biomarker veri
-
cation in a highly complex sample matrix.
The concentration-balanced internal standard
mixture not only facilitated better quantitation
accuracy of analytes with greater than 10
4
dynamic range but also improved the analyses
CVs compared using equimolar concentration
internal standard mixture. Therefore, the concen-
tration of internal standards used must be care-
fully considered when developing multiplexed
MRM-MS assays for biomarker veri
cation.
High interlaboratory assay reproducibility of 39
peptides from proteins of moderate abundance
in plasma at multisites was also demonstrated
with CVs less than 20% at four labs following
standard operation procedure.
23
Although these recent studies demonstrated
the excellent precision, reproducibility, and trans-
ferability of MRM-MS-based quantitation assay
for biomarker veri
cation, they also made it clear
that the needed pg/ml-ng/ml sensitivity
required to measure many biomarkers in plasma
is not achievable with the current technology
without any sample enrichment. Therefore,
strategies developed previously for in-depth
biomarker discovery as well as new strategies
targeted at potential biomarker veri
cation
study, quantitation accuracy could depend on
cation are
both being explored to improve the MRM-MS
Search WWH ::
Custom Search