Human platelets are anucleate blood cells that retain cytoplasmic
mRNA and maintain functionally intact protein translational capabilities.
We have adapted complementary techniques of microarray and serial
analysis of gene expression (SAGE) for genetic profiling of highly
purified human blood platelets. Microarray analysis using the
Affymetrix HG-U95Av2 approximately 12 600-probe set maximally
identified the expression of 2147 (range, 13%-17%) platelet-expressed
transcripts, with approximately 22% collectively involved in metabolism
and receptor/signaling, and an overrepresentation of genes with
unassigned function (32%). In contrast, a modified SAGE protocol
using the Type IIS restriction enzyme MmeI (generating 21-base
pair [bp] or 22-bp tags) demonstrated that 89% of tags represented
mitochondrial (mt) transcripts (enriched in 16S and 12S ribosomal
RNAs), presumably related to persistent mt-transcription in the
absence of nuclear-derived transcripts. The frequency of non-mt
SAGE tags paralleled average difference values (relative expression)
for the most "abundant" transcripts as determined by microarray
analysis, establishing the concordance of both techniques for
platelet profiling. Quantitative reverse transcription-polymerase
chain reaction (PCR) confirmed the highest frequency of mt-derived
transcripts, along with the mRNAs for neurogranin (NGN, a protein
kinase C substrate) and the complement lysis inhibitor clusterin
among the top 5 most abundant transcripts. For confirmatory characterization,
immunoblots and flow cytometric analyses were performed, establishing
abundant cell-surface expression of clusterin and intracellular
expression of NGN. These observations demonstrate a strong correlation
between high transcript abundance and protein expression, and
they establish the validity of transcript analysis as a tool for
identifying novel platelet proteins that may regulate normal and
pathologic platelet (and/or megakaryocyte) functions.