The expression patterns of surface
antigens are associated with
the differentiation status and functional characteristics of mammalian
cells. To analyze the surface antigen expression pattern in a high-throughput
manner, antibody microarrays have been developed by several groups,
including ours. This analysis can be performed using cell-binding
assays on microarrays; moreover, this approach has advantages over
conventional flow cytometry (FCM). Unlike FCM, the microarray-based
method cannot evaluate the concurrent expression of more than two
surface antigens on a single cell, and therefore, it cannot be used
for cell subset analysis. To overcome this drawback, we prepared an
antibody microarray with spots presenting co-immobilized multiple
antibodies together with spots presenting each antibody separately.
The co-immobilized spots are expected to be reactive for every surface
antigen specific to the co-immobilized antibodies. In addition, the
concept of an algebra of sets is incorporated into the derivation
of quantitative data regarding cell subsets. Here, taking cell subsets
with respect to two surface antigens as the simplest example, antibody
microarrays were prepared and initially subjected to validation studies
to verify the accuracy of cell-binding assays. Quantitative subset
analysis was performed using antibody microarrays prepared using the
anti-CD13 and anti-CD49f antibodies. For model populations that consisted
of discrete subsets, THP-1, HL-60, CCRF-CEM, and Ramos cell lines
were used because they were found by FCM to have a singular phenotype,
that is, CD13+CD49f+, CD13+CD49f–, CD13–CD49f+, and CD13–CD49f–, respectively. Five populations
were prepared by mixing these cells at various ratios and analyzed
for their subsets using microarrays. The results showed that the experimentally
determined abundance ratios of the four model subsets were in good
agreement with the predetermined abundance ratios, which provided
the proof of principle for the new method in the quantitative subset
analysis.