2003 Publication

	Read article online
	Fig. 1. NCI-60 reverse-phase protein lysate microarrays. (A) Staining with SYPRO ruby for total protein. Each row (see enlarged image at the left) consists of 10 two-fold dilutions of an NCI-60 cell line or the control pool. The pool was spottedatfourlocationstocontrol forpineffects.Concentratedpoolwasspotted at the bottom right corner of each field to serve as a registration mark for scanning. (B) CSA staining for p300 expression. (C) Negative control. (D) Representative candidate antibodies prescreened for specificity by Western blotting (20 _g per lane) with NCI-60 pool. *, bands at the predicted molecular weight. Blots 1-7 (from the left) show a single predominant band at the expected molecular weight. Blots 8-10 represent antibodies rejected for the array application because (i) the target band is fainter than other bands (lane 8); (ii) the target band is approximately equal in intensity to other bands (lane 9); and (iii) the targetbandis dominant (lane 10), but the other bands persistwhenthe lysate is diluted to a point that the target band is below saturation (data not shown).
	Fig. 2. Analysis of p300 expression. (A) An array incubated with p300 primary antibody and stained by CSA. (B) Sixty four dilution curves in eight fields on the array. y axes, P-SCAN intensity of p300 signal; x axes, log2(dilution factor). Numbers after cell line names are DI25 values. The order of cell line listing corresponds to placement on the array. (C) DI25 algorithm calculations for field 7. Broken line, the 25% level (at 43 units).
	Fig. 3. Clustered image map relating the expression levels of 52 proteins in the NCI-60 cell lines. Data were generated by using the DI25 algorithm and data mean centered across both cells and proteins. (Right Lower Inset) The difference in c-erbB2 level between MDA-MB435 cells (DI25__1.4) and MDA-N (DI25 __0.5). KRT, cytokeratins.
	Fig. 4. Transcript_protein correlation coefficients. The data were from two independent mRNA profiling platforms (cDNA and oligonucleotide arrays). Each point represents a single target protein. Molecular species were divided into two major categories according to SWISS-PROT and_or MIPS. Blue squares, cellstructure-related proteins; red squares, non-cell-structure-related proteins.
	Fig. 5. Analysis of a set of six replicate test arrays before the full array experiments. The test arrays consisted of three rows and ten columns in each of eight fields. The panels summarize the average differences among slide number, row number, field number, and pin number. The line across each diamond represents the group mean. The vertical span of each diamond represents the 95% confidence interval for that group.
	Fig. 6. Sensitivity and precision profile of the reverse-phase protein microarrays. Shown is the difference in log2 ratio intensities between MDA-MB435 and its Erb-B2 transfectant, MDA-MB435, as a function of dilution for 51 of the 52 proteins (excluding Erb-B2 itself) tested in the present study. The differences across the dilution series are small and quite uniform. The diamonds indicate 95% confidence intervals for the mean at each dilution, and the line at the middle of the diamond is the mean itself, as in Fig. 5. The first dilution is omitted because it has a larger error and is not used directly in the analysis.
	Fig. 7. The range (maximum minus minimum) and SD of DI25 values for each of 52 proteins across the 60 cell lines. These parameters reflect the pattern of differences across cell lines that we wished to measure. Demonstrated is a lack of dependence of the range and SD on protein concentration. Quantitatively, that is, the Pearson correlation coefficients for regression of range and SD on the log average DI25 are only -0.16 and -0.08, respectively.
	Table 1. Protein expression values (DI25) for 52 proteins across the NCI-60 cell lines. Identifier information includes the commercial designation of the antibody, a common name for the protein antigen (as specified by the company), and the HUGO name for the gene corresponding to the protein antigen. We translated among various types of gene/protein identifiers by using the program MatchMiner (http://discover.nci.nih.gov/matchminer; ref. 1). Additional detailed information for each antibody will soon be available by using a relational database, AbMiner (S.M., S. N., R. Rowland, U. Shankavaram, F. Washburn, D. Asin, H.K.-M., and J.N.W., unpublished work). 1. Lee, J. K., Bussey, K. J., Gwadry, F. G., Reinhold, W. C., Riddick, G., Pelletier, S. L., Nishizuka, S., Szakacs, G., Annereau, J.-P., Shankavaram, U., et al. Genome Biol., in press.

Abstract: Because most potential molecular markers and targets are proteins, proteomic profiling is expected to yield more direct answers to functional and pharmacological questions than does transcriptional profiling. To aid in such studies, we have developed a protocol for making reverse-phase protein lysate microarrays with larger numbers of spots than previously feasible. Our first application of these arrays was to profiling of the 60 human cancer cell lines (NCI-60) used by the National Cancer Institute to screen compounds for anticancer activity. Each glass slide microarray included 648 lysate spots representing the NCI-60 cell lines plus controls, each at 10 two-fold serial dilutions to provide a wide dynamic range. Mouse monoclonal antibodies and the catalyzed signal amplification system were used for immunoquantitation. The signal levels from the >30,000 data points for our first 52 antibodies were analyzed by using P-SCAN and a quantitative dose interpolation method. Clustered image maps revealed biologically interpretable patterns of protein expression. Among the principal early findings from these arrays were two promising pathological markers for distinguishing colon from ovarian adenocarcinomas. When we compared the patterns of protein expression with those we had obtained for the same genes at the mRNA level by using both cDNA and oligonucleotide arrays, a striking regularity appeared: cell-structure-related proteins almost invariably showed a high correlation between mRNA and protein levels across the NCI-60 cell lines, whereas non-cell-structure-related proteins showed poor correlation.