CD56 and RUNX1 expression in leukemic blasts of patients with acute myeloid leukemias (AMLs), defined as either CD56⁺ (n=4) or CD56⁻ (n=4) by flow cytometry. (A) Top panel: Western blot analysis. Note much stronger expression of both CD56 and RUNX1 p48 in CD56⁺ than CD56⁻ AMLs. RUNX1 N-terminal, but not RUNX1 C-terminal antibody, detects a strong additional 30-kDa band in CD56⁻ AMLs, suggesting that p30 could harbor an alternative C-terminal exon. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH), loading control. Top panel, right[b]: Reverse transcriptase–polymerase chain reaction quantification of CD56, p48, and p30 in of CD56⁺ and CD56⁻ AMLs (n=72) (Figures S4,S5). There was significant concordant expression of CD56 and p48 (P < .001, Mann-Whitney U test). p30 expression was slightly higher in CD56⁻ compared with CD56⁺ AMLs (P < .035). Bottom panel: RUNX1 isoforms and the antibodies used to discriminate them. RUNX1 (p48) and 3 new (p38a, p30, p24) RUNX1 isoforms isolated from a human CD56⁻cDNA library (Document S1). p30 and p24 harbor a new C-terminal exon, termed exon 5.4 (Ex. 5.4). Antibody symbols indicate binding regions of the commercial antisera to the N- and C-terminus of RUNX1 and of a custom-made rabbit antiserum raised against exon 5.4 (Figures S2 and 1B). TAD1 and TAD2, transactivating domains 1 and 2, respectively; RUNT (RUNT domain), DBD (DNA-binding domain). (B) Top panel: Semiquantitiative ³²P-dCTP-based reverse transcriptase–polymerase chain reaction analysis of CD56 and RUNX1 isoforms in CD56⁺ and CD56⁻ AMLs (Figure S2). CD56 and p48 expression are concordant, and there is slightly higher expression of p30 and p38a mRNA in CD56⁻ AMLs. Expression levels are given as counts per minute (cpm) of phosphoimager analysis (see also Figure S2). Bottom panel: Confirmation of differential expression of p48 and p30 isoforms by immunohistochemistry in AMLs. Examples of staining of CD56⁺ (left row) and CD56⁻ AMLs (right row) using RUNX1 C-terminal antibody (detecting p48 and p38a) and RUNX1 5.4 antibody (detecting p30 and p24). In the left row, AML M2 blasts infiltrating testis exhibit strong expression of CD56 and RUNX1 p48, whereas RUNX1 p30 expression is weak. By contrast, the right row depicts a bone marrow infiltration by CD56⁻ AML M2 blasts with lack of RUNX1 p48 but strong expression of RUNX1 p30. Representative results of 4 experiments. Slides were viewed through an Olympus d×40 microscope (Olympus, Hamburg, Germany) with a UPlan APO lens and Xylol coverslipping films (Sekura, Heppenheim, Germany). Images were acquired with an Olympus BX50 camera and Olympus DP-Soft version 5.0 (Olympus), and processed using Adobe Photoshop version 3.0 (Adobe Systems, San Jose, CA).