¹Molecular and Cell Biology and Oncology Research Institute, Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609, Singapore

Correspondence: Y Ito, E-mail: itoy@imcb.nus.edu.sg

Independent studies in many laboratories have generated a series of aliases for a catalog of proteins and DNA-binding activities, which upon further study were found to be related to the Runt-domain class of transcription factors (Table 1). The designation acute myelogenous leukemia (AML) factors (i.e. AML1, AML2 and AML3) is based on genetic studies of leukemia-related chromosomal translocations. 'Core-binding factor alpha' (CBFA) proteins were initially characterized as sequence-specific DNA-binding proteins that interact with the enhancers of retroviruses. Isolation of murine cDNAs for polyoma enhancer-binding proteins yielded the term PEBP2. Studies in bone cell biology generated yet another set of aliases, including nuclear matrix protein 2 (NMP2), osteoblast-specific complex (OBSC) and osteoblast-specific factor 2 (OSF2). It is now known that all these proteins and DNA-binding activities represent mammalian homologs of the Drosophila runt protein.

Table 1 - Aliases for runt-related DNA-binding activities abbreviation.

Full table

On November 1999, the Nomenclature Committee of the Human Genome Organization (HUGO) adopted the use of the term 'RUNX' to refer to genes encoding the runt-related proteins, based on the joint recommendation of a panel of eight investigators involved in the initial cloning of chromosomal breakpoints or cDNAs encoding these genes (see Table 2). The agreement among investigators in the field is to refer to the runt-related proteins as RUNX proteins at least once in the Abstract and/or Title of new manuscripts submitted for publication. Additional designations are at the discretion of the authors.

Table 2 - Synonyms for mammalian RUNX gene subtypes & locus.

Full table

The first mammalian runt-related protein was cloned by the identification of the breakpoint of the t(8;11) chromosomal location, which is causally related to AML. This gene was initially defined as AML1 (also known as CBFA2 and PEBP2alphaB) and received the systematic designation RUNX1. The gene that has been commonly referred to as CBFA1 (also known as AML3 and PEBP2alphaA) in the field of bone biology, has now been designated RUNX2. The third runt-related gene, which has been known as AML2/CBFA3/PEBP2alphaC, is now referred to as RUNX3. The homologous genes in mouse are presently referred to as Runx1, Runx2 and Runx3, respectively. While the numbering of RUNX proteins and their many aliases can be confusing (see Table 1), the order of the numbers can be easily remembered as the order in which the knockouts for each of the mouse Runx genes were published (Runx1/Aml1 in 1996, Runx2/Cbfa1 in 1997 and Runx3/Pebp2C in 2002). For clarity, we have opted here to avoid using differences in text type to distinguish between mouse and human or gene and protein, because the printed appearance of text is not pertinent to the RUNX nomenclature. However, as is the convention among geneticists, the gene names are normally italicized in print, whereas the protein names are in normal print.

The systematic name RUNX is an abbreviation for the term 'runt-related protein'. It was chosen as a four-letter abbreviation that is analogous to the three-letter designations for homeodomain transcription factors in vertebrates (for example, the PAX and DLX proteins are related to, respectively, the paired and distal-less genes in the fruit fly). Considering the key roles of Runt-related transcription factors in controlling development and extreme conservation of the runt-related DNA-binding domain, this analogy is appropriate. A second important factor that contributed to its selection is the fact that the term 'RUNX' has not previously been used in PubMed/NCBI database, and is therefore very useful in retrieving information unique to these proteins.

Selective utilization of distinct transcription start sites, as well as alternative mRNA splicing and translational start codons, generates a multiplicity of RUNX proteins. This diversity of protein isoforms may have biological significance and support distinct cellular functions. To facilitate accurate dissemination and interpretation of biological data, more encompassing nomenclature proposals have been discussed among RUNX investigators. These proposals cope with the similarities and differences in Runx gene promoters, proteins and mRNAs, and some of these naming suggestions have been implemented (see also Drissi H et al., 2000, J. Cell. Physiol. 184: 341–50; Bangsow C et al., 2001, Gene 279: 221–32; Levanon D et al., 2001, Gene 262: 23–33; Drissi H et al., 2002, Exp. Cell Res. 274: 323–33; Terry A et al., 2004, Gene, in press).

The adoption by the HUGO Nomenclature Committee to use this systematic name to unify the naming system for this exciting class of transcription factors has facilitated cross-referencing of papers using the RUNX keyword and has definitely been in the general interest of communication among investigators. The new nomenclature was publicly announced in July 2000 at an international meeting in Kyoto that was attended by a large segment of the community. The term RUNX is now in general use and at present more than 400 references have appeared since the new nomenclature was first adopted.