Isolation and characterization of a Pufferfish MLL (Mixed lineage leukemia)-like gene (fMll) reveals evolutionary conservation in vertebrate genes related to Drosophila trithorax

Abstract

The MLL gene is interrupted and fused to a number of partner genes as a result of chromosomal translocations in human leukemias. MLL is a very large protein with a unique domain structure and large regions of homology to Drosophila trx. To de®ne the key structural and functional domains of the MLL protein in vertebrates, we have cloned the genomic region encoding an MLLlike gene in the compact model vertebrate genome of Fugu rubripes. While the similarity between the mouse and human MLL proteins is very high, a lower overall similarity is present between the Fugu and mammalian proteins. Several new highly conserved regions were identi®ed in the portion of the protein included in the MLL leukemia-associated fusion proteins. The conserved nature of regions of similarity between vertebrate forms of MLL and the Drosophila TRX proteins, as well as other domains previously suggested to have a functional role in MLL (including the AT hooks and the DNA methyltransferase domain), was also observed. Therefore, strong evolutionary constraints limited sequence divergence within these domains. The information derived from this comparative analysis will form the basis for the functional study of the MLL protein, particularly as it relates to human leukemogenesis.

MLL2, the second human homolog of the Drosophila trithorax gene, maps to 19q13.1 and is amplified in solid tumor cell lines

“…The WBP7 domain, which overlaps AT hook 3, TRX1 and the two conserved blocks also present in fMLL (but not identi®able as such in MLL) appears to mediate, at least in vitro, interactions with WW and SH3 domains (Bedford et al, 1997), suggesting that MLL2 has evolved divergent functions. Interestingly, the comparison of MLL2 with other vertebrate homologs of Trx appears to con®rm a particular motif underlying the MTAD de®ned by alanine mutagenesis scan and further supports the existence of a proposed novel Q(L/ V)DG(V/A)DD motif, which we previously described, present in Trx, MLL and fMLL (Caldas et al, 1998;Prasad et al, 1995).…”

“…Smaller regions with identity to domains present in other proteins are also recognized in the vertebrate homologs of trithorax. These include a DNA methyltransferase domain and three AT hooks in the amino-terminus (Caldas et al, 1998;Domer et al, 1993;Tkachuk et al, 1992). The hypothesis that mammalian MLL has a regulatory function similar to trithorax has been supported by the targeted disruption of the murine Mll locus, in which animals display posterior homeotic transformations and abnormalities in HOX gene patterns of expression (Yu et al, 1995).…”

“…Another two blocks of signi®cant sequence conservation between MLL, Mll and fMLL, located between TRX1 and TRX2, are also present in MLL2 (amino acids 492 ± 504 and 532 ± 559) (see Figure 3a,b). Other blocks of homology previously identi®ed in MLL, Mll and fMLL within domain 1 (Caldas et al, 1998) are absent in MLL2.…”

“…Domain 1 corresponds to the aminoterminus of the protein inclusive of the methyltransferase (MT) homology domain and encompasses 1200 amino acids in MLL and only 900 amino acids in MLL2 (1000 if the putative ATG upstream of the 5' end of the ORF described here is real). Within this domain the blocks of homology (TRX1, TRX2 and MT) previously identi®ed in MLL, Mll and fMLL, the ortholog protein in Fugu rubripes (Caldas et al, 1998), are highly conserved in MLL2. AT hooks 2 and 3 are also present in MLL2, and a small block of amino acids with homology to AT hook 1 was also identi®ed, although G replaces P in the proposed AT hook consensus`RGRP' (see Figure 3b).…”

“…Trithorax is responsible for the maintenance, but not the initiation, of HOX gene expression (Breen et al, 1995;Breen and Harte, 1993). Comparison between the Drosophila trithorax gene and its mouse, human and pu er®sh homologs has revealed multiple conserved regions (Caldas et al, 1998;Ma et al, 1993;Tkachuk et al, 1992). These include a cluster of PHD zinc ®ngers and the carboxy-terminal SET domain.…”