Ring1A is a transcriptional repressor that interacts with the Polycomb-M33 protein and is expressed at rhombomere boundaries in the mouse hindbrain

Abstract

In Drosophila, the products of the Polycomb group (Pc‐G) of genes act as chromatin‐associated multimeric protein complexes that repress expression of homeotic genes. Vertebrate Pc‐G homologues have been identified, but the nature of the complexes they form and the mechanisms of their action are largely unknown. The Polycomb homologue M33 is implicated in mesoderm patterning in the mouse and here we show that it acts as a transcriptional repressor in transiently transfected cells. Furthermore, we have identified two murine proteins, Ring1A and Ring1B, that interact directly with the repressor domain of M33. Ring1A and Ring1B display blocks of similarity throughout their sequences, including an N‐terminal RING finger domain. However, the interaction with M33 occurs through a region at the C‐terminus. Ring1A represses transcription through sequences not involved in M33 binding. Ring1A protein co‐localizes in nuclear domains with M33 and other Pc‐G homologues, such as Bmi1. The expression of Ring1A at early stages of development is restricted to the neural tube, whereas M33 is expressed ubiquitously. Within the neural tube, Ring1A RNA is located at the rhombomere boundaries of the hindbrain. Taken together, these data suggest that Ring1A may contribute to a tissue‐specific function of Pc‐G–protein complexes during mammalian development.

The Bmi-1 oncoprotein interacts with dinG and MPh2: the role of RING finger domains

“…Others have shown co-localization of Bmi-1, RING1 and HPh1 in human SW480 cells and of Bmi-1, RING1, and M33 in U2-OS cells Schoorlemmer et al, 1997). Our results provide some insights into the nature of the molecular interactions among the PcG proteins.…”

“…The function of dinG is not immediately apparent from sequence analysis, but it has recently been shown to interact with the mouse PcG protein, M33. The dinG/ RING1B homolog, RING1, has been shown to repress reporter gene activity when targeted by fusion to LexA or Gal4, but the in vivo function of RING1 remains elusive Schoorlemmer et al, 1997).…”

“…The predicted human and murine gene products are 96% identical indicating that the 1.1 kb cDNA contained the mouse homolog of the dinG gene. Subsequently, the sequence was found to be identical to the recently described mouse RING1B cDNA (Accession #Y12783 and Y12880) (Schoorlemmer et al, 1997). We have elected to maintain the designation dinG.…”

“…The amino terminal portion of the protein contains a C-X 2 -C-X 12 -C-X-H-X 2 -C-X 2 -C-X 11 -C-X 2 -C RING ®nger motif that shares exact residue spacing and 95% identity with the RING ®nger motif of human RING1 (Lovering et al, 1993). As described by Schoorlemmer et al (1997), additional regions of signi®cant homology exist within dinG/RING1B and RING1 at the carboxy termini. With the exception of the RING ®nger domain, however, RING1 is the only protein in the database that shares signi®cant homology with dinG/RING1B.…”

“…Interestingly, dinG is also a RING ®nger protein whose RING ®nger domain shares near complete sequence identity with the protein family namesake, RING1 (Lovering et al, 1993). RING1 proteins have recently been shown to interact with other mammalian PcG proteins, Pc1, Pc2 and M33 Schoorlemmer et al, 1997), none of which bear structural similarities to Bmi-1.…”