Scientific Correspondence A Novel Link between Ran Signal Transduction and Nuclear Envelope Proteins in Plants

A novel protein domain has been identified that is shared between putative plant Ran GTPaseactivating protein (RanGAP) and a plant protein (MAF1) previously identified to be associated with the nuclear envelope. This domain is not present in RanGAPs from animals and yeast, suggesting that plant-specific protein-protein interactions might be involved in attaching RanGAP to the nuclear envelope. The nuclear envelope separates chromatin from the cytoplasm and is involved in organizing nuclear architecture. It consists of two membranes (inner and outer) that are separated by the nuclear pore complexes (NPCs). Whereas the outer membrane is generally considered an extension of the endoplasmic reticulum, the inner membrane is characterized by a specific protein composition. A number of new inner nuclear envelope proteins have recently been discovered in animals. This group now includes lamin B receptor, lamina-associated polypeptide-1, laminaassociated polypeptide-2, emerin, MAN1, otefin, and nurim (for review, see Wilson, 2000). In addition, the nuclear lamins, nuclear intermediate filament proteins, form a layer underneath the nuclear envelope and are connected to it by interactions with some of the integral membrane proteins such as lamin B receptor (Grant and Wilson, 1997). Several of these proteins have been shown to bind to chromatin, histones, and DNA, and they have been suggested to be involved in chromatin-nuclear envelope interaction during interphase. Based on their activities and localization, they are also candidates for proteins involved in nuclear envelope dynamics during open mitosis, such as the dissociation of the condensing chromatin from the nuclear envelope and the reassociation of nuclear envelope vesicles around the decondensing chromatin (for review, see Grant and Wilson, 1997). The molecular mechanism of these processes is presently not known in any organism. We have searched the higher plant sequences available in public databases, including the 88% of sequenced Arabidopsis genome, for potential homologs of the seven animal nuclear envelope proteins listed above as well as for lamin A/C and lamin B and have found no open reading frames with significant similarity to any one of them. This finding is consistent with the failure to successfully clone plant lamins, although earlier reports using animal anti-lamin antibodies indicated that proteins with some similarity to lamins are present in plants (Beven et al., 1991; McNulty and Saunders, 1992; Minguez and Moreno Diaz de la Espina, 1993). Although the Arabidopsis genome is not completed and this analysis is therefore preliminary, it appears unlikely that all nine proteins are encoded on the remaining 12% of the genome. This might imply alternatively that plants have a different composition of proteins associated with their inner nuclear envelope. Multicellular plants and animals undergo open mitosis, whereas many unicellular eukaryotes like yeast go through mitosis with their nuclear envelope intact (Grant and Wilson, 1997). If a number of nuclear envelope proteins are involved in the orchestration of open mitosis, it would be conceivable that these proteins, like the process itself, have evolved twice in the animal and plant kingdom, thus explaining the presently observed lack of homologs of the animal proteins in plants. Two plant proteins have been identified that are localized at the nuclear rim and are candidates for nuclear envelope-associated proteins. MFP1 binds matrix attachment region DNA and is a filament-like protein (Meier et al., 1996). However, unlike nuclear lamins, it does not have a typical intermediate filament protein structure, consisting of a central coiledcoil domain and globular head and tail domains. Rather, MFP1 consists of an extended coiled-coil domain that is preceded by an N terminus containing two hydrophobic, predicted transmembrane domains. The N terminus is necessary for the targeting of MFP1 to speckle-like locations at the nuclear rim, suggesting that MFP1 might be directly associated with the nuclear envelope membranes (Gindullis and Meier, 1999). MAF1 is a small novel Ser-Thr-rich protein that binds to the coiled-coil domain of MFP1. It is also located at the nuclear envelope, but in contrast to MFP1 it has a uniform distribution instead of a speckle-like pattern (Gindullis et al., 1999). We have proposed that MFP1 is involved in attaching chromatin through matrix attachment regions to the nuclear envelope (Gindullis and Meier, 1999). The potential function of MAF1 at the nuclear envelope is not known. Although both proteins are conserved among higher plants, they have no homologs in yeast or in animals, including the fully sequenced Caenorhabditis elegans genome. We had previously found * E-mail [email protected]; fax 614 –292–5379.