Yeast SMT3 (Suppressor of Mif2 temperature sensitive mutant 3) is a Mif-related prortein, and is first found in suppression Mif2 (Mitotic fidelity of chromosome transmission protein 2) mutant. So far, among animals of higher level, we have found the existence of 3 SMT3 isozymes (A,B,C). It¡¦s known that SMT3A is located on chromosome 21. And smt3 sequence of Drosophila is similar to those of yeast or human in a percentage of 55%, 73%, and 52%. We analyze 10 strains of flies with p element inserted into position 27C6-8 on chromosome 21 and found a p element mutant, P997, near the position of smt3 gene. Then we built fly¡¦s smt3 mutation stock by using P element technology, and by using PCR, we can sieve out mutant fly fast and massively. Thus we can observe the expression of smt3 during the development of fly. The P element of stock 11378 is inserted 300bp before smt3 start codon. After stock 11378¡¦s crossing with W1118, we get smt3-null mutation. Then we found that smt3-null mutation is a recessive lethal mutant, fly will die in the forth period of embryo development. At that time, the cell nucleus has divided for about 10 to 13 times, also means 90 to 120 minutes after the embryo is produced. We also found two important characteristics about dead embryo of smt3-null mutation: (1) The DNA in the embryo is scattered and couldn¡¦t be condensed in the nucleus. (2) Huge and abnormally-dispersed vacuoles are observed in the embryo.

Abstract SMT3 (Suppressor of MIF2 3 protein) was identified as a mutation suppressor in yeast centromere protein MIF2. It is also known as an ubiquitin-like protein due to the smilarities of their primary structures that is very conserved during the eukaryotic evolution. Although only one SMT3 was found in low eukaryotes such as in yeast, three members of SMT3 (SMT3A, SMT3B and SMT3C) have been identified in high eukaryotes. It has been known that SMT3C plays an important role in post-translational modification. However, the functions of SMT3A and SMT3B are not well studied yet and the relationship among the SMT3 families remains unclear. In the present study, Daxx, a Fas binding protein, was demonstrated to bind to SMT3B using yeast two-hybrid assay. It was found that the N-terminal domain of Daxx (Daxx 1) and the C-terminal domain of Daxx (Daxx 4), respecifitively, bound to all members of human SMT3 families (including SMT3A, SMT3B and SMT3C). Neverthless, mechanisms of interactions between the SMT3 families and Daxx domains remined unclear. Studies on truncated human SMT3 families have shown that two glycines on the C-terminal end of human SMT3 families were required in the interaction between SMT3 and Daxx domains, for example, SMT3A and SMT3B required C-terminal two glycines on the Daxx 4 domain where as SMT3C required C-terminal two glycines on the Daxx 1 domain. Morever, truncated SMT3C and Daxx 1 domain point mutations have also indicated that the the linkage of glycine97 of SMT3C and the lysine60 of Daxx 1, in which the SMT3C/ SUMO-1 consensus sequence £ZKXE was found. Further, SMT3C was the only member of the SMT3 families capable of self-reacting. Results also suggested that similar mechanism of interaction between SMT3A/B and Daxx 1, which is not in accordance with the model proposed in this study regarding the interaction mechanism between SMT3C and Daxx 1. Although two glycines on the C-terminal end of SMT3A/B were necessary for the interactions with Daxx 4 domains, the SMT3C/SUMO-1-consensus sequence £ZKXE was not detected in the Daxx 4 domain. It is therefore, suggested that the mechanism of the interaction between SMT3A/B and Daxx 4 is similar to that of SMT3C and Daxx 1, that may required different binding sequences that is specific for SMT3A/B.