Role of Sumo-conjugation in Development
We have recently become interested in understanding the biological role of a post-translational protein modification known as Sumo-conjugation. Sumo is a small (~100 amino acid) protein with about 20% homology to ubiquitin. Like ubiquitin, it is attached to other proteins via an isopeptide linkage between its C-terminus and e-amino groups on the target proteins. While ubiquitylation of proteins targets them for proteasomal degradation, the purpose of Sumo conjugation is not well understood. We have recently identified the Drosophila genes encoding the enzymatic machinery responsible for Sumo conjugation. This includes the genes encoding Sumo itself, the Sumo-activating enzyme (SAE1/2), the Sumo-conjugating enzyme (Ubc9), and the Sumo processing/deconjugating enzyme (Ulp1). Using a combination of biochemical, molecular, and genetic analysis our goal is to illuminate the biochemical and developmental roles of this process.
Sumo-conjugation may activate Dorsal in two ways. Our interest in Sumo-conjugation was sparked by our discovery that Dorsal is a substrate for this modification. Thus, our initial analysis of this pathway has focused on its interaction with Dorsal (Bhaskar et al., 2000; Bhaskar et al., 2002). We have found that conjugation of Sumo to Dorsal stimulates this factor in two ways. First, this modification appears to favor nuclear localization of Dorsal. Since the conjugation machinery appears to reside in the nucleus, we propose that the increase in nuclear localization may be due to an effect of Sumo on nuclear retention as opposed to nuclear import. Second, Sumo-conjugation appears to make Dorsal a more potent transcriptional activator. To explore this effect, we mapped the lysine in Dorsal to which Sumo is attached. Mutagenesis of this lysine resulted in a constitutively activated form of the factor. These findings suggest that, in its unmodified form, the Sumo-conjugation site may form part of a docking surface for a negatively acting regulatory factor. We hypothesize that either mutagenesis of the site or conjugation of Sumo to the site results in displacement of the negatively acting factor and a consequent increase in transcriptional activation.
Sumo-conjugation and the stress response. We suspect that conjugation of Sumo to target proteins may be a stress response mechanism. The first line of evidence in favor of this hypothesis derives from experiments in which we examined the total level of Sumo-conjugates in Drosophila cells before and after the cells have been subjected to environmental stresses such as heat shock or high concentrations of hydrogen peroxide. In the absence of stress, the majority of the Sumo is found in the unconjugated form, but with increasing stress high molecular weight Sumo-conjugates are formed depleting the pool of unconjugated Sumo. Apparently cellular stress produces an intracellular signal that activates the conjugation machinery.
How does Sumo-conjugation help cells cope with stress? So far our experiments suggest two answers.
(1) We have observed that Sumo-conjugation is critically required for the ability of fruit flies to mount a challenge to septic injury (Bhaskar et al., 2002). In particular, we have found that blocking Sumo-conjugation blocks the activation of genes encoding antimicrobial peptides that usually occurs in response microbial infection. This so-called innate immune response is not normally considered to be a stress response in the same sense that response to heat shock or hydrogen peroxide is considered a stress response. However, our findings suggest that mechanistic similarities may underlie these responses.
(2) We have immunopurified some of the high molecular weight conjugates that appear in response to stress and identified them by mass spectroscopy. Two of the three conjugates that we have identified are aminoacyl-tRNA synthetases. Further analysis suggests that conjugation of Sumo to these enzymes may facilitate the formation of a high molecular weight aminacyl-tRNA synthetase complex that may coordinate tRNA charging. Thus, perhaps Sumo-conjugation allows cells to cope with stress by activating tRNA aminoacylation. This may help to replenish stores of aminoacyl-tRNAs depleted by the synthesis of stress response proteins.
Multiple roles for a Sumo processing/deconjugating enzyme. Based on homology to a gene encoding a yeast Sumo-deconjugating enzyme, we previously identified a Drosophila gene (Ulp1) that encodes such an enzyme. We found that, in addition to catalyzing the cleavage of isopeptide linkages in Sumo-conjugated proteins, this enzyme can also cleave off the extension found at the C-terminal end of the primary translation product of the gene encoding Sumo. This cleavage process is required for the normal maturation of Sumo. Thus, the enzyme appears to play both positive and negative roles in the formation of Sumo-conjugates. This could explain the biphasic response curve we see when we assay the activity of Dorsal, a transcription factor that is stimulated by Sumo-conjugation, as a function of Ulp1 concentration.
Our recent studies suggest yet another function for Ulp1. We expressed a Ulp1-GFP fusion protein in Drosophila tissue culture cells and then localized the protein by fluorescence imaging. We found that the fusion protein is localized to multiple discrete foci in the nuclear membrane, perhaps representing nuclear pore complexes. This finding suggests that as a component of the nuclear pore complex, Ulp1 may play a role in the nuclear import of Sumo and/or Sumo-conjugates. To test this idea, we have examined the effect of knocking out Ulp1 activity on the subcellular localization of Sumo. We find that that in the presence of functional Ulp1 Sumo is largely nuclear, while in the absence of Ulp1 most of the Sumo relocalizes to the cytoplasm.