In this report we have shown that FOXP3 and Siva physically interact. We have mapped the binding activities to limited domains within each protein. FOXP3 and Siva-1 both independently repressed IL-2 gene expression. One mechanism by which Siva-1 represses IL-2 is via inhibition of NFκB. In contrast, FOXP3 repressed the activity of both IL-2 transactivators tested here, NFκB and NFAT. Although we did not reveal a functional interaction between FOXP3 and Siva with regards to IL-2 repression, the biophysical interaction invites questions regarding how Siva might contribute to regulation of FOXP3-expressing Treg cells.
The physical interaction was demonstrated through a standard protein-protein interaction, Co-IP assay. Two Siva isoforms have been identified and both Siva isoforms interact with FOXP3. The Siva-2 isoform lacks the second exon that encodes for the SAH and DDHR domains [28
], which were found to be dispensable for binding FOXP3 in subsequent Siva truncation Co-IP experiments. The Siva C-terminus, which encompasses both the putative B box and Zn F domains, contained FOXP3 binding activity.
The Siva C-terminus is enriched with cysteine residues that could be important to the protein's tertiary structure and function. Paired cysteine residues are associated with intramolecular disulfide bond formation. The Siva C-terminus contains six paired cysteine residues. Nestler et al
] used truncation mutants to show that Siva coordinates three zinc ions, two of which are associated with the C-terminus [42
]. Although a few groups have found that the Siva C-terminus is necessary to interact with other binding partners [43
], no one has described whether specific C-terminal point mutations block Siva function. Future site-directed mutagenesis studies of the Siva C-terminal cysteine residues could be informative to further characterize the physical interaction between Siva and FOXP3.
FOXP3's Siva-binding activity is contained within a broad central region that includes the zinc finger, the leucine zipper and the RUNX1 binding domain (amino acids 106-332). Mutants missing either the first 105 N-terminal amino acids or the C-terminal forkhead domain sustained binding to Siva-1. Further experiments are needed to define a minimal region of FOXP3 involved in binding to Siva-1. The FOXP3 leucine zipper domain is required for homodimerization, IL-2
repression, and Treg
suppressive function [11
]. Given that the leucine zipper domain is within the region that binds Siva-1, one question we are interested in addressing in the future is whether known leucine zipper IPEX mutations (Δ250, Δ251)[11
] interfere with FOXP3 binding to Siva-1.
One limitation of this study was that technical challenges obstructed our efforts to detect a biophysical interaction between endogenously expressed FOXP3 and Siva. Also, we have not yet performed sufficient experiments to address whether the biophysical interaction could be linked to any functional interaction. In order to test whether the biochemical interaction between FOXP3 and Siva-1 affects their functional interaction, a minimal region of biochemical interaction for at least one protein needs to be identified. More extensive point mutant Co-IP experiments for at least one of the proteins could be informative towards defining the relationship between the biophysical interaction and any proposed functional interaction for FOXP3 and Siva-1.
Our initial hypothesis was that FOXP3 and Siva-1 might functionally interact to repress IL-2 gene expression. We did not observe an additive repressive effect between FOXP3 and Siva-1 on IL-2 gene expression. Instead, our data suggests that FOXP3 might be a dominant suppressor that masks the repressive effects made by Siva-1 on IL-2 gene expression.
Our analysis of two IL-2
transactivators supports the assertion that Siva-1
affect shared and different IL-2
regulatory mechanisms. Consistent with previous reports, FOXP3
repressed NFκB [22
] and NFAT [13
] transactivation reporters. Siva-1
inhibited NFκB, but had no effect on NFAT in cells stimulated with PMA and Ionomycin. We observed a small, but significant additive repressive effect between FOXP3
on NFκB activity.
The NFκB-repressive effects that we observed for FOXP3
are consistent with published reports. The effects of each gene on NFκB activity may be partitioned into distinct cytoplasmic and nuclear signalling events. In resting cells, cytoplasmic IκBα sequesters the canonical NFκB subunits p65 and p50, preventing their nuclear translocation. In response to stimulus, IκBα breaks down and releases the NFκB subunits, which translocate to the nucleus [47
blocked NFκB nuclear accumulation via stabilization of cytoplasmic IκBα [36
]. On the other hand, FOXP3
blocked NFκB transactivation potential by mechanisms independent of nuclear translocation and DNA binding [13
]. Further, our group showed that FOXP3's
promoter-dependent effects on NFκB transactivating potential were independent of nuclear translocation. FOXP3 enhanced HIV-1 gene expression by increasing the amount of NFκB p65 subunit bound to the LTR (long terminal repeat) [10
]. To summarize, in separate experimental systems, inhibition of NFκB activity by Siva
has been shown to occur upstream of IκBα degradation and nuclear translocation, whereas FOXP3 inhibits NFκB activity downstream of IκBα degradation. The biophysical interaction that we have presented suggests that another regulatory mechanism could be occurring that would involve co-localization of FOXP3 and Siva-1 within the same subcellular compartment.
Given that Siva has displayed both nuclear and cytoplasmic subcellular localization [43
], the possibility remains that Siva could regulate gene transcription from within the nucleus in addition to its cytoplasmic effect on NFκB. As of yet, published reports of a nuclear specific function for Siva are lacking. Most detailed investigations of Siva function have examined Siva in relation to transmembrane signalling molecules or mitochondrial-associated apoptosis regulators thought to be present in the cytoplasm, not in the nucleus [28
]. In future studies, it will be interesting to look into whether Siva displays nucleus-specific activities such as direct binding to DNA or chromatin.