DEAF1 is a requirement for the interaction with LMO4 in vivo through mass action effects, or by recruiting other protein partners that form oligomers and/or promote molecular clustering. Alternatively, LMO4 binding to DEAF1 may inhibit XL-184 tetramerisation and prevent binding of an exportin to the DEAF NES, through direct steric inhibition, or by recruiting other partners that block binding. These hypotheses remain to be tested, although similar ‘masking effects’ of NESs have been shown for other proteins. For example in the tetrameric form of p53 tumour suppressor, the NES lies within the tetramerisation domain and nuclear export appears to be regulated by the oligomerisation state of p53. Or in the case of the APC tumour suppressor where binding of CRM1 exportin and active–Ran allow movement of the helix containing the NES within the coiled coil moiety that then unmasks the NES. An alternative explanation of our results is that DEAF1 could simply act as a nuclear localisation mechanism for LMO4. LMO4 is predominantly nuclear, although it does not have an NLS of its own. Other known binding partners of LMO4 all have NLS sequences, e.g., LIM-domain binding protein 1, CtBPinteracting protein and oestrogen receptor a and metastasis-associated gene 1. LMO4 can therefore be transported into the nucleus by the most abundant/available partner, via a piggy-back mechanism. This is commonly seen for many NLSdeficient nuclear proteins. Whichever is the case, the effect on either protein on the subcellular localisation of the other will modulate the activity of that protein, modulating the type and composition of complexes formed. Our results begin to shed light on the mechanism of some breast cancers where high levels of LMO4 are present. Under these circumstances DEAF1 may become sequestered into nuclear LMO4-DEAF1 complexes, perturbing the normal functions of DEAF1. Forming aberrant transcription complexes can prevent the formation of normal complexes. For example Rac3 GTPase, which is linked to breast cancer, cellular migration and adhesion, is transcriptionally upregulated by DEAF1 in immortalised mammary epithelial cells. Rac3 is a candidate target for LMO4:DEAF1 complexes. If these complexes are aberrantly forming due to the persistence of LMO4, it could provide a mechanism to increase cell proliferation and migration during breast oncogenesis. In contrast, the expression of hnRNP, a repression target of DEAF1 that is seen at high levels in some cancers, may be free to accumulate if other DEAF1 complexes are unable to form. Manipulating the interaction between LMO4 and DEAF1 to prevent the formation of aberrant transcriptional complexes may represent a potential novel therapeutic strategy with which to combat breast cancer. Temporal lobe epilepsy is the most common form of epilepsy in adults, and mesial TLE is often medically intractable. The majority of mTLE patients have seizures that cannot be well controlled with classical anticonvulsant drugs that target ion channels, suggesting that the pathogenesis of mTLE differs from other types of epilepsy.