No abstract

No abstract

S-adenosylhomocysteine hydrolase (AHCY) is thought to be located at the sites of ongoing AdoMet-dependent methylation, presumably in the cell nucleus. Endogenous AHCY is located both in cytoplasm and the nucleus. Little is known regarding mechanisms that drive its subcellular distribution, and even less is known on how mutations causing AHCY deficiency affect its intracellular dynamics. Using fluorescence microscopy and GFP-tagged AHCY constructs we show significant differences in the intensity ratio between nuclei and cytoplasm for mutant proteins when compared with wild type AHCY. Interestingly, nuclear export of AHCY is not affected by leptomycin B. Systematic deletions showed that AHCY has two regions, located at both sides of the protein, that contribute to its nuclear localization, implying the interaction with various proteins. In order to evaluate protein interactions in vivo we engaged in bimolecular fluorescence complementation (BiFC) based studies. We investigated previously assumed interaction with AHCY-like-1 protein (AHCYL1), a paralog of AHCY. Indeed, significant interaction between both proteins exists. Additionally, silencing AHCYL1 leads to moderate inhibition of nuclear export of endogenous AHCY.

Protein interaction networks are the basis for human metabolic and signaling systems. Interaction studies often use bimolecular fluorescence complementation (BiFC) to reveal the formation and cellular localization of protein complexes. However, large-scale studies were either far from native conditions in human cells or limited by laborious restriction/ligation cloning techniques. Here, we describe a new tool for protein interaction screening based on Gateway-compatible BiFC vectors. We made a set of four new vectors that permit fusion of candidate proteins to the N or C fragment of Venus in all fusion positions. We have validated the vectors and confirmed self-association of AHCY, AHCYL1, and galectin-3. In a high-throughput BiFC screen, we identified new AHCY interaction partners: galectin-3 and PUS7L. We also describe additional steps in protein interaction analysis, applied for AHCY-galectin-3 interaction. First, we classified the interaction in intracellular vesicles using CellCognition, machine learning free software. Then we identified the vesicles as endosomal pathway compartments, in line with known galectin-3 trafficking route. This offers a platform to rapidly identify and localize new protein interactions inside living cells, a prerequisite to validate in silico interactome data, and ultimately decode complex protein networks.

No abstract