During my PhD project, I have focused my explorations on transmembrane proteins of the secretion pathway. Single-pass transmembrane proteins localised to the endoplasmic reticulum (ER), Golgi apparatus and plasma membrane bear specific sorting motifs which are recognised by specific molecular sorting mechanisms.

While comparing two membrane proteins of the same iron-cluster binding family, it was observed that Cisd1 and Cisd2 proteins possessed an almost identical protein structure: a membrane-anchoring N-terminus, a transmembrane domain and a C-terminus carrying a KKXX motif. Cisd1 was localised to the mitochondria while Cisd2’s localisation was uncertain. Our experiments showed that Cisd2 was localised in the ER. To understand why both proteins diverged in localisation, we dissected the effect of different localisation determinants on the proteins by creating chimeric proteins. These chimeras allowed identifying the C-terminal KKXX motif and the transmembrane domain of Cisd2 as responsible for its ER localisation. We also suppose that Cisd2 could have a function in clustering ER membranes.

More generally, the transmembrane domain plays an important role in localisation of membrane proteins. Previous studies have shown that short transmembrane domains favoured an ER localisation, others have shown that they favoured a localisation in the Golgi apparatus. On the contrary, longer transmembrane domains targeted proteins to the plasma membrane. Nevertheless, very few experiments had explored how short transmembrane domains led to distinct localisation between the ER and Golgi apparatus. In this report, I have connected the dots and compared the targeting abilities of transmembrane domains for ER vs Golgi localisation. Our results based on confocal and electron microscopy suggest that topology is a major localisation determinant for targeting of transmembrane proteins. Short transmembrane domains of type I topology target proteins to the ER while short transmembrane domains of type II topology localise proteins to the Golgi apparatus. This topology parameter for protein localisation had been underestimated and deserves further explorations.