An increasing number of human disorders are being linked to mutations in components of the secretory pathway. One example is combined blood coagulation factor V and VIII deficiency, an autosomal recessive disorder leading to hemophilia due to markedly reduced levels of coagulation factors V and VIII in blood plasma. This disorder is genetically linked to the transmembrane protein ERGIC-53 and its soluble luminal interaction partner MCFD2, both of which reside in the early secretory pathway. ERGIC-53 and MCFD2 constitute a transport receptor complex required for the efficient secretion of blood coagulation factors V and VIII. The strict dependence of some secretory proteins on receptor-mediated transport illustrates the need to understand this process in detail. The characterization of transport receptors and their cognate cargo proteins is hampered by the weak and transient nature of the underlying protein-protein interactions which take place in the rather specialized luminal environment of the endoplasmic reticulum (ER). As a matter of fact, many luminal protein interactions of secretory and membrane proteins are missed by standard techniques of interaction proteomics such as affinity isolation or the yeast-two hybrid system. To overcome these substantial technical limitations, we tested if a protein fragment complementation assay (PCA) based on the yellow fluorescent protein (YFP) can be applied in vivo to capture protein-protein interactions inside the lumen of the secretory pathway. YFP PCA relies on complementing YFP from two non-fluorescent fragments (YFP1 or YFP2) which have been fused to two test proteins. If the two test proteins interact, YFP1 and YFP2 are brought into close proximity which induces the correct folding and reconstitution of fluorescent YFP. By successfully applying a YFP PCA inside the lumen of the ER, we could visualize the interaction between ERGIC-53 and its luminal interaction partners MCFD2, cathepsin Z and cathepsin C in a specific manner. Noteworthy, a direct interaction between cathepsin C and ERGIC-53 has been suspected previously but could not be established by chemical crosslinking and affinity purification-based techniques. Thus, YFP PCA is a powerful tool to capture protein interactions inside the secretory pathway. To search for additional cargo proteins of ERGIC-53, we developed a novel genomewide fluorescence complementation-based cDNA library screen. For this purpose, we constructed for the first time a cDNA-YFP1 fusion library which meets all the requirements for probing protein-protein interactions in the lumen of the secretory pathway by YFP PCA. The library was co-transfected with the YFP2-ERGIC-53 bait into mammalian COS-1 cells. Fluorescence activated cell sorting was then used to isolate yellow fluorescent COS-1 cells from which library plasmids were recovered. In a small-scale pilot screen, we identified alpha-1-antitrypsin as potential interaction partner of ERGIC-53 suggesting that ERGIC-53 might bind more cargo proteins than initially assumed. The identification of alpha-1-antitrypsin demonstrates that YFP complementation can be successfully applied to screen a cDNA library for novel protein-protein interactions. This approach should provide a firm basis to map protein interactions inside the secretory pathway in a genomewide setting. With the ability to visualize and quantify protein interactions between ERGIC-53 and its cargo in vivo, YFP PCA is a potent technique to analyze the ERGIC-53/MCFD2 transport receptor complex in more detail. Hence, we used luminal YFP complementation to establish the cargo binding properties of the ERGIC-53/MCFD2 complex and showed that ERGIC-53 can bind cathepsin Z and cathepsin C in a MCFD2-independent manner. This suggests cargo selectivity of the ERGIC-53/MCFD2 complex. While ERGIC-53 can interact with cathepsin Z and cathepsin C in the absence of MCFD2, MCFD2 is selectively required for the recruitment of blood coagulation factors V and VIII. A combination of short interference RNA-mediated ERGIC-53 knockdown, immunofluorescence-based protein localization, and tracking of metabolically labeled MCFD2 revealed a strict dependence of MCFD2 on ERGIC-53 for correct localization and intracellular retention. Our finding that MCFD2 is secreted upon a knockdown of ERGIC-53 explains the lack of MCFD2 that has been reported in ERGIC-53 deficient hemophilic patients suffering from combined blood coagulation factor V and VIII deficiency. In conclusion, this thesis provides deeper insight into receptor-mediated cargo capture by proposing cargo selectivity of the ERGIC-53/MCFD2 transport complex. Furthermore, the development of the luminal YFP PCA provides attractive and promising perspectives to analyze and screen protein interactions inside the lumen of the secretory pathway.