GFPuv is a good fluorescent reporter for target selection toward structural study
In this study, we demonstrated that GFPuv showed proper expression, folding, and monodispersity by insect cell expression (Fig. 1). GFPuv also reportedly showed highly improved fluorescence, as compared to wtGFP, when expressed in mammalian culture cells, such as Chinese hamster ovary cells.36 Therefore, GFPuv may function as a good fluorescence reporter for universal recombinant expression systems including eukaryotic cells, as well as the E. coli system, in which the protein was originally developed36 and used as a reporter.10
Notably, GFPuv exhibited favorable secreted expression, as well as cytosolic expression, in Sf9 (Fig. 1) and High Five (data not shown) insect cells. GFPuv was also reported to have good secretion efficiency by an S2 insect cell expression system.49 In this study, we further confirmed that GFPuv itself retains proper folding and monodispersity within all expression spaces in insect cells, such as the cytosol, organelles, and extracellular spaces (Fig. 1), and it properly reported the folding and oligomerization states of the target proteins, mGluR1LBD and T1RsLBD (Figs. 3, 4, 5, 3–5), localized within the various expression spaces. Secreted expression by a eukaryotic cell system is a promising production method for eukaryotic extracellular proteins and the extracellular regions of membrane proteins, because the proteins have better chances of being folded correctly, with proper posttranslational processing and modifications. Nevertheless, the GFP-based precrystallization selection has not been extensively applied to secreted expression. This might be partly because of the poor expression of the reporter GFP, as GFP produces fluorescence only by cytosolic expression in E. coli,50 and wtGFP showed insufficient secretion, even by eukaryotic expression.28, 29 Our study indicated that the application of GFPuv enables GFP-based target evaluation and screening for extracellular proteins or membrane proteins with a GFP-tag on their extracellular side, using eukaryotic secreted expression systems. However, it should be noted that the general applicability of GFPuv to secreted expression might need to be further tested, as the secretion of GFPuv was hardly observed in some cases on HEK293 expression (Takagi, personal communication).
In addition, the high fluorescence intensity from GFPuv also facilitates quantitative expression analysis in living Sf9 cells during culture (Fig. 2). It allows real-time monitoring of the production of the GFP-fusion protein for a single cell culture run, as the culture container can be kept closed, and thus is useful for quick screening of expression conditions or high-expression targets. The preparation of a standard curve or the application of internal standards for each measurement, using known amounts of GFP, will probably enable estimation of exact amounts of the samples of interest in the culture.
Secretory process and structure maturation of the extracellular domains of class C GPCRs
The extracellular domains of the class C GPCR members are important targets for structural studies as well as agonist/antagonist-binding studies, because the regions are generally responsible for binding of the major ligands for the receptors. Although some reports have described the overexpression of those regions,51–54 only the mGluRs have been analyzed successfully by crystallography.33–35 So far, conventional analyses of recombinant protein expression have often focused only on the amounts of the expressed products, evaluated by western blotting, for instance. However, to achieve successful sample preparation for structural studies, expression analyses should also address whether the recombinant proteins acquire their functional structures. We used the aforementioned GFP reporter system and examined the expression of the extracellular domains of the two class C GPCR members, mGluR and T1R, from the standpoint of their folding and oligomerization states in connection with their expression locations, such as the cellular space (cytoplasm and organelles) or the extracellular space after secretion.
In the case of mGluR1LBD, the Sf9-expressed products were secreted well, and the sample showed a narrow, monomodal elution peak (Fig. 3), indicating that the secreted mGluR1LBD was properly folded. These results suggested that mGluR1LBD achieves its proper folding and membrane trafficking by using the ubiquitous systems provided in most cells, including Sf9 cells. These results are also consistent with the fact that mGluR1LBD, expressed in insect cells, served as a successful sample for the previous crystallographic analyses.33, 34
In contrast to the situation with mGluR1LBD, T1RsLBD hardly secreted, and this problem was not solved even by the addition of the AKH signal peptide (Fig. 5), which resulted in the successful secretion of GFPuv (Figs. 1 and 2). T1RsLBD was also not secreted by mammalian cell lines, including HEK293 (Yamashita, Ashikawa, and Takagi, unpublished results). Notably, the FSEC profiles of the cellular-expressed T1RsLBD showed severe polydisperse patterns, with large aggregated fractions. These results clearly indicated that T1RsLBD, expressed under the conditions tested in this study, were unable to achieve proper folding and trafficking in Sf9 cells, and are unsuitable as samples for structural analyses even though some expression was observed in the cellular space. Interestingly, the expression of T1RsLBD-GFPuv and AKHss-T1RsLBD-GFPuv showed some differences upon western blotting and confocal microscopic observations: the former displayed less degradation and a broad distribution in the cells, while the latter included more degradation and spot-like localizations with high fluorescence intensities in the cells (Figs. 4 and 5). These results imply that the former failed in translocation to the ER, while the latter barely translocated to the ER, but was excluded from the subsequent secretory pathway by ER quality-control systems, and at least a part of them was subjected to lysosomal degradation, although further experiments are needed to define their exact localizations in the cells.
There are number of examples that cell-surface membrane proteins possess their specific folding-assistant and/or membrane-escort systems, to prevent misfolded, wrongly organized, or accidentally expressed proteins from reaching the cell surface. In the case of other class C GPCR members, for example, the surface expression of GABAB receptor is achieved through the direct interaction of the C-terminal region of the heterodimeric partner, the GB1 and GB2 subunits, resulting in the masking of the ER retention signal in GB1.55 In the transport of V2R pheromone receptors to the cell surface, the major histocompatibility complex M10 family members, together with β2-microglobulin, reportedly function as escort molecules.56 In a similar manner, T1Rs might require their own chaperones and/or membrane escort systems for proper folding and trafficking, that might specifically exist in taste-receptor expressing cells (such as taste cells), but not in other cells such as Sf9 or HEK293 cells.
The results of mGluR1LBD and T1RsLBD expression displayed the correlation between the localization of the expressed products and their folding status: the secreted sample (the secreted fraction of mGluR1LBD) was favorably folded with a monodisperse hydrodynamic state, while the samples excluded from the secretory pathway (the cellular fractions of mGluR1LBD, T1RsLBD, and AKHss-T1R1LBD) were poorly folded (Figs. 3, 4, 5, 3–5). These results imply that the maturation of the functional structures of the class C GPCR extracellular domains is not attained spontaneously, but through the secretory pathway, by receiving posttranslational-modifications and excluding of the misfolded subfractions by the quality-control systems.
It should be noted that the observed results might not be simply extrapolated to the full-length receptors. Nevertheless, the results in this study might reflect some of the structural maturation processes of cell-surface proteins with extracellular domains, including the class C GPCRs, as well as secreted proteins.