Discussion
The roles of the VWFpp, D1, and D2 domains, in both multimer assembly and storage of VWF in WPBs, are well established . Each of the two D1 and D2 domains contains a CGLC amino acid motif which is considered to be a consensus sequence of the PDI enzyme family, having a catalytic role in N-terminus interchain disulfide bonds between the D3 domains to form large multimers in the Golgi compartments . Furthermore, the propeptide serves as an intramolecular chaperone, aligning the dimeric pro-VWF subunits in the appropriate special alignment to facilitate interdimer N-terminal disulfide bond formation during multimerization and to ensure VWF helical assembly into tubules in nascent WPBs . Another probable function attributed to the VWFpp is the sorting of the mature VWF into storage granules via an intrinsic signal (linear sequence or conformation) in the trans-Golgi .
Nonetheless, here in this study, we demonstrated that VWFpp missense variants hamper the exit of VWF from the ER, resulting in VWF sequestration in the ER compartment. The present study showed that four out of the six VWFpp variants (p.Gly55Glu, p.Val86Glu, p.Trp191Arg, and p.Cys608Trp) severely impaired secretion, along with intracellular retention of these variants. Further, we illustrated that VWF proteins with the mutations residing in propeptide are accumulated in the ER, and do not appear to generate WPBs-like granules. We demonstrated an increased co-localization of these VWF mutants with the luminal ER marker, PDI, when is compared with wt.
Multiple QC mechanisms operate within the ER to ensure fidelity of protein expression in cells. These mechanisms, in general, rely on a repertoire of molecular chaperones and protein folding facilitators that promote correct protein folding and detect defective proteins for degradation . Additionally, ER-Golgi export QC is another checkpoint that ensures the secretion of correctly folded proteins through cargo receptors. By recruiting cargos for specific capture into COP I and COP II vesicles, cargo receptors can monitor the folding status of their clients before capturing into vesicles . Recently, Lopes-da-Silva et al showed that the ARF family of proteins contributes to the ER-Golgi trafficking of the VWF . In the active state, when bound to GTP, ARFs initiate coat and cargo recruitment onto intracellular membranes to conduct anterograde transportation . The switch between inactive and active state occurs through GTP/GDP exchange/binding that is mediated by guanine nucleotide exchange factors (GEFs). Lopes-da-Silva et al demonstrated that GBF1 (one of the guanine nucleotide exchange factors) reduction affects the rate of VWF movement between ER and Golgi, indirectly implicating the ARF family of proteins as the major recruitment proteins for anterograde transport of VWF . Another recent study demonstrated that the rate of the anterograde transport between ER and Golgi and consequently the morphology of the WPBs is dependent on the Sec22b-containing SNARE complexes. Sec22b is a component of ER-derived COPII-coated vesicles that direct proteins from ER to the Golgi, and interacts with Sec23/Sec24 of the COPII coating complex, which is essential for cargo recruitment in the ER. Both Sec22b silencing and Sec24 deficiency were reported to result in ER retention of the secretory proteins . Since our microscopic analysis suggested the accumulation of VWF mutants in ER, we tested the hypothesis that the misfolded VWF hampers its recruitment into COP I and COP II vesicles (mediated by ARF or Sec24) by docking one of the active ARF protein (GTP bound ARF1) and Sec24 structures onto our N-terminal VWF model. The fact that most of the variants occur at the ARF1/or Sec24-VWFpp putative interfaces on these docks strongly support the idea that the ER accumulation might be a result of these mutations interfering with the ARF/or Sec24-VWF recruitment process for anterograde transport. Nevertheless, we propose this mechanism as a putative affected QC pathway, and we can not deny involvement of the other QC checkpoints. Therefore, our structural analysis results will need to be verified by more experimental evidence in the future.
Accumulation of VWF mutants in ER was associated with reduced intracellular packaging and defect in secretion of VWF. Nonetheless, the homozygous expression of these propeptide variants still showed a minor quantity of secreted VWF, only ~ 10% of wt, as well as a trail of the oligomers (i.e. tetramers). This suggests that a small number of dimers still succeed to rescue from the ER and reach their destination in trans-Golgi. Further, all of the four causative variants showed a 65%-75% fall in the secretion of VWF when heterozygously expressed, indicating their dominant impact on secretion, along with an increased colocalization with the ER marker (though with a lesser extent compared with homozygously expressed variants). Besides, the co-expression of these VWFpp variants with wt revealed a deficiency in VWF multimer assembly, lacking large/and intermediate multimers. We consequently assume that the limited number of the mutant homodimers/or heterodimers rescued from the ER will successfully cross-over to the Golgi and will interfere with the accurate orientation of the propeptide within the dimeric bouquet and succeeding multimer elongation, based on our observations made on the VWF N-terminal model´s orientation.
Furthermore, in the current study, the biogenesis of WPBs in co-expressing cells of wt/mutants was explored by computerized morphometric analysis of the whole pseudo-WPBs population using state-of-the-art arivis Vision4D. Initially, microscopic inspections revealed a reduction in the fraction of the cells generating pseudo-WPBs in cells co-expressing wt/mutants. The computed average number of pseudo-WPBs per cell was ~7 for cells expressing wt, this number was less for the cells co-expressing p.Gly55Glu, p.Val86Glu, and p.Cys608Trp (~ 3.4-4.0 per cell). Altogether, both declines in the portion of cells forming pseudo-WPBs and their average number per cell for the later co-expressed mutants indicate their dominant-negative impact on the formation of the storage vesicles. The reduction in the number of WPBs was previously also reported for the mutations detected in type 1 and type 2A VWD, which was associated with intracellular VWF retention and decreased secretion . However, these previous reports were based on only visual inspections but not the automated quantifications as we have done in the current study.
The WPBs have a distinct rod-shaped morphology and their length varies between 0.5 and 5 µm in endothelial cells. Karampini et al have quantified the length of WPBs in cultured endothelial cells (HUVECs) and they reported the average length of ~ 2.2 µm . Furthermore, in our lab, quantification WPBs length in the blood outgrowth endothelial cells (BOECs) showed the various WPBs lengths with an average length of 2.281 µm ± 0.01 (data not shown). Likewise, here, the quantitative analysis of the whole pseudo-WPBs population in HEK293 cells showed various sizes of pseudo-WPBs (Figure 3E) with an average length of 2.180 ± 0.09 µm in cells expressing wt. The co-expressed p.Trp191Arg caused a reduction in both length and width of the WPBs-like granules. It showed also slightly augmented sphericity, which indicates the storage granules have a more rounded shape than elongated. Although co-expression of the p.Gly55Glu, p.Val86Glu, and p.Cys608Trp seems not to affect the size of pseudo-WPBs, the reduction in the computed mean intensity values implies unspecified changes in their configuration, besides their influence on the number of the generated pseudo-WPBs. Variants p.Asn211Asp and p.Gly334Glu did not demonstrate any defect in rVWF expression and function in vitro. Based on our observations, we suggest these two variants should be considered as benign variants rather than pathologic mutations.
In sum, here in this study, we demonstrated that VWFpp mutations hinder anterograde ER-Golgi VWF export, resulting in ER accumulation. Consequently, retention of the VWF inside the ER resulted in no storage vesicle formation in the homozygously transfected cells, and a decrease in the number of generated pseudo-WPBs in heterozygously transfected cells. Furthermore, the heterozygous VWFpp mutants cause changes in the size and configuration of the WPBs-like granules, with variability in the extent/type of these changes. This highlights the consequence of the VWF variants on biogenesis and configuration of WPBs, which in turn may affect the recruitment of the other WPBs’ cargos.