Formation of these structures at 2-4 days after injection of WT-MLN1 cRNA (a and b), and at later stages 5-8 days after injection (c-e), are shown. No such structures were observed in control oocytes (injected with H₂O) (f) incubated for the same periods of time (1-8 days after injection). Scale bar: 100 μm.

In (a) the Bodipy FL C₁₁-PC probe was used to assess the phospholipase activities of WT-MLN1 (closed triangles), SL-MLN1 (open squares), and the control (H₂O-injected oocytes) (+/−SEM, n=12). The activity was estimated in LE/L-containing oocyte extracts for each of these groups by measuring the increase in fluorescence intensity units (FIU) several days after injection. In (b) the K+, Na+, and Ca2+ conductances, expressed in picosiemens (pS), of WT-MLN1 (open bars) and SL-MLN1 (dark bars) are compared. The unitary conductances were 81+/−6.4 pS (+/−SEM, n=8) for K+, 77.9+/−7.8 pS (n=8) for Na+, and 30.4+/−5.5 pS (n=8). Single channel traces were taken on day 2 (c) and day 6 (d) after injection of oocytes with WT-MLN1 cRNA (top traces) or SL-MLN1 cRNA (bottom traces). The traces were taken at −90 mV in the presence of 100 mM KCl in the pipette solution. The increase in open state probability (nPo) several days after injection (e) is similar in WT-MLN1-(open triangles) and SL-MLN1-expressing oocytes (closed circles) (+/−SEM, n=12). The lysosomal exocytosis was assessed by determining the amount of N-acetyl-beta-D-glucosaminidase (NAG) (+/−SEM, n=6) released into the bath solution from untreated whole oocytes (−) and from oocytes treated for 30 min with 10 μM ionomycin (+), as shown in (f). The three pairs of bars are marked underneath from left to right for H₂O-injected oocytes (Ctrl), or oocytes expressing either WT-MLN1 (WT) or SL-MLN1 (SL). The released amount of NAG is expressed as % of its total content in the cells, as explained in Methods. (g-h). Effects of phospholipase inhibitors on the enzyme activity and on the extrusion of TVS from oocytes expressing WT-MLN1. Panel (g) shows the changes in PLA activity of WT-MLN1 in LE/L-containing oocyte extracts (estimated by measuring the fluorescence of the Bodipy FL C₁₁-PC probe) mediated by the PLA₂ inhibitors, aristolochic acid (Aris, 40 μM) and bromophenacyl bromide (BPB, 40 μM), as well as by the inhibitor of phosphatidylinositol-specific PLC, U73122 (2 μM), and the inhibitor of phosphatidylcholine-specific PLC and PLD, D609 (40 μM) (+/−SEM, n=10). Panel (h) shows that BPB (open triangles) and ArisA (closed squares) markedly reduce the percentage of whole oocytes extruding TVS, while the inhibitors of PLC and PLD (crosses and open circles) do not show significant changes in comparison to the untreated oocytes (Untr.) (closed circles)(+/−SEM, n=6).

Fibroblasts from normal subjects co-stained with the green fluorescent probe, calcein (1uM), for visualizing the cytoplasm, and with calcein in combination with LysoTracker red DND99 in (a), calcein with the LE/L membrane probe, DiIC₁₆(3) in (b and e), or DiIC₁₆(3) without calcein (f-i). (Panels c and d), GFP fluorescence of normal fibroblasts transfected with GFP-tagged WT-MLN1, showing the localization of MLN1 on the cell surfaces, particularly in long thin processes (c) and on the perimeter of bulb-shaped TVS located along a process (d). Panel (e) depicts LE/L membrane staining along the perimeter of a series of bulb-shaped TVS connected by nanotubular tethers. For panels (a-e), scale bars are 10 um. In panels (d and e) representative bulb-shaped TVS are marked with arrowheads. Panels (f-i) show the surface of normal (f and h) or MLIV (g and i) fibroblasts. For panels (f-g) the scale bars correspond to 10 um. Panels (h) and (i) represent an enlarged view of the insets in panels (f) and (g), respectively. Scale bars correspond to 2.5 um. Panel (j) shows the results of the quantitative digital image analysis (CellProfiler). The percentage of total cellular area per 2 dimensional field that is occupied by the bulb-shaped TVS was calculated. Results were normalized to show the TVS Ratio of fibroblasts from normal subjects when normalized to MLIV fibroblasts (untransfected), or compares MLIV fibroblasts transfected with cDNA constructs containing the transcripts: WT, SL, L106P and F465L, when normalized to the vector-transfected MLIV fibroblasts.

This site is located on the lumenal/extracellular side of MLN1 near its putative 1^st transmembrane domain, while the pore region is between the putative 5^th and 6^th transmembrane domains (a). The oocyte shown in (b) has been injected with GFP-tagged SL-MLN1 mutant cRNA. Oocytes injected with GFP-tagged WT-MLN1 cRNA are shown in (c) and (d). Only the green fluorescence of the GFP-tagged proteins is shown in (b-d). Both GFP-emitted green fluorescence and nonspecific red autofluorescence from the oocyte contents are displayed in (e-g). (e) oocyte expressing GFP-tagged WT-MLN1), (f) oocyte expressing GFP-tagged SL-MLN1 mutant, and H₂O-injected oocyte (g). Scale bar: 100 μm. (h) Percentage of oocytes extruding TVS during the period of 1-8 days after injection of WT-MLN1 cRNA in untreated (ML) or actinomycin D-treated oocytes (ML+ACT), or after injection of SL-MLN1 cRNA (SL) or of H₂O (H₂O, coinciding with the symbols for SL) (+/− SEM, n=6). Single oocytes from the different groups were distributed in separate wells of 96-well plates, and all the wells were checked every day under the microscope to estimate the extrusion of TVS from the oocytes. In these experiments 25-80 oocytes were injected for each of these conditions. (i) Rates of increase in GFP fluorescence intensity after injection of oocytes with WT-MLN1 (wt) or SL-MLN1 cRNAs (SL), 2, 5 or 8 days after injection (+/−SEM, n=16).

Panel (a) shows Western blots of the WT and SL fragments of MLN1 produced in vitro as described in Methods. TLC and the Bodipy FL C₅ –HPC probe were used to measure the PLA₂ activity of the WT and SL fragments of MLN1 and the reduction of this activity of WT-MLN1 by 40 μM BPB and aristolochic acid (AA), as shown in (b). In the experiments on the model membrane system the WT-MLN1 fragment mediated formation of tubulo-vesicular structures after treating liposomes that were prepared from Folch fraction I phospholipids stained with sulforhodamine-PE for 5 hr at 37°C (c), while no such structures were observed with the SL-MLN1 fragment (d). The WT-MLN1 fragment also generated tubules and vesicular enlargements from liposomes formed from asolectin phospholipids stained with NBD-PS (e-j). The SL-MLN1 fragment again failed to produce such structures in this model membrane system (k). Panel (l) summarizes the quantitative data from the experiments on the liposomes. The percentage of tubular structures per field generated under each of the different conditions was normalized in relation to that mediated by the WT-MLN1 fragment that was taken as 100%. The PLA₂ inhibitors, BPB and aristolochic acid (AA), reduced significantly the number of tubular structures (+/−SEM, n=6). Scale bars are 50 μm.

Panels (a-b), model depicting how introduction of conical shaped membrane phospholipids by PLA2s, or inverse conical shaped phospholipids by lysophosphatidic acid acyl transferase can lead to a biomechanical change in the shape of a lipid bilayer sheet (adapted from [44]). Panel (a), sagittal section of an outward membrane curvature, (b), cross-section of a tubular membrane process showing the role of membrane curvature in mediating the shape of a tubule. Panels (c-e), Proposed model for how the PLA2 could facilitate membrane remodeling during the extraversion of the lysosome and incorporation of its membrane into the TVS forming on the cell surface. (c) At the lumenal face of the neck of the fusion pore, (d) by introducing positive curvature into the LE/L derived membrane patch during exocytosis, to form an inside-out vesicle in the growing TVS (e).