Figure 1. Size separation and characterization of NLPs.

(a) Representative size exclusion chromatograms for various NLP preparations as indicated. NLPs were detected by absorption at 280 nm. Typically, fractions comprising 9–13 ml were collected (black horizontal bar). (b) Coomassie-stained SDS PAGE of NLPs. For each preparation, the amount of VAMP2 relative to ApoE was determined using densitometry. (c) Representative negative-stain EM micrographs of NLPs. The top row are SNARE-free NLPs. The bottom row are NLPs loaded with 30 v-SNARE copies. NLPs marked with * are oriented perpendicular to the imaging plane and show the flat disc structure. Scale bar = 25 nm. (d) Distribution of NLP diameters for a representative vNLP15 sample, determined from analysis of micrographs as in (c). A normal distribution fit is shown (red line). (e) Boxplot of representative NLP sizes under various conditions. NLPs containing lipid-anchored VAMP2 (vC45L, vC45M, vC45H for low, medium, and high copy numbers of C45 lipid-anchored VAMP2, bearing ~1, 4, and 15 copies) had sizes comparable to NLPs bearing similar loads of wild-type VAMP2 (vNLP1, vNLP4, and vNLP15). The activity of these NLPs was tested in an established bulk fusion assay with t-SNARE-reconstituted liposomes (Figure 1—figure supplement 1).

DOI: http://dx.doi.org/10.7554/eLife.22964.002

Figure 1—figure supplement 1. Bulk content release assay (Bello et al., 2016; Shi et al., 2012) shows that the fusion of vNLPs with t-SNARE-reconstituted small unilamellar vesicles (t-SUVs) is SNARE-dependent.

(a) Release of calcium from the liposome lumens as a function of time. t-SNARE liposomes initially entrapping calcium were mixed with vNDs. Fusion with vNDs leads to leakage of calcium through fusion pores. Binding of leaked calcium to the calcium-sensitive dye Mag-fluo-4, included in the fusion buffer, results in increased fluorescence. The maximum fluorescence was determined by adding detergent to release all entrapped calcium at the end of the reaction. Results are displayed as percentage of maximum fluorescence as a function of time. Nine independent measurements per condition were averaged. (b) Bulk calcium release rates are comparable for MSP or NLP discs bearing ~8 and ~15 copies of VAMP2, respectively, even though fusion pores can, in principle, grow to much larger sizes for NLPs. Thus, either the bulk contents release assay is insensitive to pore properties under these conditions or pores with similar properties occur in both cases. Single-pore measurements indicate that about 10% of vNLP15 pores reach large sizes (see main text). Eleven and eight independent measurements were averaged for MSP vND and vNLP15 conditions, respectively. The error bars represent standard errors of the means for both panels (a) and (b).

DOI: http://dx.doi.org/10.7554/eLife.22964.003