Design, stability, and functional characterization of photoreactive bifunctional-STX (BF-STX) in neurons and cell lines.

A. Chemical structures of STX and bifunctional-STX (BF-STX), featuring a diazirine group connected to an alkyne handle (both shown in red) for photo-crosslinking and click chemistry, respectively. Upon 7 minutes under 350 nm UV illumination in methanol (MeOH), BF-STX exhibited characteristic changes in proton resonances near the diazirine moiety, confirming photoreactivity (see Supplemental Files). BF-STX remained chemically stable in the dark. B. BF-STX rapidly attenuated µ-opioid receptor–mediated responses in POMCEGFP neurons. The µ-opioid receptor, like the GABAB receptor, is Gi/o-coupled and activates G protein-coupled, inwardly rectifying K+ (GIRK) channels in POMCEGFP neurons. In voltage-clamp recordings (Vhold= -60 mV), an EC₅₀ concentration of DAMGO (a µ-opioid receptor agonist) elicited GIRK-mediated outward currents (top trace), which were reduced by 38.7 ± 5.7 % (n=3) following a 15-minute exposure to BF-STX. This heterologous desensitization is consistent with previous observations using the BF-STX parent compound, STX, which causes a 41% attenuation in the GABAB (and µ-opioid) receptor-mediated activation of GIRK channels in POMC neurons (Qiu et al., 2003). C,D. Confocal images showing subcellular accumulation of BF-STX (10 µM, 30 min). BF-STX treated cells were subjected to UV crosslinking (2.5 min at 350 nm, on ice), cell fixation, and fluorescent labelling via copper-click reaction using picolyl-azide Alexa Fluor 488. C. In hypothalamic slices, BF-STX selectively labeled POMC neurons near the third ventricle (3V). D. BF-STX labeled mHypo43 cells.

Chemoproteomic identification of BF-STX–interacting proteins in mHypo43 cells following UV crosslinking.

A. Volcano plots showing enriched proteins identified after 30 min BF-STX incubation under UV crosslinking conditions compared with no-UV controls (left), and after competition with 3× excess unlabeled STX during UV crosslinking compared with no-UV controls (right). Proteins significantly enriched by BF-STX labeling are highlighted, with VDAC1, VDAC2, and VDAC3 among the prominent candidate targets identified after UV exposure. Competition with excess STX reduced enrichment of VDAC family proteins, consistent with competition-sensitive BF-STX labeling. B. Volcano plots showing enriched proteins identified after 5 min BF-STX incubation under UV crosslinking conditions compared with no-UV controls (left), and competition with 1× excess unlabeled STX during UV crosslinking compared with no-UV controls (right). Shorter BF-STX incubation increased enrichment and statistical significance of VDAC1, VDAC2, and VDAC3. Competitive STX treatment reduced enrichment of these proteins, consistent with BF-STX associated labeling of VDAC family members. The x-axis represents log2 (fold change), and the y-axis represents –log10 (p value).

qPCR identification of VDAC1-3 as targets in POMC neurons

A. Quantitative PCR (qPCR) amplification curves for VDAC1-3 were generated from 10-cell pools of PomcEGFP neurons in female mice. Cycle number was plotted against normalized fluorescence intensity (ΔRn) to visualize amplification, with the cycle threshold (Ct; dashed line) indicating the point at which fluorescence exceeded background levels and sample values were quantified. B. Summary bar graphs of relative expression of Vdac2, Vdac3, Vdac1 mRNA in female POMC neurons. C. Mitochondrial membrane potential determination using TMRM dye; D. Mitochondrial calcium using Rhod2 dye; and E, mitochondrial ATP using BioTracker ATP-Red live cell dye were measured in mHypo43 cells upon treatment with 50 and 100 nM STX. Data from 4-6 independent experiments are represented in the box plots. The symbols represent data from each repeat, the borders of the boxes define the 25th and 75th percentiles, with the median displayed as black lines, and error bars indicate the standard deviation from the mean. Statistical analysis was peformed using one-way ANOVA followed by the Dunnett post hoc test (**p<0.01, *p<0.05).

STX promotes voltage gating of VDAC2 and shifts the channel’s low conducting states towards anion selectivity.

A. A schematic of the experimental setup for VDAC reconstitution. In these experiments, a single VDAC spans a planar lipid membrane that separates and electrically isolates two buffer-filled compartments. The ionic current through the channel is generated by the applied voltage and constantly recorded. VDAC and STX (green diamonds) were added to the cis compartment. B. Electrophysiological recordings of the recombinant human VDAC2 reconstituted into planar lipid membranes before (Control) and after the addition of 100 and 300 nM STX to the membrane-bathing solution. The current traces, representing two VDAC2 channels, were recorded on the same membrane at the indicated voltages, which were returned to 0 mV between each subsequent voltage application. Characteristic voltage-gating behavior is seen as a stepwise current transition from the unique high-conducting or “open” state (“o” state, indicated by green dashed lines) to a low-conducting “closed” state (“c” state, indicated by red dashed lines). The lowest voltages at which gating was observed at each condition are indicated by red arrows. Here and in panel E, the dash-dotted gray lines show the zero current level (I=0). Membrane bathing solutions contained 1 M KCl buffered with 5 mM HEPES at pH 7.4. C. The normalized VDAC2 conductance as a function of the applied voltage obtained in a representative experiment with about 60 VDAC2 channels reconstituted into a planar membrane without (Control) and with subsequent additions of 50, 100, and 1000 nM of STX. Normalized conductance is defined as G/Gmax, where G is the average conductance at voltage V, and Gmax is the maximum average conductance at voltages close to 0 mV. Gating behavior was assessed by applying a triangular voltage wave of ±60 mV, 5 mHz. The addition of STX resulted in a decrease in the minimal (Gmin) normalized conductance (indicated by dashed gray lines) at the application of negative voltages (emphasized by the downward arrow), indicative of increased gating. These results, obtained in multichannel membranes, are consistent with the single-channel records in B. Other experimental conditions were as in (B). D. STX promotes VDAC2 voltage gating in a dose-dependent manner. Normalized gating response of four independent experiments, as in (C), represented as (Gmax-Gmin)/Gmin and plotted against increasing concentrations of STX. Each data point denotes the mean ± SD (n=4). The line is a ligand-binding, one-site saturation curve with Kd = 15 ± 4 nM. E,F STX affects the ion selectivity of VDAC2 voltage-induced “closed” states. E. Representative single-channel trace obtained in 1 M (cis)/ 200 mM (trans) KCl gradient before (Control, upper trace) and after (+ STX, lower trace) addition of 100 nM STX to the cis compartment at the indicated voltages. Dashed green and red lines indicate open and “closed” states, respectively; dashed gray lines indicate zero current. A planar lipid membrane was formed from DPhPC. F. I/V curves obtained from the traces, examples of which are shown in E, for the open (black circles) and three closed (triangles) states. Linear regressions (dashed lines) allow calculation of the reversal potential (ψrev, indicated by arrows) of each state (shown in the inset). A positive ψrev corresponds to anionic and a negative to cationic selectivity. The open state with conductance of 1.9 nS is anion selective (ψrev = 8.3 mV); two low-conducting states of 0.5 and 0.6 nS are also anion selective with ψrev equal to 9.4 and 3.4 mV, respectively, and the low-conducting state of 0.9 nS conductance, is non-selective (ψrev = 0.4 mV).

STX induces coordinated mitochondrial and glycolytic activation with a non-monotonic dose–response in non-serum starved mHypo43 cells.

A. Representative Seahorse XF traces of oxygen consumption rate (OCR) in non-serum starved mHypo43 cells treated with DMSO (control) or STX (1, 5, 10 nM). Sequential injections of oligomycin, FCCP, and rotenone/antimycin A were used to assess mitochondrial respiratory function. B–G. Quantification of mitochondrial respiration parameters derived from OCR measurements, including basal respiration (B), ATP-linked respiration (C), maximal respiration (D), spare respiratory capacity (E), non-mitochondrial oxygen consumption (F), and proton leak (G). STX treatment significantly increased basal respiration, ATP-linked respiration, maximal respiration, and spare respiratory capacity, with the most pronounced effects observed at 1 nM. These enhancements were maintained, though slightly attenuated, at 5 nM and 10 nM. Proton leak was significantly increased across STX-treated groups, while non-mitochondrial respiration showed modest changes. H. Representative extracellular acidification rate (ECAR) traces during the glycolysis stress test under the same treatment conditions. Sequential injections of glucose, oligomycin, and 2-deoxyglucose (2-DG) were used to assess glycolytic function. I–L. Quantification of glycolytic parameters derived from ECAR measurements, including glycolysis (I), glycolytic capacity (J), glycolytic reserve (K), and non-glycolytic acidification (L). STX at 1 nM and 5 nM significantly increased glycolysis and glycolytic capacity. At 1 nM, STX significantly increased glycolytic reserve and non-glycolytic acidification, with moderate effects observed at 5 nM. No significant enhancement was observed at 10 nM. Data are presented as mean ± SEM with individual data points shown. Statistical significance was determined by one-way ANOVA followed by Dunnett’s multiple comparison test. ns-not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

STX induces coordinated mitochondrial and glycolytic activation with a non-monotonic dose–response in serum starved mHypo43 cells.

A. Representative Seahorse XF traces of oxygen consumption rate (OCR) in 6-hr serum starved mHypo43 cells treated with DMSO (control) or STX (1, 5, 10 nM). Sequential injections of oligomycin, FCCP, and rotenone/antimycin A were used to assess mitochondrial respiratory function. B–G. Quantification of mitochondrial respiration parameters derived from OCR measurements, including basal respiration (B), ATP-linked respiration (C), maximal respiration (D), spare respiratory capacity (E), non-mitochondrial oxygen consumption (F), and proton leak (G). STX at 1 nM significantly increased basal respiration, ATP-linked respiration, maximal respiration, and spare respiratory capacity compared to control. Similar but less pronounced effects were observed at 5 nM. In contrast, 10 nM STX showed diminished or non-significant effects across most parameters. Proton leak and non-mitochondrial respiration were modestly increased at lower doses but showed no significant differences at higher concentrations. H. Representative extracellular acidification rate (ECAR) traces during the glycolysis stress test under the same treatment conditions. Sequential injections of glucose, oligomycin, and 2-deoxyglucose (2-DG) were used to assess glycolytic function. I–L. Quantification of glycolytic parameters derived from ECAR measurements, including glycolysis (I), glycolytic capacity (J), glycolytic reserve (K), and non-glycolytic acidification (L). STX at 1 nM and 5 nM significantly increased glycolysis and glycolytic capacity. At 1 nM, STX significantly increased glycolytic reserve and non- glycolytic acidification, with moderate effects observed at 5 nM. In contrast, 10 nM STX did not significantly enhance these parameters, consistent with a non-monotonic dose–response relationship. Data are presented as mean ± SEM with individual data points shown. Statistical significance was determined by one-way ANOVA followed by Dunnett’s multiple comparison test. ns-not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Primer Table