Journal: Nature Metabolism
Article Title: Insulin signalling in tanycytes gates hypothalamic insulin uptake and regulation of AgRP neuron activity
doi: 10.1038/s42255-021-00499-0
Figure Lengend Snippet: a , Knockout strategy to inactivate IR specifically in tanycytes. Animals 10–12 weeks old received i.c.v. injection of either control virus AAV-Dio2-GFP or AAV-Dio2-mKate2 (IR GFP-Tan , IR mKate2 ) or Cre-recombinase-carrying virus AAV-Dio2-iCRE-GFP or AAV-Dio2-Cre (IR ΔTan ). b , Representative images of pAKT signal in basal and lateral tanycytes of unstimulated (0 min) and 5 and 10 min post i.v. injection of 0.5 IU kg −1 insulin in NCD-fed control animals IR GFP-Tan (NCD) (top panel), tanycyte-specific IR KO animals (IR ΔTan ) (NCD) (middle panel) and HFD-fed control animals IR GFP-Tan (HFD) (lower panel). pAKT was quantified in DAPI-positive tanycyte layer. White and yellow dashed lines indicate quantified ROI in basal and lateral tanycytes, respectively. c , d , Mean intensity of pAKT signal in DAPI-positive nuclei of basal ( c ) and lateral tanycytes ( d ). c , 5 min: P (IR GFP-Tan (NCD) versus IR ΔTan (NCD)) = 0.047, P (IR GFP-Tan (NCD) versus IR GFP-Tan (HFD)) = 0.0041; 10 min: P (IR GFP-Tan (NCD) versus IR ΔTan (NCD)) = 0.0144, P (IR GFP-Tan (NCD) versus IR GFP-Tan (HFD)) = 0.0205. d , P (IR GFP-Tan (NCD) versus IR ΔTan (NCD)) = 0.0994, P (IR GFP-Tan (NCD) versus IR GFP-Tan (HFD)) = 0.0151; 10 min: P (IR GFP-Tan (NCD) versus IR ΔTan (NCD)) = 0.0031, P (IR GFP-Tan (NCD) versus IR GFP-Tan (HFD)) = 0.0309. b – d , n (0 min) = 3/IR GFP-Tan (NCD), 4/IR ΔTan (NCD), 4/IR GFP-Tan (HFD); n (5 min) = 5 mice per group; n (10 min) = 5/IR GFP-Tan (NCD), 6/IR ΔTan (NCD), 4/IR GFP-Tan (HFD). e , Knockout strategy to inactivate IR in endothelial cells. The 10–12-week-old IR fl/fl Slco1c1-CreERT2 wt/wt and IR fl/fl Slco1c1-CreERT2 tg/wt littermates received tamoxifen (10 mg d −1 , 5 d). f , Representative images of pAKT signal in lectin-positive brain cortices of unstimulated mice (0 min) and 5 min post i.v. injection of 0.5 IU kg −1 insulin in IR wt/wt and IR ΔBVEC mice. g , pAKT in lectin-positive microvessels, normalized to IR wt/wt control animals. P (5 min) = 0.0007. f , g , n (0 min) = 3/IR wt/wt , 4/IR ΔBVEC ; n (5 min) = 4/IR wt/wt , 6/IR ΔBVEC (unpaired, two-sided Student’s t -test). h , Representative images of pAKT in the ARC of unstimulated mice (0 min) and 5, 10, 20 and 30 min post i.v. injection of 0.5 IU kg −1 insulin in NCD-fed control animals IR GFP-Tan (NCD) (top panel), tanycyte-specific IR KO animals IR ΔTan (NCD) (middle panel) and HFD-fed control animals IR GFP-Tan (HFD) (lower panel). i , Quantification of pAKT-positive cells per ARC hemisphere. 5 min: P (IR GFP-Tan (NCD) versus IR ΔTan (NCD)) = 0.104, P (IR GFP-Tan (NCD) versus IR GFP-Tan (HFD)) = 0.0341; 10 min: P (IR GFP-Tan (NCD) versus IR ΔTan (NCD)) = 0.0014, P (IR GFP-Tan (NCD) versus IR GFP-Tan (HFD)) = 0.0069; 20 min: P (IR GFP-Tan (NCD) versus IR ΔTan (NCD)) = 0.007, P (IR GFP-Tan (NCD) versus IR GFP-Tan (HFD)) = 0.0023. h , i , n (0 min) = 3/IR GFP-Tan (NCD), 4/IR ΔTan (NCD), 4/IR GFP-Tan (HFD); n (5 min) = 5 mice per group; n (10 min, 20 min, 30 min) = 5/IR GFP-Tan (NCD), 6/IR ΔTan (NCD), 4/IR GFP-Tan (HFD). j , Representative images of pAKT in ARC of unstimulated mice (0 min) and 5 and 15 min post i.v. injection of 0.5 IU kg −1 insulin in tamoxifen-treated IR wt/wt and IR ΔBVEC mice. k , Quantification of pAKT-positive cells per hemisphere of ARC in treated IR wt/wt and IR ΔBVEC mice. j , k , n (0 min) = 4/IR wt/wt , 3/IR ΔBVEC ; n (5 min) = 3/IR wt/wt , 5/IR ΔBVEC ; n (5 min) = 3/IR wt/wt , 4/IR ΔBVEC . c , d , i , One-way ANOVA, Tukey post hoc test. Data are represented as the mean ± s.e.m. b , f , h , k , Scale bar, 100 µm. KO, knockout; p.o., per oral; ROI, region of interest.
Article Snippet: AAV-Dio2-mKate2 was generated to replace GFP with a far-red fluorescent protein containing a nuclear localization signal. mKate2 reporter protein sequence was amplified from pAAV-hSynapsin1-GCaMP6s-P2A-mKate2 sequence (Addgene no. 112006) using 5SpemKate2: actagtgccaccatgggtaagaagaagagaagGTGAGCGAG-CTGATTAAGGAGAAC and 3NotmKate2: gcggccgctTATCTGTGCCCCAGTTTGC-AGG primers with the High Fidelity Master PCR (no. 12140314001, Roche) kit.
Techniques: Knock-Out, Injection, Control, Virus
Addgene inc is a verified supplier 
![Effects of GCaMP6s expression level on the action-potential-related fluorescence signal. (A,B) Comparison between calcium events in units of power (fW) obtained with the 5 different constructs as indicated by the color labels, 2 weeks after injection. The schematic examples on top of the figure (in a dashed-line box; note that the image brightness of the standard deviation time series projection is adjusted for viewing and should not be considered as representative of the dynamic range during live imaging) describe extraction of the calcium fluorescence event parameters in acute IO slices. (A1) Averaged event waveforms aligned on initiation point (black dot). The shaded areas denote ± SEM. (A2) Average calcium event peaks; (A3) their cumulative distributions, with arrow and horizontal line segment highlighting the convergence of maximal values of eCa peaks between high-expressing constructs; (A4) relation between average event peaks and baseline fluorescence in each cell. (B) Same data as in (A) but normalized to F 0 for a baseline-normalized DFF values, similarly represented in boxplot (B2) , distribution plot (B3) and construct-averages represented against their respective F 0 s (B4) . (C) Same data as in (B) but peak-amplitude-normalized to compare event kinetics. (C1) averaged event waveforms; (C2) comparison of average event rise times; (C3) cumulative distributions of event rise times; (C4) individual event rise times and baseline-normalized event amplitudes show linear relation. The colored dots in (C4) represent averages for each construct data, gray dots are individual events. Colored bars in (A2,B2,C2) denote average values. Horizontal and vertical lines in (A4,B4) denote ± SEM values in each dimension. Statistics for eCa F-peak on (A2) (1-way ANOVA, [AAV9-Htr5b(3.7)-tTA/TRE vs. AAV.PHP.eB-Htr5b(3.7)-tTA/TRE] f = 1.56, p = 0.1; [AAV9-Htr5b(3.7)-tTA/TRE vs. AAV.PHP.S-Htr5b(3.7)-tTA/TRE] f = 1.9, p = 0.17).](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_3741/pmc09093741/pmc09093741__fncel-16-825056-g0005.jpg)
![Somatic calcium events are only generated by action potentials occurring in the recorded cells regardless of oscillatory status. (A) Simultaneous current-clamp and GCaMP6s imaging shows that a spike in a neighboring IO neuron (A1) is reflected as an electrophysiological “spikelet” in the oscillating cell labeled with an asterisk. The image in (A1) is a standard deviation time series projection of fluorescence time-series recording. Note that the image brightness is adjusted for viewing and should not be considered as representative of the dynamic range during live imaging. Only cells that spiked during the recording are visible. (A2) shows time-aligned fluorescence (top) and V m (bottom) traces, and the neighbor-spike-related spikelet in patched cell is indicated with an asterisk in the inset. (B) Another example of a simultaneous GCaMP6s and current-clamp recording from 3 IO neurons demonstrating presence of electrophysiological “spikelets” in a patched cell [labeled with “*” in (B1) ] linked with neighboring cell calcium events. There are three spikelets in the recording, indicated by thin vertical lines, and two of them are enlarged in insets. Current-injection-evoked spiking in the patched cell results in calcium spikes in the neighbor cells, indicative of gap junction coupling.](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_3741/pmc09093741/pmc09093741__fncel-16-825056-g0007.jpg)
![Electrophysiological features of IO action potentials are reflected in the GCaMP6s event waveform. (A) Principal component analysis of IO action potential waveforms. (A1) 60 voltage spike events detected and aligned on initiation point. (A2) Relative contributions of the principal components to the waveform variability. The 3 first PCAs explain 95%. (A3) Time-varying contributions of the 3 first PCAs, corresponding to the “calcium shoulder” (solid line), after-hyperpolarizarion (dashed line), and the sodium spike (dotted line). (B) Clustering IO calcium events (eCAs) based on the PCA analysis in (A) . (B1) k -means clustering (indicated by colored markers) of the IO action potentials shown in (A1) . (B2) Averaged electrophysiological waveforms from the clusters identified in (B1) . (B3) The average calcium event waveforms linked to the electrophysiological clusters in (B1,B2) . Note that no noticeable calcium transients are seen linked with cluster 1 that corresponds to “spikelet” events (see ). Also, the calcium events belonging liked to cluster 6 events lacking a full sodium spike are very small in amplitude. Spikes from cluster 6 are excluded from rest of analysis, and only the 46 spikes [corresponding to clusters 1–5] recorded in 13 healthy IO neurons are depicted in following. (C) Calcium event peak amplitudes (C1) and rise times (C2) strongly correlate with the electrophysiological spike widths (n = 31 events, clusters 2–5). Note that the eCa amplitude-to-spike width relation saturates with longest spikes, while rise time to spike width relation is more robust. (C3) Relation between calcium event rise time and peak amplitude. Note slight non-linearity with largest events. The colored points and black circles represent average values for respective clusters and individual observations respectively. Events from clusters 1 and 6 are not shown in (C1,C2) as they correspond to spikelets and unhealthy neuron spikes. (C3) consists of data from clusters 1–5. (D) Spike cluster diversity in different recordings. (D1) Number of spike cluster observations for each cell. Note that the “incomplete spikes” (cluster 6) and “short IO spikes” (cluster 2) are only seen in single cells each while spikes from other clusters are seen in multiple cells. (D2) Box-plot including data from clusters 2–5 providing suggestive translation scale between observed calcium event rise time (bottom scale) and the corresponding electrophysiological spike width (top scale). Calcium events with rise times longer than 150 ms correspond to “normal” IO spikes (IO-S). The “short IO spikes” (sIO-s) do not present a clear calcium shoulder and may be caused by unusual physiological state of the IO neuron, possibly due to major damage to dendrites caused by slice preparation.](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_3741/pmc09093741/pmc09093741__fncel-16-825056-g0008.jpg)
![In-vitro calcium imaging of IO subthreshold oscillations. (A,B) Example in vitro calcium imaging recordings from dense (A) and sparse (B) transfection, using the constructs as indicated above the panels. Identified IO neuron somata are indicated with dotted ROIs in time-series projection images in (A1,B1) ; note that the image brightness is adjusted for viewing and should not be considered as representative of the dynamic range during live imaging. The GCaMP6s traces corresponding to labeled ROIs are shown in (A2,B2) . In the example shown in (B) , only one cell is labeled, and 4 neighboring regions indicated by dotted squares in (B1) , are shown as ROIs representing signal from non-transfected cells. Asterisks in (A2) indicate detected calcium spikes. Note the different vertical scales in (A2,B2) , reflecting lower GCaMP6s concentration in (B) . (A3,B3) : Welch spectra for the example recordings show a sharply defined frequency peak for the oscillating cells (left). Right show the peak frequency of the example recordings (colored arrows) with respect to the entire population (gray bars) (C) The power of observed oscillations in calcium traces is stronger with higher baseline fluorescence level resulting from higher concentration of GCaMP6s. Colored lines [in C1,C3 ] and markers (C2,C4) represent data obtained with the 5 different constructs as indicated at the bottom right corner of the figure. (C1,C2) display the ranges of oscillation in terms of STO-bandwith power (C1) and oscillation peak-to-peak amplitude (C2) . Note that the oscillation amplitude increase saturates below 1% (C3) , while all of the constructs were shown to report spike amplitudes with several-fold higher . (C2,C4) display the relation between single-cell baseline fluorescence intensity (F 0 , x -axis) and oscillation power (C2) and peak-to-peak amplitude (C4) . Dotted horizontal line in (C4) represents the threshold of 0.12 % DFF for classification of the signal as “oscillating” as determined in . (D) Summary of imaging capabilities of the different constructs for reporters of IO action potentials (vertical axis) and STOs (horizontal axis). The constructs with intermediate expression levels (orange and green data points; see ) provide better resolution of spike waveforms. The constructs with highest expression (blue data) under-perform with spike amplitudes but provide high resolution of the subthreshold oscillations.](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_3741/pmc09093741/pmc09093741__fncel-16-825056-g0009.jpg)
![Recording of climbing fiber calcium events in vivo. (A1–A3) Expression of GCaMP6s in cerebellar climbing fibers with AAV9, AAV.PHP.S, AAV.PHP.eB-Htr5b(3.7)-tTA/TRE constructs, respectively. (A1-A3) are tiled and Z-projected 40x confocal images of sagittal cerebellar sections showing expression of GCaMP6s in axons in the cerebellar cortex with samples prepared for anatomical study. Expression of GCaMP6s in CF is sparse in when the viral serotype is AAV.PHP.S, while dense and widespread with AAV.PHP.eB serotype. The inset [in (B1) ] indicates the region that was imaged in with the miniscope (see schematic on top of the figure). (B) Example recording from CFs in a living mouse. (B1) Is a sagittal slice from tissue on the same brain where in-vivo calcium imaging was done. A schematic labeled beta describes the experiment procedure. (B2) , z -projection image (standard deviation) of miniscope calcium recording time series (20 second recordings per zone) as seen from the dorsal surface of the cerebellar cortex. Anterio-posterior (AP) and medio-lateral (ML) directions are indicated with white arrows. The dashed lines indicate manually-drawn ROIs for CFs, and the respective calcium traces are shown in (B3) . Parts of two of the traces [2 and 5, indicated by dashed rectangles in (B4) ] are shown with expanded time scale at the bottom of the panel. Detected events are indicated by asterisks. (C) Comparing in vivo calcium events recorded in IO axons (CFs; blue) with the somata (green). n = 70 events in 2 animals for axon recordings, 22 events in 4 animals for somata. (C1) Average in-vivo GCaMP6s transients from CFs and IO somata, aligned at initiation point. Shaded regions represent ± SEM. (C2) Comparison of calcium event amplitudes in IO somas and axons when normalized to basal fluorescence (1-way ANOVA, f = 8.19, p < 0.001). (C3) Comparison of event rise-times between IO soma and climbing fibers (1-way ANOVA, f = 5.3, p < 0.001). (C4) Comparison of the instantaneous frequency of events (inverse of inter-event interval) in the axon vs. the soma (1-way ANOVA, f = 5.35, p < 0.023). Note the slower spike rate in somatic recordings, possibly due to tissue cooling during the ventral surgery needed for somatic recordings.](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_3741/pmc09093741/pmc09093741__fncel-16-825056-g0012.jpg)
![a , b , Parametric maps of P values from paired t -test of differences in cumulative glucose metabolism over the recorded time determined by [18 F]FDG PET between insulin-stimulated (16-h fasted, i.p. 0.325 IU kg −1 insulin) and NaCl (0.9%)-injected, anaesthetized IR mKate2 and IR ΔTan mice ( n = 8/IR mKate2 ; n = 10/IR ΔTan ). Brain regions that had significantly reduced cumulative glucose metabolism in IR mKate2 were not altered in IR ΔTan mice ( a ), and brain regions that had significantly reduced signal in IR Δtan remained unaltered in IR mKate2 ( b ). Blue colour scale indicates regions where metabolism at NaCl > insulin (inhibition on insulin injection). Sagittal reference image inserts show location of corresponding coronal plates . C e / C p is the ratio of tissue and blood glucose concentrations, a blood glucose level-insensitive measure for glucose metabolism. CP, caudate putamen; BNST/LPO, bed nucleus of stria terminalis/lateral preoptic area; PAG, periaqueductal grey; RN, reticular nucleus; ZI/aSNR, zona incerta/anterior substantia nigra; pSNR, posterior substantia nigra; MV, medial vestibular nucleus; SPV, spinal vestibular nucleus. Paired, two-sided Student’s t -test. a , For IR mKate2 P (CP) = 0.0007, P (BNST/LPO) = 0.0056, P (LPO) = 0.0003, P (VMH/LH) = 0.0109, P (PAG) = 0.0015, P (RN) = 0.0097, b , For IR ΔTan P (ZI/aSNR) = 0.003, P (pSNR) = 0.0023, P (MV) = 0.0006, P (SPV) = 0.0105. Data are represented as the mean ± s.e.m. * - P <0.05, ** - P <0.01, *** - P <0.001.](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_8146/pmc08688146/pmc08688146__42255_2021_499_Fig5_HTML.jpg)
