Claims 46
From the paper Kalugin PN, et al. Simultaneous, real-time tracking of many neuromodulatory signals with multiplexed optical recording of sensors on a micro-endoscope. Neuron (2026). · claims file · checked 2026-10-22: ready
- fig1c-cells-per-gel
Imaging a MORSE hydrogel through the GRIN lens shows 100 to 200 green sensor cells spread through the full thickness of the gel.
3D two-photon imaging of the resultant hydrogel through the GRIN lens (30 planes/volume, ~600×600×270 μm 3 volume) demonstrated a uniformly distributed set of 100–200 green cells spanning the full thickness of the gel
Results, MORSE probe development and characterization, paragraph 1; Figure 1CHow it will be judged
Neither the Methods nor the code say how the cells were counted. Readings (a) the number of cell masks extracted per recording before curation (CNMF components); (b) the number of curated cell masks per recording. Reproduced when the per-probe counts of a reading lie in 90 to 220 for most probes (more than half) and the cells occupy all depths of the gel (cell centres found in the top, middle and bottom thirds of the gel thickness). Needs masks or imaging.
- fig1d-collagen-reduces-warping
Hydrogels of pure Matrigel deform substantially over tens of minutes; a mixture of Matrigel and collagen I deforms less.
The small hydrogel volume (~40 nL) on the GRIN lens tip, when composed of pure Matrigel, exhibited substantial global and local deformations over tens of minutes. We minimized these deformations using an optimized mixture of Matrigel and collagen I
Results, MORSE probe development and characterization, paragraph 2; Figure 1DHow it will be judged
Deformation is the local strain magnitude, the mean over the hydrogel volume of the magnitude of the displacement field minus its mean (Figure 1D legend, Figure S1C; morse_figs1 cells 3, 92, 97), from demons registration of Cellpose probability maps to a reference volume. With n=6 probes per recipe imaged for 9 h, reproduced when the mean deformation across probes of the 50:50 recipe stays below that of pure Matrigel over the recording (mean over the 9 h lower, and lower at the end). The paper gives no test. Needs the deformation fields or raw imaging.
- fig1f-koh-stability
Treating the silicone surface with KOH before applying the gel strongly reduces gel warping and detachment during dipping; with the optimized protocol (40:60 Matrigel:collagen I with 2.5% KOH, against 50:50 without KOH) probe stability rose from about 10% to about 65% in initial tests.
This allowed us to increase probe stability from ~10% to ~65% in initial tests ( Figure 1F )
Methods, Optimization of MORSE manufacture, paragraph 1; Figure 1F; Figure S1D-E; Results, MORSE probe development and characterization, paragraph 2How it will be judged
A probe is early warp when the 3D translation of the hydrogel (the norm of the mean displacement over the gel) exceeds 25 um in the first hour, late warp when it exceeds 25 um in the second hour, and stable otherwise (Figure 1F legend). Stable fraction over the 9 probes of 50:50 without KOH and the 24 of 40:60 with KOH. Variant (code, morse_figs1 cell 139), late warp as more than 50 um within 2 h. The code's comments (cell 126) give "~66% success" for the KOH group. The two groups differ in both recipe and KOH (Figure 1F legend), so the comparison does not separate the two. Of 9 probes, "~10%" can only be 1 (11%), recorded as a count. Needs the deformation fields or raw imaging.
- methods-revision-probe-stability
In the revision experiments more than 90% of probes were stable, 10 of 13 in vitro dipping, 9 of 9 slice and 13 of 14 in vivo CSF experiments.
In revision experiments, >90% probe stability was achieved (n=10/13 in vitro dipping experiments, n=9/9 brain slice experiments, n=13/14 in vivo CSF experiments).
Methods, Optimization of MORSE manufacture, paragraph 1How it will be judged
The paper does not define stability for the revision experiments. Read as in fig1f-koh-stability (no translation above 25 um), the only definition the paper gives, made for dipping; for slice and in vivo recordings this is a reading. Counts are exact. The ">90%" does not follow from 10/13; the counts decide the claim. Needs deformation fields or raw imaging, and a record of which recordings were revision experiments.
- fig1g-sensor-yield
Ten ten-sensor probes yielded 506 identified single-cell masks; per probe a median of 9 sensors were active (IQR 8 to 9), with on average 6.02 cells per sensor per probe.
we could assign clear sensor identities across several hundred cells from ten representative MORSE probes (median of 9 active sensors per probe, interquartile range 8–9; average of 6.02 cells per sensor per probe)
Figure 1G and legend; Results, MORSE probe development and characterization, paragraph 4How it will be judged
From SENSOR IDENTITY on the identification dipping recordings of the ten ten-sensor probes. A sensor is active on a probe when at least one cell carries its label; there is no threshold (morse_figs3_rev cells 139-143). Cells per sensor per probe is the total number of cells divided by the number of (probe, sensor) pairs with at least one cell (cells 139, 144, 506/84). Percentiles by numpy's default linear interpolation (cell 142, which prints 8.25).
- fig1g-all-ten-sensors-respond
Dipping ten-sensor probes into each ligand in turn gave rapid, strong and consistent responses in a subset of cells for each ligand, and all ten sensors were activated.
This resulted in rapid, strong, and consistent patterns of elevated fluorescence in subsets of cells evenly spread throughout the gel in response to each ligand
Results, MORSE probe development and characterization, paragraph 4; Figure 1G, Figure S3AHow it will be judged
Reproduced when each of the ten sensor labels is carried by cells on at least one of the ten probes, and the mean peak-normalized dF/F0 of each sensor's cells is highest during the well of its own ligand (Figure 1G heatmap; peak normalization by Eq. 28, mean of the highest 10% minus mean of the lowest 10%), except that the DA and NE sensors and the AVP and OXT sensors may also respond to each other's ligand (fig1h-crosstalk-pairs). Background correction by an exponential fit to the signal before activation (Methods), as in morse_figs3_rev cells 21-24.
- fig1h-crosstalk-pairs
Among the ten sensors, only the DA and NE sensors, and to a lesser extent the AVP and OXT sensors, show crosstalk in their ligand responses.
Calculating the means of the derivative time courses across each cell cluster and cross-corelating them with each other consistently demonstrated that only the DA2.9 dopamine and the NE2h norepinephrine sensors, as well as to a lesser extent the AVP0.2 vasopressin and OXT1.7 oxytocin sensors, exhibited crosstalk in their ligand responses
Methods, Quantification and statistical analysis; Figure 1H, Figure S3B; Results, MORSE probe development and characterization, paragraph 4How it will be judged
Per cell, the time derivative of the peak-normalized dF/F0 by a Savitzky-Golay filter (1-min window, order 1; Methods, which decide; morse_figs3_rev cell 216 uses 10 frames, shown as a variant), each divided by its own maximum (cell 218); the mean per sensor (cell 224); the Pearson correlation at zero lag between the ten sensor means (cell 227). The paper and the code set no threshold for crosstalk. Reproduced when the DA-NE correlation is the largest off-diagonal value, the AVP-OXT correlation is the second largest, and every other pair is lower than both. Variant (code, cell 219), a band-pass 3-minute dipping artefact subtracted before averaging.
- table1-ec50-on-target
The MORSE EC50 of each sensor for its own ligand, from consensus dose-response fits, is AVP 355 nM, OXT 2.19 uM, 5-HT 687 nM, CCK 245 nM, NE 4.04 uM, DA 1.70 uM, ACh 7.03 uM, ATP 75.2 uM, MT 22.0 nM and HA 6.39 uM.
We found that the sensors’ half-maximal effective concentrations (EC 50 , equivalent to a sensor’s apparent dissociation constant K d ) could be reliably estimated for all sensors ( Table 1 ; Figures 1I , S3C )
Table 1, EC50 (MORSE) column; Figure 1I, Figure S3CHow it will be judged
Dose-response dipping of the ten probes of Figure 1G (two mixtures at 10-fold steps; Video S6). Per cell matched to a sensor identity, background corrected by an exponentially decaying fit and peak-normalized (Methods); the plateau in each well is the mean over the third minute (Methods). Fit Eq. 9 with Hill coefficient 1, y = F/(1 + Kd/[L]), by least squares (morse_figs3_rev cell 365, Powell minimization), to the cells of all probes pooled (consensus fit, Figure 1I). Variants (code, cell 339) the plateau as the median of sorted frames 47 to 78 of the 93-frame well, peak normalization to the 6th or 7th well, a stretched exponential background, cells dropped by a noise cut, and ACh, ATP, MT and HA pooled with an older probe set (cell 365, groups_merge_rev). The Methods and the code disagree on the plateau; the Methods reading decides and the code's is shown. Each value within 10%.
- table1-ec50-off-target
The crosstalk sensors respond to the partner ligand with EC50s of 28.7 uM (AVP sensor to OXT), 247 nM (OXT sensor to AVP), 9.35 uM (NE sensor to DA) and 8.20 uM (DA sensor to NE).
How it will be judged
As table1-ec50-on-target, each sensor's cells fitted against the partner ligand's concentrations in the mixture wells, signals peak-normalized to the on-target maximum and scaled by the crosstalk ratio (morse_figs3_rev cells 365 and 368). Each value within 10%.
- table1-peak-dff-morse
On MORSE probes the peak dF/F0 of each sensor to a saturating dose of its ligand is 0.37 (AVP), 0.24 (OXT), 0.36 (5-HT), 0.40 (CCK), 0.34 (NE), 0.27 (DA), 0.37 (ACh), 0.28 (ATP), 0.30 (MT) and 0.38 (HA), and to the partner ligand 0.10 (AVP sensor, OXT), 0.05 (OXT sensor, AVP), 0.15 (NE sensor, DA) and 0.21 (DA sensor, NE).
How it will be judged
Identification dipping of the ten probes of Figure 1G (Figure S3B ligands, 10 uM AVP, 100 uM OXT, 100 uM 5-HT, 10 uM CCK, 1 mM NE, 100 uM DA, 1 mM ACh, 1 mM ATP, 10 uM MT, 1 mM HA). Per cell, dF/F0 after background correction; the cell's peak is the maximum within the third minute of the ligand's well (morse_figs3_rev cell 160); the table value is the median across cells with a positive peak (cells 248 and 263; the S3B legend draws the median). For AVP, NE and CCK the code takes the well after the ligand well (cell 248); readings (a) the ligand well, (b) the code's well. Each value within 10%.
- fig1j-ec50-consistent-across-probes
Dose-response relationships, after peak normalization and averaging over cells of the same sensor, are precise and consistent across probes.
peak-normalizing each cell’s signal and averaging across cells expressing the same sensor produced precise dose-response relationships that were consistent across probes ( Figures 1I - J , S3C )
Results, MORSE probe development and characterization, paragraph 5; Figure 1I-JHow it will be judged
Per-probe EC50s are fitted as in table1-ec50-on-target to each probe's cells alone, for probes with more than one cell of the sensor (morse_figs3_rev cell 662); Figure 1J shows their mean and SEM across 5 to 8 probes per sensor (cells 677, 685). The paper does not say what consistent means. Readings (a) for every sensor, every per-probe EC50 lies within one 10-fold dose step of the consensus EC50 (between a tenth and ten times it); (b) for every sensor, the SEM across probes is less than half the mean. Reproduced when both hold; blocked as methods_underspecified when they differ.
- figs3e-tau-on-under-one-minute
Ligands diffuse quickly through the hydrogel; almost all sensors show an onset time constant under 1 minute, and most signals plateau within 1 minute.
Almost all sensors demonstrate a t-on under 1 minute, reflecting rapid diffusion of all ten ligands through the hydrogel to reach their sensors
Figure S3E legend; Results, MORSE probe development and characterization, paragraph 5How it will be judged
Per cell, an exponential rise (x0 - xinf)*exp(-t/tau) + xinf fitted to the first three minutes in the well with the lowest concentration of its ligand above the sensor's EC50 (Methods; S3E legend; morse_figs3_rev cells 688-692, which fit frames 0 to 78 after a 31-frame median filter, set xinf to the median of frames 47 to 78, and drop tau above 200 frames). Done on the peak-normalized dF/F0 and on the Eq. 8 concentration. "Almost all sensors" read as at least 8 of the 10 sensors with a median tau-on across cells under 60 s. The S3E legend plots both fits and does not say which one its sentence is about, and the Methods treat the concentration fit as the diffusion measure; readings (a) the dF/F0 fits, (b) the concentration fits (larger by construction, S3E legend). The outcome is the one both give; if they differ, the claim is blocked as methods_underspecified.
- figs3f-depth-delay
Onset times increase only modestly with cell depth in the gel, with a positive slope for all sensors except ATP.
Notably, all but one sensor demonstrate a small increase in t-on with increasing cell depth
Figure S3F legend; Results, MORSE probe development and characterization, paragraph 5How it will be judged
The concentration tau-on of figs3e-tau-on-under-one-minute against each cell's depth, the z centre of mass of its mask measured from the front face of the gel (morse_figs3_rev cells 712-718, 9.89 um per plane), a least-squares line per sensor. Reproduced when 9 of the 10 slopes are positive (tau-on rising with distance from the front face) and the ATP slope is the negative one. "Modest" is read as a slope below 1 s per um. Needs masks with depth.
- figs2c-stabilization
Registration using probability maps of the GRAB signal stabilizes the images as well as or better than registration using the tdTomato channel, in dipping, slice and in vivo recordings.
Registration using Cellpose2.0 probability maps derived from the GRAB signals performed similarly to or better than those from tdTomato signals ( Figure S2C ; Video S4 )
Results, MORSE probe development and characterization, paragraph 3; Figure S2CHow it will be judged
Image stability is the normalized cross-correlation of the GRAB probability map at each time with the first, restricted to voxels with at least 50% cell probability at the first timepoint (S2C legend; Methods, MORSE hydrogel deformation quantification; morse_figs2 cells 3-12 use skimage match_template on maps downsampled 4-fold in x and y). Over the dipping (n=36), slice (n=11) and in vivo (n=11) recordings with both channels, reproduced when both alignments raise the mean cross-correlation above the unaligned one and the GRAB-aligned mean is at least the tdTomato-aligned mean minus its SEM at most times. Needs raw imaging.
- figs2f-cells-tracked
About 70 to 130 active sensor cells per probe can be tracked across 2 to 8 recordings in different contexts.
we could reliably track ~70–130 active sensor cells per probe across 2–8 sequential recordings in different contexts across up to 8 hours in total
Results, MORSE probe development and characterization, paragraph 3; Figure S2FHow it will be judged
The number of cells matched across recordings of the same probe (S2F, dipping n=40 recordings over 10 probes, slice n=15, in vivo n=25). Reproduced when the per-probe mean of the matched counts lies in 63 to 143 for most probes. The code's printed counts (morse_figs2_2 cell 4) range from 53 to 108, which is noted.
- fig2c-dlgn-kcl-activates-sensors
Bath application of 50 mM KCl to dLGN slices activates the NE, DA and 5-HT sensors.
Bath application of 50 mM KCl activated GRAB sensors for NE, DA, and 5-HT ( Figures 2B - C , top ; Video S7 )
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 2; Figure 2B-CHow it will be judged
In the 9 dLGN recordings, POOLED-SLICE before peak normalization. Reproduced when, in most recordings, the NE, DA and 5-HT sensors' pooled dF/F0 rise during the KCl wash-on above 1.96 times their SD over the pre-KCl period (the Methods' robustness rule applied to dF/F0). The DA rise is accounted for by NE in fig2c-dlgn-da-explained-by-ne.
- fig2c-dlgn-da-explained-by-ne
In dLGN slices the KCl responses of both the NE and the DA sensors are explained by release of NE alone, not a combination of DA and NE.
Responses to KCl at both the NE and DA sensors could be explained solely by release of NE
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 2; Figure 2C bottom, Figure S4BHow it will be judged
CONC by Eq. 19 on the pooled NE and DA sensor signals (POOLED-SLICE, CONSTANTS). Reproduced when, in most of the 9 recordings, the KCl-evoked NE concentration exceeds THRESH while the DA concentration stays below its THRESH during the KCl response window (PEAK readings). A claim of absence for DA; the paper uses the threshold, and its value is given either way.
- fig2d-dlgn-robust-counts
KCl evoked robust NE responses in 9 of 9 dLGN slices and robust 5-HT responses in 7 of 9, with robustness thresholds of 348 nM (NE), 116 nM (DA) and 62.7 nM (5-HT).
Robust responses to KCl (defined as a peak-normalized ΔF/F 0 signal increases exceeding the 95% confidence interval above pre-wash-on baseline) were detected for NE in n=9/9 slices and for 5-HT in n=7/9 slices ( Figure 2D )
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 2; Figure 2C-DHow it will be judged
POOLED-SLICE, CONC (NE and DA by Eq. 19, 5-HT by Eq. 8), THRESH over the 9 dLGN recordings, PEAK with readings (a) and (b) for the KCl window, combined with the two readings of the peak normalization in POOLED-SLICE. Counts are exact. The THRESH variant is shown. The code that made these numbers is not in the repository (morse_fig2 gives other thresholds), so no code detail binds beyond the shared methods.
- fig2e-dlgn-peak-concentrations
KCl-evoked peak concentrations at the dLGN were about 4.61 uM NE and 0.44 uM 5-HT.
KCl-evoked peak concentrations were estimated at ~4.61 μM for NE and ~0.44 μM for 5-HT ( Figure 2E )
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 2; Figure 2EHow it will be judged
PEAK of the CONC traces of fig2d-dlgn-robust-counts, averaged over recordings. Which recordings enter the mean is open. Readings (a) all 9 for both ligands; (b) all 9 for NE and the 7 robust ones for 5-HT. Each combined with the PEAK and POOLED-SLICE readings. Reproduced when both values are within tolerance under one combination.
- fig2f-dlgn-5ht-delay
In the dLGN the rise of the 5-HT sensor activation was delayed relative to the NE sensor by 111 s on average.
The rise time of activation of the 5-HT sensor was delayed relative to that of the NE sensor by an average of 111 sec ( Figure 2F )
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 2; Figure 2FHow it will be judged
TAU-RISE of the NE and 5-HT concentration traces of fig2d-dlgn-robust-counts, timed from KCl wash-on. Readings (a) the mean over the 7 recordings with both robust of the within-recording difference; (b) the mean 5-HT tau-rise over the 7 robust minus the mean NE tau-rise over all 9. Figure S4D's numbers imply mean tau-rise of about 240 s (NE) and 351 s (5-HT), whose difference is 111 s; these are reported beside the delay. Each combined with the PEAK and POOLED-SLICE readings.
- figs4de-deconvolution
Deconvolving the hydrogel transfer function sharpens the dLGN responses; the average NE and 5-HT tau-rise fall by 44.6 s (18.6%) and 63.9 s (18.2%), and the average peaks rise by 682 nM (14.8%) and 146 nM (33.3%), so true peaks occur up to 19% sooner and up to 33% higher.
This resulted in sharper and narrower peak responses for each ligand, suggesting that the true peak concentrations at the slice surface occurred up to 19% sooner and were up to 33% higher than those recorded at the sensors within the hydrogel
Figure S4D-E legends; Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 2How it will be judged
Deconvolve h(t) = (1/tau) exp(-t/tau) from each concentration trace of fig2d-dlgn-robust-counts (Methods Eq. 25-26) with tau per sensor the median of the dipping tau-on of figs3e-tau-on-under-one-minute; the code (morse_fig2_rev cell 160, morse_fig2 cell 427) uses scipy.signal.deconvolve after a Savitzky-Golay filter (20 frames, order 1). Open, whether tau comes from the dF/F0 fits or the concentration fits (S3E shows both); readings (a) dF/F0, (b) concentration. Then TAU-RISE and PEAK with and without deconvolution; percent changes are of the means across recordings (n=9 NE; 5-HT over the 7 robust, S4C-E legends). Reproduced when, under one reading, the four percentages are within tolerance; the absolute changes are reported.
- slices-no-peptide-release
KCl did not release enough AVP, OXT or HA from dLGN or PVH slices to drive their sensors; those sensor cells, present on the probes, responded only to post hoc dipping.
Most of these cells did not have matching cell masks in the slice experiment, indicating that they did not respond to bath application of NE, DA, or 5-HT, and that KCl did not elicit sufficient release of AVP, OXT, or HA from the slice to drive the respective sensors
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraphs 2 and 3; Figure 2B right, Figure S5BHow it will be judged
A claim of absence, judged on the paper's robustness rule. In the dLGN and PVH recordings followed by dipping, the AVP, OXT and HA sensor cells identified by dipping either have no mask in the slice recording or, where matched, show no KCl response above 1.96 times their pre-KCl SD (POOLED-SLICE dF/F0). Reproduced when no recording shows such a response; the largest response is given either way. Needs the dipping-only cells and the matching.
- figs5c-kcl-no-direct-effect
Dipping a probe into KCl solutions does not drive the NE or DA sensors, so the slice responses come from KCl-evoked release, not a direct effect of KCl on the sensors.
Neither sensor is driven by dipping into any KCl solution, indicating that the KCl responses on the left heatmap in B correspond to KCl-induced ligand release from the PVH slice
Figure S5C legendHow it will be judged
In the first post hoc dipping recording of the PVH probe of Figure S5B, the mean dF/F0 of the NE and DA sensor cells in each KCl well (1.56 to 50 mM) against the aCSF wells. A claim of absence, judged as the paper does on the robustness rule; reproduced when no KCl well raises either sensor above 1.96 times its SD over the preceding aCSF wells, while the DA and NE wells do. Sizes reported either way.
- figs5d-pvh-ne-alone
In the PVH, KCl activates the NE, DA and 5-HT sensors, and demixing shows that NE alone explains the NE and DA sensor responses.
quantitative estimation of NE and DA concentrations along with demixing of their combined effects at the DA sensor suggested that NE alone could explain the activation of both sensors in the PVH
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 3; Figure S5B-EHow it will be judged
As fig2c-dlgn-da-explained-by-ne, on the 6 PVH recordings with THRESH computed over them. Reproduced when the NE concentration exceeds threshold during 25 or 50 mM KCl and the DA concentration stays below its threshold, in most recordings, and the two thresholds are within tolerance. The S5D legend gives them on one example, but the Methods define one threshold per dataset (the maximum SD across its recordings), so they are the PVH dataset's thresholds.
- figs5f-pvh-robust-counts
KCl drove robust elevations of both NE and 5-HT in all 6 PVH experiments.
KCl drove robust elevations in both NE and 5-HT in n=6/6 experiments ( Figure S5F )
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 3; Figure S5FHow it will be judged
Robustness on the pooled NE and 5-HT sensor dF/F0 (S5F plots dF/F0 with thresholds; 5-HT was not converted to concentration, S5D legend), the threshold being 1.96 times the largest pre-KCl SD across the 6 recordings (THRESH before conversion). In 4 of the 6 the 5-HT identity is inferred from responding to KCl but not to DA or NE (S5F legend); the dataset's labels are used.
- figs5g-pvh-dose-dependence
NE sensor cells in the PVH were reliably driven by 25 mM KCl, while 5-HT sensor cells needed 50 mM KCl to be fully activated.
While NE sensor cells were reliably driven by 25 mM KCl, 5-HT sensor cells consistently required a higher dose of KCl (50 mM) to be fully activated ( Figure S5G )
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 3; Figure S5GHow it will be judged
Per recording, the peak pooled dF/F0 of each sensor at each KCl dose. The code (morse_fig2 cell 642) takes the median of the last 10 frames before the end of each lower dose, and the peak for 50 mM, with hand-set dose ends; the main reading is PEAK of dF/F0 within each dose period (from that dose's onset to the next onset), the code's as variant. "Reliably driven" and "fully activated" are not defined. Readings (a) the NE peak at 25 mM exceeds the threshold of figs5f-pvh-robust-counts in all 6, and the 5-HT peak at 25 mM is below it in all 6 and above it at 50 mM; (b) the NE peak at 25 mM is at least half its 50 mM peak, and the 5-HT peak at 25 mM is below half its 50 mM peak, in all 6.
- figs5i-pvh-ne-concentration
The average NE concentration at the PVH with 50 mM KCl was 21.4 uM, nearly five times that at the dLGN.
The average NE concentration detected at the PVH in response to 50 mM KCl was 21.4 μM – nearly five-fold higher than at the dLGN
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 3; Figure S5F, H, IHow it will be judged
PEAK of the CONC NE trace (Eq. 19) during the 50 mM dose period over the 6 PVH recordings, mean across recordings (S5I, not deconvolved). The ratio is to the mean of fig2e-dlgn-peak-concentrations; 21.4/4.61 = 4.6, which "nearly five-fold" states. The ratio's expected value is that quotient, not a printed number.
- fig2h-lc-ne-da-peaks
Optogenetic stimulation of LC releases NE and DA together with distinct time courses; DA reaches its peak (248 nM on average) before NE (347 nM).
Quantification of NE and DA release while accounting for sensor crosstalk demonstrated that both ligands were released simultaneously and with distinct time courses, with the DA concentration reaching its peak (248 nM, on average) before the NE concentration (347 nM
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 4; Figure 2H, 2K, Figure S4FHow it will be judged
POOLED-SLICE and CONC (Eq. 19) on the LC recordings; PEAK in the 5 to 12.5 min window (morse_fig2_rev cell 139); mean over the 5 recordings with robust NE (cells 145 and 157 print 347 and 248 nM for these). The paper does not say how "before" is measured. Readings (a) the DA TAU-RISE earlier than the NE TAU-RISE (fig2l-lc-ne-delay, Figure 2L); (b) the time of the maximum of the mean DA concentration trace in the window earlier than that of NE. The outcome is the one both give; if they differ, the claim is blocked as methods_underspecified.
- fig2i-lc-increments
After each of the first several optogenetic trains, NE and DA concentrations rose in discrete steps of 100 to 300 nM, so 20 s between trains let the released ligands equilibrate with the hydrogel.
In fact, discrete 100–300 nM increments in the concentrations of both ligands were evident after each of the first several trains of optogenetic stimulation
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 4; Figure 2I legendHow it will be judged
The example probe of Figure 2H-I. The NE and DA concentration (CONC, Eq. 19) with a 5-frame median filter (morse_fig2_rev cell 205), train onsets from the stimulation record (45 trains 20 s apart). The increment is open. Readings (a) the code's, concentration from 4 to 15 frames after the train onset minus the value 4 frames after (cells 224-231); (b) the value at the next train's onset minus that at this train's onset. Over the first 5 trains, reproduced when the increments of both ligands lie in 90 to 330 nM. The legend's "100-200 nM" (90 to 220 nM) is reported as a second range. The paper names no number of trains; "several" is read as 5.
- fig2jk-lc-robust-counts
Of five LC slices with robust NE elevations, four also had robust DA elevations and none had robust 5-HT elevations; thresholds were 148 nM (NE), 84.3 nM (DA) and 30.3 nM (5-HT).
Out of five brain slices with robust elevations in NE, four also had robust elevations in DA, while none had robust elevations in 5-HT ( Figures 2J - K )
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 4; Figure 2H, 2J-KHow it will be judged
THRESH as the Methods give it (SD of the pooled peak-normalized signal before stimulation) decides; the code's variant (morse_fig2_rev cells 120, 144, 146, the SD of the unnormalized pooled dF/F0 over the first 5 min) is shown, and is the one whose printed values match the paper's thresholds. Which recordings form the "dataset" over which the maximum SD is taken is open (the Methods do not define it); readings (a) the LC recordings, (b) the code's 12 quantified opto recordings (LC and DRN; the paper mentions no DRN optogenetic recordings). PEAK in 5 to 12.5 min. The code picks the five NE responders by hand (cells 116, 183, 195); here a slice is an NE responder when its NE peak exceeds the threshold. Counts are exact.
- fig2l-lc-ne-delay
In LC slices the rise of NE concentration lagged the rise of DA by 31 s, or 1 to 2 optogenetic trains.
Rise times of NE concentration were consistently delayed relative to DA concentration by 31 sec, or 1–2 trains of optogenetic stimulation ( Figure 2L )
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 4; Figure 2LHow it will be judged
TAU-RISE from stimulation onset of the NE and DA concentration traces, in the 4 slices with robust NE and DA (morse_fig2_rev cells 170-174 print 1.432 and 0.920 min, 30.7 s). "Consistently" is read as the NE tau-rise later than the DA tau-rise in each of the 4.
- slices-probe-stability
MORSE probes stayed stable for over an hour of slice recording (46 of 48 probes) and over the following 1.5 or more hours of dipping (15 of 17).
these hydrogel probes remained stable for over an hour during slice recordings (46/48 probes) and over a subsequent 1.5+ hours of post hoc dipping (15/17 probes)
Results, MORSE detects regionalized neuromodulator release in brain slices, paragraph 4How it will be judged
Stability as in fig1f-koh-stability (hydrogel translation not above 25 um), which the paper defines only for dipping. Counts are exact. Not in the code. Needs all 48 slice recordings with deformation fields or raw imaging.
- fig3df-ip-responses
In awake mouse CSF, IP AVP, OXT and 5-HT each evoked a response in the corresponding sensor within at most 10 to 15 min, followed by washout.
Upon peripheral delivery of each ligand, we reliably observed a specific corresponding sensor response in the CSF within at most 10–15 minutes, followed by washout
Results, Real time, quantitative tracking of CSF composition with MORSE in vivo, paragraph 2; Figure 3B-FHow it will be judged
POOLED-INVIVO on the three-sensor recordings (n=11 mice in the legend; the code uses 10, morse_fig3_rev cell 96). Onset is open. Readings (a) TAU-RISE of the per-mouse trace within 0 to 45 min; (b) the first time after injection at which the trace exceeds 1.96 times the largest pre-injection SD across mice (the Methods' threshold rule). Reproduced when, for each ligand, the onset is at most 16.5 min in every mouse (or in the mean trace, reported both), the mean peak dF/F0 of the matching sensor in 0 to 45 min exceeds that of the other two sensors (specificity, read from Figure 3D-F), and the mean trace at 45 min is below its peak by more than the threshold (washout). The code prints tau-rise of 11 to 36 min for older data (morse_fig3 cells 142-150), which is noted.
- figs6ef-photometry-kinetics
Single-sensor fiber photometry probes in CSF show AVP and OXT time courses after IP injection with kinetics similar to MORSE two-photon recordings.
This approach demonstrated similar kinetics of CSF AVP and OXT time courses following IP injection as compared to the MORSE technique involving two-photon imaging ( Figures 3D - E )
Results, Real time, quantitative tracking of CSF composition with MORSE in vivo, paragraph 2; Figure S6D-F, Figure 3D-EHow it will be judged
Photometry at 50 Hz, a 2-s median filter, an exponential background fitted over the tens of minutes before activation, dF/F0 against the 15 min before each IP injection (Methods). n=4 mice per sensor. Kinetics compared through TAU-RISE of the mean traces (photometry against MORSE of fig3df-ip-responses). "Similar" is not defined; readings (a) the photometry tau-rise lies within the range of per-mouse MORSE tau-rise; (b) the two mean tau-rise differ by less than 5 min. Not in the code. Needs the photometry recordings.
- fig3g-ten-sensor-yield
With ten-sensor probes in CSF of four mice, up to 9 sensors per probe were recovered (median 8.5), with on average 3.8 cells per sensor per probe.
Across four animals, we were able to successfully recover simultaneous signals from up to 9 distinct sensors per probe ... 8.5 sensors; average of 3.8 cells per sensor) ( Figure 3G )
Results, Real time, quantitative tracking of CSF composition with MORSE in vivo, paragraph 3; Figure 3GHow it will be judged
SENSOR IDENTITY on the four ten-sensor in vivo recordings; a sensor is recovered on a probe when at least one cell carries its label; cells per sensor per probe is total cells over (probe, sensor) pairs with a cell (morse_fig3_rev cells 192-203 print sensors per probe 9, 8, 9, 5 and 119/31 = 3.84).
- fig3h-each-sensor-responds-ip
Each of the ten sensors responded after IP injection of its ligand, with kinetics that differ across ligands.
Each sensor demonstrated a response following IP injection of the respective ligand, with diverse response kinetics across ligands that likely reflect differential permeability from blood into CSF ( Figure 3H )
Results, Real time, quantitative tracking of CSF composition with MORSE in vivo, paragraph 3; Figure 3HHow it will be judged
POOLED-INVIVO for each sensor around the IP mixture that contains its ligand (and the ICV 5-HT), on the ten-sensor probes (Figure 3H uses 5 probes, morse_fig3_rev cells 337-346, 31-frame median filter). Reproduced when, for every sensor recorded, the mean dF/F0 after its ligand's injection exceeds 1.96 times the largest pre-injection SD across probes (Methods threshold rule) within 0 to 45 min, and the TAU-RISE of the sensors' mean traces span more than a factor of 2 (diverse kinetics). Sensors with no cell in any probe are listed.
- fig4-ip-csf-concentrations
IP delivery raised CSF concentrations to tens of nanomolar for 5-HT and hundreds of nanomolar for AVP and OXT.
We found that IP delivery drove CSF levels in the tens of nanomolar (5-HT) or hundreds of nanomolar (AVP, OXT) range ( Figures 4A , D , G )
Results, Real time, quantitative tracking of CSF composition with MORSE in vivo, paragraph 3; Figure 4A, D, G and 4B, E, H leftHow it will be judged
CONC-INVIVO on the mice with ICV of the same ligand (AVP 5, OXT 6, 5-HT 5; morse_fig3_rev cell 224). Which value is meant is open. Readings (a) the mean across mice of the value at 30 min after injection, a single timepoint (Figure 4B, E, H left; code cells 294-297); (b) the mean across mice of the peak within 0 to 45 min. Reproduced when, under one reading, 5-HT lies in 9 to 109 nM and AVP and OXT each in 90 to 1099 nM.
- fig4-ex-vivo-elevations
In naive mice, IP injections raised serum levels of the injected ligands 30 min later, and raised AVP and OXT in CSF as well.
We found that the IP injections drove elevated levels in serum, with AVP and OXT increases also detected in CSF.
Results, Real time, quantitative tracking of CSF composition with MORSE in vivo, paragraph 3; Figure 4B-C, E-F, H-IHow it will be judged
ELISA (AVP, OXT) and mass spectrometry (5-HT, 5-HIAA) of cisterna magna CSF and cardiac serum 30 min after IP ligand (n=3 AVP, n=3 OXT, n=4 5-HT) or saline (n=2, 2, 4), as in the Figure 4 legends; values as tabulated (the code loads AVP_all_samples.csv, OXT_all_samples.csv, vandy_core.csv, morse_fig3_rev cells 298-323). No test is reported. Readings (a) the IP group mean exceeds the saline mean; (b) every IP sample exceeds every saline sample. Reproduced when both hold for serum AVP, OXT and 5-HT and for CSF AVP and OXT; not reproduced when neither does; blocked as methods_underspecified when they differ. Needs the ELISA and mass spectrometry tables.
- fig4-ex-vivo-two-orders-lower
CSF concentrations measured ex vivo were over two orders of magnitude lower than those MORSE estimated in vivo 30 min after the same IP injection.
Notably, the levels quantified in these ex vivo CSF samples were over two orders of magnitude lower than those estimated in vivo by MORSE at the same 30-minute timepoint following IP injection.
Results, Real time, quantitative tracking of CSF composition with MORSE in vivo, paragraph 3; Figure 4B, E, HHow it will be judged
For AVP and OXT, the mean across mice of the MORSE 30-min value (fig4-ip-csf-concentrations reading a) divided by the mean ELISA CSF value of the IP group; reproduced when both ratios exceed 100. The sentence follows the one on AVP and OXT increases in CSF, and does not say whether 5-HT is included. For 5-HT the ratio of the MORSE 30-min value to the mean mass spectrometry CSF value of the IP group (Figure 4H) is reported, or to the detection limit (0.050 pg/uL, about 0.28 nM; Methods) where values lie below it; the paper does not say whether they do. Not computed in the code (a note in morse_fig3_rev cell 334).
- figs6j-serum-dipping-matches-elisa
AVP concentrations estimated by MORSE dipping into serum samples ex vivo were similar to ELISA values from the same kind of samples.
AVP concentration estimated by MORSE ex vivo by dipping in serum samples was similar to that obtained with ELISA
Results, Real time, quantitative tracking of CSF composition with MORSE in vivo, paragraph 3; Figure S6G-JHow it will be judged
A ten-sensor probe dipped in serum from IP AVP mice (n=3); AVP concentration by Eq. 8 from the AVP sensor signal calibrated in the same recording (S6H). The code takes twice the median over the last part of each well (morse_fig2 cell 56, 2*median of frames 104 to 156) and prints 13.9, 31.0 and 39.5 nM (morse_fig3_rev cell 320); the factor 2 is not explained, so readings with and without it are shown. "Similar" is not defined; readings (a) the MORSE and ELISA ranges overlap; (b) the ratio of means lies within a factor of 10 (against the "two orders of magnitude" contrast the paper draws). Reproduced when both hold; not reproduced when neither does; blocked as methods_underspecified when they differ.
- fig5ab-spontaneous-5ht
In 3 of 7 recordings of undisturbed mice, CSF 5-HT rose spontaneously within the first 60 to 90 min, in all 5-HT sensor cells and in none of the other sensor cells.
In a subset of experiments (n=3/7), we observed a spontaneous elevation in CSF 5-HT that was evident in all 5-HT sensor cells (but none of the other sensor cells) within the first 60–90 min of the recording
Results, MORSE detects endogenous CSF signals at concentrations that impact downstream signaling, paragraph 1; Figure 5A-BHow it will be judged
The 7 three-sensor recordings with 2 h of quiet imaging before IP injections. The code picks the 3 by hand (morse_fig3 cells 232-243, l_vals 1, 2, 5) and gives no rule. Reading, an elevation is a rise of the per-mouse median 5-HT sensor dF/F0 (F0 the mean of the lowest 1/15 of values, 51-frame median filter, code cell 232) above 1.96 times its SD over the first 15 min, peaking within the first 90 min and before the first IP injection, with every 5-HT cell rising above its own baseline and no AVP or OXT sensor median doing so. Count exact.
- fig5cd-spontaneous-5ht-dynamics
The spontaneous CSF 5-HT fluctuations were as large as or larger than responses to IP 5-HT, and rose and decayed over tens of minutes.
These fluctuations were of similar magnitude or larger than the signals we observed in CSF following IP 5-HT delivery (c.f. Figure 3F ) and slowly rose and decayed over tens of minutes ( Figures 5C - D )
Results, MORSE detects endogenous CSF signals at concentrations that impact downstream signaling, paragraph 1; Figure 5C-DHow it will be judged
On the 3 recordings of fig5ab-spontaneous-5ht, the per-mouse mean 5-HT trace; FWHM as the last minus the first time above half the maximum, tau-off as the last time above max/e minus the time of the peak (morse_fig3 cells 238-243, which print FWHM 79.8, 94.6 and 87.6 min and tau-off 91.1, 84.8 and 76.1 min). Reproduced when the FWHM and tau-off of each lie between 10 and 99 min (tens of minutes) and each peak dF/F0 is at least 0.9 times the mean peak dF/F0 of the 5-HT sensor after IP 5-HT in fig3df-ip-responses (same F0 rule across both, readings with the Figure 3 F0 and with the Figure 5 F0).
- fig5gh-chp-calcium
IP and ICV 5-HT drove calcium responses of similar amplitude in choroid plexus epithelial cells, both lasting at least 60 min.
Both routes of 5-HT administration drove similar amplitudes of calcium activity in the ChP epithelial cells, and both responses persisted for at least 60 minutes following delivery.
Results, MORSE detects endogenous CSF signals at concentrations that impact downstream signaling, paragraph 2; Figure 5G-HHow it will be judged
GCaMP6f in FoxJ1-Cre::Ai95D mice, ICV n=3 and IP n=6. A square ROI over several epithelial cells, the mean signal; an exponential fitted to the signal before injection and subtracted where it decays; dF/F0 against the first 5 min (Methods, ChP signal processing). The code (morse_fig3 cells 334-405) median filters over about 100 s and divides by the mean over the 15 min before injection; both F0 readings are shown. Peak = maximum after injection. Code prints peaks ICV 0.72, 0.75, 1.47 and IP 0.58, 0.50, 0.34, 0.61, 1.28, 0.59. "Similar" read as the IP and ICV mean peaks within a factor of 2 of each other. Persistence read as the mean dF/F0 at 60 min after delivery still above 1.96 times the pre-injection SD, for both routes. Needs the ChP traces.
- fig6-chp-grab-similar-time-courses
GRAB sensors expressed in the choroid plexus epithelium report IP 5-HT, AVP and OXT with time courses similar to those seen with MORSE in CSF.
In subsets of mice, IP delivery of 5-HT, AVP, or OXT drove associated sensor signals in ChP with similar time courses as observed previously with MORSE
Results, MORSE detects endogenous CSF signals at concentrations that impact downstream signaling, paragraph 3; Figure 6C, E, G against Figure 3F, D, EHow it will be judged
ChP GRAB dF/F0 as in fig5gh-chp-calcium (5-HT 5 injections in 4 mice, AVP 9 in 4, OXT 3 in 2), TAU-RISE per injection (morse_fig3 prints 5-HT 1.9 to 6.8 min, AVP 4.4 to 42.9, OXT 4.2 to 29.3). "Similar" is not defined; readings (a) the mean ChP tau-rise of each ligand lies within the range of per-mouse MORSE tau-rise of fig3df-ip-responses; (b) the mean tau-rise differ by less than a factor of 2. Needs the ChP traces.
- fig6c-ip-5ht-reaches-chp-fast-enough
IP 5-HT reached the apical choroid plexus membrane fast enough to explain the choroid plexus calcium response.
Furthermore, IP 5-HT reached the apical ChP membrane ( Figure 6C ) fast enough to explain the ChP calcium response ( Figure 5H )
Results, MORSE detects endogenous CSF signals at concentrations that impact downstream signaling, paragraph 3; Figure 6C, Figure 5HHow it will be judged
TAU-RISE after IP 5-HT of the ChP GRAB-5-HT2h signal (Figure 6C right, 5 injections) against the GCaMP6f signal (Figure 5H right, 6 mice). The paper does not say what "fast enough" means. Readings (a) the mean GRAB tau-rise is not later than the mean GCaMP tau-rise plus its SEM; (b) the mean GRAB signal exceeds 1.96 times its pre-injection SD (the Methods' threshold rule) no later than the mean GCaMP signal does. The outcome is the one both give; if they differ, the claim is blocked as methods_underspecified. The code's printed values (below) fail reading (a), so it may not be what the authors meant. The code prints GRAB 6.8, 5.5, 1.9, 5.0, 3.8 min and GCaMP 1.4, 4.4, 0.6, 1.3, 1.6, 3.0 min (morse_fig3), which is noted; the code's exclusion of one GCaMP mouse from the mean is shown as a variant.
- figs6a-no-implant-inflammation
MORSE probe implantation caused no asymmetry in microglial (Iba1) or astrocytic (GFAP) staining, nor in diffusion of labeled BSA from CSF into tissue.
No Iba1 or GFAP signal asymmetry between left and right hemispheres induced by MORSE probe implantation into one of the cannulae was noted in any animals (n=12) across the span of the ventricular system.
Figure S6A-C legendsHow it will be judged
Histology of 12 mice. A claim of absence with no test in the paper; reproduced when the Iba1 and GFAP intensities on the implanted side are within the range of the other side in every mouse, and the BSA profile (S6C) is symmetric. Needs histology images, which the paper does not say are deposited.
Pipeline
- Paper
- done (1 of 1)
- Claims
- done
- Claims check
- done (ready)
- Attempt
- todo
- Questions
- none
- Verified
- none