Claims 38
From the paper Anxa1+ dopamine neuron vulnerability defines prodromal Parkinson’s disease bradykinesia and procedural motor learning impairment · claims file · checked 2026-10-21: ready
- fig1b-anxa1-downregulated-rnaseq
In bulk RNA-seq of sorted midbrain DANs, both early-stage mitochondrial PD models (cMfn2KO-early and MP-early) show significant downregulation of the Anxa1 transcript among DAN subtype markers.
Differential expression analysis of genes enriched in midbrain DAN subtypes revealed that both PD models showed a shared profile of significant early-stage downregulation of the Annexin A1 (Anxa1) transcript
Results, Anxa1 is a marker of an early PD vulnerable SNc DAN subpopulation, paragraph 1; Figure 1A and 1BHow it will be judged
Figure 1B marks markers downregulated at FDR < 0.05. The pipeline (aligner, counting, differential expression package) is not given in the text and is open (the Acknowledgements credit NBIS staff with the RNA-seq analysis; reference 41, Filograna et al. 2021, is the group's earlier cMfn2KO data). Reproduced when Anxa1 is downregulated against its controls at FDR < 0.05 in both models in the paper's own differential expression results, if GEO holds them. Otherwise the readings are DESeq2 and edgeR with default settings on the deposited counts, and the claim is judged only where both give the same outcome. The paper names one deposit for its RNA-seq data, GSE282753 (Code and data availability), and no other place for the cMfn2KO counts; reference 41 is cited for the mouse model only. Where the deposit lacks the cMfn2KO-early data, that half cannot be tested and the claim stays blocked; the MitoPark half is run and reported. The paper does not say how technical libraries of a mouse are combined. The legend counts mice (MP-early n = 5), so the mouse is the unit, and libraries are summed per mouse before the test (reading a); treating each library as a sample does not fit the legend's n and is shown only as a variant. GEO holds more mice than the legend's 5 per model; which were used is not given, and all are used.
- fig1c-anxa1-aldh1a1-overlap
About 99% of Anxa1+ DANs co-express Aldh1a1, while about 40% of Aldh1a1+ DANs express Anxa1.
We found that ∼99% of Anxa1+ DANs co-expressed Aldh1a1, while ∼40% of the Aldh1a1+ DANs expressed Anxa1
Results, Anxa1 is a marker of an early PD vulnerable SNc DAN subpopulation, paragraph 2; Figure 1C; Figure S1How it will be judged
Counts of Th+ cells by Anxa1 and Aldh1a1 status, pooled over the sections and mice of Figure 1C and Figure S1 (control mice). Whether the fractions pool cells or average mice is open; reproduced when both values fall within tolerance under either reading.
- fig1e-ventral-snc-location
Anxa1+ DANs lie in the ventral SNc, with a small fraction in the VTA; they are lateral in anterior SNc and progressively more medial in posterior planes.
we found that the Anxa1+ DANs were localized in the ventral part of the SNc, and a small fraction was found in the VTA region ... Anxa1+ DANs were localized laterally in the anterior portions of the SNc, with their distribution becoming progressively more medial towards more posterior coronal planes
Results, Anxa1 is a marker of an early PD vulnerable SNc DAN subpopulation, paragraph 2; Figure 1C to 1EHow it will be judged
From cell coordinates (X-Y per section, Methods, Anatomical tracing) with the medial lemniscus as the SNc-VTA border. Reproduced when the median dorsoventral position of Anxa1+ DANs is ventral to that of Aldh1a1+/Anxa1- and Th-only DANs, most Anxa1+ DANs lie in SNc, and their median mediolateral distance from the midline decreases from the most anterior to the most posterior section. The paper gives no measure or test; this rule is set here.
- fig1h-mp-early-anxa1-loss
In MP-early mice the number of Anxa1+ DANs in the ventral tier of the SNc and the fraction of Anxa1+ DANs among Th+ SNc cells are significantly reduced, while the VTA Anxa1+ DANs are not reduced.
These experiments revealed a significant reduction in the number of Anxa1+ DANs in the ventral tier of the SNc in MP-early mice ... we observed a significant decrease in the fraction of Anxa1+ DANs relative to the total Th+ cells in the SNc, whereas no reduction was found in the Anxa1+ DANs of the VTA
Results, Anxa1 is a marker of an early PD vulnerable SNc DAN subpopulation, paragraph 3; Figure 1F to 1H; Figure S1How it will be judged
Per mouse, the proportions of the three SNc subpopulations (Anxa1+ Aldh1a1+ Th+, Aldh1a1+ Anxa1- Th+, Th+ only) among Th+ cells; two-way ANOVA genotype by subtype with Sidak's post hoc, as in the legend of Figure 1H. Reproduced when the interaction has p < 0.05 and the Anxa1+ proportion is lower in MP-early mice at Sidak p < 0.05. The legend does not say which subtype carries p = 0.0452 and which p < 0.0001; both are reported. The reduced count of Anxa1+ DANs in the ventral tier (Figure 1G, a dorsoventral line graph) has no test in the text; it is reported per mouse and not judged. The VTA part (Figure S1) is a claim of absence whose test is in the supplement, which was not read; it is judged with the same two-way ANOVA and Sidak test on VTA cells, reproduced when the Anxa1+ fraction does not differ at p < 0.05.
- fig1j-ipsc-anxa1-reduced
In iPSC-derived DANs from PD patients with SNCA triplication, ANXA1 mRNA is significantly reduced against healthy donors, while ALDH1A1 is not significantly altered.
While ALDH1A1 expression was not significantly altered, we observed a significant reduction in ANXA1 expression in iPSC-DANs carrying the SNCA-trp mutation
Results, Anxa1 is a marker of an early PD vulnerable SNc DAN subpopulation, paragraph 4; Figure 1I and 1JHow it will be judged
RT-qPCR normalized to B2M (Methods). The legend names no test and no n. The Methods' two-group tests are (a) the two-sided unpaired t-test per gene and (b) multiple two-sided unpaired t-tests with Holm-Sidak adjustment over the genes compared (Methods); both are applied to the normalized expression per line or culture. Reproduced when, under a reading, ANXA1 is lower in SNCA-trp at p < 0.05 and ALDH1A1 does not differ at p < 0.05. Whether the unit is a donor line or a differentiation is open; both are run where the data allow.
- fig1m-6ohda-anxa1-loss
After a mild unilateral 6-OHDA lesion, the Anxa1+ fraction of SNc DANs is significantly lower on the lesioned side than on the intact side.
In this model, we observed a significant decrease in the Anxa1+ fraction in the SNc compared to the intact side
Results, Anxa1 is a marker of an early PD vulnerable SNc DAN subpopulation, paragraph 5; Figure 1K to 1MHow it will be judged
Per mouse and side, the proportions of the three SNc subpopulations among Th+ cells; RM two-way ANOVA side by subtype with Sidak's post hoc, as in the legend. The residual df of 9 with 4 mice fits side as the repeated factor and subtype as a between factor (12 mouse-subtype rows, 12 - 3 = 9); that layout is used, and the fully repeated layout is a variant. Reproduced when the interaction has p < 0.05 and the Anxa1+ proportion is lower on the lesioned side at Sidak p < 0.05.
- figs1-6ohda-partial-th-loss
The low-dose 6-OHDA model gives a partial loss of about 50% of Th+ DANs.
This approach established a mild degeneration model, resulting in partial loss (∼50%) of Th+ DANs
Results, Anxa1 is a marker of an early PD vulnerable SNc DAN subpopulation, paragraph 5; Figure S1How it will be judged
Loss is one minus the ratio of Th+ SNc cells on the lesioned side to the intact side. The text does not say how mice are combined; readings are (a) per mouse, then the mean over mice, and (b) counts pooled over mice. Whether SNc alone or SNc and VTA are counted is shown in Figure S1, which was not read; both readings are run, with (a) and (b) for each.
- fig1p-catg7ko-anxa1-loss
In 24-week-old cAtg7KO mice with Lewy body-like inclusions the SNc shows a specific, significant reduction of Anxa1+ DANs.
In the SNc of cAtg7KO mice, we observed a specific and significant reduction in Anxa1+ DANs
Results, Anxa1 is a marker of an early PD vulnerable SNc DAN subpopulation, paragraph 5; Figure 1N to 1PHow it will be judged
As in fig1h-mp-early-anxa1-loss, two-way ANOVA genotype by subtype with Sidak's post hoc. Reproduced when the interaction has p < 0.05, the Anxa1+ proportion is lower in cAtg7KO at Sidak p < 0.05, and no other subtype is lower at p < 0.05 (the legend reports one post hoc difference only).
- figs1-vta-no-effect
In the 6-OHDA and cAtg7KO models the Anxa1+ fraction in the VTA is not significantly changed.
Notably, analyses of the VTA in both models revealed no significant effects on the Anxa1+ fraction
Results, Anxa1 is a marker of an early PD vulnerable SNc DAN subpopulation, paragraph 5; Figure S1How it will be judged
A claim of absence. The test is in Figure S1, which was not read; the paper's tests for the matching SNc panels are an RM two-way ANOVA (6-OHDA) and a two-way ANOVA (cAtg7KO) with Sidak's post hoc, and these are applied to VTA proportions. Reproduced when, in each model, the Anxa1+ fraction does not differ at Sidak p < 0.05; the size of the difference is reported.
- fig2c-snr-dendrites
Anxa1+ DANs extend long dendrites deep into the anterior and medial SNr, with a dendron-like structure in anterior SNr.
We observed that Anxa1+ DANs extend long dendrites deep into the anterior and medial portions of the SNr
Results, Specialized organization of Anxa1+ DANs, paragraphs 1 and 3; Figure 2C to 2E and 2IHow it will be judged
Anatomical images; the paper gives no measure or test. Reproduced when eYFP+ dendrites are present in SNr and their density is higher in anterior than posterior and in medial than lateral SNr sections of the images. This rule is set here, as the paper gives none.
- fig2f-dorsal-cp-projection
Anxa1+ DANs project densely to the most dorsal domain of the CP, with a density gradient enriched anteromedially, and sparsely to the core and medial shell of the ACB.
we found that Anxa1+ DANs formed a dense projection to the most dorsal domain of the CP (dCP) and some axonal labeling in the core and medial shell of the ACB ... Anxa1+ axonal labeling revealed a pronounced density gradient enriched in the anteromedial regions of the dCP
Results, Specialized organization of Anxa1+ DANs, paragraph 1; Figure 2F to 2HHow it will be judged
eYFP optical density in coronal striatal sections (OD, Figure 2 abbreviations). The paper gives no test. Reproduced when the dorsal band of CP has higher eYFP density than central CP and ACB in each section, and the dCP density is higher in anterior and medial than posterior and lateral portions. Region boundaries are drawn from the atlas; the band depth is open. This rule is set here, as the paper gives none.
- figs2-evoked-da-dcp-ccp
Optogenetic activation of Anxa1+ DANs evokes DA release in the dCP at varying intensities, and the evoked release in the cCP has a lower amplitude.
Optogenetic activation of Anxa1+ DANs at varying intensities elicited DA release in the dCP ... optogenetically evoked DA release in the cCP exhibited a lower amplitude compared to the dCP signal
Results, Specialized organization of Anxa1+ DANs, paragraph 2; Figure S2How it will be judged
dLight1.3b z-scored signal aligned to ChRmine stimulation (1 s at 40 Hz, Methods). The intensities, window, test and n are in Figure S2, which was not read; all are open. Reproduced when the peak of the mean evoked response after stimulation onset exceeds the pre-stimulus baseline in dCP and is lower in cCP than in dCP, per recording, by a two-sided unpaired t-test (the Methods' two-group test) at p < 0.05. Peak within 0 to 1 s and within 0 to 2 s of onset are both run. These windows are set here (the train lasts 1 s, Methods); whether dCP and cCP were recorded in the same or in different mice is not given, and a paired t-test is run as well where the same mice hold both. The main text says only "at varying intensities"; which powers enter the comparison, the unit (session or mouse) and the pre-stimulus baseline window are not given (Figure S2 not read). The readings are (a) each session's mean over all trains, all powers pooled, the session as unit; (b) each power alone, the session as unit. A per-mouse summary is reported, not judged (two mice per site). The baseline is the mean z-score over (i) 1 s and (ii) 2 s before train onset; both are run.
- fig2i-striosomal-features
Anxa1+ DAN axons in dorsolateral CP are enriched in MOR+ striosomes and the subcallosal streak, suggesting a nigro-striato-nigral loop.
eYFP axonal labeling in the dorsolateral CP was enriched in the mu-opioid receptor (MOR) immunoreactive striosomes and subcallosal streak
Results, Specialized organization of Anxa1+ DANs, paragraph 3; Figure 2IHow it will be judged
Images only; no measure or test is given. Reproduced when eYFP density inside MOR+ patches is higher than in the surrounding matrix of dorsolateral CP. This rule is set here.
- fig2n-6ohda-axon-ablation
After low-dose 6-OHDA, degeneration of eYFP+ Anxa1+ SNc neurons is accompanied by near-complete loss of their eYFP+ axons in the dCP across the anteroposterior axis.
Degeneration of the eYFP+ SNc neurons was accompanied by near-complete ablation of the corresponding eYFP+ axons in the dCP across the anteroposterior axis
Results, Specialized organization of Anxa1+ DANs, paragraph 4; Figure 2J to 2NHow it will be judged
eYFP optical density in dCP, lesioned against intact side, per section (Figure 2N, n = 5 mice). No test is given. "Near-complete" is not quantified; reproduced when the lesioned dCP density is below 10% of the intact side (background subtracted) at every anteroposterior level in the mean over mice, partially reproduced when it is reduced at every level but not below 10%. The 10% bound is set here, as the paper gives none; the measured ratio is reported.
- fig3d-anxa1-ramp-to-touch
Anxa1+ and DAT+ axons in dCP ramp up before a water drop, peak at contact with it, and return quickly to baseline during water consumption.
Recordings showed that the Anxa1+ as well as the DAT+ axons in dCP exhibited a ramping activity that peaked upon contact with the water droplet, followed by a rapid return to baseline during water consumption
Results, Anxa1+ DANs do not encode motivational value, paragraph 1; Figure 3DHow it will be judged
Z-scored signals aligned to the water drop touch; per recording, the mean over touches of the AUC in three windows; RM one-way ANOVA (Geisser-Greenhouse corrected, as the fractional degrees of freedom show) with Sidak's post hoc, the recording as unit (legend). The first window is the baseline (legend of 3F-G). The window edges are drawn only in Figure 3D, which was not seen; they are open. Readings are (a) 1 s windows, -2 to -1, -1 to 0 and 0 to 1 s from touch, the width the paper uses for motif AUC in Figure 4, and (b) 2 s windows, -4 to -2, -2 to 0 and 0 to 2 s. Reproduced when, under a reading, the Anxa1 ANOVA has p < 0.05 and AUC is larger in the second window than the third at Sidak p < 0.05, and the DAT ANOVA has p < 0.05 with the second window larger than the first. A reader may be asked for the edges from the figure. The event is "the water drop touch" (legend) and "contact with the water droplet" (Results); the paper names no event types. Readings are (i) correct (drop retrieved) touches only and (ii) every contact with a drop (correct, drop and semidrop); empty-spout events are not touches of a drop and are excluded in both. Each window reading is run with both. The paper gives no rule for which recordings enter (legend n of 10, 4, 13 and 11) and does not say whether "Artifacts and instances of signal intensity loss were excluded" (Methods) drops whole recordings or stretches of signal. Every recording of the group and site with at least one event is used; the n used is reported against the legend's, and a per-mouse summary is reported, not judged.
- fig3e-dlight-dcp-ccp-reward
DA release in dCP follows the same ramp-and-drop dynamics as Anxa1+ axons, while in cCP it shows a classical reward-related profile, ramping before contact and peaking during consumption before returning to baseline.
Similarly, the recorded DA signals in the dCP showed the same dynamics ... In contrast, recordings from the cCP revealed a typical reward-related phasic DA release profile, showing a ramping signal before droplet contact with a peak during reward consumption that then returned to baseline
Results, Anxa1+ DANs do not encode motivational value, paragraph 1; Figure 3EHow it will be judged
As in fig3d-anxa1-ramp-to-touch, with the same window, event and recording readings. Reproduced when, under a reading, dLight-dCP has ANOVA p < 0.05 and the second window larger than the third at Sidak p < 0.05, and dLight-cCP has ANOVA p < 0.05 with both the second and third windows larger than the first at Sidak p < 0.05.
- fig3fg-auc-change-from-baseline
Against the baseline window, the legend lists three AUC changes, the second window of DAT GCaMP-dCP (p = 0.0308) and of dLight-cCP (p = 0.0028) and the third window of dLight-cCP (p = 0.0148), with only the dLight-cCP second window marked significant; the other groups and windows are not listed.
How it will be judged
Per recording, the AUC of the second and of the third window minus that of the first; one-sample two-tailed t-test against 0, Bonferroni corrected over the eight tests (four groups, two windows), window, event and recording readings as in fig3d-anxa1-ramp-to-touch. Only dLight-cCP second carries an asterisk in the legend; whether the printed p are corrected, and over how many tests, is open. The single asterisk fits printed raw p with Bonferroni over eight tests (all groups and windows) or over four (the groups, per window); both families are run. Reproduced when, under a window reading, the dLight-cCP second-window change is positive and significant after Bonferroni under a family, and the other two listed changes are positive; their raw and corrected p and the other five tests are reported.
- fig3i-dat-not-anxa1-self-stimulation
Optogenetic activation of DAT+ neurons strongly reinforces self-stimulation in the first two days, while activation of Anxa1+ DANs does not.
We found that optogenetic activation of DAT+ neurons strongly reinforced self-stimulation during the first two days of the task ... In contrast, optogenetic activation of the Anxa1+ DANs did not support self-stimulation behavior
Results, Anxa1+ DANs do not encode motivational value, paragraph 2; Figure 3I and 3JHow it will be judged
Days 1 and 2 (right port active). Active pokes per mouse over the two days (sum or mean give the same ranks, so the choice cannot change the outcome). Preference is not defined in the text; readings are (a) active over total pokes and (b) active minus inactive pokes. Kruskal- Wallis over the three groups with Dunn's post hoc. Reproduced when, under a reading of preference, both Kruskal-Wallis tests have p < 0.05, DAT-ChRmine exceeds both other groups at Dunn p < 0.05 for active pokes and for preference against Anxa1-ChRmine, and Anxa1-ChRmine does not differ from Ctrl-Tdtomato.
- fig3kl-reversal-dat-only
When the active port is reversed, DAT-ChRmine mice quickly switch to the newly active left port, while Anxa1-ChRmine mice develop no significant preference for the active port in the last reversal days.
When active and inactive port assignment was reversed, DAT-Cre mice quickly reversed their responding from the inactive (right) to the active (left) port ... mice did not develop any significant preference for the active port during the last days of the reversal
Results, Anxa1+ DANs do not encode motivational value, paragraph 2; Figure 3K and 3LHow it will be judged
The legend does not say which group the RM two-way ANOVA and Friedman test belong to; F(4, 32) fits 9 mice over five days, so the reading is DAT-ChRmine, days -1 and 1 to 4 (the other reading, all six days, is run as well). Per group, RM two-way ANOVA side by day on left and right pokes with Sidak's post hoc, and Friedman test on preference over days with Dunn's post hoc, preference read as in fig3i. Reproduced when for DAT-ChRmine the interaction has p < 0.05 and the Friedman test has p < 0.05 with preference changed from day -1 to day 1, and, as a claim of absence, Anxa1-ChRmine shows no left against right difference at Sidak p < 0.05 on reversal days 3 and 4.
- fig3o-tetx-no-motivation-deficit
Mice with silenced Anxa1+ DANs show no significant change in total nosepokes in the FR5 or PR stages.
Mice with genetic silencing of Anxa1+ DANs exhibited no significant changes in the total number of nosepokes during FR5 or PR tasks
Results, Anxa1+ DANs do not encode motivational value, paragraph 2; Figure 3M to 3OHow it will be judged
A claim of absence. The legend names no test and no n; the Methods' test for two groups is the two-sided unpaired t-test, used here. Total pokes per mouse over the two FR5 days and over the two PR days, Anxa1-TeTx against control. Reproduced when neither comparison reaches p < 0.05; the difference and its size are reported.
- fig4b-speed-correlation-dcp
Anxa1+ axonal activity in the dCP correlates positively with linear speed during open field exploration, above a shuffled control, and DA release in dCP shows a similar correlation.
We found a significant positive correlation between Anxa1+ axonal activity in the dCP and linear speed, and this correlation was similar when we recorded DA release in the dCP
Results, Anxa1+ DAN projections are activated during vigorous actions, paragraph 1; Figure 4BHow it will be judged
Per recording, the correlation over the 20 min session of the z-scored FP signal with linear speed, against the same with the signal shuffled; two-sided paired t-test over recordings (legend). Pearson's r, the paper's R. Open, (1) the shuffle, read as (a) a random permutation of samples and (b) a circular shift by a random lag; (2) the time base, the FP signal at 60 fps averaged onto the 30 fps video frames, a standard choice; (3) speed as in the header. Reproduced when, under a reading, both groups have a mean correlation above 0 and above the shuffle at p < 0.05. The printed p equal the exact two-sided Wilcoxon signed-rank p when all 9 (or 8) recordings lie on one side (2/512 and 2/256); the Wilcoxon test is run as a variant. DAT GCaMP-dCP (5 mice, 7 recordings) has no p in the legend and is reported.
- fig4b-speed-correlation-ccp-negative
DA release in the cCP correlates negatively with locomotion speed.
The recorded DA release in cCP instead showed a negative correlation with locomotion speed
Results, Anxa1+ DAN projections are activated during vigorous actions, paragraph 1; Figure 4BHow it will be judged
As in fig4b-speed-correlation-dcp. Reproduced when the mean dLight-cCP correlation is below 0 and below the shuffle at p < 0.05 under a reading (p = 0.016 equals the exact Wilcoxon p 2/128 for 7 recordings on one side; Wilcoxon as a variant).
- fig4c-vame-motifs
Unsupervised VAME analysis of open field pose identifies 37 kinematic motifs, grouped into action categories.
This unsupervised analysis of postural dynamics identified 37 kinematic motifs, which we categorized into seven communities of actions
Results, Anxa1+ DAN projections are activated during vigorous actions, paragraph 2; Figure 4C; Methods, Behavioral analysisHow it will be judged
A count of the distinct motif labels in the paper's motif labelling (header). The tolerance is 0, as a label set is counted exactly. The text names seven communities while the legend of Figure 4C names six categories (Stationary/Sniffing, Turn, Rear, Groom, Walk, Sniff/Walk) and Figure 5G five; the category of each motif is drawn only in Figure 4C and is taken from the dataset if it holds it, otherwise it is open.
- fig4e-anxa1-high-speed-motifs
Anxa1+ axon activity is associated with the onset of high-speed motifs regardless of action type, as is DA in dCP, while DA in cCP rises strongly after the onset of low-speed motifs.
This analysis demonstrated that the activity of Anxa1+ projections was associated with high-speed motifs independently of the action type, and this profile was also observed in DA signals recorded in the dCP ... However, DA signals recorded in the cCP demonstrated a different pattern with strong activation after the onset of low-speed motifs
Results, Anxa1+ DAN projections are activated during vigorous actions, paragraph 2; Figure 4D and 4E; Figure S3How it will be judged
Mean z-scored signal aligned to each motif onset (first frame of a run of that label), motifs ranked by mean linear speed. Judged by fig4f-auc-correlations and fig4h-auc-motif-speed; in addition, reproduced for "independently of the action type" when, within at least two action categories, the Spearman correlation of Anxa1 AUC (0 to 1 s) with motif speed is positive, and for cCP when its AUC correlates negatively with motif speed. Spearman within categories is set here, as the paper gives no test for this sentence.
- fig4f-auc-correlations
Across motifs, the onset AUC of Anxa1+ terminal activity correlates positively with that of dLight in dCP (R = 0.70) and negatively with that of dLight in cCP (R = -0.52).
the area under the curve (AUC, 0 to 1 sec) calculated for each motif from Anxa1-terminal activity showed a positive correlation with the AUC of dLight signals recorded in the dCP (Figure 4F) and a negative correlation with the AUC of dLight signals recorded in the cCP
Results, Anxa1+ DAN projections are activated during vigorous actions, paragraph 2; Figure 4F and 4GHow it will be judged
Per motif, the AUC from 0 to 1 s after onset of the z-scored signal (legend), then Pearson's R over motifs between groups. Open, how traces are averaged; readings are (a) all onsets of all recordings of a group pooled, then the mean trace, and (b) the mean per recording, then the mean over recordings. Whether a pre-onset baseline is subtracted is open; readings are (i) the AUC of the z-scored trace as is and (ii) minus its mean over -1 to 0 s. Reproduced when, under a reading, both R are within tolerance and p < 0.01.
- fig4h-auc-motif-speed
The onset AUC of Anxa1+ terminal activity is highly correlated with mean motif linear speed across motifs (R = 0.83).
Furthermore, the AUC from Anxa1-terminal activity was highly positively correlated with motif speed
Results, Anxa1+ DAN projections are activated during vigorous actions, paragraph 2; Figure 4HHow it will be judged
AUC readings as in fig4f-auc-correlations; motif speed is the mean linear speed over the frames of the motif (legend of 4H), over the Anxa1 recordings or over all mice of the motif set, both run. Reproduced when R is within tolerance and p < 0.01 under a reading.
- fig5c-tetx-locomotion
Anxa1-TeTx mice show moderate reductions in linear and angular speed in the open field, with a slight increase in immobility time.
we found that Anxa1-tetanus mice showed moderate reductions in linear and angular speed, accompanied by a slight increase in immobility time
Results, Anxa1+ DAN silencing produces a prodromal PD bradykinetic profile, paragraph 1; Figure 5B to 5EHow it will be judged
Per mouse, mean linear speed, mean angular speed and immobility time over the open field session; two-sided unpaired t-test against Ctrl-Tdtomato (legend). Speed and immobility as in the header. Reproduced when linear and angular speed are lower at p < 0.05 and immobility is higher; partially reproduced when one speed is lower at p < 0.05. Which panels carry asterisks is in the figure, not seen (the legend gives only the key *p < 0.05, **p < 0.01, ***p < 0.001, with n = 18 Ctrl-Tdtomato and 14 Anxa1-TeTx); the p of immobility is reported.
- fig5e-early-models-locomotion
The mild locomotor phenotype and acceleration distribution of Anxa1-silenced mice resemble the early-stage PD models (mild 6-OHDA, MP-early), consistent with a bradykinetic and not an akinetic/rigid profile.
The mild locomotor phenotype and acceleration distribution pattern in Axna1-silenced mice was similar to the early-stage PD models suggesting that loss of Anxa1+ DANs is linked primarily to a bradykinetic
Results, Anxa1+ DAN silencing produces a prodromal PD bradykinetic profile, paragraph 1; Figure 5C to 5F; Figure S4How it will be judged
No test of similarity is given. For each experimental group, the change of mean linear speed, mean angular speed and immobility against its own control (in percent of the control mean), and the acceleration density (legend of 5F). Reproduced when the changes of Anxa1-TeTx lie within the range spanned by 6OHDA-low and MP-early, or closer to theirs than to MP-middle's, for at least two of the three measures. The rule is set here; the values, including those of MP-middle, are reported.
- fig5g-pc1-stage
In a PCA of motif usage, PC1 separates the groups by PD stage, with Anxa1-TeTx mice between the early and middle PD models, and PC2 separates middle-stage MitoPark mice from the others.
We found that the first principal component (PC1) effectively separated the groups according to PD stage (Figure 5G). The motif usage profile of Anxa1-tetanus mice positioned them between early and middle PD models along the PC1 axis ... PC2 separated middle-stage MitoPark mice from the other groups
Results, Anxa1+ DAN silencing produces a prodromal PD bradykinetic profile, paragraph 2; Figure 5GHow it will be judged
PCA of per-mouse motif usage time (37 motifs) over all mice of the eight groups. Open, the scaling; readings are (a) standardized usage (Prism's default) and (b) centred only; usage as fraction of session time (equal sessions make seconds equivalent). The sign of a PC is arbitrary and is set so that MP-middle has the largest group mean. Reproduced when, under a reading, the group mean PC1 of Anxa1-TeTx lies between the means of the early models (6OHDA- low, MP-early) and MP-middle, controls lie on the low side, and the MP-middle mean on PC2 lies farther from the mean of all other groups than any other group's mean.
- fig5h-walk-stationary-loadings
Walk and stationary motifs lie at opposite extremes of PC1 in the loading plot.
“Walk” and “stationary” motifs were prominently represented at the opposite extremes of the PC1 axis
Results, Anxa1+ DAN silencing produces a prodromal PD bradykinetic profile, paragraph 2; Figure 5HHow it will be judged
Loadings of PC1 per motif, readings of fig5g. Reproduced when, under a reading, the five highest-loading motifs on one end are mostly walk motifs and the five on the other end are mostly stationary motifs. Motif categories from the dataset, else open (fig4c-vame-motifs).
- fig5i-pc1-motif-speed
Motif PC1 values correlate negatively with mean motif speed (R = -0.762), while PC2 shows no significant correlation (R = -0.1494, p = 0.3772), nor do PC3 and PC4.
linear regression analysis between motif speed and PC1 eigenvalues revealed a significant negative correlation (Figure 5I). In contrast, we did not observe any significant correlation between the eigenvalues of PC2 (Figure 5I), PC3, PC4 and motif speed
Results, Anxa1+ DAN silencing produces a prodromal PD bradykinetic profile, paragraph 2; Figure 5I; Figure S5How it will be judged
Per motif, the PC1 (and PC2 to PC4) loading, called eigenvalue in the paper, against mean motif linear speed over all mice; Pearson R. The printed p = 0.3772 for R = -0.1494 fits 37 motifs. A PC's sign is arbitrary, so the magnitude is judged and the sign reported, with PC1 oriented as in fig5g. Reproduced when, under a reading of fig5g, |R| for PC1 is within tolerance with p < 0.0001, and PC2, PC3 and PC4 each have p >= 0.05.
- fig5j-tetx-five-motifs
Anxa1-silenced mice show significant downregulation of five high-speed motifs (42, 34, 16, 0 and 25), not of all vigorous high-speed motifs.
Specifically, we found that Anxa1-silenced mice displayed significant downregulation of five high-speed motifs (42, 34, 16, 0 and 25)
Results, Anxa1+ DAN silencing produces a prodromal PD bradykinetic profile, paragraph 3; Figure 5JHow it will be judged
Per mouse, usage time of each of the 37 motifs; Anxa1-TeTx against Ctrl-Tdtomato, two-sided unpaired t-test per motif, Holm-Sidak adjusted over the 37 motifs at alpha 0.05 (legend). Motif numbers are the paper's labels. Reproduced when exactly these five motifs are reduced at adjusted p < 0.05; partially reproduced when at least three of them are and no more than one other motif is changed.
- fig5m-walk-motif-network
Anxa1+ DAN silencing has no major effect on the overall structure of motif sequences, but the reduced high-speed walk motifs form a small network of high-probability transitions (motif 0 to 16 to 34).
We did not observe any major effects in the overall structure of motif sequences after Anxa1+ DANs silencing (Figure S5); however, we identified that the significantly reduced high-speed “walk” motifs formed a small network with high-probability transition connections (motif 0 -> motif 16 -> motif 34
Results, Anxa1+ DAN silencing produces a prodromal PD bradykinetic profile, paragraph 4; Figure 5K to 5M; Figure S5How it will be judged
Transition probability from one motif to the next different motif, from the motif label sequence (self-transitions excluded, as in VAME's transition matrix; including them is a variant), pooled over control mice (Ctrl-Tdtomato). Reproduced when the probability from 0 to 16 is the highest from motif 0 among walk motifs, and that from 16 to 34 the highest from motif 16 among walk motifs. The legend's n = 3 mice for Figure 5M does not match n = 14 in Figure 5C, and the "overall structure" test is in Figure S5, not read; the Fisher's exact tests per transition on pooled transition counts, Benjamini-Hochberg adjusted, are run on all mice and reported.
- fig6c-maze-locomotion
In the arrow maze, Anxa1-silenced mice show only a small decrease in linear speed on the fourth day and no change in immobility time.
we found that Anxa1-silenced mice displayed only minor changes in locomotion, showing a small decrease in linear speed on the fourth day of the test without any changes in immobility time
Results, Anxa1+ DANs are necessary for procedural motor learning, paragraph 1; Figure 6C; Figure S6How it will be judged
Per mouse and day (days 1 and 4), mean linear speed and immobility time over the 10 min session; RM two-way ANOVA group by day with Sidak's post hoc. Reproduced when the speed group effect has p < 0.05 with Anxa1-TeTx slower, the Sidak test is p < 0.05 on day 4 only, and, as a claim of absence on the same test, immobility shows no group effect and no Sidak difference at p < 0.05.
- fig6d-maze-preference-learning
Control mice learn to prefer the rewarded arm over four days, while Anxa1-silenced mice show no increase in preference for the rewarded arm.
Control mice successfully learned the basic task strategy over the four days, as evidenced by a significant increase in the preference for the rewarded arm ... However, mice with silenced Anxa1+ DANs failed to show any increase in the preference for the rewarded arm
Results, Anxa1+ DANs are necessary for procedural motor learning, paragraph 1; Figure 6B and 6D leftHow it will be judged
Per mouse and day, the number of rewarded transitions (centre to the left, rewarded, end zone) minus error transitions (centre to the right end zone), from the Bonsai end-zone entries, days 1 and 4; RM two-way ANOVA group by day with Sidak's post hoc. The legend's "Day 1 vs. Day 2" is read as the first against the fourth day, the days plotted. Reproduced when the interaction has p < 0.05, controls increase from day 1 to day 4 at Sidak p < 0.05, Ctrl exceeds Anxa1-TeTx on day 4 at Sidak p < 0.05, and Anxa1-TeTx do not increase at p < 0.05.
- fig6d-maze-efficiency
Control mice improve task efficiency (reward relative to distance traveled) over the four days, while Anxa1-silenced mice do not.
a significant increase in task efficiency quantified as the received reward relative to the traveled distance ... and did not improve their motor efficiency in the task
Results, Anxa1+ DANs are necessary for procedural motor learning, paragraph 1; Figure 6D right; Figure S6How it will be judged
Per mouse and day, rewards received divided by distance traveled (Results), days 1 and 4; whether the central initiation drops count as rewards is open, readings (a) lateral rewards only and (b) all drops. RM two-way ANOVA group by day with Sidak's post hoc. Reproduced when, under a reading, the interaction has p < 0.05, controls rise from day 1 to day 4 at Sidak p < 0.05, Anxa1-TeTx do not, and Ctrl exceeds Anxa1-TeTx on day 4 at Sidak p < 0.05. The legend's two group differences (p = 0.0027 and p < 0.0001) imply a difference on day 1 as well; it is reported.
- fig6j-reaching-accuracy
Control mice improve reaching accuracy over four sessions; Anxa1-TeTx mice have normal accuracy in the first session but fail to improve, and Anxa1-Arch mice show reduced accuracy during the last three sessions with optogenetic inhibition.
As expected, control mice improved their reaching accuracy over the four training sessions ... In contrast, Anxa1-tetanus mice exhibited normal accuracy during the first session but failed to refine their grasping movements in subsequent sessions ... as Anxa1-Arch mice showed a significant reduction in reaching accuracy during the last three sessions
Results, Anxa1+ DANs are necessary for procedural motor learning, paragraph 2; Figure 6H to 6J; Figure S6How it will be judged
Per mouse and session, the fraction of grasp attempts with the preferred forelimb whose measured distance from the target is within a threshold (Results; distances measured by hand in Kinovea). The threshold is given as < 0.2 cm in the Results and < 0.1 cm in the legend of 6J; both are run. RM two-way ANOVA group by day on days 1 and 4 with Sidak's post hoc. Reproduced when, under a reading, the interaction has p < 0.05, controls improve from day 1 to day 4 at Sidak p < 0.05, controls exceed both Anxa1-TeTx and Anxa1-Arch on day 4 at Sidak p < 0.05, and Anxa1-TeTx does not differ from controls on day 1. The residual df of 14 fits 17 mice with both days, one fewer than 8 + 4 + 6; which mouse was left out is not given, so all 18 are used and each single exclusion is shown as a variant. The paper does not define a grasp attempt by event type. It describes a reaching attempt as start, "grasping upon contact with the water droplet" and retraction (Results), and Figure 6F shows on- and off-target efforts. The readings are (a) every labelled reach with a distance, and (b) all but empty-spout and unknown events. The preferred forelimb is not defined; readings are (i) per mouse over all its sessions, (ii) per session. Day 1 and day 4 are the first and fourth of four consecutive training sessions after habituation (Methods); the paper does not map these to session labels, so the first and fourth training sessions of each mouse are used. For Arch, the laser is off on day 1 and on during days 2 to 4 (Methods, Results); a session labelled otherwise is reported as departing from the paper. The legend has one control group (Ctrl, n = 8) in a single ANOVA with both silenced groups, so the controls are one pooled group; the paper does not say from which cohort they come, and controls of both cohorts are pooled. Mice recorded after the preprint (22 December 2024) cannot be in the figure; the main run uses mice recorded before it, all mice with both days are a variant.
Pipeline
- Paper
- done (1 of 1)
- Claims
- done
- Claims check
- done (ready)
- Attempt
- todo
- Questions
- none
- Verified
- none