This notice corrects errors in the published PLOS ONE article reporting altered stress and fear responses in the VPA rat model of autism. The correction addresses two issues: an incorrect Funding statement and several supporting information files (S1 Table and S1–S6 Figs) that were uploaded incorrectly. The publisher provides the corrected Funding statement and the properly uploaded supporting files with this notice.
The corrected Funding statement indicates that the work was supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, Brazil; grant #88882.376056/2019–1 to DH), was partially supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq; grant #383539/2025–0 to DH), and by CNPq grants #010186/2018–2 and #421955/2023–6 to RNRP. These are the funding attributions reported in the correction. No other funding changes or additional funders are listed in the correction notice.
The following supporting files were re-uploaded with corrected versions: S1 Table and S1 Fig through S6 Fig. The correction notice provides links and DOIs for each corrected supplementary file. The legends included with each re-uploaded file summarize experimental design details, sample sizes, statistical tests used, and the primary observations relevant to the supplementary material. The summaries below reproduce the information supplied in those file legends.
S1 Table provides a summary of animals and litters used across experiments. Offspring from multiple litters were distributed across experimental conditions to reduce confounding by litter identity. Within each experimental paradigm, independent litters were assigned to control (CTL) and VPA groups so that no single litter contributed animals to both treatment conditions. The table file (TIFF) is provided as the corrected version in the supporting information.
S1 Fig documents how vocalizations recorded during the emotional contagion paradigm were assigned to demonstrator (DEM) or observer (OBS) animals based on ultrasonic vocalization (USV) power intensity. Thresholds were empirically established by comparing USV power recorded with a microphone placed in the same chamber as the vocalizing animal or in the opposite chamber. The legend reports mean USV power comparisons during maternal separation (P08 and P14) and during fear conditioning (P35–P38), with Mann–Whitney tests indicating significant differences in those calibration data.
Vocalizations with power > −76 dB were classified as originating from the microphone chamber (DEM), while those < −85 dB were classified as originating from the opposite chamber (OBS). Calls with power between −76 and −85 dB were excluded to minimize ambiguity; calls within this interval constituted 9.8% of all vocalizations recorded in the emotional contagion experiment (2086 of 21,388 calls), representing 8.4% of CTL calls and 11.2% of VPA calls. The corrected S1 Fig file is provided in TIFF format.
S2 Fig presents the design schematic for the touch sensitivity experiment and reports the number of fecal pellets released on each experimental day by female and male animals. For CTL animals, there were no significant sex differences across days (p > 0.05). Similarly, for VPA-treated animals no significant sex differences were detected (p > 0.05). Sample sizes reported in the legend are CTL: n = 18 (9 females, 9 males) and VPA: n = 21 (10 females, 11 males). The file is provided as corrected TIFF supporting information.
S3 Fig includes measures of tactile and nociceptive thresholds. The legend reports that overall tactile and nociceptive thresholds showed no significant group differences (p > 0.050), while at least one comparison yielded a highly significant result (* p < 0.001) indicated in the file. The legend also notes a sex-related difference in tactile thresholds (& p = 0.007). Reported sample sizes were nCTL = 30 (15 females, 15 males) and nVPA = 28 (15 females, 13 males). Statistical tests cited include paired Student’s t-test and Mann–Whitney test. The corrected S3 Fig TIFF is included in the supporting information.
S4 Fig provides a schematic of the fear memory retrieval experiment: animals were placed in the testing apparatus for 5 minutes before return to the home cage. The legend reports the percentage of animals displaying freezing during fear conditioning memory retrieval and cites no between-group difference in freezing prevalence during retrieval (p > 0.05). However, the percentage of time spent freezing increased from the fear conditioning baseline (B0) to the memory retrieval session (p < 0.001) within groups. Tests cited include Fisher’s exact test and Student’s t-tests. Sample sizes reported were CTL: n = 27 (13 females, 14 males) and VPA: n = 28 (14 females, 14 males). The corrected S4 Fig is available as TIFF.
S5 Fig shows sex-specific percent-time-freezing data across experimental blocks for CTL and VPA animals in the emotional contagion paradigm. The figure legend indicates repeated-measures two-way ANOVA was used and lists group sample sizes for NoC+ and NoV+ conditions (each including both females and males). A 20% reference line is shown in the figure; full numerical details and statistical outputs are contained in the corrected TIFF file.
S6 Fig reports the behavioral and vocal responses of demonstrator (DEM) animals during the emotional contagion paradigm. The design schematic describes a 4-minute exploration period followed by five electrical shocks to DEM animals, with an additional 4 minutes of post-shock observation. Key points in the legend include:
Foot shock exposure significantly increased the prevalence of freezing behavior in both shocked demonstrator groups (dC+ and dV+) compared with their non-shocked controls (dC- and dV-), independent of prenatal treatment. No significant differences in freezing prevalence were found between dV+ and dC+ or between dV- and dC-.
Among responsive individuals, dC+ and dV+ groups showed a sharp increase in freezing time from block B1 and a sustained plateau from B2 to B5, whereas non-shocked controls (dV- and dC-) maintained baseline-like freezing.
Shocked demonstrators had a higher prevalence of vocalization than non-shocked controls. The legend reports (1) no significant differences in vocal engagement between dV+ vs. dC+ or between dV- vs. dC-; (2) significant within-treatment differences when comparing shocked vs. non-shocked groups at specific blocks; and (3) increased vocalization prevalence relative to baseline in shocked groups from block B2 onward.
The total number of USVs increased in shocked groups relative to their non-shocked controls from block B4 onward. Within-group increases relative to baseline are reported for dV+ (blocks B2–B5), dC+ (blocks B2 and B4–B5), and dV- (block B1). No within-group changes were observed for dC- controls.
Acoustic properties of calls—call duration and principal frequency (F0)—did not differ between dV+ and dC+ animals. Because dV- and dC- emitted few vocalizations (median 3 calls per block), their acoustic properties were excluded from these analyses.
The legend lists the statistical tests used (Fisher’s exact test, repeated-measures two-way ANOVA, Friedman test, Mann–Whitney test) and provides sample sizes for the various groups. The corrected S6 Fig TIFF is included with the correction.
The correction cites the original article: Hashiguchi D, Dias AL, Romcy-Pereira RN. Altered stress and fear responses in the VPA rat model of autism: Behavioral dissociation across tactile, nociceptive, and social contexts. PLoS One. 2026;21(7):e0353839. The correction itself is published in PLoS One (2026) with DOI 10.1371/journal.pone.0357445 and the correction notice includes the corrected funding statement and links to the corrected supporting information files. All descriptions above are derived from the text of the published correction and the legends of the re-uploaded supplementary files; no additional experimental data have been added beyond what the correction notice reports.