Intranasal administration of exosomes derived from mesenchymal stem cells ameliorates autistic-like behaviors of BTBR mice

MSC-exo were characterized by NanoSight. The average size was 114 ± 2.9 nm, and the average concentration was 3.81 × 10⁸ particles/μL (Fig. 1a, b). Western blot analysis indicates that the MSC-exo express CD9 and CD63, while we could not detect it in the MSC lysate (not shown). In contrast, Calnexin was undetectable in the MSC-exo and found in the MSC lysate, indicating for the purity of the exosomes [18] (Fig. 1c).

Mice were tested for male to male interaction before and after MSC-exo or saline administration. Intra-subject analysis showed that the BTBR MSC-exo mice group had spent significantly more time in social interactions after the treatment whereas the BTBR saline and C57BL saline groups did not show any change (paired t test, t₅ < 0.0001). Comparison analysis between the groups showed that the BTBR mice of both groups had spent significantly less time engaging in social interaction in comparison to C57BL basal behavior (ANOVA1, F_2,18 = 14.4, p < 0.01, Bonferroni). Exosome treatment dramatically increased social interactions in the BTBR MSC-exo group in comparison to BTBR mice treated with saline and is comparable to the C57BL mice group (ANOVA1, F_2,16 = 9.44, p < 0.01, Bonferroni) (Fig. 2a, Additional files 1 and 2).

During the social interaction test, the repetitive behavior was also measured. Intra-subject analysis showed that the BTBR MSC-exo-treated group spent significantly less time in repetitive behaviors while both the BTBR saline and C57BL saline mice groups did not show any changes in comparison to their basal behaviors (paired t test, t₅ < 0.001). A comparison analysis between groups showed that before treatment, there was no difference between BTBR MSC-exo and BTBR saline basal behaviors, and that both groups significantly differed from the basal behavior of the C57BL saline group (ANOVA1, F_2,18 = 13.71, p < 0.001, Bonferroni). After treatment, The BTBR MSC-exo group was not significantly different from the BTBR saline group and the C57BL group (Fig. 2b). In repetitive behavior tests that were outside of the context of social interaction, BTBR MSC-exo spent significantly less time in repetitive grooming and digging in intra-subject analysis compared to their basal behaviors (t₅ < 0.05). In groups, comparisons showed that they were significantly different from the BTBR saline group but not from the C57BL saline group (ANOVA1, F_2,18 = 13.83, p < 0.01, Bonferroni) (Additional file 5: Figure S1). Importantly, C57BL MSC-exo mice did not present any behavioral differences in neither social interactions, antisocial interactions, nor repetitive behaviors during social contact (Additional file 5: Figure S2A). In comparison between MSC-exo to exosomes isolated from neuronal stem cells (NSC-exo), we found that only the MSC-exo-treated group presented a significant increase in social interaction, a reduction in repetitive behaviors, and a significant increase in communication while NSC-exo-treated mice did not present the same behavioral differences compared to saline-treated group (ANOVA1, F_2,14 = 4.28, p < 0.05, Bonferroni). Importantly, MSC-exo and saline groups were also used as a biological replication of the social interaction with different seven mice per group (Additional file 5: Figure S3).

In general, male mice emitted a large number of complex ultrasonic vocalizations (USVs) when interacting with adult females (Fig. 3a, experimental demonstration). As seen qualitatively from the spectrograms in our study, BTBR MSC-exo vocalizations became more complex and longer compared to the BTBR saline group, making them more similar to C57BL (Fig. 3b). During the first 5 min of interaction with females, BTBR saline mice emitted 317 ± 39.4 syllables, BTBR MSC-exo emitted 571 ± 74 (180% more), and C57BL emitted 854.5 ± 65.2 syllables (Fig. 3c, ANOVA1, F_2,18 = 19.2, p < 0.001, Bonferroni). Classification of syllables revealed significant differences between the syllable types used by BTBR saline and BTBR MSC-exo (Fig. 3d, e). For simple syllables, BTBR saline used 89%, BTBR MSC-exo used 57%, and C57BL saline used 60% (ANOVA1, F_2,18 = 49.44, p < 0.001, Bonferroni). For complex syllables, BTBR saline used 2%, BTBR MSC-exo used 23%, and C57BL saline used 10% (ANOVA1, F_2,18 = 30.61, p < 0.001, Bonferroni). For down syllables, BTBR saline used 7%, BTBR MSC-exo used 16%, and C57BL saline used 26% (ANOVA1, F_2,18 = 25.23, p < 0.001, Bonferroni). For up syllables, BTBR saline used 1%, BTBR MSC-exo used 1%, and C57BL saline used 6% (ANOVA1, F_2,18 = 41.22, p < 0.001, Bonferroni).

There was no significant difference between groups in time spent sniffing the female’s genitals or faces, meaning that the effect seen in the ultrasonic vocalizations was not impacted by the pheromones of the females (Fig. 3f). C57BL mice that were treated with MSC-exo showed similar number of syllables compared to C57BL saline group (Additional file 5: Figure S2B). In comparison between MSC-exo and NSC-exo, only the BTBR MSC-exo group presented a significant increase in number of syllables compared to BTBR saline group (ANOVA1, F_2,14 = 9.44, p < 0.01, Additional file 5: Figure S3B). This result also used as biological replication of the vocalizations test.

We have performed pup-retrieval tests for MSC-exo- or saline-treated mothers, naïve virgins, and experienced virgins [19, 20]. The C57BL saline mothers retrieved all pups (15/15), and their total average time for retrieval was 10.4 s ± 1.08 s. In contrast, only one of eight BTBR saline mothers retrieved two pups (in 156 s), without bringing them back to the nest. Altogether, only 2/24 pups were retrieved. After treatment, BTBR MSC-exo mothers retrieved all pups (18/18) at a mean time of 25.24 s ± 5.7 s (ANOVA1 on time of retrieval, F_2,48 = 332.5, p < 0.0001). C57BL saline trained virgins retrieved all pups (15/15) with a total average time of 38.3 s ± 3.3 s. BTBR saline trained virgins mostly did not retrieve the pups besides for one female that had retrieved two pups with a mean time of 152 s. Altogether, 2/21 pups were retrieved by BTBR saline trained virgins. BTBR MSC-exo trained virgins retrieved 7/9 pups at a total average time of 83.3 ± 16.02 s (ANOVA1 on time of retrieval, F_2,42 = 29.47, p < 0.0001). Neither the C57BL naïve virgins (0/21) nor the BTBR naïve virgins had retrieved the pups (0/21) (Fig. 4c, Additional files 3 and 4). We had assumed that the MSC-exo naïve virgins would also not retrieve the pups.

MAESTRO whole brain imaging was used to visualize PKH26-labeled MSC and MSC-exo. After intranasal and intravenous administration, MSC-exo can be visualized fluorescently while MSC cannot be detected (Fig. 5a). Immunostaining indicated that MSC-exo penetrate the brain parenchyma and are found in the cells of the tissue (Fig. 5b, c).

The ability of MSC-exo to cross the blood-brain barrier and to be uptake by cells in the brain is critical for the therapeutic effect. To block the MSC-exo from being uptake by the cells in the brain and to use them as control, MSC-exo were treated with protease-k (ProtK) in order to remove the membrane proteins. Immunostainings of the tissue indicated that the MSC-exo-protk delivered intranasally have not been uptake into the cells at the same efficiency as MSC-exo (Fig. 6a as compared to Fig. 5b). To demonstrate the effectiveness of ProtK in degrading the proteins on the exosomes’ membrane, we used Western blot analysis to CD9 and CD63 since they are known to be overexpressing on the membrane of the exosomes and are commonly used for exosomes detection and characterizations [42]. The Western blot showed deletion of the membrane proteins CD9 and a reduction of CD63 (Fig. 6b). NanoSight analysis showed no significant change in the number and size of MSC-exo-protK compared to MSC-exo (Fig. 6c). BTBR mice that were treated with MSC-exo-protK did not demonstrate behavioral differences in social interaction and repetitive behaviors (Fig. 6d).

Human MSC were purchased from Lonza (cat:PT-2501, Basel, Switzerland). Cells were cultured and expanded as previously described [42]. Before exosome collection, the cells were cultured in exosome-free platelets for 3 days and this medium was then collected.

The exosomes were purified by taking the culture fluid and centrifuging it for 10 min at 300g. The supernatant was recovered and centrifuged for 10 min at 2000g. Once again, the supernatant was recovered and centrifuged for 30 min at 10,000g. The supernatant was filtrated through a 0.22-μm filter and centrifuged for 70 min at 100,000g. The pellet containing the exosomes and proteins was washed in PBS and then centrifuged for 70 min at 100,000g. The pellet containing the purified exosomes was re-suspended in 200 μm of sterilized PBS. Each centrifugation occurred at 4 °C [43].

NanoSight technology (Merkel technologies LTD, Israel) was used to characterize the size and concentration of the exosomes (3.81 × 10⁸ particles/μL). Lysates of MSC-exo were subject for Western blot using SDS polyacrylamide gel. Proteins were transferred to Immobilon®-P membranes (Millipore, Amsterdam, The Netherlands), incubated in 5% milk for 1 h, and probed overnight at 4 °C with CD9 antibody (ABCAM), CD63 (ABCAM), and Calnexin (ABCAM). After three washes in TBS-Tween 20, the membranes were incubated with the secondary antibody for 1 h and re-washed.

Exosomes were labeled with PKH26 (Sigma-Aldrich) [44, 45]. PKH26 (2 μL) in 500 μL diluent was then added to 50 μL exosomes in PBS for 5 min of incubation. Exosomes were suspended in 70 ml PBS and were centrifuged for 90 min at 100,000g at 4 °C. The pellet was suspended in 200 μL of PBS.

Adult BTBR male mice (6–7 weeks) were given 5 μL of labeled exosomes or labeled MSC via intranasal administration (N = 2 per condition). Another adult BTBR male mice (6–7 weeks) were given 100 μL of intravenously with labeled exosomes or labeled MSC. (The tail vain was warmed using water and exosomes were directly injected to the vain, no anesthetics used, N = 2 per condition.) The number of exosomes per mouse was equal between the intranasal/intravenous administrations (19.05 × 10⁸ particles) as well as the number of MSC. For 24-h post administration, mice were perfused and fixated with PBS and 4% paraformaldehyde (PFA). The brains were incubated in PFA for 24 h followed by 30% sucrose for 48 h and stored at 4 °C. Whole brain fluorescence imaging was taken with Maestro CRi, excitation filter 523, and emission filter 560. For immunostaining analysis, the brains were frozen in chilled 2-methylbutane (Sigma-Aldrich), stored at 4 °C, and subsequently sectioned into slices measuring 10 lm. Slides were incubated with blocking solution (5% goat/donkey serum, 1% BSA, 0.5% Triton X-100 in PBS) for 1 h. Thereafter, slides were incubated overnight at 4 °C with primary antibody in blocking solution (mouse anti-NueN, 1:500, Abcam) and secondary antibody in blocking solution (goat anti-mouse Alexa 488, 1:500, Molecular Probes, Invitrogen) for 1–2 h at room temperature. Next, nuclei were counterstained with DAPI (1:500; Sigma-Aldrich). Sections were ultimately mounted with fluorescent mounting solution (Fluoromount-G, Southern Biotech), covered with a cover slide, and sealed with nail polish.

Two hundred microliters of PKH26 labeled exosomes were incubated with 7 μL proteinase-K (Roche Diagnostica GmbH, Germany) and 750 μL BPS in 55 °C for 10 min [45]. Proteinase-K inhibitor (7 μL) (phenylmethanesulfonyl fluoride, Sigma) was added to the solution and was suspended for 2 h in 70 ml BPS for ultracentrifugation. Finally, the pellet (MSC-exo-protk) was re-suspended in 200ul PBS. For behavioral treatment, non-labeled MSC-exo-protk were used. Proteinase-k-treated MSC-exo were characterized with NANO-sight and Western blot.

Mice were placed under a 12-h light/12-h dark condition and grown in individual ventilated cages with access to food and water ad libitum. All experimental protocols were approved by the Tel Aviv University Committee of Animal Use for Research and Education. All methods were performed in accordance with relevant guidelines and regulations. BTBR mice were bred from adult pairs originally purchased from The Jackson Laboratory (Bar Harbor, ME). At 4 weeks of age, the first cohort of littermate male mice was randomly assigned for basal behavioral experiments followed by intranasal administration of saline (BTBR saline, N = 7) or MSC-exo (BTBR MSC-exo, N = 7). For the pup retrieval experiment, the first cohort of littermate female mice was randomly assigned at 5 weeks of age to saline or to MSC-exo. At the end of the treatment, random groups of females were placed at breeding colonies with BTBR male mice (1 male 2 females, BTBR mothers). At days 1–2 post whelping, random groups of virgin females were placed with the mothers for 4 days to gain experience (BTBR experienced virgins). The rest of the mice were left at their home cage (BTBR naïve virgins) (Additional file 5: Table S1 for number of females in each group).

C57BL mice were bred from adult pairs originally purchased from The Jackson Laboratory (Bar Harbor, ME). At 4 weeks of age, the first cohort of littermate male mice was randomly assigned for basal behavioral experiments followed by intranasal administration of saline (C57BL saline, N = 7) or MSC-exo (C57BL MSC-exo). For the pup retrieval experiment, the first cohort of littermate female mice was randomly assigned to saline at 5 weeks of age. At the end of the treatment, random groups of females were placed at breeding colonies with C57BL male mice (1 male 2 females, C57BL mothers). On days 1–2 post-whelping, random groups of virgin females were placed with the mothers for 4 days to gain experience (C57BL experienced virgins). The rest of the mice remained at their home cage (C57BL naïve virgins) (Additional file 5: Table S1 for number of females in each group).

For comparison between NSC-exo and MSC-exo, another group of mice was randomly assigned. They were raised as previously described and treated with intranasal administration of MSC-exo (N = 7), NSC-exo (N = 5), or saline (N = 7). For comparison between saline and MSC-exo+protK in behavioral experiments, another group of mice was randomly assigned. They were raised as previously described and treated with intranasal administration of MSC-exo+protK (N = 8) or saline (N = 6). Complete summary of all the mice in all the experiments is found in Table S2.

Mice were treated with intranasal MSC-exo or saline for 12 days, 5 μL a day, and every other day (total of 30 μL per mouse). Administration was done using a gentle pipette of 2.5 μL per nostril with no anesthetics.