The clinical work package of PRISM, a collaborative research project funded by the EU public-private Innovative Medicine Initiative, identified a transdiagnostic and pathophysiological relationship between default mode network (DMN) integrity and social dysfunction. For testing the causality between the quantitative variation in DMN integrity and social dysfunction, we developed a preclinical backtranslation approach for disturbing DMN integrity and applied cutting-edge technologies such as functional ultrasound imaging and automated analysis of social behaviour in groups of mice5. Here we show that focal demyelination in the DMN leads to a decrease in global DMN connectivity and subsequently to impaired social behaviour and increased anxiety in mice. Reversibility of these observations suggests that integrity of the forceps minor, a fibre tract connecting important regions of the DMN in the prefrontal brain, plays a significant role in social functioning. In summary, this back-translation study demonstrates a causal pathophysiological relationship between DMN integrity and social dysfunction. The findings can support precision psychiatry approaches for the development of effective treatment of social dysfunction.
Article
Forceps minor control of social behaviour
https://doi.org/10.21203/rs.3.rs-4162770/v1
This work is licensed under a CC BY 4.0 License
You are reading this latest preprint version
Social relationships, interactions, and overall social behaviour are essential for evolutionary survival, particularly in primates [1,2]. It has been suggested that the neocortex of primates expands during evolution due to social pressure experienced within groups [3]. However, the increased degree of specialization and complexity of the ‘social brain’, which has evolved to address ecological challenges at a group-level [4], may result in increased vulnerability to pathogenic interventions. Interestingly, early indicators of various neuropsychiatric diseases, such as social withdrawal, loneliness, and interpersonal dysfunction, can be observed before the main symptoms appear [5]. These specific indicators could function as transdiagnostic markers, facilitating early detection of psychiatric diseases and enabling the provision of personalized treatment.
The EU-funded Psychiatric Ratings using Intermediate Stratified Markers project (PRISM) [6] is focusing on biological parameters related to Schizophrenia (SZ), Alzheimer’s Disease (AD), Major Depressive Disorder (MDD) and social functioning. Although the pathophysiology behind these diseases is heterogenous, their symptomatology in relation to social dysfunction is very similar [7,8,9,10]. Based on these findings, one hypothesis of PRISM describes that dysfunction of the Default Mode Network (DMN) [11], could potentially link social impairment with these neuropathophysiologies. The DMN primarily operates during wakeful rest [12] and is also involved in social behaviour, as evidenced by its overlap with the ‘social brain’ [2,13]. The core regions within the human brain’s DMN include the ventral medial prefrontal cortex (vMPFC), posterior cingulate/retrosplenial cortex (PCC/Rsp), inferior parietal lobule (IPL), lateral temporal cortex (LTC), dorsal medial prefrontal cortex (dMPFC) and the hippocampal formation (HF) [12].
In addition to various clinical observations describing an implication of DMN in social functioning [14],[15], a PRISM-study [7] demonstrated a correlation between altered DMN connectivity and social dysfunction in SZ, AD, and healthy control (HC) participants. Notably, the rostromedial prefrontal cortex exhibited an association between social malfunction and abnormal white-matter integrity, regardless of previously diagnosed pathology or presumed healthiness. Another study [16] linked perceived social support with reduced mean diffusivity during diffusion magnetic resonance imaging in the forceps minor (fmi), an interhemispheric fibre tract that forms the genu of the corpus callosum (CC) (Fig.1a) and links both orbitofrontal cortices (OFCs). As an evolutionary conserved brain structure [17], the fmi is also part of the DMN [18]. One study previously demonstrated the involvement of a dysfunctional fmi in neuropsychiatric diseases, as adults with high functioning autism displayed a reduced volume in the fmi and body of the CC [19].
To substantiate the clinical findings of a correlation between social dysfunction and reduced DMN white matter integrity in a preclinical setting, we developed a mouse model of impaired fmi integrity (Fig. 1b). To this end, we focused on manipulating the fmi-forming white matter tracts and specifically targeted the myelinating oligodendrocytes.
Lysophosphatidylcholine (LPC, Lysolecithin) is an endogenous lysophospholipid [20] and its levels have been shown to correlate with ageing [21], obesity [22] atherosclerosis, and inflammatory diseases [23]. Furthermore, LPC is widely used for studying focal de- and remyelination in the CNS [24,25,26,27]. After local injection, LPC integrates into cellular membranes and non-specifically disrupts myelin lipids [28]. Locally injected LPC is cleared after 24 hours and remyelination of the lesion site begins around day 10 post-injection.
Functional ultrasound [29] (fUS) and radio-frequency identification (RFID)-assisted SocialScan [30] were applied to investigate the functional and behavioural outcome of transient fmi demyelination. The social arena enabled us to monitor complex behaviours of individual mice within a social group during the demyelination and remyelination phases. The behavioural data were complemented by fUS imaging of DMN connectivity. The preclinical dataset supports the main clinical findings of PRISM by demonstrating that transient fmi demyelination impairs DMN integrity, leading to a reversible reduction of social interactions. The findings offer new options for developing precision psychiatry treatments of social dysfunction.
To investigate whether LPC-induced demyelination in the fmi impacts connectivity between prefrontal brain regions, we recorded brain activity by high-resolution fUS on day seven post-LPC injection (Fig. 1c). In a three-step scan, a volume was recorded, covering the bilateral orbital area, secondary motor cortex, prelimbic area and agranular insular area (Fig. 1d-e). These brain regions, except for the agranular insular area, are part of the DMN. A pair-wise analysis between all regions, based on Pearson’s correlation, is depicted in matrix Fig. 1f. Compared to the control group (n = 12), LPC individuals (n = 14) exhibited overall reduced correlation patterns. Specifically, correlation between the left prelimbic and the right orbital area was significantly decreased in LPC treated animals (*p < 0.05), as well as the interaction between right secondary motor and left prelimbic area (*p < 0.05). Interestingly, the left secondary motor area appears to be predominantly affected by the treatment. Correlations between this region and all other scanned DMN connected areas are significantly compromised. Furthermore, correlation values between this region and the orbital area even withstand FDR-correction, while the insular region, a non-DMN region, remains unaffected. By transforming the correlation coefficients into z-scores using Fisher’s Z-transformation, a group-level connectivity matrix also revealed the effect of LPC treatment on functional connectivity (Supplementary Fig. S1 online). As already demonstrated by Pearson’s correlation, the most significant effect was observed between left secondary motor cortex and all DMN connected regions. Overall, animals injected with LPC exhibit reduced functional connectivity between DMN sub-regions in the coronal plane through the fmi.
A more detailed correlation analysis, focusing on the global DMN connectivity, including the orbital area, prelimbic area and secondary motor cortex in both hemispheres, is significantly reduced after demyelination of the fmi (**p < 0.01) (Fig. 1g). Connectivity between left and right orbital area through the fmi (Fig. 1h), is not significantly decreased in LPC individuals (p = 0.3136).
Comparing the strength of filtered CBV signals in the left secondary motor cortex and the right orbital area revealed a strong linear relationship between these values in control animals (r = 0.64, p < 0.0001) (Fig. 1i and Supplementary Fig. S1 online). After LPC-induced DMN impairment, blood flows in the left secondary motor cortex and the right orbital area change and their interdependence is lost (r = 0.1, n.s.) (Fig. 1j and Supplementary Fig. S1 online).
To investigate the effect of focal demyelination in fmi on behaviour, a second, independent cohort of mice was analysed in the social arena between day five and ten post-injection (Fig. 2a,b).
We focused our analyses on the peak activity period of the dark phase and calculated the mean values of sniffing and approaching for each individual animal in two three-hour time blocks. Social sniff did not show a significant difference between LPC and control group in the first three hours of the dark phase. In the subsequent period until midnight, the animals injected with LPC significantly reduced their sniffing time (***p < 0.001) (Fig. 2c). Focusing on social approach, a coherent pattern compared to social sniff can be observed. In the first three hours, animals in both cohorts are approaching other individuals at a similar frequency. In the second time block of three hours again, the animals with fmi demyelination significantly reduced their approaching times (**p < 0.01) (Fig.2d).
To substantiate this finding, we focused on the behaviour of individual animals. Z-scores, calculated based on the control group, were analysed in 30-minute intervals between 6 p.m. and midnight. The matrices (Supplementary Fig. S2 online) display the mean z-scores for each animal during the demyelination phase (day 5-10). Both matrices, representing social sniff and approach, exhibit similar patterns. In comparison to the control mean value, LPC animals appeared to reduce their social interaction time, with the reduction becoming more prominent as time progressed.
Next, we calculated the location of an animal within a predefined area. Although animals often use nests for sleeping and withdrawal, they spend a significant amount of time in the food zone. However, social interactions were detected mainly in the centre and the periphery, areas that were frequently crossed during their activity phase. Time spent in the centre of the social arena revealed a significantly reduced stay of LPC animals compared to controls (*p < 0.05) (Fig. 2e). In contrast, the time spent in the periphery was not significantly different (p = 0.9189) (Fig. 2f). Total nesting time, time in locomotion, food zone and water area were not significantly different (Supplementary Fig. S3 online). However, nesting time with another cage mate, a behaviour referred to as ‘social stay in nest’ is reduced in LPC animals (Fig 2g) but did not reach significant difference (p = 0.0579) in contrast to the outcome during remyelination phase (Fig 3d). For comparison, all analysed parameters are displayed as normalised values in a spider web plot (Fig. 2h). This representation illustrates that fmi demyelination leads to a decrease in social interactions between the animals without impairing their behaviour in general.
Next, we investigated whether the reduction of functional connectivity and impaired social behaviour is reversible after remyelination of the fmi. The same cohorts that had already been studied in the social arena during the demyelination phase, were monitored for five additional days (Fig 3a). From day twelve to seventeen post-LPC injection, the decreased centre time was not observed any longer (Fig. 3b,c) and all other measures of general behaviour did not differ between the groups (Supplementary Fig. S3 online).
In the next step, social parameters such as sniffing and approaching were analysed during the remyelination phase. Unlike during the demyelination phase, when the time for sniffing and approaching in the LPC-group was reduced between 9 p.m. and midnight, the same animals spend similar amounts of time with their cage mates as controls do (Fig. 3e,f). Furthermore, the z-scores based detailed analysis of each individual mouse during this six-hour period revealed a constant pattern of behaviour during the 6 hours observation period and no decrease in the second half anymore (Supplementary Fig. S2 online). Intriguingly, the decrease in the parameter social stay in nest (Fig. 3d) reached statistical significance (*p < 0.05), while the overall time spend in nests was comparable to controls (p = 0.8194) (Supplementary Fig. S3 online). Thus, it appears that LPC-treated animals prefer staying alone in nests. The spider web plot (Fig. 3g) displays again the analysed parameters as normalised values. This representation illustrates that all behaviours except the complex social behaviour “social stay in nests” returned to normal.
Body weights were measured throughout the experiment, but no significant differences have been observed (Supplementary Fig. S3 online).
To confirm that the brain connectivity impairment after fmi demyelination is transient, a coronal fUS scan, including the orbital area, was performed for 60 minutes on day 20 post injection (Fig. 3a). In fact, functional connectivity of the DMN subregions in the scanned plane did not differ significantly between LPC-treated and saline treated animals. Except for the left insular region and right secondary motor cortex, all correlation coefficients in the prefrontal volume scan shown in the matrix were not significantly different between control and LPC group (Fig.4a and Supplementary Fig. S1 online). A more detailed view into correlation analysis between DMN subregions (Fig. 4b) (p = 0.5350) and orbital areas (Fig. 4c) (p = 0.3829) revealed no difference anymore. Moreover, the scatter plot of filtered CBV-signals in animals which underwent de-and subsequent remyelination shows that the interdependence of regional blood flows in the left secondary motor cortex and the right orbital area is reconstituted (Fig. 4d).
To confirm the effect of LPC injections on fmi myelination, we examined brain tissue slices by histological Luxol-Fast-Blue (LFB) staining (Fig. 4e) and 3D-light sheet fluorescence microscopy (Supplementary Fig. S4 online). We evaluated the percentage of myelin loss in coronal slices compared to control animals injected with 0.9 % NaCl. On day seven post-surgery, in the middle of demyelination phase, LFB staining was performed on brains previously scanned by fUS, revealing a significant decrease in volume of myelinated area (***p < 0.0001) (Fig. 4f). In a preliminary experiment, we cleared brains two days post unilateral LPC (Supplementary Fig. S4 online) or Saline (Supplementary Fig. S4 online) injection and stained them with fluorescent myelin antibody directly conjugated to a dye. Compared to controls, a large void volume can be observed in the expected location of the fibre tract. 3D-visualization of the prefrontal brain confirmed correct injection site in all brains. On day 20 post injection, no significant differences in myelination of vehicle and LPC-treated animals were found (p = 0.4367) (Fig. 4g), which agrees with previously published data (Plemel et al., 2018). 3D-evaluation using the Light-sheet-microscope confirmed the recovery of fibre tracts (Supplementary Fig. S4 online).
This study was designed to back-translate clinical findings into a pre-clinical animal model. Human studies, conducted as part of the PRISM project and aiming to identify biological underpinnings for social dysfunction, revealed a disruption in white matter integrity in individuals with low social functioning. This was specifically observed in fmi, a fibre tract in the DMN [7,16,31]. Our approach involved the local injection to the fmi of the demyelinating phospholipid lysophosphatidylcholine (LPC), to reproduce the impairment of white matter integrity in mice observed in the clinical study. The lipotoxin LPC targets lipid-rich myelin membranes, resulting in focal reduction of myelin thickness [32]. In contrast, the number of axons and axon integrity appears to be unaffected. Spontaneous remyelination, which begins with oligodendrocyte precursor cells (OPC) recruitment to the focal lesion site 72 h post-injection [28], provides a foundation for studying the reestablishment of brain connectivity and behaviour. Given that various neuropsychiatric diseases exhibit microstructural white matter abnormalities [7,33], we considered LPC-induced manipulation of the fmi as a relevant model for back-translating the clinical findings.
We used functional ultrasound (fUS), to investigate the consequences of LPC-induced fmi demyelination on activity and connectivity changes in the brain based on neuro-vascular coupling [29]. Compared to functional magnetic resonance imaging (fMRI), fUS detects microvascular fluctuations with remarkable sensitivity thereby providing high spatiotemporal resolution [29]. In our study, we decided to scan a coronal volume covering the orbital area, secondary motor cortex, prelimbic area and insular region in both hemispheres during the demyelination phase on day 7 and during the remyelination phase on day 20. According to [28,32], myelin wrapping is restored at day 20 post LPC injection. We selected a coronal scan, assuming that bilateral manipulation of the fmi would predominantly affect frontal circuitries. Due to the limitation of a linear fUS probe, we could only scan frontal areas, leaving global effects on whole brain dynamics unexplored. Nevertheless, we were able to confirm our hypothesis that focal reduction of fmi integrity leads to a reduction in functional connectivity between frontal brain regions. Delving deeper, even though no significant decrease of connectivity could be detected between the left and right orbital area, where fmi forming fibre tracts end, we observed a trend of impairment. Nevertheless, analysis of all prefrontal areas involved in the DMN network, including orbital, prelimbic and secondary motor cortex regions, revealed a reduction in interregional correlations of hemodynamic fluctuations between these regions. Interestingly, in our case, symmetric, bilateral LPC-injection appeared to affect both hemispheres differently. The most significant decrease in DMN connectivity was seen in the left secondary motor cortex, while in the contralateral side, the effect was less pronounced. An explanation for the reproducibly different bilateral findings could be asymmetric functional connectivity as described in the human brain [34]. During resting state, the mean connectivity shows a symmetry of around 95 %, thereby revealing higher correlations in DMN regions of the left hemisphere [35]. Therefore, any manipulation could affect one side more severe, compared to the contralateral one. In line with a previous published study in mice [36], we observed a diminishing effect on brain activity during the resting state in the demyelination phase and recovery of the system after remyelination. Unexpectedly, we observed a decrease in connectivity between the right insular and left prelimbic region at day 20 post LPC injection. However, this effect is unlikely to be a result of the LPC injection, as there is no significant decrease in functional connectivity in the insular cortex during the demyelination phase. The observed restoration of connectivity aligns with literature regarding the myelopathy of LPC-induced demyelination, which suggests complete remyelination at day 14 post injection [28]. The restoration of the disrupted DMN connectome to control levels made it possible to study the recovery of social behaviour in the same animals which have been analysed during the demyelination stage. Furthermore, the lack of a significant difference between NaCl injected animals and LPC injected animals’ DMN connectivity after remyelination verifies that the effects observed during demyelination phase in the fUS cohort are due to the LPC-induced disruption.
Next, we sought to understand the impact of DMN hypoconnectivity on social functioning in mice. Unlike previous studies that used the 3-chamber or sociability test to examine social behaviour in mice, our setup provides long-term and individual monitoring in a semi-natural environment [30]. The data collected post-LPC induction showed a significant decrease in the social parameters of sniffing and approaching. Specifically, between 3 and 6 hours post switching to the dark phase of the light cycle, we observed a decrease in social interest, potentially indicating a loss of motivation to initiate social contact with their batch mates. This dysfunction was resolved once remyelination was underway or completed. According to Peleh and colleagues, mice, being nocturnal animals, are most interactive during dark phases, particularly in the first half of the night (6 p.m.-midnight) [37]. Focusing on this time interval, only the second half is affected, whereas in the first half social interaction of the animals is not reduced.
In mice, the potential role of the medial prefrontal cortex (mPFC) in social interaction was previously explored in a study by Lee and colleagues, where a subset of mPFC neurons showed increased discharge rates correlating with approaching behaviour towards an unfamiliar mouse [38]. Consistent with these findings, Jennings and colleagues identified distinct neuron populations in the OFC, also located in the frontal brain, that react to social stimuli [39]. These findings indicate that both DMN-regions, the mPFC and OFC, are somehow linked to social behaviour. The impact of manipulating the fmi fibre tract with LPC on sociability, has previously been studied [40]. However, in this study, no difference between LPC and control groups has been found using the social interaction test to measure the social interest of mice. In contrast to our study, social interaction only has been observed for 30 minutes on the seventh day post-surgery, which may have been too short to detect differences between the groups. In agreement with these findings, our data did not indicate significant changes in social behaviours during the first 3h time interval of the dark phase. One explanation for this delayed effect could be attributed to the high activity levels and exploratory behaviour of the mice during the early hours of the night. This particularly active phase could potentially interfere with the consequences of fmi demyelination on social behaviour. Moreover, Xiao and colleagues directly injected into the genu of the CC, accepting the possibility of affecting non-fmi tracts. Therefore, discrepancies between both observations could also be due to the extension of manipulated areas.
Our data are further substantiated by extending the assessment of behaviours on group level to the behavioural analysis of individual animals. Z-score heat maps of behavioural differences of individual animals during the first 6 hours of the night phase show cooler colours in the second half of the observation period. Despite of inter-individual differences, the observation of decreasing social interaction is shared by all individual animals with fmi demyelination. Most importantly, after remyelination, the reduced social approach and sniffing was reversed, reinforcing our hypothesis of fmi involvement in social behaviour. Nonetheless, social behaviour is complex, and the full behavioural spectrum has not been re-established until d17 post LPC lesion. In contrast to direct social interaction parameters sniffing and approach, the time spent in a nest together with a cage mate is reduced even after remyelination. The longer lasting effect on more complex social behaviours may be explained by thinner and probably less efficient myelin sheaths after remyelination [41].
Impaired microstructure of DMN backbone fibre tracts also has been associated with anxiety, another commonly diagnosed neuropsychiatric symptom [33]. Young, healthy individuals displaying high trait anxiety showed specific alterations in the inferior fronto-occipital fasciculus, corpus callosum, and most important in this context, the fmi-tract. We therefore hypothesized that fmi-manipulation should also influence anxiety-like behaviour. Due to the large size of the social arena (80x60cm), the surface area can be considered as an unconventional open field. During the long observation period the mice build nests close to the walls and in corners, but the open centre is considered anxiogenic [42]. Consistent with human findings, animals with fmi demyelination spent significantly less time in the centre of the social arena compared to controls. Importantly, after remyelination, the anxiety-like behaviour of avoiding the centre is restored, suggesting that fmi manipulation was the underlying cause. In humans, constant exposure to anxious situations can increase the risk for developing social anxiety [43]. Moreover, both symptoms, general anxiety and social anxiety can coexist. Therefore, it is plausible to speculate that these symptoms in animals, indicated by reduced time in the centre and social behaviour parameters, can influence each other, leading to a worsening circle.
In addition to social behaviour, non-social behaviour can also be analysed in the social arena. In terms of locomotion, no significant differences were observed, although a slight reduction in LPC animals compared to controls was detected. A lack of motivation can negatively influence movement behaviour and therefore the opportunity to meet and interact with cage mates. However, as summarized in the spider web plots, both groups display similar levels of ambulation and activity in general. Also, eating and drinking behaviour, measured as time spend in the food zone or in water area and body weight, was comparable in both groups. These results consolidate our hypothesis that disturbing the structural integrity of the fmi by transient demyelination predominantly affects outcomes of social behaviours. This study provides new insights into DMN functioning, the influence of fmi on sociability and contributes to develop innovative treatment strategies for precision psychiatry.
Acknowledgement
We thank M.J.H. Kas for his comments and reading the manuscript.
Author contributions
B.H. and H.M.M developed the scientific concept of the project. B.H. and F.S designed the experiments and wrote the manuscript. F.S., E.H., T.I., and E.K. contributed to the experiments and data analysis.
Competing interests
F.S., E.H., T.I., E.K., H.M.M.,and B.H. received salaries from Boehringer Ingelheim Pharma GmbH & Co. KG
The project leading to this manuscript has received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 115916. This Joint Undertaking receives support from the European Union’s Horizon 2020 research and innovation programme and EFPIA. This publication reflects only the author’s views and neither the IMI 2JU nor EFPIA nor the European Commission are liable for any use that may be made of the information contained therein.
All data supporting the findings of this study are available within the paper and its Supplementary Information.
- Holt-Lunstad, J., Smith, T. B., & Layton, J. B. Social Relationships and Mortality Risk: A Meta-analytic Review. PLoS Med 7(7): e1000316 (2010). https://doi.org/10.1371/journal.pmed.1000316
- Porcelli, S. et al. Social brain, social dysfunction and social withdrawal. Neuroscience & Biobehavioral Reviews, 97, 10–33 (2019). https://doi.org/10.1016/j.neubiorev.2018.09.012
- Dunbar, R. I. M. & Shultz, S. Evolution in the Social Brain. Science, 317, 1344-1347 (2007). https://doi.org/10.1126/science.1145463
- Dunbar, R. I. M. The social brain hypothesis and its implications for social evolution. Annals of Human Biology, 36, 562-572 (2009). https://doi.org/10.1080/03014460902960289
- Kas, M. J. et al. A quantitative approach to neuropsychiatry: The why and the how. Neuroscience & Biobehavioral Reviews, 97, 3–9 (2019). https://doi.org/10.1016/j.neubiorev.2017.12.008
- PRISM Precision medicine - PRISM 2 (prism2-project.eu)
- Saris, I. M. J. et al. Social dysfunction is transdiagnostically associated with default mode network dysconnectivity in schizophrenia and Alzheimer’s disease. The World Journal of Biological Psychiatry, 23, 264–277 (2022). https://doi.org/10.1080/15622975.2021.1966714
- Honda, Y., Meguro, K., Meguro, M. & Akanuma, K. Social Withdrawal of Persons With Vascular Dementia Associated With Disturbance of Basic Daily Activities, Apathy, and Impaired Social Judgment. Care Management Journals, 14 (2013). https://doi.org/10.1891/1521-0987.14.2.108
- Teo, A. R. et al. Social withdrawal in major depressive disorder: a case-control study of hikikomori in japan. Journal of Affective Disorders, 274, 1142–1146 (2020). https://doi.org/10.1016/j.jad.2020.06.011
- Winograd-Gurvich, C., Fitzgerald, P. B., Georgiou-Karistianis, N., Bradshaw, J. L. & White, O. B. Negative symptoms: A review of schizophrenia, melancholic depression and Parkinson’s disease. Brain Research Bulletin, 70, 312–321 (2006). https://doi.org/10.1016/j.brainresbull.2006.06.007
- Andrews-Hanna,J.R., Smallwood, J. & Spreng,R.N.. The default network and self-generated thought: component processes, dynamic control, and clinical relevance. Ann N Y Acad Sci. 1316, 29–52 (2014). https://doi.org/10.1111/nyas.12360
- Buckner, R. L., Andrews‐Hanna, J. R. & Schacter, D. L. The Brain’s Default Network. Annals of the New York Academy of Sciences, 1124, 1–38 (2008). https://doi.org/10.1196/annals.1440.011
- Mars,R.B. et al. On the relationship between the “default mode network” and the “social brain”. Front Hum Neurosci. 2012; 6: 189 (2012). https://doi.org/10.3389/fnhum.2012.00189
- Li, W., Mai, X. & Liu, C. The default mode network and social understanding of others: what do brain connectivity studies tell us. Frontiers in Human Neuroscience, 8, 74 (2014). https://doi.org/10.3389/fnhum.2014.00074
- Spreng, R. N. et al. The default network of the human brain is associated with perceived social isolation. Nature Communications, 11, 6393 (2020). https://doi.org/10.1038/s41467-020-20039-w
- Costanzo, A. et al. Social Health Is Associated With Tract-Specific Brain White Matter Microstructure in Community-Dwelling Older Adults. Biological Psychiatry Global Open Science, 3, 1003–1011 (2023). https://doi.org/10.1016/j.bpsgos.2022.08.009
- Dodero, L. et al. Neuroimaging Evidence of Major Morpho-Anatomical and Functional Abnormalities in the BTBR T+TF/J Mouse Model of Autism. PLoS ONE, 8, e76655 (2013). https://doi.org/10.1371/journal.pone.0076655
- Jandric, D., Lipp, I., Paling, D., Rog, D., Castellazzi, G., Haroon, H., Parkes, L., Parker, G. J. M., Tomassini, V. & Muhlert, N. Mechanisms of Network Changes in Cognitive Impairment in Multiple Sclerosis. Neurology, 97, e1886–e1897 (2021). https://doi.org/10.1212/wnl.0000000000012834
- Thomas, C., Humphreys, K., Jung, K.-J., Minshew, N. & Behrmann, M. The anatomy of the callosal and visual-association pathways in high-functioning autism: A DTI tractography study. Cortex, 47, 863–873 (2011). https://doi.org/10.1016/j.cortex.2010.07.006
- Zhao, Z. & Xu, Y. An extremely simple method for extraction of lysophospholipids and phospholipids from blood samples. Journal of Lipid Research, 51, 652–659 (2010). https://doi.org/10.1194/jlr.d001503
- Semba, R. D. et al. Low plasma lysophosphatidylcholines are associated with impaired mitochondrial oxidative capacity in adults in the Baltimore Longitudinal Study of Aging. Aging Cell, 18, e12915 (2019). https://doi.org/10.1111/acel.12915
- Barber, M. N. et al. Plasma Lysophosphatidylcholine Levels Are Reduced in Obesity and Type 2 Diabetes. PLoS ONE, 7, e41456 (2012). https://doi.org/10.1371/journal.pone.0041456
- Zhou, X. et al. Identification of Lysophosphatidylcholines and Sphingolipids as Potential Biomarkers for Acute Aortic Dissection via Serum Metabolomics. European Journal of Vascular and Endovascular Surgery, 57, 434–441 (2019). https://doi.org/10.1016/j.ejvs.2018.07.004
- Baydyuk, M. et al. Tracking the evolution of CNS remyelinating lesion in mice with neutral red dye. Proceedings of the National Academy of Sciences, 116, 14290–14299 (2019). https://doi.org/10.1073/pnas.1819343116
- Miron, V. E. et al. M2 microglia and macrophages drive oligodendrocyte differentiation during CNS remyelination. Nature Neuroscience, 16, 1211–1218 (2013). https://doi.org/10.1038/nn.3469
- Morris, A. D. & Kucenas, S. A Novel Lysolecithin Model for Visualizing Damage in vivo in the Larval Zebrafish Spinal Cord. Frontiers in Cell and Developmental Biology, 9, 654583 (2021). https://doi.org/10.3389/fcell.2021.654583
- Yamazaki, R. et al. Macroscopic detection of demyelinated lesions in mouse PNS with neutral red dye. Scientific Reports, 11, 16906 (2021). https://doi.org/10.1038/s41598-021-96395-4
- Plemel, J. R.et al. Mechanisms of lysophosphatidylcholine‐induced demyelination: A primary lipid disrupting myelinopathy. Glia, 66, 327–347(2018). https://doi.org/10.1002/glia.23245
- Macé, E. et al. Functional ultrasound imaging of the brain. Nature Methods, 8, 662–664 (2011). https://doi.org/10.1038/nmeth.1641
- Peleh, T., Bai, X., Kas, M. J. H. & Hengerer, B. RFID-supported video tracking for automated analysis of social behaviour in groups of mice. Journal of Neuroscience Methods, 325, 108323 (2019). https://doi.org/10.1016/j.jneumeth.2019.108323
- Saris, I. M. J. et al. Cross-disorder and disorder-specific deficits in social functioning among schizophrenia and alzheimer’s disease patients. PLoS ONE, 17, e0263769 (2022a). https://doi.org/10.1371/journal.pone.0263769
- Wallace, V. C. J., Cottrell, D. F., Brophy, P. J. & Fleetwood-Walker, S. M. Focal Lysolecithin-Induced Demyelination of Peripheral Afferents Results in Neuropathic Pain Behavior That Is Attenuated by Cannabinoids. The Journal of Neuroscience, 23, 3221–3233 (2003). https://doi.org/10.1523/jneurosci.23-08-03221.2003
- Lu, M., Yang, C., Chu, T.,& Wu, S. Cerebral White Matter Changes in Young Healthy Individuals With High Trait Anxiety: A Tract-Based Spatial Statistics Study. Frontiers in Neurology, 9, 704 (2018). https://doi.org/10.3389/fneur.2018.00704
- Wang, B. et al. Brain asymmetry: a novel perspective on hemispheric network. Brain Science Advances, 9, 56–77 (2023). https://doi.org/10.26599/bsa.2023.9050014
- Raemaekers, M., Schellekens, W., Petridou, N. & Ramsey, N. F. Knowing left from right: asymmetric functional connectivity during resting state. Brain Structure and Function, 223, 1909–1922 (2018). https://doi.org/10.1007/s00429-017-1604-y
- Cerina, M. et al. The quality of cortical network function recovery depends on localization and degree of axonal demyelination. Brain, Behavior, and Immunity, 59, 103–117 (2017). https://doi.org/10.1016/j.bbi.2016.08.014
- Peleh, T. et al. Cross-site Reproducibility of Social Deficits in Group-housed BTBR Mice Using Automated Longitudinal Behavioural Monitoring. Neuroscience, 445, 95–108 (2020). https://doi.org/10.1016/j.neuroscience.2020.04.045
- Lee, E. et al. Enhanced Neuronal Activity in the Medial Prefrontal Cortex during Social Approach Behavior. The Journal of Neuroscience, 36, 6926–6936 (2016). https://doi.org/10.1523/jneurosci.0307-16.2016
- Jennings, J. H., et al. Interacting neural ensembles in orbitofrontal cortex for social and feeding behaviour. Nature, 565, 645–649 (2019). https://doi.org/10.1038/s41586-018-0866-8
- Xiao, G. et al. Interference of commissural connections through the genu of the corpus callosum specifically impairs sensorimotor gating. Behavioural Brain Research, 411, 113383 (2021). https://doi.org/10.1016/j.bbr.2021.113383
- Melchor, G. S., Khan, T., Reger, J. F. & Huang, J. K. Remyelination Pharmacotherapy Investigations Highlight Diverse Mechanisms Underlying Multiple Sclerosis Progression. ACS Pharmacology & Translational Science, 2, 372–386 (2019). https://doi.org/10.1021/acsptsci.9b00068
- Prut, L. & Belzung, C. The open field as a paradigm to measure the effects of drugs on anxiety-like behaviors: a review. European Journal of Pharmacology, 463, 3–33 (2003). https://doi.org/10.1016/s0014-2999(03)01272-x
- Fung, K. & Alden, L. E. Social anxiety compared to depression better accounts for enhanced acquisition of self-reported anxiety toward faces paired with negative evaluation in a conditioning task. Journal of Experimental Psychopathology, 11 (2020). https://doi.org/10.1177/2043808719888309
Mice
Six weeks old C57BL/6J male mice were ordered from Charles River Research Models and Services Germany, housed in groups of four. Mice were maintained under standard conditions (12h/12h dark/light cycle, temperature 20-24 °C, humidity 45-65 %, and ad libitum access to food and water) in IVC GM 500 open cages. Following a one-week acclimation period, surgeries were conducted, and mice remained in their designated groups of four throughout the experiment. For fUS demyelination experiment, 14 LPC and 12 controls were scanned. The behaviour cohort consisted of 16 LPC against 16 control animals. Out of those, 12 LPC and 11 controls mice proceeded to be scanned using fUS to assess the remyelination effect. Additionally, three animals were used for the light sheet fluorescence microscope (LSFM). All experiments and procedures were carried out in accordance with relevant guidelines and regulations and with the approval of the responsible governmental animal ethics committee (Regierungspräsidium Baden-Württemberg, Tübingen, Germany). All methods are reported in accordance with ARRIVE guidelines (https://arriveguidelines.org) for the reporting of animal experiments.
Stereotaxic surgery and RFID-transponder implantation
1% LPC (Sigma-Aldrich, Germany) was dissolved in physiological (0.9 %) NaCl solution by sonification, before being split into aliquots and stored at -20°C prior to surgery. Seven weeks old mice were subcutaneously injected with 10 mg/kg meloxicam (Metacam®, Boehringer Ingelheim, Germany) 20 min before being anaesthetized in a chamber with 4 % isoflurane in oxygen. After shaving the scalp, animals were fixed in a stereotaxic frame (David Kopf Instruments, USA) and anaesthesia was maintained at 1.5 % isoflurane in oxygen. Heart rate, respiratory rate and blood oxygen saturation levels were monitored throughout the surgery using a biometric paw-attached modular system (PhysioSuite, Kent Scientific Corporation, USA). Temperature was maintained at approximately 37°C using a homeothermic warming pad and rectal thermometer (PhysioSuite, Kent Scientific Corporation, USA). Bupivacaine (Carbostesin, Aspen Germany GmbH, Germany) was directly injected under the scalp as a local anaesthetic. After five minutes an incision was made using a scalpel. Based on stereotaxic mouse brain atlas44, bilateral focal lesion was induced by 1% LPC injections at the following coordinates: AP = + 1.78, ML = ± 1.00, DV = - 2.75. For unilateral lesions induced in the preliminary experiment, the respective coordinates were AP = + 1.78, ML = + 1.00, DV = - 2.75. Lysolecithin was sonicated before injection for 30 min and 500 nl of the solution was injected bilaterally using a 2 µl Hamilton syringe at a rate of 100 nl/min. To prevent solution’s spread, the syringe was lowered to DL = - 2.80 to create a pocket and finally injected at DL = - 2.75. After injection, the syringe was kept at the coordinate for five minutes, then retracted to DL = - 2.70 for an additional five more minutes. Controls received physiological (0.9 %) NaCl solution. Open drilling holes were closed with Bone wax (Ethicon, Germany). After stitching the skin, an ISO compliant RFID transponder (1.25 mm x 8 mm), already prepared in a sterile syringe, was subcutaneously implanted in the lower back close to the tail after disinfecting the site. For perioperative maintenance of body fluid, mice received 0.5 mL saline before and after surgery. As postoperative treatment, Meloxicam was given on three consecutive days and animals’ health was scored for an additional two days.
fUS imaging session
Sedation in mice was induced using 4 % isoflurane in oxygen in a separate chamber. After shaving the scalps of the animals, they were fixed in a stereotactic frame (David Kopf Instruments, USA) while reducing isoflurane concentration to 1 %. Throughout the scan, oxygen level and heart rate were constantly monitored, and body temperature was maintained at 37°C by using PhysioSuite homeothermic warming equipment. Meloxicam (0.05 mg/kg) and a bolus injection of dexmedetomidine hydrochloride (0.067 mg/kg, Tocris, USA) was subcutaneously injected. While progressively reducing isoflurane concentration to 0 %, sedation maintenance was reached by constant subcutaneous dexmedetomidine infusion (0.2 mg/kg/h, 5 mL/kg/h flow rate). After application of ultrasonic gel on the scalp, ultrasound probe was positioned ~1mm above animal’s head. We used the same Iconeus One scanner provided by Iconeus (Paris, France), as described in previous publications31,45,46. The linear probe records Doppler images at 2.5 Hz frame rate. In each session, the prefrontal area of the animal was scanned by imaging a volume divided into three planes by steps of 0.2 mm. To ensure the right position of the probe, an angiographic scan of 30 coronal slices was generated (steps of 0.2 mm) in advance. Based on the Brain Positioning System47 the target area, containing bilaterally the orbital area, the prelimbic area, the infralimbic area, and the insular region was selected. Acquisition started 10 min after isoflurane concentration was decreased to 0 %. Total recording time was 60 min. After completing the imaging session, animals were directly perfused. The scan was performed either on day 7 (demyelination phase, mid of social tracking) (Fig. 1c) or day 20 post injection (remyelination phase, after social tracking) (Fig. 3a). For more detailed information about fUS processing, we refer to the study of Ionescu46.
Social arena experiment
On the fifth day post-surgery (Fig.2a), the functionality of the RFID-transponder was verified, and social arenas (RFID-assisted SocialScan) were set up as described5. Due to the construction flexibility, three nests filled with nesting material (one large and two small) were affixed to the arena. Additionally, the arena included two water bottles, food hoppers filled with weighted pellets, and bedding material, evenly distributed (Fig.2b). Mice (16 LPC animals vs. 16 controls) in groups of four were housed in the arena for a duration of 12 days, with 12 hours light/12 hours dark cycle and the dark phase beginning at 6 p.m., while their behaviour was constantly monitored. All results obtained between day 5 (starting time: 6 p.m.) to day 10 (end: 6 p.m.) post injection, were included into demyelination phase, all results between day 12 (starting time: 6.p.m.) and day 17 (end: 6 p.m.) into remyelination. The period of 48 hours in between was determined as transition phase and was therefore excluded from analysis. Upon completion of the experiment, the body weight was recorded, and the animals were returned to standard cages in their respective groups of four.
Perfusion and Luxol Fast Blue Staining
Animals were injected intraperitoneal (i.p.) with an overdose of Narcoren (Pentobarbital 10mg/kg), were transcardially perfused and their brains taken out into 4 % PFA for ~16 hours. Those brains were transferred to 30 % Sucrose and then cut into 30 µm slices on a microtome. Slices were stored in PBS containing 0.01 % Thimerosal. The formalin-fixed tissue was washed in distilled water and positioned on slides. After an incubation time of 24 hours at room temperature in Luxol Fast Blue (LFB) Solution, slices were again washed in distilled water and differentiated several times in Lithium Carbonate Solution (0.05 %). For further differentiation, sections were dipped into ethanol (EtOH) (70 %) to receive slices where grey-matter is colourless and white-matter remains blue. Afterwards, sections were washed two times in distilled water and incubated in Cresyl Echt Violet (0.1%) for 2 min. Before quickly dehydrating sections in absolute alcohol, they were washed quickly in distilled water. By using EntellanTM (Sigma-Aldrich) sections were mounted with cover slips and imaged via light microscope. Images were further processed by ImageJ, where percentage of demyelination was calculated based on ratio of black pixels in the region of interest.
Perfusion, MACS Clearing and Immunohistochemistry
In a preliminary experiment, three animals used for Light-sheet fluorescence microscopy were injected unilaterally with the respective solution. Two animals received LPC and were killed two days or 20 days post-injection, the third one was injected with Saline and killed two days after surgery. They were injected i.p. with an overdose of Narcoren, perfused with ice-cold heparinised (5 U/mL, ThermoFisher Scientific, Germany) PBS at 7mL/min for 5 minutes before receiving 4% PFA for a further 8 minutes to achieve increased tissue clarity. Whole brains were cleared according to the MACS (MXDA-based Aqueous Clearing System) clearing protocol, as this is compatible with lipophilic dyes, and allows for whole-brain rapid clearing48. Permeabilization solution, 10X Antibody Staining Solution and Clearing Solution all came pre-prepared as part of the MACS clearing kit (Miltenyi Biotec, Germany).
Brains were collected in EDTA buffer before transferral to 4% PFA for 2h. Then they were washed in PBS before incubation for 48h in 5mL MACS Permeabilization Solution under slow continuous rotation at room temperature. MBP recombinant antibody (anti-human/mouse, 1:50, REA1154, Miltenyi Biotec, Germany) was diluted in Antibody Staining Solution and brains were placed in the solution and left on a horizontal shaker for 2 weeks at 37˚C. Brains were then washed twice in Antibody Staining Solution at room temperature overnight under slow, continuous rotation, then once more for 24h.
Absolute ethanol was prepared using activated molecular sieves (3A) and a 0.22 μm filter unit before dilution to a series of 30%, 50%, 70%, and 90%, with 2% Tween® 20. 50mL tubes were filled completely with 30% ethanol/Tween® 20 solution and a brain was placed inside each and left overnight under continuous rotation at 28˚C before transferral to the higher concentration of ethanol/Tween® 20, and this was repeated with all dilutions. To ensure complete sample dehydration, the absolute (≥99.8%) ethanol solution was changed once, and the 24h incubation step repeated twice. Finally, brains were transferred to 5mL MACS Clearing Solution and incubated 24h at room temperature, under continuous rotation. Before imaging, brains were temporarily stored at room temperature in MACS imaging solution.
Light Sheet Fluorescence Microscopy
In a preliminary experiment, the cleared and immunostained whole mouse brains, injected unilaterally with the respective solution, were imaged in the horizontal plane. One brain per acquisition was mounted using two screws holding it at the frontal part and cerebellum. The brain was then immersed in the imaging chamber filled with MACS imaging solution (refraction index 1.556) of the LaVision Biotec Ultramicroscope II (Miltenyi Biotec, DE). The laser beams were aligned horizontally with the brain, using motors in the x, y, and z direction. Afterwards, an MI Plan DC57 4x objective was fitted (Miltenyi Biotec, DE) with a numerical aperture of 0.35. Images were acquired by generation of multiple z-stacks at 0.6x magnification with laser sheets from both sides at 567nm wavelength to give an overview of whole brain immunostaining. Acquired stacks were generated by ImSpector (Version 4.0.360, LaVision BioTec, Germany) and saved as TIFF images for each channel and plane separately. After stitching those stack images via the ImageJ macro BigSticher, a 3D visualization of the immunostained brains was generated.
Data processing and statistical analysis
fUS
Regional Power Doppler signal was extracted from 8 different brain regions from both hemispheres46. After cleaning the raw signal of artifacts49, the Doppler signal was filtered for sequences between 0.01-0.2 Hz applying the second-order Butterworth filter. Afterwards, functional connectivity matrices, displaying mean values for both groups, were generated, taking correlations between all time courses of selected regions. By using Fisher’s Z-transformation, Pearson’s r correlation coefficient was further processed to z-scores, allowing group-level analysis.
Social arena
The social arena was set up as described previously5. It included three attached nests to provide withdrawal and sleeping opportunities (Fig. 2a). Centre, periphery, corners, food, and water zones were fixed. Primary behavioural data were generated by the software package RFID-Assisted SocialScan. Pre-defined parameters for velocity, angle of movement, and distances5 were applied to compute non-social and social behaviours including time spent in zone, locomotion, approach, sniffing, following, contact and fighting. Individual animals were tracked by their implanted RFID-transponder. The primary behaviour data sets were extracted and processed in MATLAB. The total of all detected events in a three-hour time window on each day was used to extract numbers for all social parameters and locomotion. Then, means were calculated for each individual mouse during demyelination and remyelination phases. For all other parameters, including time spent in nest, food zone, water, centre and periphery, the total of all detected events over 24 hours was taken, followed by the mean for both recording sessions. The term 'social stay in nest' is used to denote the time spent with one or more cage mates inside nests. This duration is calculated by summing up all such events that occur within a day in a single box. The graphs in Fig. 2g and Fig. 3d display twenty values for each treatment group. These values represent the behaviour observed in the four boxes per treatment group over a period of five days during each monitoring phase.
Behavioural z-scores were calculated for individual observation (X) based on standard deviation (σ) and mean value (μ) of the control group. Those z-scores were generated as a mean value over each five days period.
The spider web plots in this study represent all analysed parameters. For behaviours such as sniffing, approaching, following, fighting, and contact, we used the values depicted in the corresponding graphs from 9 p.m. to midnight. For social stay in nest, centre, periphery, nest, food zone and water, we used the values depicted in the corresponding graphs collected over a 24-hour period. After normalising the individual values, we computed the means and displayed them including significance differences for each treatment group in the spider web plots.
Statistical analysis
All statistics were processed in GraphPad Prism or MATLAB. For fUS data, a dynamic analysis was used by applying a sliding window and the following two-sample t-tests for each time slot to achieve higher accuracy in detecting an effect46. Each t-test was corrected by False Discovery Rate (FDR) correction50. For processing the CBV signal for the fUS-example animals, a simple linear regression was performed. For the social arena task, results were displayed as mean values (± SEM) and further processed by using one-way analysis of variance (ANOVA) corrected via Bonferroni post-hoc test. Only significant differences between both groups in the specific time phase were displayed. To analyse differences in spending time in a specific area in 24 hours, an unpaired t-test was performed, as well as for LFB-staining. Significance levels were set at p < 0.05 indicated by (*), p < 0.01 (**) and p < 0.001 (***).
Methods references
44 Paxinos, G., & Franklin, K. B. J. The Mouse Brain in Stereotaxic Coordinates (Vol. 5). Academic Press ISBN: 9780128161586 (2019).
45 Grohs-Metz, G., Smausz, R., Gigg, J., Boeckers, T. & Hengerer, B. Functional ultrasound imaging of recent and remote memory recall in the associative fear neural network in mice. Behavioural Brain Research, 428, 113862 (2022). https://doi.org/10.1016/j.bbr.2022.113862
46 Ionescu, T. M., Grohs-Metz, G., & Hengerer, B. Functional ultrasound detects frequency-specific acute and delayed S-ketamine effects in the healthy mouse brain. Frontiers in Neuroscience, 17, 1177428 (2023). https://doi.org/10.3389/fnins.2023.1177428
47 Nouhoum, M. et al. A functional ultrasound brain GPS for automatic vascular-based neuronavigation. Scientific Reports, 11, 15197 (2021). https://doi.org/10.1038/s41598-021-94764-7
48 Zhu, J. et al. MACS: Rapid Aqueous Clearing System for 3D Mapping of Intact Organs. Advanced Science, 7, 1903185 (2020). https://doi.org/10.1002/advs.201903185
49 Brunner, C. et al. Whole-brain functional ultrasound imaging in awake head-fixed mice. Nature Protocols, 16, 3547–3571 (2021). https://doi.org/10.1038/s41596-021-00548-8
50 Benjamini, Y. & Hochberg, Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society, Series B (1995). https://doi.org/10.1111/j.2517-6161.1995.tb02031.x
Competing interest reported. Employment: F.S., E.H., T.I., E.K., H.M.M.,and B.H. received salaries from Boehringer Ingelheim Pharma GmbH & Co. KG