Short term changes in the proteome of human cerebral organoids induced by 5-methoxy-N,N-dimethyltryptamine

Dimethyltryptamines are hallucinogenic serotonin-like molecules present in traditional Amerindian medicine (e.g. Ayahuasca, Virola) recently associated with cognitive gains, antidepressant effects and changes in brain areas related to attention, self-referential thought, and internal mentation. Historical and technical restrictions impaired understanding how such substances impact human brain metabolism. Here we used shotgun mass spectrometry to explore proteomic differences induced by dimethyltryptamine (5-methoxy-N, N-dimethyltryptamine, 5-MeO-DMT) on human cerebral organoids. Out of the 6,728 identified proteins, 934 were found differentially expressed in 5-MeO-DMT-treated cerebral organoids. In silico systems biology analyses support 5-MeO-DMT’s anti-inflammatory effects and reveal a modulation of proteins associated with the formation of dendritic spines, including proteins involved in cellular protrusion formation, microtubule dynamics and cytoskeletal reorganization. Proteins involved in long-term potentiation were modulated in a complex manner, with significant increases in the levels of NMDAR, CaMKII and CREB, but a reduction of PKA and PKC levels. These results offer possible mechanistic insights into the neuropsychological changes caused by the ingestion of substances rich in dimethyltryptamines.


Introduction
Dimethyltryptamines are naturally occurring hallucinogenic molecules hypothesized to be involved in spontaneous altered states of consciousness, such as dreams, free imagination and insightful creativity (Strassman 2001; Barker et al. 2012). N,N-dimethyltryptamine (N,N-DMT) and bufotenine (5-HO-DMT) have been traditionally used as entheogens by Amerindians (McKenna 2004;Ott 2001)

as major active ingredients of a brew called
Ayahuasca and the Virola snuff (Holmstedt & Lindgren 1967). The popularity of Ayahuasca as part of religious ceremonies continues to spread in South America and other countries (Labate & Feeney 2012), possibly motivated by its strong antidepressant effects (Osório Fde et al. 2015;Sanches et al. 2016). Chronic Ayahuasca ingestion has been associated with cognitive gains and structural brain changes in areas related to attention, self-referential thought, and internal mentation (Bouso et al. 2012;Bouso et al. 2015).
The search for the molecular mechanisms underlying the effects of dimethyltryptamines showed that N,N-DMT and 5-methoxy-N,Ndimethyltryptamine (5-MeO-DMT), two closely related metabolic products, can act as systemic endogenous regulators of inflammation and immune homeostasis through both 5-hydroxytryptamine receptors (5-HTRs) and sigma 1 receptors (σ-1Rs) (Szabo et al. 2014;Fontanilla et al. 2009). Under severe hypoxia, N,N-DMT robustly increased the survival of cultured human cortical neurons in vitro, monocyte-derived macrophages, and dendritic cells acting through σ-1Rs (Szabo et al. 2016). The direct evidence of neuro-immune communication and neuroregenerative effects of N,N-DMT and 5-MeO-DMT greatly enhanced expectations for dimethyltryptamine research.
Our limited understanding of the physiological activity of dimethyltryptamine and other classic psychedelic substances is caused by legal restrictions on such research (Nutt et al. 2013) but also by the lack of adequate experimental models (Vollenweider & Kometer 2010;Hanks & González-Maeso 2013;de la Torre & Farré 2004). In the past few years, considerable progress has been made regarding the neural differentiation of human pluripotent stem cells into mature neurons and cerebral organoids (Kelava & Lancaster 2016). Human neural progenitors (hNPC) are useful cell systems for high-throughput screening due to their homogeneity, with little complexity and limited differentiation potential. On the other hand, cerebral organoids are complex three-dimensional (3D) culture systems with multiple cell types that selforganize into various brain regions similarly to those in vivo, including the cerebral cortex, ventral forebrain, midbrain-hindbrain boundary, and hippocampus (Lancaster & Knoblich 2014;Qian et al. 2016). A comparison of gene expression programs of human fetal neocortex and in vitro cortical development by single-cell RNA sequencing found remarkable similarities (Camp et al. 2015). Cerebral organoids are being widely applied to humanspecific biological questions and purposes, since they have been proved to be a good system for drug testing. They may well recapitulate effects in the human nervous system, particularly related to plasticity and growth (Garcez et al. 2016;Lancaster & Knoblich 2014), and they circumvent problems of discrepancies in metabolic pathways occurring in translational studies involving animals. The development of such model offers an exciting new range of opportunities to investigate the molecular responses of human neural tissue to psychoactive substances.
Here we analyzed the effect of 5-MeO-DMT on neural cells and human brain organoids. By employing mass spectrometry-based proteomics to analyze cerebral organoids, we managed to investigate effects in a large scale and in an unbiased manner and to get insights into the molecular mechanisms and biochemical pathways involved (Martins-de-Souza 2012). Our results show that 5-MeO-DMT modulates long term-potentiation (LTP), in addition to morphogenesis and maturation of dendritic spines, while inhibiting neurodegeneration and cell death.

Human neural progenitor cells are unaffected by 5-MeO-DMT
First, we examined the effects of 5-MeO-DMT in hNPCs (detailed characterization in (Dakic et al. 2016). hNPCs showed basal expression of σ-1Rs but not 5-HT 2A and 5-HT 2C ( Fig. 1A and B). Using a high content screening analysis, we tested the effects of 5-MeO-DMT (23 nM to 7.11 µM) upon cell death, proliferation and differentiation of hNPCs. There was no evidence of change in cell death or proliferation in response to 5-MeO-DMT ( Fig. 1C and D). In addition, by quantifying some aspects of dendritic branch complexity, we measured neural arborization based on MAP2 staining of young neurons exposed or not to 5-MeO-DMT. Despite a slight trend, there were no statistically significant differences in the measured parameters ( Fig. 1 E, F, G and H).

Human cerebral organoids express 5-MeO-DMT receptors
The lack of alterations in cell death, proliferation or differentiation/arborization could be due to the low cellular diversity and lack of complex interactions among different cell types. Thus, we challenged human cerebral organoids, which recapitulate better the complexity and function of in vivo neural circuitry.
Basal immunostaining of ionotropic receptors AMPA and NMDA, characteristic of glutamatergic synapses, along with the neuronal marker MAP2 was observed in 45-days-old cerebral organoids ( Fig. 2A-D), as previously described (Sartore et al. 2017). Glial cells (GFAP+) are also present in organoids, as shown in Fig. 2E. Interestingly, in contrast with hNPCs, we were able to detect the expression of 5-HT 2A via PCR and/or immunostaining, as well as of σ-1Rs, the primary pharmacological molecular targets for 5-MeO-DMT. As shown in Figure

5-MeO-DMT alters the proteome of human cerebral organoids
Due to the complexity of the organoid system, we decided to cast a much wider net to detect potentially important 5-MeO-DMT effects. By analyzing the proteome of organoids with or without treatment, we were able to look for changes in the expression of a considerable number of proteins, in an unbiased approach. Thus, to resolve the proteome of human neural tissue under the effect of 5-MeO-DMT, we analyzed 45-days-old cerebral organoids after 24-hour treatment (Fig. 3A). A total of 144,700 peptides were identified at a false discovery rate (FDR) below 1%. These led to the identification of 6,728 unique proteins by, at least, two unique peptides present in no less than two out of three biological replicates analyzed. Notably, there was an overlap of 99% of identified proteins among all treatment groups (Fig. 3B), demonstrating the robustness of the method. From these commonly identified proteins, we found 934 differentially expressed (using a -2 < Log 2 ratio > 2 cut-off), comprising 360 downregulated and 574 upregulated proteins, when comparing 5-MeO-DMT and vehicle groups. Functional enrichment for combined up and downregulated proteins predicted the biological functions of those changes. Regarding diseases or functions, using prediction effect analysis (-2 < z-score > 2.0 is significant for inhibition/activation) (Fig. 3C), we observed a significant activation score for dendritic spine and cellular protrusion formation, microtubule and cytoskeletal organization, and also mild activation of T lymphocyte differentiation. On the other hand, biological functions such as neurodegeneration, cell death, and brain lesion were predicted to be inhibited.

5-MeO-DMT leads to inhibition of NF-κB signaling pathway
Among the canonical pathways identified are nuclear factor of activated T-
Our work also revealed that only a 24h-treatment with 5-MeO-DMT, i.e., a single dose, modulates specific signaling molecules identified as key players in LTP, a classic mechanism of learning and memory (Malenka & Bear 2004).
Based on in silico predictions using proteomics data, modulation of these signaling molecules by 5-MeO-DMT would produce a complex regulation of LTP. One possibility is that LTP may be augmented in some cell types and inhibited in others, leading to a mixed profile.
Additionally, we observed major downregulation of mGluR5 after treatment Binding of ephrin-Bs to EphB receptors initiate bidirectional signaling, which, by altering actin cytoskeleton, lead to changes in dendritic spine shape, size, and number (Klein 2009). It was shown that EPHB2 interacts with intersectin and activates its GEF activity in cooperation with N-WASP, which in succession activates the Rho-family GTPase Cdc42 and spine morphogenesis (Irie & Yamaguchi 2002). N-WASP is a critical regulator of Arp2/3-mediated actin polymerization (Takenawa & Miki 2001). Henkemeyer and colleagues demonstrated that triple EphB1,2,3-deficient hippocampal neurons have abnormal formation of actin clusters along dendrites, impairing normal dendritic spine formation in vivo (Henkemeyer et al. 2003). Meanwhile, in vitro, knockdown of EphB2 alone is sufficient to reduce synapse density (Kayser et al. 2006). Postnatal re-expression of EphB2 in slice cultures from animals lacking EphB1-3 is sufficient to rescue dendritic spine defects (Kayser et al. 2006). Although EphB signaling has a clear role in dendritic spine morphogenesis through kinase domain activity, it can also regulate activity-dependent synaptic plasticity interacting with both NMDA (Takasu et al. 2002) and AMPA receptors (Kayser et al. 2006). Literature shows that σ- metalloproteinase (Bozdagi et al. 2007) and integrins (Shi & Ethell 2006;Dityatev & Schachner 2006). An upregulation of integrins, as we observed here in 5-MeO-DMT-treated organoids, was also found in major depression patients who presented good response to antidepressants, suggesting the importance of this class of proteins in brain plasticity ). One more protein significantly downregulated is srGAP, an intracellular signaling molecule with a role in processes underlying synaptic plasticity, higher cognitive function, learning, and memory (Endris et al. 2002).
Finally, we also found important functions such as neurodegeneration, cell death, and brain lesion inhibited by 5-MeO-DMT. These neurorestorative and cellular protective effects are expected after activation of σ1R (Frecska et al. 2013;Szabo et al. 2016). σ1R agonists exert neuroprotective effects by regulating intracellular calcium levels (Mueller et al. 2013), preventing expression of pro-apoptotic genes (Tchedre & Yorio 2008) and protecting mRNA of anti-apoptotic genes, such as Bcl-2.
Fast antidepressants also have strong effect on synaptic plasticity, reversing functional and structural synaptic deficits caused by stress. A typical example of this group is ketamine, an hallucinogenic, non-competitive NMDA glutamate receptor channel antagonist, which causes an improvement in mood ratings within hours, as opposed to weeks as in typical antidepressants . Ketamine increases mammalian target of rapamycin complex 1 (mTORC1) signaling, via activation of Protein kinase B (PKB or Akt) and ERK. mTOR signaling than boosts synaptic protein synthesis, spine stability and function in the prefrontal cortex (Duman & Aghajanian 2012;Duman et al. 2016;Gerhard et al. 2016).
Therefore, the pattern of proteins altered after 5-MeO-DMT treatment points to robust actions on synaptic plasticity and improvement of cell survival.
Taken together our data offer a possible mechanistic insight into the neural changes produced by the chronic ingestion of substances containing dimethyltryptamines.

Materials and methods
Human embryonic stem cells BR1 lineage of human embryonic stem cells (hESCs) (Fraga et al. 2011) was cultured in mTeSR1 media (Stemcell Technologies) on Matrigel (BD Biosciences) -coated surface. The colonies were manually passaged every seven days and maintained at 37°C in humidified air with 5% CO 2 .

Human neural progenitor cells
To induce hESCs towards neural differentiation, we adapted Baharvand and coworkers' protocol (Baharvand et al. 2007;Dakic et al. 2016  days of treatment and cells were allowed to differentiate for 3 more days. On day 7 cells were fixed for immunocytochemistry.

High Content Analysis
All images were acquired on Operetta high-content imaging system (Perkin Elmer, USA). For proliferation, incorporated EdU was detected with Alexa Fluor 488 using Click-iT EdU kit (C10351, Invitrogen, Carlsbad, USA) following the manufacturer's instruction. Total number of cells was calculated by nuclei stained with 1 mg/mL of DAPI (4',6-diamidino-2-phenylindole). S phase was determined by percentage of total cells labeled with EdU. Images were acquired with a 10x objective with high numerical aperture (NA).

For cell death analysis, cells were labeled with LIVE/DEAD®
Viability/Cytotoxicity Kit (Thermo Fisher Scientific). This kit contains two probes: calcein AM and ethidium homodimer (EthD-1). The first one allows measuring of intracellular esterase activity and second one plasma membrane integrity. Mix of probes was done in DMEM/F-12 (without phenol red, Life Technologies), together with the cell-permeant nuclear dye Hoechst. After incubation for 30 min at 37ºC and 5% CO 2 , the dye cocktail was replaced by new medium and live cell imaging was performed using temperature and CO 2 control option (TCO) of Operetta, set to 37°C and 5% CO 2 at 10x magnification. Quantification analyses were normalized to the number of cells

Differentiation into cerebral organoids
Differentiation of hESCs into cerebral organoids was based on previously described protocol (Lancaster et al. 2013;Sartore et al. 2017

RNA Isolation and PCR Analysis
Total RNA was isolated using the GeneJET RNA Purification Kit (Thermo

Sample preparation
Sample lysates were thawed and centrifuged at 10,000 x g for 10 min at 4 °C.
The supernatant was collected and subjected to quantification by Qubit® 3.0 Fluorometer (Thermo Fisher Scientific). Each sample (50 µg) was subjected to a SDS-PAGE gel electrophoresis. Gel lanes were sliced and digested in gel overnight as previously described (Maccarrone et al. 2014). Generated peptides were dried in a SpeedVac concentrator and stored at -80 o C prior to shotgun mass spectrometry analyses.

Liquid chromatography-mass spectrometry
Qualitative and quantitative proteomic analysis were performed on a 2D-LC-MS/MS system with ion-mobility-enhanced data-independent acquisitions (Souza et al. 2017 Samples were all run in technical and biological triplicates.

Database search and quantification
Raw data was processed with Progenesis ® QI version 2.1 (Waters) and proteins were identified. Quantitative data was processed using dedicated algorithms and searching against the Uniprot human proteomics database (version 2015/09), with the default parameters for ion accounting and quantitation (Li et al. 2009). The databases used were reversed "on the fly" during the database queries and appended to the original database to assess the false-positive identification rate. The following parameters were considered in identifying peptides: 1) Digestion by trypsin with at most one missed cleavage; 2) variable modifications by oxidation (M) and fixed modification by carbamidomethyl (C); 3) false discovery rate (FDR) less than 1%. Identifications that did not satisfy these criteria were not considered.

In silico analysis
Protein networks and canonical pathways associated with differentially expressed proteins were identified using software Ingenuity Pathway Analysis (IPA, Ingenuity Systems, Qiagen, Redwood, CA, USA; www.ingenuity.com).
This software uses curated connectivity information from the literature to determine interaction network among the differentially expressed proteins and canonical pathways in which they are involved. Here, we have considered information from nervous system tissues and cells, immune cells and stem cells. The significant biological functions are based on Fisher's exact test.
Multiple correlation hypothesis are based on Benjamini-Hochberg (B-H) approach using 1% FDR threshold, the significance of the IPA test is expressed as p-values (Schubert et al. 2015).