Comprehensive analysis of 42 psilocybin-producing fungal strains reveals metabolite diversity and species-specific clusters

Abstract

Psilocybin-producing fungi have garnered attention due to accumulating evidence regarding the therapeutic potential of their principal component psilocybin. This diverse group of fungi harbors a wealth of less-studied metabolites, however, thus far most research has addressed them as a cohesive group. By optimizing an approach for extraction and analysis, we examined the metabolomes of 42 distinct fungi strains and show that the breadth and diversity of metabolites within and between 9 species. We integrated and validated the reproducible and reliable extraction of fruiting bodies followed by chromatographic separation, quantification and identification of their known and yet to be identified secondary metabolites. The optimal extraction of fruiting bodies for high yield of indole alkaloids was achieved using a 1:20 tissue:solvent ratio, 25:75 H₂O:MeOH (pH = 9), for 1.5 h, followed by the quantification of 8 tryptophan-derived indolamines by HPLC–DAD and the identification of putative metabolite hydroxypsilocybin by HPLC–MS/MS. The metabolomic analysis revealed the diversity of metabolites within and between species. Finally, we developed and present a method that mimics the in vivo process of dephosphorylation that occurs upon ingestion for in vitro setups. Overall, our study summarizes a standardized approach for both in vitro and in vivo studies involving psilocybin-producing fungi, showcasing the unique metabolome of each strain and the rich diversity of these fungi, encompassing promising pharmaceutical potential.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-97710-z.

Introduction

Psilocybin-producing fungi have been used for centuries for their perceived sacred and healing powers within various indigenous societies^1,2. They have been used as consciousness altering drugs (entheogens) during transcendental ceremonies by pre-Columbian Mesoamerica inhabitants³, as well as in healing ceremonies^1,4. In recent years, the medical interest in these fungi has been reignited as evidence accumulates regarding their therapeutic potential in treating different diseases^5–9, mainly on psilocybin^5,10–16, their major secondary metabolite^3–7. Fungal secondary metabolites have shown promising efficacy not only in mental health disorders^17,18, but also in inflammation^19,20, neuroplasticity^8,21 and have been suggested to modulate the brain-gut axis^22,23. Secondary metabolites play diverse biological roles that contribute to the survival of the organism^15,16. In nature, they are important for the interaction of the producing fungi with the ever-changing environment²⁴, but in the context of human use, these metabolites can be harnessed for medicinal purposes to treat human diseases and disorders^25,26. Secondary metabolites from psilocybin-producing fungi are a promising source of non-toxic compounds for drug development efforts^26,27, but only a fraction of these fungi species have been studied.

The medicinal properties of these fungi are attributed to psilocybin (4-phosphoryloxy-N,N-dimethyltryptamine), the prodrug to the psychedelic psilocin (4-hydroxy-N,N-dimethyltryptamine)²⁸. Psilocybin is the phosphorylated and less-reactive form of psilocin, both are indolamine derivatives of the amino acid tryptophan, differed by a phosphate ester at the C-4 position. Once ingested, psilocybin is metabolized and converted into psilocin through dephosphorylation²⁸. Characteristically, this dephosphorylation occurs upon injury of the psilocybin-rich tissues of the fungi. It is apparent by a bluish-turquoise hue²⁹, which develops due to the oligomerization of oxidized psilocin and other indole alkaloids³⁰.

Many strains of fungi produce psilocybin, the common species belong to the genera Psilocybe and Panaeolus^31–37, and a small number of additional species belong to other genera^38–41. These strains differ greatly in their morphology⁴², and moreover, in their metabolome. Current biomedical research focuses only on a few bioactive components^4,43, while neglecting a plethora of other tryptophan-derived indolamines with high therapeutic potential. To investigate these indolamines in vitro and in vivo, researchers require standardized sample extraction and metabolite identification protocols. Existing extraction methods encompass a wide spectrum of organic solvents, pH ranges, durations, solvent composition and tissue to solvent ratios, and temperatures^44–50; making it challenging to determine the optimal approach. Moreover, most methods are designed to isolate and quantify only a few metabolites with an emphasis on psilocybin and psilocin. Furthermore, due to the complexity and diversity of the secondary metabolites in these fungi^26,51, common analytical techniques are lacking in the characterization of these important compounds. It is crucial to develop reliable and efficient methods for extracting and analyzing these fungal extracts consistently. This is particularly important as newer studies open new research avenues into the therapeutic properties of a large array of indole alkaloids^22,52,53.

To elucidate the diversity of secondary metabolites within and between psilocybin-producing fungal species, in this study we designed and validated an integrated approach for the reproducible and reliable extraction of sporophores (fruiting bodies), followed by accurate identification and quantification of their known metabolites, particularly tryptophan-derived indoleamines, and comparison of their entire metabolomes. In addition, we established an in vitro dephosphorylation methodology, essential for the investigation of biotransformed metabolites in in vitro settings.

Results

Establishment of optimal extraction parameters

As the therapeutic potential of psilocybin-producing fungi is gaining momentum and new data accumulate, there is an increasing need for establishing methodologies for robust, accurate and reproducible extraction that allows for comprehensive characterization of their full metabolomic profiles and preserve the bioactive metabolites with potential therapeutic properties. The initial step requires the transformation of the fresh fruiting bodies to a uniform, stable and dissolvable extract that meets the requirements of both in vitro and in vivo experimental designs. To optimize the extraction procedure, we chose three distinct psilocybin-producing species^31,42: Psilocybe cubensis BG, Panaeolus cyanescens Vietnam and Psilocybe tampanensis TampG. The different species exhibited fruiting bodies with a wide range of sizes, shapes and colors (Fig. 1A). Once harvested, fruiting bodies are immediately dried to inhibit spoilage by microorganisms. Drying the fresh fruiting bodies at only 40 °C in vacuum conditions drastically speeds up the drying process and reduces oxidation⁵⁴ and possible dephosphorylation⁵⁵. Most extraction protocols employ temperatures of 50–70 °C⁵⁶, which may cause the degradation of target metabolites. Grinding to a fine powder is also essential as it increases the surface area in contact with the extraction solvent. The final product of the process we developed is an extracted, homogenized and lyophilized powder.

Fig. 1.

Establishment of optimal extraction methodologies. (A) Phenotypical differences in fruiting bodies of Psilocybe cubensis BG, Panaeolus cyanescens Vietnam and Psilocybe tampanensis TampG when fresh (top) and after drying (bottom). Dried fruiting bodies are ground and extracted, the solvent is removed and the sample is completely dried by lyophilization (right). To capture images, the fungi were positioned on the same black background and photographed individually. (B–D) Fungi were assessed by HPLC–DAD for the concentration (%w/w) of the major indolamine metabolites psilocybin and psilocin, and the minor metabolite baeocystin following extraction with different (B) time intervals, (C) tissue to solvent ratio, and (D) H₂O:MeOH solvent ratio at 100:0, 50:50, 25:75 and 0:100, respectively. (E) Extracted tissue of the three fungi was re-extracted two additional times and the relative levels of psilocybin, psilocin and baeocystin were assessed with HPLC–DAD. (F) The fruiting bodies of the three fungi were extracted for 1.5 h at 1:20 (tissue:solvent) with 25:75 (H₂O:MeOH) solvent where the H₂O (pH 6.8) was buffered with ammonium acetate to pH = 9. Then, the relative levels of psilocybin, psilocin and baeocystin were assessed with HPLC–DAD. Results are presented as mean (n = 3) ± %RSD.

Existing methods integrate a wide range of experimental conditions^44–50. Most methods rely on prolonged agitation of dry tissue in methanol (MeOH), a commonly used solvent useful for the extraction of a broad range of compounds^57,58 due to its polar and non-polar solvency properties and low boiling point, which makes it easily evaporated from the solute⁵⁹. H₂O aids in the extraction of hydrophilic metabolite groups. The use of alkaline-buffered H₂O also may avoid unwanted dephosphorylation of psilocybin and possibly other phosphorylated metabolites. This solvent mixture ensures the extraction of both lipophilic and hydrophilic metabolites.

To define the optimal extraction conditions for high yield of indole alkaloids, we tested three parameters that affect the final metabolomic makeup of the extract^{45,48–50,56,60,61}: the duration of extraction and the ratio of solvent:tissue, to ensure efficient and rapid recovery without degradation processes such as the dephosphorylation of psilocybin to psilocin; as well as different degrees of polarity for the extraction solvent (H₂O:MeOH), which directly affects compound solubility (Fig. 1B–D). To assess the optimal conditions, we analyzed the levels of the two major indolamines, psilocin and its prodrug psilocybin, and one minor indolamine, baeocystin. We found the major metabolites are sufficiently extracted already at 30 min, with a trend of decline in the extraction efficiency occurring between 6 and 24 h. Baeocystin could not be recovered from Psilocybe cubensis BG, but a similar trend of decreased extraction efficiency after 30 min was evident in the other two fungi (Fig. 1B). The tissue:solvent ratio demonstrated a trend of slight increase in Panaeolus cyanescens Vietnam when extraction was performed with more solvent, while in the other fungi changing the ratio had no effect (Fig. 1C). Hence, a ratio of 1:20 is sufficient to extract indole alkaloids from a wide variety of fungi. Regarding the solvent mixture, we expected a difference in solubility due to the phosphoryloxy, methyl and hydroxy chemical groups of each metabolite. Baeocystin was recovered in MeOH, assumingly due to less dephosphorylation into norpsilocin, while psilocybin and psilocin were more soluble in 25:75 H₂O:MeOH (Fig. 1D). The 25:75 ratio yields relatively high concentrations of metabolites without altering the chemical composition. Altogether, the optimal extraction parameters for high yield of indole alkaloids in their form as in the fruiting bodies were defined as tissue:solvent ratio of 1:20 in 25:75 H₂O:MeOH for a duration of 1.5 h.

To validate the efficiency of the extraction, the same tissue of the three fungi was re-extracted two additional times under identical conditions (Fig. 1E). The second extraction yielded 10–12% psilocybin and less than 10% psilocin and baeocystin relative to the amount yielded initially. The third extraction yielded less than 5% for all three metabolites. These results confirmed less than 10% of metabolites remained in the extracted tissue after the first extraction, negating the need for repeated extractions.

The pH of the H₂O used in the extraction solvent is slightly acidic, which can contribute to the dephosphorylation of psilocybin through chemical hydrolysis and possible enzymatic routes²⁸. Therefore, we compared the optimal extraction conditions using 25:75 H₂O:MeOH buffered to pH = 9 (Fig. 1F). Raising the pH with ammonium acetate increased the extraction efficiency of all three metabolites in Psilocybe cubensis BG. No change was detected for Panaeolus cyanescens Vietnam, and an opposite small but significant trend was found in Psilocybe tampanensis TampG.

Overall, using a relatively small volume (1:20 tissue to solvent, respectively) of 25:75 H₂O:MeOH (pH = 9) solvent mix and a short extraction duration of only 1.5 h, resulted in an efficient and consistent final homogenized and lyophilized powder, which was fourfold concentrated relative to the original biomass.

Separation, quantification and identification of fungi metabolites

The known biosynthetic pathway, originating from the amino acid tryptophan and leading to the known major and minor tryptophan-derived indoleamines^62–64, is shown in Fig. 2A. We identified different metabolites in the pathway, which were previously characterized by four transformation processes: decarboxylation⁶⁵, oxidation⁶², methylation⁶⁶ and phosphorylation⁶⁷. (These are orchestrated in the fungi by the enzymes PsiD, PsiH, PsiM and PsiK; respectively.

Fig. 2.

Separation, quantification and identification of known and unknown indolamine-type metabolites. (A) Structural formulas and chemical transformations of known tryptophan-derived indolamine metabolites. (B) HPLC–DAD chromatograms of Psilocybe cubensis BG, Panaeolus cyanescens Vietnam and Psilocybe tampanensis TampG and a mix of 8 standards (S1–S8): norbaeocystin, baeocystin, aeruginascin, psilocybin, norpsilocin, psilocin, tryptophan and tryptamine; respectively. Axes represent the abundance (mAU) and the RT (min) at a wavelength of 275 nm. (C–D) HPLC–MS/MS chromatogram showing the (C) standards of norbaeocystin (S1), psilocybin (S4) and psilocin (S6); and (D) Putative hydroxypsilocybin found in Psilocybe cubensis Golden Teacher. Full MS scan was performed in the range of 50–1,000 m/z. (E–F) MS/MS fragmentation spectra (50–325 m/z range) of (E) Psilocybin (S4) and putative hydroxypsilocybin.

Due to their shared indole-type structure, the UV absorption characteristics of these small metabolites are very similar, making the identification of overlapping peaks challenging. We achieved high separation based on the difference in minor structural modifications, using a dedicated C18 reversed-phase column with high reproduction rate. The metabolites of tryptophan, tryptamine, norbaeocystin, norpsilocin, baeocystin, psilocin, psilocybin and aeruginascin were consistently identified in different concentrations in all psilocybin-producing fungi. Therefore, we used analytical standards for these 8 metabolites as part of the standardized high-performance liquid chromatography coupled with a photodiode array detector (HPLC–DAD) quantification method (Fig. 2B). The separation of the 8 metabolites was evident by well-defined, narrow peaks with clear baseline separation, and minimal tailing or fronting. The analytical method validation parameters are presented in Supplementary Table S1.

While our HPLC–DAD approach is useful for high throughput dereplication of known metabolites in fungal extracts, it cannot serve in the identification of new unknown metabolites. For that purpose, we utilized HPLC coupled with tandem mass spectrometry (MS/MS). The successful implementation of the separation method is evident by the separate narrow peaks of norbaeocystin at m/z 257.0686, psilocybin at m/z 285.0999 and psilocin at m/z 205.1335 (Fig. 2C). Once precursor ions are identified, they are fragmented and the resulting MS/MS spectra are explored for product ion patterns that provide valuable structural information. The mass spectrometry characteristics of identified indole alkaloids are presented in Supplementary Table S2. For example, a newly identified compound produced by Psilocybe cubensis Golden Teacher, putatively named⁶⁸ hydroxypsilocybin⁶⁸ (Fig. 2D), was assigned as a psilocybin derivative with an additional oxygen atom based on the observed 15.9949 m/z mass increase. The [M + H]⁺ ion at 301.0937 m/z closely matches the theoretical mass of putative hydroxypsilocybin (adduct [M + H]⁺ at 301.0948 m/z with mass error -3.65 ppm). Fragmentation in the MS/MS spectrum (Fig. 2E, F) provided validation, as all major fragments of psilocybin showed an equivalent 15.9949 m/z increase (250.1328 m/z, 160.0749 m/z, and 250.0410 m/z shifted to 221.1271 m/z, 176.0696 m/z, and 256.0355 m/z; respectively) confirming the presence of the additional oxygen. The fragment of N,N-dimethyl at 58.0652 m/z remained unmodified. The exact position of the additional oxygen remains to be determined. Thus, the fragmentation of indole-based metabolites provides data regarding the functional group identities and possible modifications of unknown similar metabolites, and enables us to predict chemical transformation products, and then verify their presence in the fungal extract through the analysis of their MS/MS spectra. These methods were validated for linearity, analytical limits, precision, and robustness with a maximum 10% shift in chromatographic separation.

Clustering of secondary metabolite arrays according to fungal specie

Following the establishment of the optimal extraction and identification methodologies, we selected the highest diversity strains of as many species as possible and extracted a total of 42 psilocybin-producing fungi strains derived from 9 distinct species according to morphological identification⁴² (Table 1). As the taxonomic accuracy of species determination according to morphological identification may be unreliable³², all 42 samples were also cross-referenced with internal transcribed spacer (ITS) identification (Table 1, rightmost columns). The fungal metabolomes were analyzed with HPLC–MS/MS (the metadata is available on Open Science Framework, see Data availability). To understand the variability in the content of metabolites between the different species and also within each specie, we utilized unsupervised Principal Component Analysis (PCA) to investigate the metabolomic similarities among the different strains. As the 42 strains were cultivated under the same controlled conditions, it is likely the variance in the metabolomes is driven by differences in secondary metabolites. The first two principal components explaining 25.4% (PC1 15.9%, PC2 9.5%) of the variance were retained and visualized using a biplot (Fig. 3³² The metabolic clustering correlated better to the ITS identification than to the classic morphological one. According to ribosomal ITS identification, cubensis and natalensis were both identified as cubensis, while mexicana, galindoi and tampanensis were identified as tampanensis. The Panaeolus species formed a distinct cluster in comparison to the major cluster of the Psilocybe species (as indicated by PC1 in the x-axis). Within the Psilocybe genera, cubensis clustered separately from the other Psilocybe species (as indicated by PC2 in the y-axis).

Table 1.

Genetically stable fungi strains classified according to sequenced ITS.

#	Genus	Species	Strain	ITS species classification	Identity (%)	Query cover (%)	GenBank accession
1	Panaeolus	cyanescens	Australia	Panaeolus cyanescens	100.00	93	PV069779
2	Panaeolus	cyanescens	British Virgin Island (BVI) V1	Panaeolus cyanescens	97.99	69	PV069780
3	Panaeolus	cyanescens	British Virgin Island (BVI) V2	Panaeolus cyanescens	96.05	84	PV069781
4	Panaeolus	cyanescens	British Virgin Island (BVI) V3	Panaeolus cyanescens	99.73	97	PV069782
5	Panaeolus	cyanescens	Goliath	Panaeolus cyanescens	98.89	97	PV069783
6	Panaeolus	cyanescens	PanMex	Panaeolus cyanescens	99.71	97	PV069784
7	Panaeolus	cambodginiensis	Sandose V1	Panaeolus cyanescens	99.73	97	PV069785
8	Panaeolus	cambodginiensis	Sandose V2	Panaeolus cyanescens	99.72	95	PV069786
9	Panaeolus	cambodginiensis	Sandose V3	Panaeolus cyanescens	99.73	97	PV069787
10	Panaeolus	cyanescens	Vietnam	Panaeolus cyanescens	99.18	97	PV069788
11	Psilocybe	caerulescens	Caro V1	Psilocybe caerulescens	99.72	96	PV069789
12	Psilocybe	caerulescens	Caro V2	Psilocybe caerulescens	98.44	96	PV069790
13	Psilocybe	cubensis	A-OD-SQ	Psilocybe cubensis	99.73	96	PV069791
14	Psilocybe	cubensis	B +	Psilocybe cubensis	99.73	97	PV069792
15	Psilocybe	cubensis	BG	Psilocybe cubensis	98.92	96	PV069793
16	Psilocybe	cubensis	Blue Meanies	Psilocybe cubensis	99.46	97	PV069794
17	Psilocybe	cubensis	Coda	Psilocybe cubensis	99.46	96	PV069795
18	Psilocybe	cubensis	Golden Teacher	Psilocybe cubensis	99.73	97	PV069796
19	Psilocybe	cubensis	GW	Psilocybe cubensis	99.46	97	PV069797
20	Psilocybe	cubensis	Ishmael V1	Psilocybe cubensis	99.73	95	PV069798
21	Psilocybe	cubensis	Ishmael V2	Psilocybe cubensis	99.46	97	PV069799
22	Psilocybe	cubensis	KSSS (UFO)	Psilocybe cubensis	99.19	96	PV069800
23	Psilocybe	natalensis	LS	Psilocybe cubensis	98.51	93	PV069801
24	Psilocybe	cubensis	Mr.Peanut	Psilocybe cubensis	99.73	96	PV069802
25	Psilocybe	natalensis	Natal V1	Psilocybe cubensis	99.19	97	PV069803
26	Psilocybe	natalensis	Natal V2	Psilocybe cubensis	99.17	95	PV069804
27	Psilocybe	natalensis	Natal V3	Psilocybe cubensis	99.19	96	PV069805
28	Psilocybe	cubensis	PE	Psilocybe cubensis	99.73	97	PV069806
29	Psilocybe	cubensis	Rusty Whyte	Psilocybe cubensis	100.00	93	PV069807
30	Psilocybe	natalensis	Swedish	Psilocybe cubensis	98.92	97	PV069808
31	Psilocybe	natalensis x cubensis	Yellow umbo	Psilocybe cubensis	97.49	97	PV069809
32	Psilocybe	cubensis	Yonis	Psilocybe cubensis	100.00	94	PV069810
33	Psilocybe	hoogshagenii	Convexa (VIVA)	Psilocybe subtropicalis	99.72	97	PV069811
34	Psilocybe	tampanensis	A Strain	Psilocybe tampanensis	99.72	96	PV069812
35	Psilocybe	galindoi	ATL#7 V1	Psilocybe tampanensis	99.72	97	PV069813
36	Psilocybe	galindoi	ATL#7 V2	Psilocybe tampanensis	99.72	96	PV069814
37	Psilocybe	tampanensis	Beulah	Psilocybe tampanensis	99.71	81	PV069815
38	Psilocybe	tampanensis	Mexicana	Psilocybe tampanensis	99.44	96	PV069816
39	Psilocybe	tampanensis	Pollock	Psilocybe tampanensis	100.00	94	PV069817
40	Psilocybe	tampanensis	TampG V1	Psilocybe tampanensis	98.10	96	PV069818
41	Psilocybe	tampanensis	TampG V2	Psilocybe tampanensis	99.72	97	PV069819
42	Psilocybe	tampanensis	TampG V3	Psilocybe tampanensis	99.44	97	PV069820

Fig. 3.

Principle Component Analysis (PCA) of 42 psilocybin-producing fungi strains. The fruiting bodies of 42 fungi were extracted for 1.5 h at 1:20 (tissue:solvent) with 25:75 (H₂O:MeOH) solvent. Then, extracts were subjected to HPLC–MS/MS analyses (n = 3) and the metabolomic data are presented in the PCA score plot. Colors represent fungi specie according to morphological identification; shapes represent fungi specie according to ITS classification. X-axis represents the first principal component and y-axis represents an uncorrelated second highest variance component. Psilocybe cubensis BG, Panaeolus cyanescens Vietnam and Psilocybe tampanensis TampG are indicated.

In vitro dephosphorylation

A main process upon ingestion of psilocybin-producing fruiting bodies is the dephosphorylation of psilocybin by alkaline phosphatase and other nonspecific esterases in the gastrointestinal tract and the liver^69–71. Even though this dephosphorylation occurs in in vivo experimental settings in which fungal extracts are administered orally, most in vitro systems lack this complexity and psilocybin is not transformed into psilocin. To overcome this obstacle, we developed an in vitro dephosphorylation method based on the use of homogenates and purified preparation^28,72. We utilized purified Alkaline Phosphatase (AP) to mimic the dephosphorylation occurring by this enzyme in the body (Fig. 4A). The calculated reaction rate (Fig. 4B) to the complete and rapid turnover of psilocybin was found as 60 ng/min/U. We validated the method on an extract of Psilocybe cubensis BG (Fig. 4C). The original extract contained 2.8% w/w psilocybin and 0.7% w/w psilocin, and after in vitro dephosphorylation the percent of psilocybin dropped below LOQ, while that of psilocin increased to 3.5% w/w.

Fig. 4.

In vitro dephosphorylation of fungal extracts. (A) Dephosphorylation of psilocybin to psilocin by AP at 37 °C releases the phosphate group (blue) from the oxygen in position C-4 of the indolamine structure, resulting in its hydrolysis. (B) Enzyme kinetics graph displaying Michaelis–Menten curve for the conversion of psilocybin to psilocin (n = 3). (C) An extract of Psilocybe cubensis BG was diluted in AP buffer and added 1 µl of AP (20 U) working at a conversion rate of 1200 ng/ml psilocybin per 1 min. The reaction was halted by adding equal volume EtOH to the reaction mix after an amount of minutes corresponding to the quantified psilocybin concentration. Samples from the original extract (green) and the dephosphorylated one (blue) were quantified according to our HPLC–DAD method at 263 nm. The peaks of psilocybin and psilocin are indicated.

Discussion

Psilocybin-producing fungi present a wide range of diverse strains harboring a wealth of secondary metabolites with potential bioactivity. However, current research efforts focus mainly on psilocybin and psilocin, leading to the misconception that these fungi are a uniform group that produces similar bioactive compounds. To illustrate the diversity of metabolites within and between species, we analyzed the metabolomes of 42 distinct fungi strains (Table 1) and showed the unique metabolome of each strain (Fig. 3). This diversity in chemical profiles could offer valuable insights for future pharmaceutical efforts.

One of the major issues we faced going into this study was the lack of standardization due to the manifold of extraction and analytical methods. Moreover, most of these optimized the parameters only for very few strains of psilocybin-producing fungi. We addressed this by developing an approach for the extraction, identification, and quantification of their metabolome with an emphasis on tryptophan-derived indolamines (Fig. 1). The extraction method we outlined above was verified on different strains and species of psilocybin-producing fungi, thus ensuring its applicability across diverse strains. The optimized parameters were carefully determined to maximize the recovery of both major and minor indoleamines, as well as other lipophilic and hydrophilic metabolites, while minimizing their degradation. Furthermore, the resulting lyophilized powder is preferable to crude extracts, as it likely has higher purity and stability, more readily weighed, and easily soluble⁷³. This approach that preserves the integrity of the fungal metabolome highlights the significant variability in metabolite profiles observed among different species and strains.

Interestingly, our research revealed that the pH of the extraction solvent significantly affects the dephosphorylation of phosphorylated metabolites (Fig. 1F). Therefore, it is essential to note that the metabolite profile in the extract may not reflect the one in the fungi without this adjustment. This observation suggests the presence of fungal enzymes, possibly phosphatases, that remain active in 75% methanol. Moreover, the differential responses to pH among species hint at the existence of species-specific phosphatases or varying enzyme concentrations. These findings could have important implications for basic mycology and potential therapeutic applications. In accordance with that, we required an in vitro method that mimics the process of dephosphorylation that occurs in vivo, but not necessarily by cells in culture. The adjustment of metabolite research to in vitro setups is an additional step that can be done after extraction (Fig. 4).

The analytical methodologies (Fig. 2) achieved two main objectives: the high-throughput quantification of known metabolites according to analytical standards by HPLC–DAD, and the identification of new metabolites using HPLC–MS/MS, as shown for putative hydroxypsilocybin. The method was optimized to address the challenges of indole alkaloids, which have subtle chemical differences and share similar UV absorbance profiles. Achieving clear chromatographic separation was required to distinguish these metabolites without coupling to a mass spectrometer. We systematically evaluated a variety of C18 column types to identify the most effective for achieving hydrophilic separation and at the same time ensuring reproducibility, minimizing tailing effects and achieving consistent peak resolution across diverse fungal extracts.

Our analyses of Psilocybe cubensis BG, Panaeolus cyanescens Vietnam, and Psilocybe tampanensis TampG revealed significant chemical profile differences. Notably, Psilocybe cubensis BG exhibited the highest psilocybin content, while Panaeolus cyanescens Vietnam had the highest psilocin levels. Psilocybe tampanensis TampG showed equal amounts of both compounds. The UV chromatogram indicated that the two Psilocybe species shared more similarities with each other than with the Panaeolus species, which contained higher amounts of minor metabolites and numerous unidentified peaks. This points to differences not only in the quantities of secondary metabolites among psilocybin-producing fungi but also in their overall composition. For example, in Psilocybe cubensis Golden Teacher we identified the putative hydroxypsilocybin. This psilocybin derivative serves as an example to the utility of the approach in metabolite discovery, demonstrating its potential to identify novel compounds. Employing this approach can allow the creation of a metabolomic library that compares hundreds of different psilocybin-producing fungi, as was previously done for other notable therapeutic substances^74–76.

The scientific community addresses psilocybin-producing fungi collectively as “magic mushrooms”, ignoring species-specific bountiful and complex metabolomes. Our research aimed to extract and analyze the complete range of metabolites found in psilocybin-producing fungi, not just psilocybin or psilocin, to provide a comprehensive understanding of fungal chemical composition. Relying only on isolated compounds or their synthetic counterparts overlooks the potential synergistic effects present in the entire metabolome, a phenomenon known as polypharmacology⁷⁷. When the 42 fungi metabolomes were visualized as PCA, it revealed the high variability between psilocybin-producing strains. Therefore, they cannot be addressed as a cohesive group, rather, each is a producer of unique metabolites that may display distinct biochemical and pharmaceutical activities.

Conclusion

Psilocybin-producing fungi encompass a wide array of strains rich in secondary metabolites with potential therapeutic applications. In this study, we described a simple validated approach for the extraction, identification, and quantification of their metabolome. The resulting final product is a powder-extract suitable for in vitro and in vivo assays. Our approach revealed significant variability among psilocybin-producing strains. Hence, each of these has a distinct metabolome with unique pharmaceutical potential, and must not be referred to as a uniform therapeutic source.

Experimental section

Chemicals and reagents

HPLC grade water (H₂O), MeOH, and acetonitrile (MeCN) were obtained from Bio-Lab Chemicals (Jerusalem, Israel); Trifluoroacetic acid (TFA) was purchased from Sigma-Aldrich (Rehovot, Israel); and used for sample extraction and HPLC analysis. HPLC–MS-grade H₂O, MeOH, ammonium acetate and Acetic acid (AA) were purchased from Merck (Darmstadt, Germany) and used for HPLC–MS/MS analysis. AP 20 U/µL (11,097,075,001) was purchased from Sigma-Aldrich (Rehovot, Israel).

Indoleamine metabolite standards

The analytical standards (> 98%) Psilocybin (4-PO-DMT; 15,695), Psilocin (4-hydroxy DMT; 15,696), Norpsilocin (4-HO-T; 23,586), Baeocystin (4-PO-MT; 31,583), Norbaeocystin (4-PO-T; 31,584) and Aeruginascin (4-PO-Me-DMT; 31,582) were purchased from Cayman Chemical Company, L-Tryptophan (Trp; 51,145), and Tryptamine (T;76,706) were purchased from Sigma-Aldrich. All standards were diluted in MeOH, and the standard mix was kept in amber vials at − 20 °C.

Cultivation and preparation of fruiting bodies

Fungi were identified according to Stamets 1996⁴² taking into account color, cap size, stipe, gills and veil; and cultivated by PsygaBio (Caesarea, Israel) in 2 L ‘shoeboxes’ on a substrate mix based on coconut coir or horse manure as the major nutritional ingredient, at a temperature of 20–25 °C, high humidity environment (80–85%), and a light/dark cycle of 12 h. The fruiting bodies were harvested and immediately dried for 24 h at 40 °C in an industrial food dehydrator (Foshan Dalle Technology, China), followed by 2 h at 40 °C in a vacuum oven (Holland Green Science, Netherlands) to a ≤ 10% moisture content according to Precisa’s moisture analyzer (Precisa Precision Balances, Switzerland). Dried fruiting bodies were kept in vacuum bags in the dark at 4 °C. Samples (20 g) were ground to homogenous fine powder using an electric grinder (Morphy Richards, UK).

Ribosomal DNA-based specie identification

The DNA from dried and ground fruiting bodies (1 g) was extracted with Cetyltrimethylammonium bromide (CTAB)⁷⁸ lysis buffer using a modified protocol⁷⁹, where the incubation at 65°C lasted 30 min and an additional cleanup step with AMPure XP Beads (Beckman Coulter, IN, USA). The primers were as follows⁸⁰: ITS1 (fwd)—5′ TCCGTAGGTGAACCTGCGG 3′; ITS2 (rev)—5′ GCTGCGTTCTTCATCGATGC 3′; ITS3 (fwd)—5′ GCATCGATGAAGAACGCAGC 3′; and ITS4 (rev) 5′ TCCTCCGCTTATTGATATGC 3′. The resulting DNA samples were quantified using Qubit 4 fluorometer (Invitrogen; MA, USA) and the ITS region of the ribosomal DNA^80,81 was amplified and sequenced by Hylabs (Rehovot, Israel). Sequences were aligned using nblast (NCBI) against the nonredundant standard database to determine the ITS classification species. The full list of fungi is presented in Table 1.

Sample extraction

A representative of each cluster, Psilocybe cubensis BG, Panaeolus cyanescens Vietnam and Psilocybe tampanensis TampG, was selected to demonstrate the effects of different parameters in the optimal extraction assessment due to their confirmed distinct ITS classification. Exactly 1 g from each ground sample were added 50 mL HPLC-grade solvent solution of H₂O (pH = 6.8) and MeOH at ratios of 100:0, 50:50, 25:75 and 0:100, respectively. In some of the extractions, the H₂O was buffered by ammonium acetate 0.1M to a final pH = 9, as indicated. Extraction was performed for 0.5–24 h, in amber vials. Samples were extracted by agitation on an orbital shaker (200 rpm) at 25 °C, then centrifuged at 16.4 g for 15 min. For the validation, the pellet was re-extracted two additional times. The supernatant was separated and filtered under vacuum through Whatman filter paper number 4 (Whatman plc, UK) using a Buchner funnel and the MeOH was evaporated under reduced pressure at 45 °C using a rotary evaporator (Laborata 4000; Heidolph Instruments & Co. KG, Germany). Then, samples were frozen at − 80 °C followed by lyophilization (Labconco; Missouri, USA) to complete dryness and stored in the dark at 4 °C. The extraction using 1:20 tissue:solvent ratio, 25:75 H₂O:MeOH (pH = 9), for 1.5 h, was performed on additional 39 genetically stable isolates from 9 species (Table 1).

HPLC–DAD separation and quantification

Absolute quantifications of Psilocybin, Psilocin, Norpsilocin, Baeocystin, Norbaeocystin, Aeruginascin, L-Tryptophan and Tryptamine were performed with their respective standards using UltiMate 3000 HPLC with photodiode array detector (Thermo Fisher Scientific). Ten-point calibration curves were prepared in the range of approximately 0.01 to 500 µg/ml. The calibration curves were linear for all the analytes (R² > 0.99). The limit of quantification was 0.01 µg/ml. Precision was predetermined according to maximum relative standard deviations (RSDs) < 15%. Lyophilized extracts were dissolved in MeOH:H₂O (95:5) to reach a concentration of 20 mg/ml. The extract was filtered through 0.22 µm PTFE syringe filter (Thermo Fisher Scientific) and loaded into amber injection vials. The chromatographic separation was achieved using a ZORBAX Eclipse XDB-C18 column (3.5 μm, 4.6 × 150 mm i.d., Agilent), with a Rapid Resolution HT guard column (cartridge 3.5 μm, 4.6 × 15 mm, Agilent), and a mobile phase gradient (Line A: 0.1% TFA in water, Line B: MeOH, and Line C (wash solution): MeCN:H₂O (9:1 v/v), all solvents were of HPLC grade). The gradient program was established as follows: initial conditions were 98% A for 5 min, lowered to 50% A until 20 min, then raised to 98% A until 24 min. Between batch injections, line C was used to wash the column for 1 h. A flow rate of 1 ml/min was used, the column temperature was 40 °C ± 1 °C and the autosampler temperature was 4 °C ± 2 °C. The injection volume was 1 μL. Data acquisition was performed at 263 nm and 200–400 nm photodiode array.

HPLC–MS/MS metabolome discovery and analyses

Metabolome analyses were performed by HPLC-HRMS/MS (UltiMate 3000) coupled with Q Exactive™ Focus Hybrid Quadrupole-Orbitrap MS (Thermo Fisher Scientific, Bremen, Germany) using a high-resolution system in positive and negative ionization modes. Full MS acquisition was conducted over 50–1,000 m/z at a resolution of 70,000 with a 1.6 m/z isolation window including a second stage of tandem mass spectrometry in data-dependent MS/MS mode. Each extract was prepared in triplicates at a concentration of 1 mg/ml. The chromatographic separation was achieved using a ZORBAX Eclipse XDB-C18 column (1.8 μm, 2.1 × 150 mm i.d., Agilent), with a Rapid Resolution HT guard column (cartridge 1.8 μm, 2.1 × 5 mm, Agilent), and a mobile phase gradient (Line A: 0.1% AA in water, Line B: MeOH, all solvents were of LC–MS grade). The gradient program was as described for HPLC–DAD.

In vitro dephosphorylation

The enzymatic activity rate of purified bovine intestine AP (11,097,075,001; Roche, Switzerland) was calibrated by adding 1 U to 6 U to the psilocybin standard or to fruiting bodies extracts. These were diluted to psilocybin concentration range 35–2200 ng/ml. The reaction at 37 °C was halted after 10 min by adding 1:1 EtOH. The concentrations of psilocybin and psilocin in the dephosphorylated samples were quantified by HPLC–DAD, and the reaction rate was calculated according to the Michaelis–Menten equation⁸². An extract of Psilocybe cubensis BG was diluted in AP buffer to 30 µg/ml of psilocybin per reaction, 1 ml of the diluted sample was added 1 µl of AP (20 U) at 37 °C for 25 min, then psilocybin and psilocin were quantified with HPLC–DAD.

Data analyses

Data were analyzed with GraphPad Prism 8.0 (GraphPad Software Inc., San Diego, CA, USA). PCA was performed using Orange Data Mining 3.36.2 (University of Ljubljana, Slovenia) to explore clustering patterns in the dataset. Chemical structures were produced using ChemDraw 22.2.0 (PerkinElmer Informatics, Inc., Waltham, MA, USA). The processing of HPLC–DAD data was performed with Chromeleon 7 (Thermo Scientific, Bremen. Germany) and that of HPLC–MS/MS data with TraceFinder software v. 5.0 (Thermo Scientific, Bremen. Germany) and FreeStyle 1.8 SP2 (Thermo Scientific, Bremen. Germany).

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

AA

Acetic acid

AP

Alkaline phosphatase

CTAB

Cetyltrimethylammonium bromide

HPLC

High-performance liquid chromatography

HPLC–DAD

High-performance liquid chromatography coupled with a photodiode array detector

HPLC-HRMS/MS

High-performance liquid chromatography coupled with high-resolution tandem mass spectrometry

ITS

Internal transcribed spacer

LC–MS

Liquid chromatography mass spectrometry

LOQ

Limit of quantification

MeCN

Acetonitrile

MeOH

Methanol

MS/MS

Tandem mass spectrometry

PCA

Principal Component Analysis

Pcu.BG

Psilocybe cubensis BG

Pcy.V

Panaeolus cyanescens Vietnam

PT.TampG

Psilocybe tampanensis TampG

RT

Retention time

RSDs

Relative standard deviations

TFA

Trifluoroacetic acid

Author contributions

JC and DM, Conceptualization and design; LM and IT, Methodology; JC, Acquisition of data; JC, SP, LS, YL, AS and DM, Analysis and interpretation of data; JC, SP and YL Figure preparation; JC, SP and DM Writing, review, and/or revision of the manuscript; DM, Study supervision.

Funding

This study was supported by grant 2031221 from PsygaBio.

Data availability

The datasets generated during the current study are stored on the Open Science Framework and available online at http://osf.io/jek2u. The sequencing data that support the findings of this study have been deposited in GenBank PV069779-PV069820.

Declarations

Competing interests

D.M. is a co-founder of Cannasoul Analytics, which is a shareholder in PsygaBio. Y.L., L.M. and I.T. are employees of PsygaBio. The rest of the authors have declared that no conflict of interest exists.

Consent for publication

All authors give their consent to publish this manuscript.