Comparative metabolic profiling of wild Cordyceps species and their substituents by liquid chromatography-tandem mass spectrometry

Cordyceps is a genus of ascomycete fungi and used widely in fungal drugs. However, in-depth studies of the metabolites of wild Cordyceps species and their substituents are lacking. In this study, a liquid chromatography-tandem mass spectrometry (LC-MS/MS)-based metabolomics analysis was carried out to comprehensively profile the metabolites in wild Chinese Cordyceps species (Ophiocordyceps sinensis (Berk.) G.H. Sung, J.M. Sung, Hywel-Jones and Spatafora 2007) from Naqu (NCs) and Yushu (YCs) and their substituents including artificially cultivated Cordyceps species (CCs) and mycelia. A total of 901 metabolites were identified in these samples, including lipids, amino acids, nucleosides, carbohydrates, organic acids, coenzymes, vitamins, alkaloids and their derivatives. Univariate and multivariate statistical analyses revealed remarkable differences and significantly different metabolites among them. Seventy amino acid-relevant metabolites were analyzed quantitatively in four samples for the first time. The four samples contained abundant L-glutamic acid and oxidized glutathione as well as multiple unique amino acid-relevant metabolites (e.g., 3-chloro-L-tyrosine, 6-aminocaproic acid, L-theanine, anserine, γ-glutamyl-cysteine). Collectively, our study provides rich metabolic information of wild Cordyceps species and their substituents, which could facilitate their quality control and optimal utilization.

However, Cordyceps species are rarely found in the wild, and their substituents are desired (Li et al., 2019). The fermented mycelia of wild Cordyceps species are popular, and have been made into medical and functional products in the forms of capsules, tablets, or granulations (Dong et al., 2015;Zhao et al., 2020). Besides, Sunshine Lake Pharma (Dongguan, China) has undertaken artificial cultivation of Chinese Cordyceps species. Ophiocordyceps sinensis (Berk.) G.H. Sung, J.M. Sung, Hywel-Jones & Spatafora 2007 (known previously as Cordyceps sinensis) is a fungus that grows on insects in the family Ophiocordycipitaceae (Li et al., 2019). O. sinensis is a form of wild Cordyceps species. The annual yield of cultured Cordyceps species (CCs) reaches tens of tons, accounting for ≥20% of the total natural resource (Li et al., 2019). However, whether wild Cordyceps species (e.g., O. sinensis) and such substituents have chemical homogeneity is controversial.
Metabolomics analysis has become a powerful approach for large-scale study of metabolites in biological samples (Patti et al., 2012). In general, two analytical methods are used to conduct metabolomics analysis: nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS) (Perez et al., 2021). Compared with MS, NMR provides unique structural information. Nonetheless, it has limited sensitivity when measuring a large number of metabolites with a wide dynamic range Perez et al., 2021). With regard to MSbased metabolic profiling, gas chromatography (GC) and liquid chromatography (LC) are hyphenated to MS for more accurate identification of metabolites. GC-MS often requires further derivatization of metabolites, and makes them more volatile and more thermally stable (Smart et al., 2010;Zhang et al., 2020). Nevertheless, LC-MS can detect thermally labile metabolites directly without derivatization, and is used more widely for metabolite profiling. LC-MS-based metabolomics analysis has been applied to analyze a certain class of metabolites (e.g., lipid, nucleoside, nucleobase, nucleotide, protein) (Fan et al., 2006;Mi et al., 2016;Cheng et al., 2017;Zhang et al., 2018;Jia et al., 2019;Lin et al., 2022) or dozens of metabolites Zhang et al., 2015;Chen et al., 2018;He et al., 2019;Yao et al., 2019) in wild Cordyceps species and/or their substituents. However, few studies have profiled hundreds of metabolites in wild Cordyceps species or cultivated Cordyceps species and their mycelia for more comprehensive assessment of their chemical composition and alterations.
In this study, we leveraged liquid chromatographyelectrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS) to comprehensively survey metabolites in wild Cordyceps species (O. sinensis) from Naqu (Tibet, China; NCs) and Yushu (Qinghai, China; YCs), CCs (O. sinensis), and mycelia (Hirsutella sinensis X.J. Liu, Y.L. Guo, Y.X. Yu & W. Zeng 1989) from Bailing capsules (BLs). Using univariate and multivariate statistical analyses, chemical differences among NCs, YCs, CCs, and BLs were evaluated. Furthermore, differential metabolites among four samples were screened out and enrichment analysis using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database was undertaken. Besides, 70 amino acid-relevant metabolites in these four samples were analyzed quantitatively. Taken together, this work provides a holistic insight into the metabolic differences in wild Cordyceps species and their substituents. Liquid chromatography-tandem mass spectrometry for metabolic profiling of wild and cultivated species of Cordyceps and mycelia

Materials and reagents
To extract metabolites, 50 mg of each sample was first weighed and homogenized with steel balls at 30 Hz for 3 min. Then, 1 ml of a methanol:water mixture (7:3) was added, and the mixture was vortex-mixed for 5 min followed by centrifugation (12,000 rpm, 10 min, 4°C). Subsequently, the supernatant Frontiers in Pharmacology frontiersin.org 02 (400 μl) was collected and stored at −20°C overnight. Finally, the supernatant was centrifuged again (12,000 rpm, 3 min, 4°C), and 200 μl of supernatant pipetted for LC-MS.
A Triple Quad ™ 6500+ MS (Sciex, Framingham, MA, United States) instrument was operated in electrospray ionization in positive (ESI+) and negative (ESI−) modes with the following parameters: positive ion spray voltage, 5500 V; negative ion spray voltage, −4500 V; source temperature, 500°C; ion source gas 1, 50 psi; ion source gas 2, 50 psi; curtain gas, 25 psi; collision gas, high. A specific set of multiple reaction monitoring (MRM) transitions was recorded for each period according to the metabolites eluted within this period.

Liquid chromatography-tandem mass spectrometry for quantifying amino acidrelevant metabolites
Similarly, 50 mg of each sample was weighed and homogenized with steel balls at 30 Hz for 3 min. Then, 500 µl of a methanol:water (7:3) mixture containing [2H2]L-threonine was added. The mixture was vortex-mixed for 5 min followed by centrifugation (12,000 rpm, 10 min, 4°C). Subsequently, the supernatant (300 μl) was collected and stored at −20°C for 30 min. Finally, the supernatant was centrifuged again (12,000 rpm, 10 min, 4°C) and 200 μl of the supernatant was pipetted for LC-MS.
Acquisition of MS data was undertaken on a Triple Quad ™ 6500+ MS system (Sciex) in MRM mode with the following parameters: positive ion spray voltage, 5500 V; negative ion spray voltage, −4500 V; source temperature, 550°C; ion source gas 1, 50 psi; ion source gas 2, 50 psi; curtain gas, 35 psi; collision gas, medium.

Statistical analyses
Principal component analysis (PCA) and orthogonal partial least squares-discriminant analysis (OPLS-DA) were conducted in R (R Institute for Statistical Computing, Vienna, Austria) using the statistics function "prcomp" and "MetaboAnalystR" package, respectively. Hierarchical clustering analysis (HCA) and the Pearson correlation coefficient were carried out by the R package "ComplexHeatmap."

Results and discussion
Liquid chromatography-tandem mass spectrometry-based wide-targeted metabolomics analysis for the characterization of metabolites in wild Cordyceps species and their substituents LC-MS-based strategies for metabolomics analysis can be targeted or untargeted (Patti et al., 2012). Untargeted metabolomics analysis is dependent upon high-resolution MS to measure metabolites without bias. Typically, targeted metabolomics analysis uses MRM on MS/MS to analyze a priori defined metabolites with high sensitivity and quantitative reproducibility. In this work, a LC-MS/MS-based wide-targeted metabolomics analysis was employed to comprehensively identify metabolites and assess their alterations in wild Chinese Cordyceps species (O. sinensis CCs, and BLs. Table 1 shows that the chemical constituents of samples included lipids, amino acids, nucleotides, carbohydrates, organic acids, vitamins, and their derivatives or other small molecules. Notably, more lipids, amino acids, organic acids, and their derivatives were detected. Specifically, in ESI + mode, 453, 466, 453, and 452 metabolites were detected in NCs, YCs, CCs, and BLs, respectively. In ESI− mode, 389, 385, 379, and 382 metabolites were detected in NCs, YCs, CCs, and BLs, respectively ( Figure 1). Detailed information of metabolites detected in NCs, YCs, CCs, and BLs is displayed in Supplementary Table S2. A Venn diagram highlighted that 753 out of 901 metabolites were detected in these four samples ( Figure 1B). Importantly, some widely used bioactive constituents or biomarkers were identified in all samples, such as nucleosides (e.g., cordycepin, adenosine), ergosterol, D-mannitol, and amino acids (e.g., tryptophan, glutamic acid) Yue et al., 2013;Zhao et al., 2014;Liu et al., 2015). Besides, 1, 2, 2, and 29 metabolites were only detected in NCs, YCs, CCs, and BLs, respectively ( Figure 1B; Table 2). Collectively, we detected a wide variety of metabolites with large amounts in wild and cultivated Cordyceps species and mycelia, which facilitated comprehensive assessment of their differences.

Multivariate statistical analyses for assessment of differences between wild Cordyceps species and their substituents
To characterize differences between wild Cordyceps species and their substituents comprehensively, multivariate statistical analyses (PCA, OPLS-DA, clustering analysis, correlation analysis) were employed. Metabolites measured in QC samples with relative standard deviation <30% were used for subsequent analyses (Supplementary Table S2). Results of PCA ( Figure 2A) and OPLS-DA ( Figures 2B-F) clearly showed distinct sample clusters corresponding to NCs, YCs, CCs, and BLs, respectively. Besides, clustering analyses ( Figure 2G) and correlation analyses ( Figure 2H) indicated differential metabolites among NCs, YCs, CCs, and BLs. Notably, the differences between CCs and wild Cordyceps species Hormones and hormone related compounds 4 Others 11 Frontiers in Pharmacology frontiersin.org (NCs and YCs) were less than those between BLs and wild Cordyceps species. The similarity between CCs and NCs was higher than that between CCs and YCs ( Figure 2H). Taken together, these results suggested that metabolic differences were present among NCs, YCs, CCs, and BLs.

Differential metabolites between wild and cultivated species of Cordyceps
Similarly, 356 significant differential metabolites (95 were up-regulated and 261 were down-regulated) were found between CCs and NCs, and 452 remarkably altered metabolites (147 were up-regulated and 305 were down-regulated) were found between CCs and YCs (VIP ≥1 and FC ≥ 2 or ≤0.5) (Figures 4A,B). Among them, there were 258 overlapped differential metabolites (including lipids, amino acids, nucleotides) between CCs and wild Cordyceps species ( Figure 4C) and they are emboldened in Supplementary Tables S4, S5. Of note, the content of ergosterol was lower in CCs than in both types of wild Cordyceps species. Besides, six sugar acids (D-glucoronic acid, L-gulonolactone, D-glucarate, L-arabinonic acid-1,4-lactone, mucic acid, gluconic acid), and two sugar derivatives (N-acetyl-Dglucosamine, methyl beta-D-galactopyranoside) detected in wild and cultivated species of Cordyceps displayed no differences. Moreover, enrichment analyses using the KEGG database clearly showed that more significant differential metabolites between CCs and wild Cordyceps species were associated mainly with purine metabolism (Figures 4D,E). These results showed that CCs and wild Cordyceps species had different chemical constituents. On the one hand, this finding will provide promising indicators to discriminate artificially cultivated Cordyceps species from wild Cordyceps species . On the other hand, this finding may encourage study of whether these differential metabolites between CCs and wild Cordyceps species are relevant to differences in their pharmacology.

Differential metabolites between bailing capsules and wild Cordyceps species
Compared with NCs, 625 metabolites were altered significantly in BLs (VIP ≥1; FC ≥ 2 or ≤0.5; 334 were up-regulated; 291 were down-regulated) ( Figure 5A; Supplementary Table S6). Besides, 598 significantly different metabolites (324 were up-regulated and 274 were down-regulated) were found between BLs and YCs ( Figure 5B; Supplementary Table S7). Figure 5C shows 522 overlapped differential metabolites between BLs and wild Cordyceps species, and they are emboldened in Supplementary Tables S6, S7. There were more differential metabolites (including lipids, amino acids, nucleotides) between BLs and wild Cordyceps species (Figures 5A,B) than those between CCs and wild Cordyceps species (Figures 4A,B). For the well-known bioactive constituents, the cordycepin content was higher in BLs than in wild Cordyceps species, whereas the ergosterol content was lower in BLs. Likewise, these differential metabolites between BLs and wild Cordyceps species could facilitate discrimination of adulteration in the processed products of wild Cordyceps species (Zhang et al., 2020). Besides, enrichment analyses using the KEGG database showed that differential metabolites between BLs and NCs and between BLs and YCs were located to different pathways ( Figures  5D,E), which indicated that wild Cordyceps species from different habitats had metabolomic differences.

Quantitative analyses of amino acidrelevant metabolites in wild Cordyceps species and their substituents
Amino acids are important chemical and bioactive constituents in Cordyceps species (Zhang et al., 1991;Liu et al., 2015;Zhang et al., 2022) but have not been   Differential analyses between NCs and YCs. (A) Volcano plot of differential metabolites between NCs and YCs. Differential metabolites were screened based on VIP ≥1 and FC ≥ 2 or ≤0.5. The fold change was the ratio of NCs to YCs. (B) Enrichment analysis of differential metabolites between NCs and YCs using the KEGG database. The top-20 enriched metabolic pathways (with p-values) are displayed.

Cordyceps samples Unique metabolites Subclasses
Frontiers in Pharmacology frontiersin.org 07 characterized comprehensively previously. Here, 70 amino acids and their relevant metabolites in four samples were analyzed quantitatively. Figures 6A,B show the quantitative results of some amino acids and their derivatives. Figure 6C displays the quantitative results of 11 small peptides. Compared with widely targeted metabolic analyses (Supplementary Figure S2), most of these amino acid-relevant metabolites detected by targeted quantitative analyses showed consistent trends in alteration across four samples ( Figure 6). However, 11 amino acids and their derivatives were not detected by widely targeted metabolic analysis (nicotinuric acid, 5-hydroxy-tryptophan, homoserine, 3-iodo-L-tyrosine, 3-chloro-L-tyrosine, N-acetylaspartate, L-leucine, L-theanine, argininosuccinic acid, 2-aminobutyric acid, 4-acetamidobutyric acid) (Supplementary Figures S2A,B). Targeted quantitative analyses showed such undetected metabolites with relatively low abundance ( Figures 6A,B). Additionally, absolute quantitative results showed that L-glutamic acid and oxidized glutathione had a high level in wild and cultivated species of Cordyceps and mycelia ( Figures 6B,C). Notably, the content of L-glutamine and L-ornithine was significantly higher in wild and cultivated species of Cordyceps than that in BLs ( Figure 6A). Furthermore, some unique amino acid-relevant metabolites were noted: 1) 3-chloro-L-tyrosine was not Differential analyses between CCs and wild Cordyceps species. (A) Volcano plot of differential metabolites between CCs and NCs. Differential metabolites were screened based on VIP ≥1 and FC ≥ 2 or ≤0.5. The fold change was the ratio of CCs to NCs. (B) Volcano plot of differential metabolites between CCs and YCs. Differential metabolites were screened based on VIP ≥1 and FC ≥ 2 or ≤0.5. The fold change was the ratio of CCs to YCs. (C) Venn diagram of differential metabolites between CCs and wild Cordyceps species. (D) Enrichment analysis of differential metabolites between CCs and NCs using the KEGG database. The top-20 KEGG enriched metabolic pathways (with p-values) are displayed. (E) Enrichment analysis of differential metabolites between CCs and YCs using the KEGG database. The top-20 KEGG enriched metabolic pathways (with p-values) are displayed.
Frontiers in Pharmacology frontiersin.org 08 detected in CCs and BLs, but was present in wild Cordyceps species ( Figure 6A); 2) 6-aminocaproic acid was not detected in YCs but was present in NCs, CCs, and BLs ( Figure 6B); 3) L-theanine was not detected in wild and cultivated species of Cordyceps but was present in BLs ( Figure 6B); 4) anserine and γ-glutamyl-cysteine were not detected in CCs and BLs, respectively ( Figure 6C).

Conclusion
In summary, this work performed a comparative metabolic profiling to comprehensively characterize metabolites and assess their alterations in wild Cordyceps species (O. sinensis and their substituents. LC-MS/MS-based widely targeted approach measured 901 metabolites in Cordyceps samples, including lipids, amino acids, nucleosides, carbohydrates, organic acids, coenzymes, vitamins, alkaloids and their derivatives. Univariate and multivariate statistical analyses revealed metabolic differences among wild Cordyceps species from different habitats, cultivated Cordyceps species, and mycelia, and covered all the detected metabolite classes. Enrichment analyses using the KEGG database clearly showed differential metabolic pathways among four samples. Importantly, some amino acid-relevant metabolites were found to be unique to wild Cordyceps species (e.g., 3-chloro-L-tyrosine) or their substituents (e.g., L-theanine). These differences revealed among wild and cultivated Cordyceps Differential analyses between BLs and wild Cordyceps species (A) Volcano plot of differential metabolites between BLs and NCs. Differential metabolites were screened based on VIP ≥1 and FC ≥ 2 or ≤0.5. The fold change was the ratio of BLs to NCs. (B) Volcano plot of differential metabolites between BLs and YCs. Differential metabolites were screened based on VIP ≥1 and FC ≥ 2 or ≤0.5. The fold change was the ratio of BLs to NCs. (C) Venn diagram of differential metabolites between BLs and wild Cordyceps species. (D) Enrichment analysis of differential metabolites between BLs and NCs using the KEGG database. The top-20 enriched pathways (with p-values) are displayed. (E) Enrichment analysis of differential metabolites between BLs and YCs using the KEGG database. The top-20 enriched pathways (with p-values) are displayed.
Frontiers in Pharmacology frontiersin.org

FIGURE 6
Absolute quantitative results of some amino acids and their derivatives (A,B) and small peptides (C) in wild and cultivated Cordyceps species and BLs.
Frontiers in Pharmacology frontiersin.org 10 species and mycelia could facilitate rational utilization and better QC.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.