Transcriptome analysis of Tetrahymena thermophila response to exposure with dihydroartemisinin

Abstract

Dihydroartemisinin (DHA) is a derivative of artemisinin and is toxic to parasites. We used the Tetrahymena thermophila (T. thermophila) as a model to explore DHA toxicity. Results showed that low concentration of DHA (20 μmol/L) promoted cell proliferation, whereas high concentrations of DHA (40–1280 μmol/L) inhibited that. Appearance of nucleus was pycnosis by laser scanning confocal microscope. DHA significantly elevated activities of SOD and GSH-Px (P < 0.01) and MDA was markedly increased at high level but decreased at low level (P < 0.01). Further results of transcriptome in T. thermophila treated with different concentration DHA group (0, 20, 160 μmol/L) showed that differentially expressed genes (DEGs) were involved in oxidation-reduction and metabolism of exogenous substances indicated oxidative stress stimulation. Kyoto Encyclopedia of Genes and Genomes showed that DEGs were involved in the cytochrome P450-mediated metabolism of exogenous substances, glutathione metabolism and ABC transport. Remarkably, DNA replication was significantly enriched in low concentration DHA, energy metabolism related pathways and necrotic process were considerably enriched in high concentration DHA. The results of RT-qPCR of 13 DEGs were the same as that of transcriptome, in which the expression of GST and GPx family genes were significantly altered after exposed to high-DHA group. DHA induced oxidative stress damage through disturbing with energy. However, detoxification pathways in T. thermophila to resist oxidative damage and cell alleviated low concentration DHA stress by regulating antioxidant enzyme. This study provides good practice on pharmacological mechanism of artemisinin-based drugs in antiparasitic.

1. Introduction

Artemisinin (ART), a sesquiterpene lactone with an endoperoxide group [1,2], is extracted from the plant Artemisia annua L. found in China. Dihydroartemisinin (DHA) is a main derivative synthesized by introducing a hydroxyl group into the structure of ART [3]. ART and a series of its derivatives, such as DHA, artemether, and arteether, have been reported to be anti-malarial drugs in the first-line treatment for many years [[4], [5], [6]]. Artemisinin derivatives mainly worked on the membrane structure of the malaria parasite, and induced cytotoxicity by producing reactive oxygen species (ROS), which can cause depolarization of mitochondria and plasma membrane [7]. It was also found that DHA contained an endoperoxide group which was critical to the generation of free carbon-centred intermediates that could kill malarial parasite via oxidative stress [8]. Apart from malarial, artemisinins had broad-spectrum inhibitory effects on pathogenic parasites, such as protozoan parasites schistosomiasis [9]. Establishment of the pharmacodynamic screening model take an important role on identifing the pharmacological mechanism of artemisinins action on parasites.

Tetrahymena thermophila, a free-living ciliate in the freshwater around the globe, is an ideal model organism for research on exogenous chemicals such as drugs [10] and inorganic substances [11]. Furthermore, many known gene families were identified in T. thermophila, which related with environmental stresses. Studies have found that it is the presence of 44 genes encoding functional P450 genes [12], 165 genes encoding ATP-binding cassette (ABC) transporter [13], and 70 genes encoding putative GST [14] in T. thermophila. It was inferred that evolution alteration of T. thermophila exposed to diverse xenobiotics to rapidly respond to stress.

The classification status of T. thermophila is similar to that of ciliated parasites such as fish melon worm and Trichodinids, which presents the opportunity to consider non-parasitic protists T. thermophila as a great substitute for parasitic protists. To determination of the thermal dynamic mechanism of artemisinin-based drugs, Shen explored the influence of artemisinin on T. thermophila bio-thermal activity as well as its activity of bio-thermal characteristics by microcalorimetry [15]. Our previous studies discovered that DHA had toxic effect on T. thermophila and antioxidant system work to survive against DHA exposure [16]. With the rapid development of high-throughput sequencing technology, omics analyses have been widely used in the mechanism of drugs treatment on parasites disease in recent years [[17], [18], [19]]. It has a significant meaning in pharmacodynamic screening model based on high throughput transcriptome sequencing technology as well as to actively explore into the pharmacological mechanism of artemisinin-based drugs.

However, little is known about the biological processes of artemisinin compounds on T. thermophila from molecular function. This paper aimed to detect the toxic effects of DHA on T. thermophila and assess its growth inhibitory, phenotypic observation, oxidative stress and gene roles. Our study provides a new way for the exploration of drugs with anti-parasites mechanisms.

2. Materials and methods

2.1. Strains and cultural conditions

T. thermophila strain B2086.2 was provided by Cornell University. T. thermophila cells were grown in Neff medium (weight/vol: 0.25% peptone, 0.5% glucose, 0.25% yeast extract, and 0.9 g/dL FeCl₃ added by 0.1% volume ratio) at 30 °C for 48 h with shaking 150 rpm. The cells in the logarithmic phase were taken to carry out the subsequent analysis.

2.2. Evaluation of the half inhibitor concentration value (IC₅₀)

T. thermophila were cultured in conical flask containing 10 mL medium and DHA was treated with 0, 10, 20, 40, 80, 160, 320, 640, 1280 μmol/L. Only medium was added as blank group, 0.1% DMSO was added as negative control group. Cells with each group were inoculated to a final concentration of 10⁴ cells/mL and incubated for 48 h, respectively. After treatment, cell viability was detected with CCK-8 kit and calculated by the following formula.

cell viability (%) = (A₂ − A₁)/(A₂−A₀) × 100%

where A₁, A₂, A₀ are the absorbance of the experiment, DMSO negative control group and blank group, respectively. The IC₅₀ was calculated by SPSS 19. Each treatment was repeated three times. The results represented the average of three independent experiments.

2.3. Morphologic observation in T. thermophila

T. thermophila was exposed on DHA at different concentrations (0, 20, 40, 80, 160, 320 μmol/L) with the above culture conditions. Then, 10 mL of cell suspension was centrifuged at 367 rpm for 3 min at 4 °C and the supernatant was washed with phosphate buffered saline (PBS) for three times. After fixation with 2% paraformaldehyde (PFA), cells were stained with 200 μL mixed dye (2.5 mL PBS, 2.5 μL PI, 12.5 μL FITC and 1.5 μL DAPI) for 30 min at 37 °C in the dark. To observation from different angles, cell morphology with three colors was identified under high-times optic microscope, cell count with DAPI was identified under low-times optic microscope. Made slides were observed by laser confocal microscope (Leica TCS SP8, Germany).

2.4. Measurement of antioxidant indices

Based on the IC₅₀, 20 μmol/L DHA (low concentration group, LG) and 160 μmol/L DHA (high concentration group, HG) were identified as the action concentration in the subsequent experiment, which was lower and higher than IC₅₀, respectively. 0.1% DMSO was added to 5 mL of medium containing T. Thermophila as control group. Similar to the above procedure for incubation, cells were lysed with RIPA Lysis Buffer (Solarbio, China), followed by centrifugation at 8500 rpm for 10min at 4 °C. The content of lipid peroxidation product malonaldehyde (MDA) was measured by the thiobarbituric acid method. The content of glutathione peroxidase (GSH-Px) was measured by colorimetry. The activity of superoxide dismutase (SOD) was measured by a water-soluble tetrazolium salt (WST-1). The procedure was carried out following the instruction of assay kits (Nanjing Jiancheng Bioengineering Institute, China), respectively.

2.5. Extraction of total RNA

As described above experimental group, total RNA was extracted with RNesy Plus Mini kit (QIAGEN, Germany) according to the manufacturer's recommendations. The contamination and degradation of RNA were analyzed with 1.5% agarose gels. And the purity of RNA was determined by a Nanodrop 2000c spectrophotometer (Thermo Fisher, MA, USA). The total RNA was measured using Qubit®2.0 fluorometer (Life Technologies, CA, USA). The sample was assayed by RNA Nano 6000 assay kit, using an Agilent Bioanalyzer 2100 system (Agilent Technologies, CA, USA) to detect the integrity of RNA. A total of 5 μg from the qualified RNA was used as input sample for the following analysis.

2.6. Transcriptome analysis

The generated procedure includes library construction, quality detection, sequencing and so on. The sequencing libraries were established for the Illumina® sequencing platform (NEB, USA). After the library completion, Agilent 2100 system detect the preliminary quality of library. After the inserted fragment met expectations, the library was accurately quantified by the Q-PCR. Subsequently, the prepared library was sequenced following paired-end on an Illumina Hiseq 2000 platformt (HiSeq/MiSeq). To reduce interference caused by invalid data, Fastp software was then performed to filter and assess the quality of the raw reads. Hisat2 software (https://github.com/infphilo/hisat2) was used to compare the filtered reads to the genome alignment [20,21], and the sam file from each sample was obtained under comparison.

2.7. Differentially expressed genes analysis

The amount of reads count in each sample were measured by FPKM (Fragments Per Kilobase per Million mapped reads). DEGs were analyzed using edgeR [20,21]. The P-value was adjusted by BH method for multiple hypothesis testing corrections, and the FDR was the correction results. Genes with FDR<0.05 and |log2 (fold-change)| >1 were assigned as DEGs. The function and pathway analysis of the screened differential genes were carried out by Gene Ontology (GO, http://www.geneontology.org) and Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.kegg.jp/), respectively. There is a significant enrichment in GO or KEGG when P < 0.05.

2.8. Validation of RT-qPCR analysis

To verify the results of transcriptome analysis, 13 gene expression levels of T. thermophila under DHA stress were evaluated by RT-qPCR. cDNA was synthesized by total RNA from the three different groups using PrimeScript™ RT reagent kit with gDNA Eraser (Takara, Japan). PCR reaction was prepared to a total volume of 20 μL using an SYBR ® Premix Dimer Eraser™ (Takara, Japan). The next reaction was conducted in a CFX96TM real-time system (BIO-RAD, USA). The specific primers were shown in Table S1. The GAPGH was used as a reference gene to analysis in the process. The relative expression level of each gene was described by the 2^–ΔΔCt method [22]. Correlation analysis based on Pearson's correlation coefficient was applied between RT-qPCR and transcriptome.

2.9. Statistical analysis

Each measurement was conducted in triplicate. Data from cell viability assays and antioxidant index tests were analyzed using ANOVA by SPSS 25.0. Differences were calculated by LSD. P < 0.05 was considered statistically significant. All values were expressed as the mean ± standard deviation (SD).

3. Results

3.1. Effects on T. thermophila proliferation

The cell viability affected by different concentrations of dihydroartemisinin are shown in Fig. 1A. Compared with the control group, low concentration DHA (20 μmol/L) presented stimulation to T. thermophila growth, and high concentration DHA (40–1280 μmol/L) inhibited the growth of T. thermophila (P < 0.01). The half inhibition concentration IC₅₀ was described in terms of the logit of cell inhibition rate as the ordinate and the lg of DHA concentrations as the abscissa. IC₅₀ of DHA was 85 μmol/L (Fig. 1B).

Fig. 1.

T. thermophila grown in dihydroartemisinin (DHA) stress condition at 48 h. (A) Effects of different concentrations DHA on cell activity. The difference between the experimental group with the superscript and the control group is significant (**, P < 0.01). (B) The IC50 was calculated by drawing S-curve, the logit of cell inhibition rate as the ordinate and the lg of DHA concentrations as the abscissa.

3.2. Morphological changes of T. thermophila under DHA

Red cells with PI, green cell contents with FITC, blue cell nucleus with DAPI were viewed through the oil microscope (Fig. 2A). During the 48 h incubation period, the T. thermophila in the control group showed drop-shaped appearance, dark blue nucleus with an oval. Difference of morphology was not obvious in 20 and 40 μmol/L treatment compared with the normal cells. Under high concentration DHA stress (80, 160, 320 μmol/L), there is a change that the appearance of cells was shrunken and round, the characteristic of nucleus with granular fluorescence was pycnosis and was smaller than the other groups in shape. However, it was hard to tell the different properties of the green cytoplasm among groups from the photos. With the increase of DHA concentration, the number of cells decreased (Fig. 2B).

Fig. 2.

Fluorescence analysis of T. thermophila morphology exposed to different concentrations dihydroartemisinin (DHA) after 48 h by laser scanning confocal microscope. (A) Morphological changes in cell (Scale bar, ×189). (B) Morphological changes in nucleus (Scale bar, ×40). Note: Ⅰ, Control group; Ⅱ, Ⅲ, Ⅳ, Ⅴ, Ⅵ are experimental groups with the DHA concentrations of 0, 20, 40, 80, 160, and 320 μmol/L, respectively.

3.3. Effects on oxidative stress of T. thermophila

To study the effect of oxidative stress on T. thermophila with DHA, SOD, MDA and GSH-Px were investigated and the results were as followed (Fig. 3). The SOD and GSH-Px between different treatment groups were higher than the control group (P < 0.05) (Fig. 3A and C). Compared with the control group, MDA level increased significantly in high concentration group (P < 0.05) and decreased significantly in low concentration group (P < 0.05) (Fig. 3B).

Fig. 3.

Antioxidant indices for T. thermophila exposure to dihydroartemisinin. (A) SOD, (B) MDA, (C) GSH-Px. Results were expressed as mean ± SD (n = 3). The difference between the experimental group with the superscript and the control group is significant (*, P＜0.05; **, P < 0.01, ***, P < 0.001).

3.4. Quality control of the transcriptome

Clean reads obtained by filtering low-quality data were statistically analyzed in the base content and quality distribution (Table 1). According to the data quality statistics, about 231 million clean paired reads with Q20 > 97% and Q30 > 92% were made from 9 libraries (Table S2). About 97% of the reads from each library were fully qualified to match reference genes. Overall, the quality of sequencing data can be used for the subsequent analysis.

Table 1.

Summary statistics of transcriptome sequencing.

Sample	Clean paired reads	Clean bases(G)	Q20 (%)	Q30 (%)	GC content (%)	Clean data ratio (%)
CG1	26502468	7.88	98.65	94.74	36.15	93.51
CG2	23568226	7.01	98.76	95.03	34.1	94.13
CG3	29403122	8.76	98.69	94.84	35.92	94.17
LG1	27122889	8.06	98.57	94.48	35.28	93.13
LG2	30686194	9.14	98.71	94.9	35.84	93.9
LG3	25039536	7.46	98.81	95.15	33.99	94.63
HG1	21902946	6.52	98.66	94.76	36.36	94
HG2	23622744	7.02	98.64	94.71	35.63	93.39
HG3	23337003	6.91	97.73	92.59	34.85	86.8

3.5. Identification of DEGs under DHA stress

Based on gene expression, the relationship between samples and genes was obtained by hierarchical cluster analysis (Fig. 4A). The results indicated that DEGs were well distinguished for DHA treatment on T. thermophila among the high group, low group and control group. The total DEGs in HG vs. CG were 2561 (Fig. 4B), including 1286 downregulated and 1275 upregulated genes. In LG vs. CG, the expression level of 1917 genes significantly changed, among which 854 downregulated genes and 1063 upregulated genes. There were 116 DEGs in LG vs. HG, including 78 downregulated genes and 38 upregulated genes. Notably, the results indicated that DHA can induce the expression of transcriptome in T. thermophila. DEGs test in statistics were described with volcano map in the HG vs. CG, LG vs. CG and LG vs. HG, respectively. The absolute value of the log2 (FC) was less differentiating at the above samples, but the number of DEGs were great in the HG vs. CG and LG vs. CG, suggesting most of the transcripts were altered by different DHA exposure (Fig. 4C).

Fig. 4.

(A) Hierarchical cluster map based on the relation with expressed genes and samples. The color from red to blue represents the correlation degree from high to low. HG, LG, CG represents high concentration DHA treatment group (160 μmol/L), low concentration DHA treatment group (20 μmol/L) and control group (0.1% DMSO), respectively. The same as the following. (B) The number of differentially expressed genes (DEGs) under different level of dihydroartemisinin (DHA) stress compared to that of the control group. (C) Volcano plots of DEGs in different groups. (C-1), (C-2) and (C-3) are experimental groups with HG vs. CG, LG vs. CG and LG vs. HG, respectively. Each dot represents a gene, red dots are DEGs in statistical tests and black dots are unobvious difference. log2 fold change <0 is down-regulated gene and >0 is up-regulated gene. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3.6. GO enrichment analysis of DEGs

To verify the DEGs involved in functional annotations under DHA treatment, GO analysis was performed. GO was divided into three types: biological process, molecular function, and cellular component. The top 10 DEGs with three categories were listed in the three groups respectively (Table 2). Both HG vs. CG and LG vs. CG were mostly involved in the metabolic processes (such as small molecule, organic acid, carboxylic acid, and lipid) in terms of biological process; in oxidoreductases activity and ATPase activity in terms of molecular functions; and in mitochondria and its membrane in terms of cellular component.

Table 2.

Top GO enrichment analysis of DEGs between different samples.

Group	Ontology	GO Term	Count	P-value
HG vs. CG	Biological processes	Small molecule metabolic process	176	8.83E−14
		Organic acid metabolic process	112	8.07E−10
		Oxoacid metabolic process	110	1.03E−09
		Carboxylic acid metabolic process	107	4.15E−09
		Monocarboxylic acid metabolic process	67	1.58E−08
		Lipid metabolic process	113	2.99E−08
		Generation of precursor metabolites and energy	33	4.33E−07
		Lipid transport	45	8.45E−07
		Cellular lipid metabolic process	91	1.08E−06
		Quinone metabolic process	12	3.90E−06
	Cellular components	Mitochondrial part	90	1.58E−14
		Mitochondrion	105	1.22E−10
		Mitochondrial membrane	54	7.34E−10
		Organelle inner membrane	39	2.35E−07
		Mitochondrial inner membrane	34	7.15E−07
		Mitochondrial membrane part	18	3.46E−05
	Molecular functions	Oxidoreductase activity	141	4.04E−16
		Cofactor binding	96	1.07E−09
		Coenzyme binding	58	3.47E−07
		ATPase activity, coupled to transmembrane movement of substances	68	2.28E−06
		Primary active transmembrane transporter activity	68	1.19E−05
		P–P-bond-hydrolysis-driven transmembrane transporter activity	68	1.19E−05
		Antioxidant activity	18	1.29E−05
		Oxidoreductase activity, acting on a sulfur group of donors	19	2.16E−05
		ATPase activity, coupled to movement of substances	70	2.27E−05
		Flavin adenine dinucleotide binding	23	2.57E−05
LG vs. CG	Biological processes	Small molecule metabolic process	151	2.36E−11
		Organic acid metabolic process	90	5.29E−05
		Oxoacid metabolic process	88	5.85E−05
		Carboxylic acid metabolic process	85	0.0002
		Monocarboxylic acid metabolic process	53	0.0006
		Lipid transport	38	0.0010
		Lipid metabolic process	88	0.0015
		Quinone metabolic process	11	0.0014
		Small molecule biosynthetic process	50	0.0014
		Cellular lipid metabolic process	74	0.0018
	Cellular components	Mitochondrial part	77	5.19E-14
		Mitochondrion	93	3.73E−12
		Mitochondrial membrane	43	8.06E−08
		Mitochondrial membrane part	17	5.99E−06
		Mitochondrial inner membrane	27	1.67E−05
		Organelle inner membrane	30	2.00E−05
		Extrinsic component of mitochondrial inner membrane	5	0.0004
		Mitochondrial matrix	17	0.0007
	Molecular functions	Oxidoreductase activity	110	1.95E−11
		Cofactor binding	73	2.03E−06
		ATPase activity, coupled to transmembrane movement of substances	57	5.30E−06
		Protein disulfide oxidoreductase activity	10	1.15E−05
		Oxidoreductase activity, acting on a sulfur group of donors	17	1.81E−05
		Primary active transmembrane transporter activity	57	2.14E−05
		P–P-bond-hydrolysis-driven transmembrane transporter activity	57	2.14E−05
		ATPase activity, coupled to movement of substances	58	5.76E−05
		Electron transfer activity	12	0.0001
		Disulfide oxidoreductase activity	12	0.0001
LG vs. HG	Biological processes	Potassium ion transport	8	3.67E−06
		Response to oxidative stress	8	7.63E−06
		Monovalent inorganic cation transport	8	2.53E−05
		Potassium ion transmembrane transport	6	3.08E−05
		Cation transport	11	3.22E−05
		Metal ion transport	9	4.74E−05
		Retinal metabolic process	2	9.34E−05
		Pigment biosynthetic process	4	0.0002
		Regulation of ion transmembrane transport	9	0.0002
		Regulation of transmembrane transport	9	0.0002
	Molecular functions	Oxidoreductase activity	21	9.01E−10
		Glutathione peroxidase activity	6	1.75E−09
		Antioxidant activity	8	3.13E−09
		Phospholipid-hydroperoxide glutathione peroxidase activity	4	8.41E−08
		Peroxidase activity	6	2.28E−07
		Oxidoreductase activity, acting on peroxide as acceptor	6	2.28E−07
		Transmembrane transporter activity	20	6.04E−05
		Coenzyme binding	9	8.50E−05
		ATPase activity, coupled to transmembrane movement of substances	10	0.0001
		Dihydrokaempferol 4-reductase activity	2	0.0001
	Cellular components	Photoreceptor inner segment	2	0.0015
		Integral component of pigment granule membrane	2	0.0062
		Extrinsic component of cytoplasmic side of plasma membrane	1	0.0300
		Receptor complex	2	0.0364
		Plasmodesma	3	0.0382
		Pigment granule	2	0.0410
		Presynapse	2	0.0458

Of the biological processes, potassium transport and response to oxidative stress were significantly enriched in LG vs. HG. The molecular functions mainly focused on oxidoreductase activity, glutathione peroxidase activity, and the antioxidant activity. In terms of cell composition, DEGs were mainly concentrated in cell part and membrane part. It found that the number of enriched terms in LG vs. HG was less than that of else two groups, which might be related to a smaller number of DEGs.

3.7. KEGG pathway analysis of DEGs

To understand the DEGs involved in metabolic pathways, KEGG analysis was performed. The top 30 pathways with low P value between different groups were shown (Fig. 5). The DEGs in HG vs. CG group were mapped to 318 KEGG pathways, of which 23 were significantly enriched (P < 0.05) (Fig. 5A). There were 318 KEGG pathways involved in LG vs. CG, and 29 of them were significantly enriched (P < 0.05) (Fig. 5B). Both HG vs. CG and LG vs. CG had enhanced metabolism of other amino acids (glutathione metabolism), lipid metabolism (arachidonic acid metabolism, fatty acid elongation), amino acid metabolism (alanine, aspartate and glutamate metabol, d-Glutamine and d-glutamate metabolism), energy metabolism (oxidative phosphorylation), xenobiotics biodegradation and metabolism (metabolism of xenobiotics by cytochrome P450) and membrane transport (ABC transporters). And DEGs were distributed in 16 KEGG pathways in the LG vs. CG, including the metabolism of xenobiotics by cytochrome P450, glutathione metabolism, ABC transporters, p53 signaling pathway, etc. (Fig. 5C). It indicated that DHA was highly sensitive to these pathways, which be induced even at low dose. Notably, in the basis of the DEGs involved in the enrichment, low lever DHA promoted DNA replication process, of which 10 upregulated and 3 downregulated. While high lever DHA caused apoptosis.

Fig. 5.

KEGG pathway enrichment analyses in HG vs. CG (A), LG vs. CG (B) and LG vs. HG (C). The Y-axis was used for the KEGG pathway. The X-axis was used for the rich factor. The size of point indicate the number of genes which enriched in a particular pathway. And low p-values indicates a large degree of enrichment.

3.8. The verification results of the RT-qPCR

There were 13 DEGs from the treatment groups selected for the validation of RT-qPCR (Fig. 6A), including seven genes with significantly upregulated and six genes with significantly downregulated. Among them, the expression of genes encoding GST (TTHERM_00516420) and GPx family protein (TTHERM_00895650, TTHERM_00895660) were increased obviously by DHA, which was consistent with the results in the GO analysis (Fig. 6A, Table S1). Then, the fold changes drawn from RT-qPCR were compared with the transcriptome sequencing results. The linear analysis showed that the gene expression differences given by the two methods were highly correlated (R² = 0.8849), as shown in Fig. 6B.

Fig. 6.

Relative expression levels of thirteen DEGs under DHA stress. (A) Histogram of relative gene expression levels. (B) Correlation analysis of fold change values based on RT-qPCR and RNA-seq data. R² represented the correlation coefficient.

4. Discussion

4.1. Toxic effect of T. thermophila induced by DHA

Artemisinin contains a unique peroxide bridge, which can combine with ferrous ions to form oxygen free radicals, which gradually damage cell structure and function through oxidation leading to cell death. Meanwhile, DHA drugs may lead to the significant behavior, physiology and biochemistry changes in different parasites [[23], [24], [25]]. In this study, we found that compared with the control group, the high concentration DHA can inhibit cell growth. Interestingly, the DNA replication process was generally upward by KEGG analysis under low DHA stress, and the necroptosis process was markedly upregulated under high DHA stress. The results are similar with previous studies such as low dose of Panax japonicus could produce growth for T. thermophila [26]. Similarly, it was reported that the growth and metabolism of T. thermophila were inhibited under the high levels of artesunate and DHA; however, above that can be promoted under the low levels [15]. Numerous researchs have found in the past which indicate that the hormetic dose-response model, which is characterized by a low-dose stimulation and a high-dose inhibition [27]. We can speculate that DHA exposure caused T. thermophila damage has a certain threshold. The transcriptome analysis suggests that oxidative stress is an important factor of DHA toxicity to T. thermophila. The degree of oxidative stress damage is dose-dependent. The concentration of DHA caused T. thermophila damage might have a certain threshold. Below the threshold, T. thermophila mainly used antioxidant enzymes activity to eliminate the toxicity of free radicals and the stability and functional integrity of the nuclear were not damaged according to the observation of cell morphology. Under cellular stress, ROS and lipid peroxidations generation in T. thermophila can increase the level of MDA [[28], [29], [30]], even damage DNA [10]. In this study, antioxidant system was inhibited along with the increase of MDA, nuclear condensation and cell membrane damage [31] under high concentration DHA stress. However, the mechanism of this phenomenon should be further studied.

Under normal situations, organisms can remove ROS to protect cells from injury. The genes of Ca²⁺-ATPase and thioredoxin glutathione reductase (TGR) were regarded as classic genes for insecticidal target. When activated, artemisinins can work as an inhibitor of Ca²⁺-ATPase in P. falciparum [32]. Some compounds, such as oxadiazole 2-oxides, isoxazolones and phosphoramidites, have been regarded as agents to inhibit TGR, which plays a crucial role in the redox balance of the parasite from Schistosoma [33]. We found that expression of genes coding TGR and calcium-transporting ATPase activity were reduced at high DHA stress. It inferred that the exposure to high level DHA in the T. thermophila might interfere the redox balance of cell and interact with the target genes, partially leading to cell growth inhibition.

4.2. DHA act on T. thermophila by interfering with energy metabolism

Mitochondria is a vital organelle, which has been reported as sensitive target in T. thermophila exposed to external stimulation [[34], [35], [36]]. Cytotoxicity of artemisinin drugs associated with mitochondrial damage as shown in previous study [37]. Under DHA stress, DEGs were mainly concentrated in mitochondria and its membrane according to the cellular components, suggesting that encoded protein might affect the oxidative phosphorylation level by acting on the transcription and translation of mitochondrial genes. Meanwhile, pathways with tricarboxylic acid cycle and glycolysis/gluconeogenesis were dramatically inhibited based on the KEGG analysis in DHA treatment groups. This observation is consistent with our previous study for mitochondrial membrane depolarization in T. thermophila under high level DHA exposure [16]. The artemisinin drugs can damage the membranes structure of mitochondria and nucleus, along with the phenomenon on mitochondrial swelling and nuclear fragmentation, resulting in cytotoxicity. Meanwhile, it was also reported that other drugs have the toxic effects to interfere with energy metabolism in the organism. For example, A-viniferin could kill Raillietina echinobothrida in chickens by inhibiting enzymes related to energy metabolism [38]. By transcriptomic and proteomic analysis, researchers have revealed that octadecanoic acid-3,4-tetrahydrofuran diester, a compound from Chinaberry, could kill mites through interference with energy metabolism [39]. In our study, the expression level of glucose metabolism and energy metabolism related genes were down-regulated by GO analysis. In addition, the expression of selected FBP (TTHERM_00058590) a major rate-limiting enzymes for the metabolic pathway of gluconeogenetic was decreased by RT-qPCR and was consistent with the above. These results indicated that excessive oxidative stress induce respiratory injury of mitochondria in T. thermophila, which might be the partial reason for cell death. But the mechanism of DHA to the mitochondria still need further study.

4.3. DHA act on T. thermophila by regulating cytochrome P450, GST and ABC transporters

Glutathione peroxidases (GPXs) genes [40] were important antioxidant family in organisms and help to control the excessive production of free radicals to protect cells from oxidative stress. In this study, we selected oxidative stress-related genes to explore the changes of expression in T. thermophila exposure with high DHA concentrations. It found that the expression levels of GST and GPx family genes increased with DHA exposure, as the accumulation of MDA content significantly increased, which was an indicator of oxidative stress degree. It further indicated that other detoxifying pathways worked to prevent T. thermophila from the damage of lipid peroxidation and peroxide stress.

Moreover, metabolism of xenobiotics by cytochrome P450, glutathione metabolism and ABC transporters in this study has shown significantly enrichment under DHA treatment on T. thermophila. The metabolism process of xenobiotics has undergone three phases: modification, conjugation, and excretion [41]. Therefore, the resistance of oxidative stress damage under DHA treatment partially attribute to regulation of the above mentioned three pathways in T. thermophila cells. It may be inferred that the metabolism process of DHA in T. thermophila based on the previous study and the results of this experiment, detoxification enzymes (such as cytochrome P450 family, drug metabolism enzyme) joined in the control of metabolism of xenobiotics as the first defense [42], which can defend against hydrophobic xenobiotic compounds by adding oxygen atoms to turn them into a partially soluble state in water. Subsequently, GST gene family can participate in the resistance to the oxidative damage induced by DHA in T. thermophila. Finally, dissolved xenobiotic molecules were excreted from cells via ATP-binding cassette (ABC) transporter superfamily [13]. Further results of metabolomics detection showed that multiple amino acid metabolic pathways were enriched by KEGG pathway analysis, suggesting it may be an important regulatory pathway. Therefore, the potential marker genes in amino acid metabolism would be studied in combination with amino acid content, to further understand the mechanism of amino acid metabolism of T. thermophila under DHA stress.

5. Conclusion

In summary, we evaluated the toxicity of DHA in T. thermophila B2086.2 and identified that DHA can inhibit cell growth. It is first report that DHA exposure with T. thermophila induced oxidative stress, which cause damage to nucleus, and increase of antioxidant enzymes activity along with the disturbance to metabolic processes and molecular functions. These toxic effects may work by interfering with energy metabolism. Meanwhile, detoxification pathways played important roles in the protection against oxidative damage under DHA stress in T. thermophila, such as metabolism of xenobiotics cytochrome P450 and glutathione metabolism, ABC transporter pathway. But the molecular mechanisms of action still need further investigations. This research provides the basic data for further exploration to mechanism of DHA toxicity in the parasites.

Author contribution statement

Houjun Pan: Conceived and designed the experiments; Contributed reagents, materials, and analysis the data.

Yuguo Liu: Conceived and designed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, and analysis the data; Wrote the paper.

Tiantian Fang: Conceived and designed the experiments; Performed the experiments; Wrote the paper.

Meiling Deng: Performed the experiments; Analyzed and interpreted the data.

Bin Zhang: Analyzed and interpreted the data.

Funding statement

Mrs Yuguo Liu was supported by Traditional Chinese Medicine Scientific Research Project of Traditional Chinese Medicine Bureau of Guangdong [20232098], Guangdong Medical Research Foundation [B2019156].

Houjun Pan was supported by China Agriculture Research System of MOF and MARA [CARS-45], Natural Science Foundation of Guangdong Province [2017A030313172].

Data availability statement

Data will be made available on request.

Declaration of interest's statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article.