Introduction

The term ‘ribotoxic stress response (RSR)’ was coined in 1997 by Iordanov et al.. The researchers demonstrated that drugs and toxins that interfere with the region of 28S ribosomal RNA (rRNA) critical for amino acid-tRNA binding, peptidyl transfer, and translocation, activate p38 and JNK, which are stress- and mitogen-activated protein (MAP) kinases1. The RSR is one such pathway that is activated by ribosomal impairment, which can disrupt protein synthesis, increase inflammatory signaling, and even lead to cell death if left unresolved2. The RSR has an important physiological role in regulating cell-fate decisions, cytokine expression, and balancing apoptosis and cell survival during immune responses3,4,5,6. RSR triggers have been identified, including translation inhibitors, ribotoxins, chemotherapeutics, and UV radiation. These agents cause structural damage to rRNA, impair ribosome function, and lead to the strong activation of p38 and JNK. The initial signaling triggers a proinflammatory response, but if the duration or magnitude of RSR signaling is sufficient, it can eventually lead to cell death. In addition to the RSR, there are two other main translation surveillance pathways, Ribosome-associated Quality Control and Integrated Stress Response, which also play important roles in maintaining cellular homeostasis.

RNA degradation is a crucial process in gene expression regulation that controls coding and non-coding RNA levels in response to developmental and environmental cues7,8. It also plays a vital role in eliminating defective RNAs that can be produced during the maturation of functional transcripts from their precursors. The control of RNA quality prevents the accumulation of aberrant non-coding transcripts or the translation of defective messenger RNAs (mRNAs), which could potentially cause deleterious effects. Studies have shown that RNA degradation can prompt death execution, as malfunction of the protein synthesis machinery may occur following rRNA degradation9. Viral and endogenous double-stranded RNA (dsRNA) can also trigger programmed RNA degradation by the 2–5 A/RNase L complex, termed 2-5AMD. This response switches global protein synthesis from homeostasis to production of interferons10. RNase L cleaves a broad spectrum of host RNAs, including rRNAs, tRNAs, mRNAs, and Y-RNAs. Although the cleavage of rRNA does not inhibit the activity of the ribosome, recent studies have shown that mRNA levels in the cell can be reduced up to ∼90% due to RNase L, causing a global effect termed 2-5AMD10,11. Interestingly, mRNA decay fragments via the nonsense mediated decay (NMD)12, may be translated to produce short peptides that could be important for antiviral activity or others13.

The histone H3 N-terminal protein domain, also known as N-tail, is subject to regulation through various posttranslational modifications such as methylation, acetylation, phosphorylation, and proteolytic cleavage14. Cleavage is observed to be more frequent at gene promoters that are bound and repressed by JMJD5, which suggests that H3 N-tail cleavage plays a role in gene expression regulation15. There have been many studies demonstrating the importance of histone H3 clipping in various biological processes, such as human embryonic stem cell lines (ESCs) and human peripheral blood mononuclear cells (PBMCs), regulation of monocyte-to-macrophage differentiation, a novel signature of human neutrophil extracellular traps, and regulation of myoblast differentiation15,16,17,18,19. In addition to the well-known cysteine protease Cathepsin L19, glutamate dehydrogenase and MMP-2 have been identified as novel H3 N-tail proteases20,21. These proteases primarily cleave at amino acid residue A21, but multiple other sites have also been found at amino acid residues: T22, K23, A24, R26, and K27 17. The potential effects of histone tail clipping may lead to altered electrostatic interactions between histones and DNA and remove existing histone marks, either activating or repressive, which may erase epigenetic memory and alter the regulation of cell-type-specific gene transcription.

For centuries, in traditional medicine, Taraxacumreportedly exhibited analgesic, angiogenic, anti-allergic, anticarcinogen, anticoagulant, anti-inflammatory, antirheumatic, antioxidant, cholagogue, choleretic, and diuretic properties, among others22,23,24,25,26. Taraxacum-induced cell death could be due to the cleavage of 28S rRNA. This cleavage may lead to apoptosis-associated rRNA cleavage and activate signals such as JNK, p38, and RPS6 27. It is possible that Taraxacum-induced 28S rRNA cleavage may have similar biological consequences to an RSR, such as apoptosis. In a previous study, we investigated the potential antitumor effects of aqueous extracts from two species of Taraxacum, T. mongolicum and T. formosanum, on HeLa human cervical cancer cells and various types of liver cancer cells27. Our results showed that the extracts suppressed cellular proliferation and transcription while inducing necrosis, late apoptosis, and endoplasmic reticulum (ER) stress in HeLa cells. Another study focused on the potential antitumor effects of these species on three human breast cancer cell lines: the MDA-MB-231 TNBC cell line, and the ZR75-1 and MCF-7 hormone-dependent breast cancer cell lines28. The study found that T. mongolicum exhibited cytotoxic effects against MDA-MB-231 cells, including the induction of apoptosis, suppression of cellular proliferation, colony formation, migration, disruption of the mitochondrial membrane potential, and downregulation of the oxygen consumption rate. In contrast, T. formosanum induced ribotoxic stress in MDA-MB-231 and ZR-75-1 cells, while T. mongolicum did not.

Mitochondrial depolarization occurs when there is a loss of mitochondrial membrane potential, which is often linked to impaired mitochondrial function. This disruption can result in cellular dysfunction and may eventually lead to cell death29. Autophagy is a cellular survival pathway that helps cells compensate for nutrient depletion and ensures the appropriate degradation of organelles30. Mitophagy is a related process that regulates mitochondrial number and health by subjecting excessive or damaged mitochondria to autophagic degradation31. Evidence suggests that mitochondria can also significantly influence the autophagic process32. Mitophagic malfunction can lead to the accumulation of dysfunctional mitochondria, making the cell more susceptible to mitochondrial membrane permeabilization and apoptosis. In turn, apoptotic executers can inactivate pro-autophagic proteins, inhibiting autophagy and leading to cell death33. In this study, we aimed to address three important issues related to the effect of Taraxacum on the quality of 28S rRNA, such as cleaved rRNA fragments, affects the stability of mRNAs or/and protein synthesis, and then to identify the rescue reagents. We tried to elucidate whether the working mechanisms of RSR, rRNA cleavage, and histone H3 N-tail clipping have linked each other or related to each other. With the detailed information obtained from our study, we hope to elucidate the mechanisms of gene regulation from cleaved rRNAs and N-truncated histone H3 to mRNAs, and their respective cellular death and survival stresses.

Results

The verification of the cleavage of rRNAs and the ribotoxic stress response by T. formosanum via the RIN value and RNA sequencing analysis in human cervical HeLa cells

The integrity of RNA molecules is of paramount importance for experiments that try to reflect the snapshot of gene expression at the moment of RNA extraction. The advent of microcapillary electrophoretic RNA separation provides the basis for an automated high-throughput approach, in order to estimate the integrity of RNA (RIN) samples in an unambiguous way with an Agilent 2100 bioanalyzer34. RIN values range from 10 (intact) to 1 (totally degraded) based on adaptive learning tools. The data of microcapillary electropherogram are divided into different areas and peaks, and the analysis software supporting the Agilent 2100 bioanalyzer can generate RIN values and 28S/18S ratios based on the test results. Our previous results showed that rRNA degradation could be detected with RNA agarose gel electrophoresis27. We applied microcapillary electrophoretic RNA separation to calculate RIN values and 28S/18S ratios of T. formosanum and T. mongolicum treated HeLa cells. We compared the general RNA agarose gel electrophoresis (Fig. 1A) and microcapillary electrophoretic RNA separation (Fig. 1B) in T. formosanum treated HeLa cells. The red single- and double-arrow in Fig. 1A indicate cleaved 28S rRNA and 18S rRNA, while the red single- and double-arrow in Fig. 1B indicate intact 28S rRNA and 18S rRNA. We observed the cleavage of 28S rRNA and 18S rRNA when the concentrations of T. formosanum and T. mongolicum were over 7 mg/mL. We calculated the ratio of 28S/18S and RIN value for individual T. formosanum concentration (Fig. 1C). The ratio of 28S/18S decreased from 2.1 to 0.1, and the value of RIN decreased from 8.5 to 5.0. These data suggest that the quality of rRNAs was disrupted by T. formosanum and T. mongolicum extract-treated HeLa cells. Our current data support that T. formosanum extract might have the ability to degrade 28S rRNA and 18S rRNA in HeLa cells via the reliable microcapillary electrophoretic RNA separation plus RIN value.

Fig. 1
figure 1

Effects of T. formosanum and T. mongolicum on the RNA quality in HeLa cells. (A-C) HeLa cells were treated for 24 h with the indicated concentrations of T. mongolicum or T. formosanum, after which their cell lysates were subjected to (A) total RNA agarose gel electrophoresis analysis; (B) total RNA microcapillary electrophoresis analysis; (C) microcapillary electropherograms were generated by Agilent 2100 bioanalyzer. The green single-arrow is the cleaved 28S rRNA and the double-arrow is the cleaved 18S rRNA. The red single-arrow is 28S rRNA and the double-arrow is 18S rRNA for the measure of RIN value. M, bp: the molecular weight of rRNA.

Full size image

Higher amounts of T. mongolicum and T. formosanum-induced rRNA degradation had been found to suppress most mRNAs, including well-known internal mRNA control GAPDH and β-actin, while their proteins were expressed well in HeLa cells27. To further understand the relationship between rRNA degradation and mRNA stability, lower amounts of T. mongolicum or T. formosanum were combined with de novo mRNA synthesis inhibitor actinomycin D (Act D) in HeLa cells to analyze mRNA stability. Using the total RNA agarose gel analysis, there was no apparent difference in each group (Fig. 2A). However, during the 4-hour window with Act D treatment, through RT-PCR analysis of target mRNA expression, we detected a reduction in the abundance of stable mRNAs, including p21, C/EBP-homologous protein (CHOP), and interleukin-6 (IL-6) (Fig. 2B, compared with lanes 1, 5, and 9). However, the treatment of HeLa cells with T. mongolicum and T. formosanum induced the expression of p21, CHOP, and IL-6 mRNA. In the western blotting analysis (Fig. 2C), we found that hypoxia-inducible factor 1 alpha (HIF-1α) protein might suppress cyclin B1 mRNA as a repressor and induce CHOPmRNA as an activator based on previous literature35,36. Additionally, histone H3 chipping fragments were induced by T. mongolicum and T. formosanum, and the changes of H3P (phosphorylation of histone H3 serine 10) and demethylation of histone H3 lysine 79, H3(K79), proteins. Our data showed that the signals of H3P were suppressed with the increasing amounts of T. mongolicum and T. formosanum, but the signals of dimethyl H3(K79) were not affected. However, we need the optimal condition of T. mongolicum and T. formosanum to elucidate their effects on the stability of some mRNAs, including internal mRNA control GAPDH.

Fig. 2
figure 2

Effects of T. mongolicum and T. formosanum on rRNAs, mRNAs, and proteins in HeLa cells. (A-C) HeLa cells were treated for 24 h with the indicated concentrations of T. mongolicum or T. formosanum in the presence of various time of 1µM Act D treatment, after which their cell lysates were subjected to (A) total RNA agarose gel electrophoresis analysis; (B) RT-PCR analysis using primer pairs against the indicated mRNAs, with GAPDH mRNA as the mRNA loading and internal control; (C) western blot analysis using antibodies against the indicated proteins, with ACTN as the protein loading control (supplementary file 1). M, bp: the molecular weight of rRNA. M, kDa: the molecular weight of protein.

Full size image

It has been demonstrated that certain agents inducing structural damage to rRNA and impair ribosome function can trigger a robust activation of a proinflammatory response3,4,5,6. In this study, we applied RNA sequencing analysis to compare T. formosanum extract-treated HeLa cells in terms of gene expression. The analysis revealed 893 upregulated genes and 509 downregulated genes (Fig. 3A). Among the highlighted upregulated genes were small nuclear RNAs (snRNAs). Most snRNAs are involved in the splicing of introns from pre-mRNA as part of either the major or minor spliceosome8. The major spliceosome features U1, U2, U4, U5, and U6 snRNPs. U1 snRNP protects pre-mRNAs from premature cleavage and polyadenylation37. All snRNA genes are named with the root symbol “RNU” for “RNA, U# small nuclear.” In this study, two highlighted downregulated genes were POLR2A and RNU6ATAC35P. The POLR2A gene encodes the largest subunit of RNA polymerase II, while RNU6ATAC35P is a small RNA molecule found in the cell nucleus that is involved in the processing of pre-mRNA snRNA. It is worth noting that human spliceosomal snRNA sequence variants generate variant spliceosomes38. One of the most conserved mechanisms by which cells sense viral infection is through the detection of dsRNAs by a set of receptors in the innate immune system that trigger antiviral and inflammatory immune responses39. Gene set enrichment analysis (GSEA) by T. formosanum showed that hallmark TNF-α signaling via NF-κB, Inflammatory response, and IL6-JAK-STAT3 signaling were all affected (Fig. 3B).

Fig. 3
figure 3

The RNA sequencing analysis of T. formosanum extract-treated HeLa cells. (A) Volcano plot and highlighted up- and down-regulated genes. (B) Gene set enrichment analysis for TNF-α signaling via NF-κB, Inflammatory response, and IL6-JAK-STAT3 signaling.

Full size image

The rescuing mechanisms of caffeine for the cleavage of rRNAs by T. formosanum in human cervical HeLa cells

We utilized various inhibitors to rescue the 28S rRNA cleavage induced by 5 mg/mL T. formosanum for 24 h, 10 mg/mL T. formosanum for 8 h, and 7 mg/mL T. formosanum for 16 h, including signaling kinase inhibitors (such as p38 kinase inhibitor SB203580, MAPK inhibitor PD98059, AKT/PI3K inhibitor LY294002, JNK inhibitor, ribosomal S6K inhibitor BI-D1870, and EGFR tyrosine kinase inhibitor AG1478), antioxidant ellagic acid, de novo mRNA synthesis inhibitor Act D, proteasome inhibitor MG132, lysosome inhibitor ammonium chloride (NH4Cl), p53 transcription activity inhibitor PFT-α, apoptosis inhibitor Z-VAD-FMK, calpain inhibitor II ALLM, adenosine receptor antagonist caffeine, Na+/K+ ATPase ligand ouabain, and anti-inflammatory agent resveratrol (Fig. 4A-D). Our data suggest that caffeine might be one of the choices for rescuing the 28S and 18S rRNA cleavage induced by T. formosanum in HeLa cells.

Fig. 4
figure 4

The drug screening procedure was to rescue the cleavage of 18S and 28S rRNA by T. formosanum in HeLa cells. (A-D) HeLa cells were treated for the indicated concentrations of T. formosanum and various potential drugs treated with indicated time, after which their cell lysates were subjected to total RNA agarose gel electrophoresis analysis. The red single-arrow is the cleaved 28S rRNA and the double-arrow is the cleaved 18S rRNA. M, bp: the molecular weight of rRNA.

Full size image

We further elucidated the rescue mechanisms of caffeine in T. formosanum extract-treated HeLa cells. The RNA quality index and 28S/18S ratio decreased RIN value from 7.8 to 4.8 and 28S/18S ratio from 1.4 to 0.3 by T. formosanum (Fig. 5A). However, caffeine restored both RIN and the 28S/18S ratio to 6.3 and 0.4, respectively. The total RNA agarose gel electrophoresis also revealed alterations in 28S and 18S rRNA fragments induced by T. formosanum, and these changes were reversed in the presence of caffeine (Fig. 5B). However, the RT-PCR data exhibited patterns consistent with those presented in our previous publication27, indicating that several target mRNAs, including GAPDH and β-actin, were challenging to detect in the presence of T. formosanum (Fig. 5C). In this suppression of T. formosanum, caffeine appeared to be ineffective in rescuing any mRNA. Notably, we observed interesting data in the western blotting analysis (Fig. 5D). The effects of T. formosanumon cyclin D1, HIF-1α, and histone H3 clipping fragments were reversed with the addition of caffeine. Moreover, we observed consistent findings regarding the specific effects of caffeine on p53 and cleaved PARP proteins with our previous publication27.

Fig. 5
figure 5

The rescue effects of caffeine on the T. formosanum-induced 18S and 28S rRNAs cleavage in HeLa cells. (A-D) HeLa cells were treated with 7 mg/mL T. formosanum for 16 h alone or in the presence of pretreated with indicated amounts of caffeine for 2 h, after which their cell lysates were subjected to (A) total RNA microcapillary electrophoresis analysis and microcapillary electropherograms were generated by Agilent 2100 bioanalyzer. (B) total RNA agarose gel electrophoresis analysis; (C) RT-PCR analysis using primer pairs against the indicated mRNAs, with GAPDH mRNA as the mRNA loading control and 18S rRNA as an internal control; (D) western blot analysis using antibodies against the indicated proteins, with ACTN as the protein loading control. M, bp: the molecular weight of rRNA. M, kDa: the molecular weight of protein.

Full size image

Based on the defect of rRNAs, we analyzed the effects of T. formosanum on ribosomal proteins, including ribosomal proteins of large subunit and small subunit (RPLs and RPSs) using the RNA sequencing data. We summarized up- and down-regulated RPLs and RPSs by T. formosanum and rescued by caffeine (Fig. 6A). We then selected some RPLs (RPL3, RPL4, and RPL6) and RPSs (RPS2 and RPS26) for the verification of the rescue effect of caffeine with PCR gel agarose electrophoresis (Fig. 6B). Most of them were consistent with the analysis of RNA sequencing data.

Fig. 6
figure 6

Effects of T. formosanum on ribosomal protein genes in HeLa cells. (A and B) HeLa cells were treated with 7 mg/mL T. formosanum for 16 h alone or in the presence of pretreated with indicated amounts of caffeine for 2 h, after which their cell lysates were subjected to (A and B) the list of up- and downregulated of (A) ribosomal protein genes (B) and RT-PCR analysis using primer pairs against the selected mRNAs, with β-actin mRNA as the mRNA loading and internal control. The green single-arrow is one of the target ribosomal proteins for further verification.

Full size image

Our Figs. 2 and 5 suggested that it is an important issue of the treatment concentration and time of T. formosanum to address its effects on rRNAs, mRNAs, and proteins levels in HeLa cells. We failed to observe the rescue effect of caffeine on T. formosanum-suppressed mRNA expressions, even we could observe the rescue effect of caffeine on specific proteins (Fig. 5). It was also a puzzle for higher amounts of T. formosanum suppressed GAPDH and β-actin mRNAs, not both proteins. Hence, we sought to appropriate conditions of T. formosanum for the monitoring changes of all interesting mRNAs and proteins. In Fig. 7A, the value of RIN was 6.3 compared with 4.8 in the same treatment condition of 7 mg/mL T. formosanum-treated HeLa cells for 16 h (compared Fig. 1C with Fig. 5A). Wild-type SRSF3 (serine/arginine rich splicing factor 3) (SRSF3 WT) mRNA is changed to exon 4-included premature SRSF3 (SRSF3 PTC) mRNA by caffeine. SRSF3 is an important splicing factor for the expression of p53 protein α form and hypoxia inducible factor 1 alpha (HIF-1α)40. In Fig. 7B and C, T. formosanum suppressed SRSF3, p53, and cyclin B1 mRNAs and proteins and induced CHOP mRNA and protein. However, T. formosanum suppressed HIF-1α, eIF2α, and poly(ADP-ribose) polymerase (PARP) mRNAs, but increased HIF-1α, p-eIF2α, and cleaved PARP (cPARP) proteins. T. formosanum induced H3 clipping and γH2AX. proteins. Increasing amounts of caffeine differentially rescued or reversed the effect of p53β, HIF-1α, cyclin B1, CHOP, p-eIF2α, cPARP, H3 clipping, and γH2AX. proteins by T. formosanum. The rescue of T. formosanum-induced H3 clipping fragment by caffeine, accompanying with the decrease of H3P (phosphorylation of serine 10). Caffeine suppressed rRNA cleavage enzyme RNase L mRNA and H3 clipping enzyme Cathepsin L. Our data revealed that HIF-1α was a repressor for cyclin B1 and activator for CHOP in HeLa cells, consistent with literatures. Caffeine regulated cyclin B1 and CHOP mRNAs and proteins via the function role of HIF-1α.

Fig. 7
figure 7

The rescue effects of caffeine on the T. formosanum-dependent mRNAs and proteins in HeLa cells. (A-C) HeLa cells were treated with 7 mg/mL T. formosanum for 16 h alone or in the presence of pretreated with indicated amounts of caffeine for 2 h, after which their cell lysates were subjected to (A) total RNA agarose gel electrophoresis analysis; (B) RT-PCR analysis using primer pairs against the indicated mRNAs, with 18S rRNA as an internal control; (C) western blot analysis using antibodies against the indicated proteins, with ACTN as the protein loading control. The red single-arrow is the cleaved 28S rRNA and the double-arrow is the cleaved 18S rRNA. M, bp: the molecular weight of rRNA. M, kDa: the molecular weight of protein.

Full size image

To test whether Taraxacum induces histone H3 clipping in different cells. We treated human HeLa cervical cancer, human liver cancers, including Huh6, Huh7, HepG2, and Hep3B, human L0-2 normal liver cell line with different resources of Taraxacum. We identified the clipping fragment of histone H3 with two antibodies against H3 and H3P(S10) using the western blotting analysis (Fig. 8). Our data showed that H3 clipping fragments were found in HeLa, Huh6, and HepG2 cells treated with T. formosanum (fresh extract). The clipping status determined the sensitivity to H3P antibody.

Fig. 8
figure 8

Effects of Taraxacum on histone H3 status in various cells. HeLa, Huh6, Huh7, HepG2, Hep3B, and L02 cells were treated for 24 h with the indicated concentrations of T. mongolicum, T. formosanum, or Herba Taraxaci. These cell lysates were subjected to western blot analysis using antibodies against the indicated proteins, with ACTN as the protein loading control. M, kDa: the molecular weight of protein.

Full size image

The effects of caffeine on necrosis, apoptosis, autophagy, and mitochondrial depolarization by T. formosanum in human cervical HeLa cells

Our previous data showed that T. formosanum induced cell death, including necrosis and apoptosis27. Caffeine enhanced the cleaved PARP fragments by T. formosanum (Fig. 7C). We further analyzed the effect of T. formosanum on necrosis and apoptosis using 7-AAD/Annexin V flow-cytometry analysis (Fig. 9). Our data demonstrated that T. formosanum induced 88% necrotic and 13% late apoptotic populations and increasing amounts of caffeine significantly suppressed necrotic populations to 82%, p(ANOVA) = 1.9 × 10−5, and significantly enhanced late apoptotic populations 18%, p(ANOVA) = 7.5 × 10−6. T. formosanum and caffeine both had no effect on early apoptotic population (Fig. 9A). Caffeine had no apparent effect on the sum of necrotic and apoptotic populations by T. formosanum in HeLa cells. However, increasing amounts of caffeine had enhanced effects on the populations of Annexin V. This finding could be verified with the PE-Annexin V M1 gating data (Fig. 9B and C).

Fig. 9
figure 9

The effects of caffeine on the T. formosanum-dependent necrosis and apoptosis in HeLa cells. HeLa cells were treated with 7 mg/mL T. formosanum for 16 h alone or in the presence of pretreated with indicated amounts of caffeine for 2 h. Collected cells were subject to the flow-cytometry analysis for (A) necrosis and apoptosis analysis with 7-AAD/Annexin V kit. (B and C) Cells stained with PE-Annexin V were gated and plotted. The results are representative of three independent experiments. #p > 0.05, *p < 0.05, and ***p < 0.001 (Student’s t-test).

Full size image

In addition to necrosis, apoptosis, and ER stress, we further analyzed the effects of T. formosanum on autophagy and mitochondrial membrane potential and examined whether caffeine had the ability to reverse its effects. We first analyzed the effect of T. formosanum on autophagy using the acridine orange dye with the flow-cytometry. We observed T. formosanum suppressed the endogenous autophagy (the formation of autophagolysosome), but caffeine reversed the suppression to relative endogenous level with the increasing amounts in HeLa cells (Fig. 10A and B). Then, we examined the effect of T. formosanum on mitochondrial membrane potential using the JC-1 dye with the flow-cytometry. Our data showed that T. formosanum disrupted mitochondrial membrane potential by the decreased red JC-1 aggregates and increased green JC-1 monomers (Fig. 10C). The red/green ratio apparently was decreased T. formosanum and rescued by caffeine in a dose-dependent manner. Lipidated LC3 (LC3B II) plays a crucial role in the closure of autophagosomes and facilitates the docking of specific cargos and adaptor proteins, such as p62. Our Fig. 10D western blotting data showed that the changes in p62 levels rather than with LC3BII/I in the presence of T. formosanum or in combination with caffeine correlated with the autophagy percentage we determined (reflecting autophagolysosome presence through acridine orange, Fig. 10B).

Fig. 10
figure 10

The effects of caffeine on the T. formosanum-dependent autophagy and mitochondrial membrane potential in HeLa cells. HeLa cells were treated with 7 mg/mL T. formosanum for 16 h alone or in the presence of pretreated with indicated amounts of caffeine for 2 h. Collected cells were subject to the flowcytometry analysis for (A and B) autophagy and (C) mitochondrial membrane potential with AO and JC-1 dyes, respectively. The results are representative of three independent experiments. #p > 0.05, *p < 0.05, **p < 0.01, and ***p < 0.001 (Student’s t-test). (D) These cell lysates were subjected to western blot analysis using antibodies against the indicated proteins, with ACTN as the protein loading control. M, kDa: the molecular weight of protein.

Full size image

Discussion

The integrity of RNA molecules is crucial for experiments that aim to reflect the snapshot of gene expression at the moment of RNA extraction. The advent of microcapillary electrophoretic RNA separation now provides the basis for an automated high-throughput approach to estimate the RIN value in an unambiguous manner using an Agilent 2100 bioanalyzer34. In this study, we used both the ratio of 2 S/18S and RIN to confirm RNA molecule integrity. Our data showed that the quality of mRNA was the most easily degraded material compared to rRNA and protein. With the help of the RIN values, we could elucidate the effect of T. formosanum on the cleavage of rRNAs which was related to RSR. In addition to the activation of p38 and JNK signaling pathways, we identified the hallmark major spliceosome, ribosomal proteins, TNF-α signaling via NF-κB, inflammatory response, and IL6-JAK-STAT3 signaling with the analysis of RNA sequencing. Caffeine was the only screening drug to rescue the effect of T. formosanum on RSR mediated through the blockade of rRNA cleavage, the reduction of HIF-1α, CHOP, H3 clipping, and γH2AX proteins, and downregulation of RNase L and cathepsin L mRNAs. The rescue effects of caffeine were not mediated through cell death, but partly through autophagy, ER stress, and mitochondrial membrane potential.

In this study, we showed data that the quality of rRNA, such as cleaved 28S–18S rRNA fragments, affects the stability of mRNAs or/and protein synthesis, was disrupted by Taraxacummediated through RSR. The cleaved rRNA fragments might serve as non-coding RNAs or long non-coding RNAs, as well as N-truncated histone H3, involved into the epigenetic regulation in cells8,14,39,41. Many studies demonstrated the importance of histone H3 clipping in various biological processes, such as ESCs and PBMCs, regulation of monocyte-to-macrophage differentiation, a novel signature of human neutrophil extracellular traps, and regulation of myoblast differentiation15,16,17,18,19. One of the most conserved mechanisms by which cells sense viral infection mediated through dsRNA related RSR to trigger antiviral and inflammatory immune responses3,4,5,6. Our data showed that caffeine had the ability to rescue the effects of T. formosanum on the rRNA cleavage and histone H3 N-tail clipping events. Our findings suggest that it is important to identify whether the working mechanisms of rRNA cleavage and histone H3 N-tail clipping are mediated through similar or common enzymes. Hence, the differentiation of human THP-1 cells from monocyte to macrophage by phorbol 12-myristate acetate or rat C2C12 myoblast cells by 2% horse serum should be a good platform to examine the event timing of rRNA cleavage and histone H3 N-tail clipping in cells.

The RSR is activated by ribosomal impairment, which can disrupt protein synthesis, increase inflammatory signaling, and even lead to cell death if left unresolved2,42. RNase L of 2-5AMD cleaves a broad spectrum of host RNAs and the cleavage of rRNA does not inhibit the activity of the ribosome, whereas it might reduce ∼90% mRNA levels in the cell10. In addition, mRNA decay fragments via the NMD are to produce short peptides or functional truncated proteins43. In current study, we failed to provide direct evidence of RNase L involved in the cleavage of 28S rRNA or Cathepsin L involved in the clipping of histone H3 induced by T. formosanum and the rescue by caffeine. Compared with the well-known cleaved at amino acid residue A21 of histone H3, our western botting data showed that the N-truncated fragment was detected with dimethyl H3(K79) antibody, not H3P antibody. Therefore, the RIN value, the status of histone H3, and the alternative splicing of SRSF3 provide the hint on the quality of total RNA, the epigenetic status of H3, and the SRSF3 isoform for the expression levels of various mRNAs and proteins. It should be a challenge to identify non-coding RNAs from the cleavage site(s) of rRNAs and histone H3 and enzymes for the cleavage of rRNAs and the clipping of histone H3.

Our current data support the notion that T. formosanum induces apoptotic and necrotic cell death by disrupting the mitochondrial membrane potential and inhibiting autophagy formation. However, caffeine, a rescue agent, can reverse the effects of T. formosanum on mitochondrial membrane potential and autophagy. Nevertheless, caffeine fails to decrease cell death, resulting in an increasing percentage of late apoptosis and a decreasing percentage of necrosis. It would be interesting to investigate the order of autophagy, mitochondrial dysfunction, or apoptosis in the presence of T. formosanum, caffeine, or both combinations. It will elucidate the mechanisms of gene regulation from cleaved rRNAs and N-truncated histone H3 to mRNAs, and their respective cellular death and survival stresses.

Cellular stresses can lead to damage of RNAs by modifying, mutating, or cleaving them, resulting in the production of non-functional proteins that can be harmful to cells44. Ribosome collisions caused by various RNA damages are a key regulator of specific stress response pathways. For example, low doses of the translational elongation inhibitor anisomycin increase the frequency of ribosome collisions, triggering cellular recovery through integrated stress response. On the other hand, high doses of anisomycin lead to persistent ribosome stalling, causing apoptosis through the activation of MAPK signaling cascades due to 28S rRNA damage1,42. Three mRNA surveillance pathways, NMD, no-go decay, and non-stop decay, play crucial roles in resolving translation problems that can arise from faulty mRNAs or defective ribosomes41,45,46. Our current data indicate that T. formosanum induces a RSR, resulting in 28S rRNA cleavage to degrade most of mRNAs but which still partly express their proteins. T. formosanum also modulates ribosomal proteins and splicing pathways. Hence, investigating the involvement of non-coding RNA, ribosome collision, and mRNA surveillance pathways by T. formosanum in cells is a task for future research.

In summary, we first used the RIN value and 28S/18S ratio to confirm the Taraxacum-induced rRNA cleavage. In addition to the activation of p38 and JNK signaling pathways, we identified the hallmark major spliceosome, ribosomal proteins, TNF-α signaling via NF-κB, inflammatory response, and IL6-JAK-STAT3 signaling with the analysis of RNA sequencing. Caffeine was the only screening drug to rescue the effect of T. formosanum on RSR mediated through the blockade of rRNA cleavage, the reduction of HIF-1α, CHOP, H3 clipping, and γH2AX proteins, and downregulation of RNase L and cathepsin L mRNAs. The rescue effects of caffeine were not mediated through cell death, but partly through autophagy, ER stress, and mitochondrial membrane potential. By providing detailed information on Taraxacum-induced rRNA cleavage and N-truncated histone H3’s mechanisms of gene regulation, we hope to understand their respective cellular death and survival stresses.

Materials and methods

Preparation of aqueous extracts of Taraxacum

The dried whole T. mongolicum plants were purchased from the Yu Hong Biotechnology Company (Chiayi, Taiwan, ROC). T. formosanum is distributed mainly in the coastal areas of north Taichung in Taiwan. The whole T. formosanum plants were collected and identified by Dr. Rong-Chi Yang, Adjunct Associate Professor, School of Traditional Chinese Medicine, Chang Gung University (Taoyuan, Taiwan, ROC). T. mongolicum and T. formosanum of this study complied with the National Defense Medical Center’s, Taiwan’s, and international guidelines, including the IUCN Policy Statement on Research Involving Species at Risk of Extinction and the Convention on the Trade in Endangered Species of Wild Fauna and Flora. One hundred grams of the dried whole T. mongolicum and T. formosanumplants were crushed and prepared as previously described27. All Taraxacum powders were dissolved in boiled ddH2O to the desired final concentration (mg/ml) for cell treatment previously outlined27.

Cell culture and chemicals

HeLa, HepG2, Hep3B, Huh6, Huh7, and L02 cells were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Invitrogen, USA). Actinomycin D, acridine orange, AG1478, ALLM, ammonium chloride, BI-D1870, caffeine, ellagic acid, LY294002, MG132, ouabain, PD98059, PFT-α, resveratrol, SB203580, and Z-VAD-FMK were purchased from Sigma Aldrich (MO, USA).

Western blot analysis

Cells were lysed in Radio-Immunoprecipitation Assay buffer (100 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.1% SDS, and 1% Triton 100) at 4 °C. Protein concentrations were measured using a DC Protein Assay Kit (Bio-Rad Laboratories, USA). Proteins in aliquots of the lysate were separated by SDS-PAGE and electro-transferred to PVDF membranes (Immobilon-P; Millipore, Bedford, MA, USA) using a Bio-Rad Semi-Dry Transfer Cell. The blots were then incubated with primary antibodies against α-actinin (ACTN), p53, cyclin B1, p62 (Santa Cruz Biotechnology, USA); HIF-1α, CHOP, eIF2α, peIF2α, PARP, H3, H3P, Di-methyl-H3 (K79), LC3B (Cell Signaling, USA); γH2AX, cyclin D1, and p21 (Abcam, Cambridge, UK); SRSF3 (WH0006428M8, Sigma-Aldrich). Thereafter, the blots were incubated with HRP-conjugated secondary antibodies (anti-mouse IgG, AP192P; and anti-rabbit IgG, AP132P, Merck-Millipore). The immunoreactive proteins were detected using ECL™ Western Blotting Detection Reagents and Amersham Hyperfilm™ ECL (GE Healthcare, USA) as previously described47.

Reverse transcription-polymerase chain reaction (RT-PCR)

Total RNAs were isolated from HeLa cells using TRIzol reagent (Invitrogen). Reverse transcription for first strand cDNA synthesis was carried out using MMLV reverse transcriptase (Epicentre Biotechnologies, USA) with 1 µg of total RNA for 60 min at 37 °C. The PCR reactions were run on a Veriti Thermal Cycler (Applied Biosystems, USA), as previously described47. The PCR primers are listed below.

Primer name

Sequence (5’◊3’)

p53

Forward: 5’-CTCTGACTGTACCACCATCCACTA-3’

Reverse: 5’-GAGTTCCAAGGCCTCATTCAGCTC-3’

cyclin D1

Forward: 5’-ATGGAACACCAGCTCCTGTGCTGC-3’

Reverse: 5’-TCAGATGTCCACGTCCCGCACGTCGG-3’

p21

Forward: 5’-CTGAGCCGCGACTGTGATGCG-3’

Reverse: 5’-GGTCTGCCGCCGTTTTCGACC-3’

CHOP

Forward: 5’-CATTGCCTTTCTCCTTCGGG-3’

Reverse: 5’-GCCGTTCATTCTCTTCAGCT-3’

GAPDH

Forward: 5’-CTTCATTGACCTCAACTAC-3’

Reverse: 5’-GCCATCCACAGTCTTCTG-3’

18S

Forward: 5’-CAGCCACCCGAGATTGAGCA-3’

Reverse: 5’-TAGTAGCGACGGGCGGTGTG-3’

β-actin

Forward: 5’-GTGGGGCGCCCCAGGCACCA-3’

Reverse: 5’-CTCCTTAATGTCACGCACGATTTC-3’

Cyclin B1

Forward: 5’-GTTGATACTGCCTCTCCAAG-3’

Reverse: 5’-CTTAGTATAAGTGTTGTCAGTCAC-3’

HIF-1α

Forward: 5’-GAACCTGATGCTTTAACT-3’

Reverse: 5’-CAACTGATCGAAGGAACG-3’

IL-6

Forward: 5’-ATGAACTCCTTCTCCACAAGCGC-3

Reverse: 5’-CTACATTTGCCGAAGAGCCCTCA-3’

H3

Forward: 5’-ATGGCTCGCACTAAGCAAACTG-3’

Reverse: 5’-TCACGCCCTCTCTCCGCGGA-3’

RPL3

Forward: 5’-GTACTGCCAAGTCATCCGTG-3’

Reverse: 5’-CGGTGATGGTAGCCTTTCTG-3’

RPL4

Forward: 5’-GATACGCCATCTGTTCTGCC-3’

Reverse: 5’-CGCCAAGTGCCGTACAATTC-3’

RPL6

Forward: 5’-GCAAGAGGGTGGTTTTCCTG-3’

Reverse: 5’-GAGCAAACACAGATCGCAGG-3’

RPS2

Forward: 5’-GAGGCTACTGGGGGAACAAG-3’

Reverse: 5’-GTGTGGGTCTTGACGAGGTG-3’

RPS26

Forward: 5’-GAAGGAACAATGGTCGTGCC-3’

Reverse: 5’-TTACATGGGCTTTGGTGGGG-3’

SRSF3

Forward: 5’-ATGCATCGTGATTCCTGTCCATTG-3’

Reverse: 5’-CTATTTCCTTTCATTTGACCTAGATC-3’

p53β

Forward: 5’-ATGGAGGAGCCGCAGTCAGAT-3’

Reverse: 5’-TTTGAAAGCTGGTCTGGTCCTGA-3’

eIF2α

Forward: 5’-ACCTCAGAATGCCGGGTCTA-3’

Reverse: 5’-GTGGGGTCAAGCGCCTATTA-3’

PARP

Forward: 5’-CAGAAGTACGTGCAAGGGGT-3’

Reverse: 5’-GGCACTTGCTGCTTGTTGAAG-3’

RNase L

Forward: 5’-TGGACGGCTGGTCCTCTATG-3’

Reverse: 5’-CTGTGGGTTTGGGGGAAATGC-3’

Cathepsin L

Forward: 5’-GATGGAGGAGAGCAGTGTGG-3’

Reverse: 5’-GCAGCCTTCATTGCCTTGAG-3’

Fluorescence-activated cell sorting (FACS), apoptosis, autophagy, and mitochondrial membrane potential analysis

Early and late stages of apoptotic cells were evaluated by a fluorescein Phycoerythrin (PE)-Annexin V Apoptosis Detection Kit (BD Biosciences) according to the manufacturer’s protocol. Cells are stained with PE-Annexin V as well as 7-Amino-Actinomycin (7-AAD) to determine the effects of indicated drugs (T. formosanum and caffeine) on early apoptosis (PE-Annexin positive and 7-AAD negative), late apoptosis (PE-Annexin positive and 7-AAD positive), and necrosis (PE-Annexin negative and 7-AAD positive), as previously described47.

The acidic compartments of the cells were detected using acridine orange (Sigma, Cat. No. A8097) staining and measured by flow cytometry. As the protonated form of acridine orange accumulates inside acidic vesicles, it is a marker for the final steps of the autophagy process. Briefly, the cells were treated with the indicated lidocaine dosages for 24 h, stained with acridine orange (1 µg/mL) for 20 min at 37 °C, and then trypsinized for harvesting. Afterwards, the cells were washed once with PBS, resuspended in 400 µl of PBS, and then analyzed via flow cytometry (FACSCalibur, BD, Biosciences). The excitation wavelength was 488 nm and fluorescence were detected at 510–530 nm (green fluorescence, FL1) and 650 nm (red fluorescence, FL3). The data were analyzed using the CellQuest™ software program. The percentage of autophagy cells was calculated based on the number of cells present in the upper-left and upper-right quadrants, as previously described48.

Mitochondrial membrane potential was monitored using the MitoScreen (JC-1) kit (BD Pharmingen). After being treated with drugs, dead and live cells were collected, and JC-1 solution was added prior to the 15 min incubation. The cells were then washed twice with a binding buffer. Each sample was evaluated using the FACSCalibur flow cytometer and Cell Quest Pro software, as previously described49.

RNA sequencing and analysis

Total RNA of each sample was extracted using TRIzol Reagent and RNeasy Mini Kit (Qiagen). Total RNA of each sample was quantified and qualified by Agilent 2100/2200 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA) and NanoDrop (Thermo Fisher Scientific Inc.). 1 µg total RNA with a RIN value above 6.5 was used for following library preparation34. Next generation sequencing library preparations were constructed according to the manufacturer’s protocol and previously described50. The poly(A) mRNA isolation was performed using Poly(A) mRNA Magnetic Isolation Module or rRNA removal Kit. The mRNA fragmentation and priming were performed using First Strand Synthesis Reaction Buffer and Random Primers. First strand cDNA was synthesized using ProtoScript II Reverse Transcriptase and the second-strand cDNA was synthesized using Second Strand Synthesis Enzyme Mix. The purified double-stranded cDNA by beads was then treated with End Prep Enzyme Mix to repair both ends and add a dA-tailing in one reaction, followed by a T-A ligation to add adaptors to both ends. Size selection of Adaptor-ligated DNA was then performed using beads, and fragments of ~ 400 bp (with the approximate insert size of 300 bp) were recovered. Each sample was then amplified by PCR using P5 and P7 primers, with both primers carrying sequences which can anneal with flow cell to perform bridge PCR and P5/P7 primer carrying index allowing for multiplexing. The PCR products were cleaned up using beads, validated using an Qsep100 (Bioptic, Taiwan), and quantified by Qubit 3.0 Fluorometer (Invitrogen, Carlsbad, CA, USA). Then libraries with different indices were multiplexed and loaded on an Illumina HiSeq/Novaseq instrument according to manufacturer’s instructions (Illumina, San Diego, CA, USA). Sequencing was carried out using a 2 × 150 paired-end configuration; image analysis and base calling were conducted by the HiSeq Control Software + OLB + GAPipeline-1.6 (Illumina) on the HiSeq instrument. The sequences were processed and analyzed by GENEWIZ.

Differential Gene expression analysis and Gene Set Enrichment Analysis (GSEA)

Differential expression analysis used the DESeq2 Bioconductor package, a model based on the negative binomial distribution. the estimates of dispersion and logarithmic fold changes incorporate data-driven prior distributions, Padj of genes were set < 0.05 to detect differential expressed ones. To comprehensively evaluate gene expression changes related to T. formosanumand caffeine in HeLa cells, we performed ranked gene list analyses using GSEA software version 4.3.2. Genes were ranked based on the t-statistic and analyzed through pre-ranked analysis using the gene set enrichment analysis software, along with the Hallmark gene set from MSigDB v2023.1.Hs. For detailed software settings, please refer to the previous article5152.

Statistical analysis

The values are expressed as the mean ± SD of at least three independent experiments. All the comparisons between groups were conducted using the Student’s t-tests and comparison among multiple groups was conducted using analysis of variance (ANOVA) with SPSS 20.0 for Windows (SPSS, Chicago, IL). The statistical significance was set at p < 0.05.