roles of mmu pir 003399 in microcystin-leucine arginine ... · roles of mmu_pir_003399 in...
TRANSCRIPT
Roles of mmu_piR_003399 in Microcystin-Leucine
Arginine-Induced Reproductive Toxicity in the
Spermatogonial Cells and TestisLing Zhang,*,† Xiannan Meng,*,† Zou Xiang,‡ Dongmei Li,*,†,1 andXiaodong Han*,†,2
*Immunology and Reproduction Biology Laboratory & State Key Laboratory of Analytical Chemistry for LifeScience, Medical School, Nanjing University, Nanjing, Jiangsu 210093, China and; †Jiangsu Key Laboratory ofMolecular Medicine, Nanjing University, Nanjing, Jiangsu 210093, China; and ‡Department of HealthTechnology and Informatics, Faculty of Health and Social Sciences, The Hong Kong Polytechnic University,Hung Hom, Kowloon, Hong Kong, China1To whom correspondence should be addressed. Fax: þ86-25-83686497. E-mail: [email protected]. 2Fax: þ86-25-83686497. E-mail: [email protected].
ABSTRACT
Our previous work has demonstrated that microcystin-leucine arginine (MC-LR) is a potent toxin for the reproductivesystem of male mice and it exerts cytotoxicity in spermatogonial cells, resulting in the constitutional and functionalchanges of the mouse testis. The present study was designed to investigate the functions of P-element-induced wimpy(piwi)-interacting RNAs (piRNAs) in MC-LR-induced reproductive toxicity in male mice. We observed an increase in themmu_piR_003399 level in spermatogonial cells and mouse testicular tissues following treatment with MC-LR. Moreover, ourdata confirmed that cyclin-dependent kinase 6 (CDK6) was the target gene of mmu_piR_003399. Increases in theconcentration of mmu_piR_003399 were correlated with the reduced expression of CDK6 both in vitro and in vivo.mmu_piR_003399 induced cell cycle arrest at the G1-phase, down-regulated sperm counts and sperm motility, andcompromised sperm morphology. On the contrary, suppressing the expression of mmu_piR_003399 could substantiallyattenuate MC-LR-induced pathology in mice including cell cycle arrest, reduced mature sperm counts, sperm viability lossand abnormal sperm morphology. Furthermore, our data supported that mmu_piR_003399 existed in mouse serum andplasma, and its level was increased in MC-LR-treated mice. In conclusion, we demonstrate that mmu_piR_003399 plays acrucial role in regulating MC-LR-induced reproductive toxicity.
Key words: microcystin-leucine arginine; piRNA; mouse; cyclin-dependent kinase 6; serum; plasma.
Microcystins (MCs) are produced by various cyanobacterial spe-cies in eutrophic water and are responsible for the potential poi-soning of animals and human beings. Conventional drinkingwater treatments, such as flocculation, coagulation, and sandfiltration, can remove cyanobacterial cells. However, they donot efficiently remove the dissolved MCs (Pitois et al., 2001).Microcystin-leucine arginine (MC-LR) is the most frequentlyreported and most toxic congener. Previous studies have con-firmed that MC-LR exerts toxic effects to the liver (Clark et al.,2007), brain (Kist et al., 2012), heart (Milutinovic et al., 2006),
kidney (Qin et al., 2012), and intestine (Zegura et al., 2008). It wassubsequently found to be toxic to reproductive organs especiallythe male reproductive system using zebrafish (Su et al., 2016),medaka fish (Trinchet et al., 2011), mouse (Chen et al., 2017) andrat (Chen et al., 2013) models. Our previous work supports thatMC-LR contributes to the viability loss in rat primary spermato-gonial cells even at a very low concentration (nmol/l [nM] scale)in vitro (Zhou et al., 2012). We also observed reduced prolifera-tion of GC-1 cells (a mouse cell line that shares characteristicsof a developmental stage between type B spermatogonia and
VC The Author 2017. Published by Oxford University Press on behalf of the Society of Toxicology.All rights reserved. For Permissions, please e-mail: [email protected]
1
TOXICOLOGICAL SCIENCES, 2017, 1–12
doi: 10.1093/toxsci/kfx209Advance Access Publication Date: October 12, 2017Research Article
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
primary spermatocytes) after exposure to MC-LR at higherconcentrations (lmol/l scale) (Zhou et al., 2014). Furthermore,MC-LR induces enhanced apoptosis and impaired proliferationof spermatogenic cells in vivo (Zhou et al., 2013). Nevertheless,the molecular mechanisms of the reproductive toxicity of MC-LR remain to be clarified. It is thus worthy to investigate themechanism underlying MC-LR-mediated toxicity induction ingerm cells.
P-element-induced wimpy (piwi)-interacting RNAs (piRNAs)are a novel class of small noncoding RNAs of 26–32 nucleotides(nt) in length (Slotkin and Martienssen, 2007). The piRNAs aremodified at 30 termini by 20-O-methylation, and processed bypiwi-clade argonaute proteins. In mammals, piRNAs are highlyabundant in the testis, the male gonad for spermatogenesis.Mutation of the piRNA pathway genes in mice, including thoseencoding the piwi family and accessory proteins, leads to malesterility (Deng and Lin, 2002). The piwi proteins belong to theargonaute family, and are specifically expressed during the de-velopment of the animal germline. The piwi/piRNA machineryfunctions in transposon silencing, maintenance of genome in-tegrity, DNA methylation, epigenetic regulation, and germlinedevelopment (Hirakata and Siomi, 2016). Nevertheless, fewstudies have been conducted in view of the function of piRNAsin the mouse reproductive system after exposure to environ-mental toxins.
Some previous studies on various tumor types confirmedthat dysregulated piRNAs are involved in carcinogenesis (Nget al., 2016). In addition, our previous study has demonstratedthat the expression piRNAs is altered in the testis of male off-spring of the mice that have been exposed to MC-LR. We specu-lated that MC-LR could induce passive reproductive toxicity inthe male offspring through regulating piRNA levels (Zhang et al.,2017). However, the underlying molecular mechanisms associ-ated with piRNAs and the MC-LR-induced reproductive toxicityremain unclear. Thus, the aim of the present study was to in-vestigate the effect of MC-LR on piRNA expression in the GC-1cells and the testis, and whether this could provide a mechanis-tic explanation for MC-LR-induced reproductive toxicity.
MATERIALS AND METHODS
Main chemicals. MC-LR with a purity of �95% was purchasedfrom Alexis Biochemicals (Lausen, Switzerland). MC-LR was dis-solved in methanol at 10 mg/ml followed by dilution to 1 mmol/l in PBS as the stock solution which was kept at �20 �C and usedwithin 6 months. Dulbecco’s Modified Eagle’s medium (DMEM),penicillin, streptomycin sulfate and Bouin’s solution were pur-chased from Sigma-Aldrich (St Louis, Missouri, USA). The FBSwas purchased from Dojindo Molecular Technologies(Kumamoto, Japan). The dual-luciferase reporter assay systemwas purchased from Promega (Madison, Wisconsin, USA). SYBRGreenER qRT-PCR kit and lipofectamine 2000 were purchasedfrom Invitrogen (Shanghai, China). qRT supermix for qRT-PCRwas obtained from Vazyme (Jiangsu, China). Stem-loop RTprimer and primer pairs were purchased from Genscript(Jiangsu, China) (Supplementary Tables 1 and 2). The cel-miR-39and its primers were obtained from Ribobio (Guangzhou,China). Mouse anticyclin-dependent kinase 6 (CDK6), rabbitantipiwi-like RNA-mediated gene silencing 2 (mili), rabbit anti-piwi-like RNA-mediated gene silencing 1 (miwi), mouse antireti-noblastoma tumor suppressor protein (Rb), rabbit antiphospho-Rb (p-Rb), and rabbit antiGAPDH were purchased from Abcam(Cambridge, Massachusetts, USA). Horseradish peroxidase-conjugated goat-antirabbit/mouse IgG purchased from Boster
(Wuhan, China) was used as the secondary antibody. ThepiRNA inhibitor, piRNA mimic, their negative controls (NCs),lentiviral vector-pGLV-CDK6-GFP-puro (LV-CDK6), lentiviral vec-tor-NC (LV-NC) and lentiviral vector-pGLV-piRNA-GFP-puro (LV-piRNA) were purchased from Genechem (Shanghai, China)(Supplementary Table 3).
Animals and treatment. A 7-week-old male BALB/c mice (specificpathogen-free) (18–22 g) were purchased from the ExperimentalAnimal Center of the Academy of Military Medical Science,Beijing, China. The animal experiments were performed accord-ing to the Guide for the Care and Use of Laboratory Animals(The Ministry of Science and Technology of China, Nanjing,China, 2006) and all experimental protocols were approved un-der the animal protocol number SYXK (Su) 2009-0017 by theAnimal Care and Use Committee of Nanjing University.
Mice were randomized into 8 groups (n ¼ 6 for each)(Supplementary Table S4). 4 groups of mice were each intraperi-toneally injected with MC-LR at 7.5, 15, or 30 lg/kg body weight,or an identical volume of 0.9% saline every day for 2 weeks. Micein the other 4 groups each first received 8 ll of LV solution con-taining LV-CDK6, LV-NC or LV-piRNA (Supplementary Table 3)via efferent duct injection (Supplementary Figure 1). The con-centration of the LV solution was 1 � 109 transducing units/ml.1 week later, the mice in groups 5 and 6 were intraperitoneallyinjected with MC-LR at 15 lg/kg body weight daily for 2 weeks.
Mouse plasma and serum samples were collected frominfraorbital plexus of the mice and stored at room temperatureat least for 30 min. The plasma and serum were then obtainedby centrifugation (3000 rpm, 10 min). Testicular tissues were col-lected immediately after blood collection.
Cell culture and treatment. Mouse GC-1 cells were obtained fromProf. Zuoming Zhou (Nanjing Medical University, Jiangsu,China). GC-1 cells were grown at 37 �C in a humidified 5% CO2
atmosphere in DMEM containing 10% FBS. Cells were subcul-tured every 2 days. GC-1 cells were cultured either in the ab-sence or presence of MC-LR at various concentrations (62.5, 125,250, 500, 1000, or 2000 nM). Control cells were treated with anidentical volume of the diluent. All the in vitro experimentswere repeated 3 times.
The cells were transfected with piRNA mimic, mimic control,inhibitor and inhibitor control using Lipofectamine 2000 for 6 haccording to the manufacturer’s transfection protocol. Next, thecells were incubated without or with MC-LR (500 nM) for 24 h inDMEM supplemented with 10% FBS.
For preparing LV-infected GC-1 cells, approximately 5 � 105
cells were plated into each well of 6-well culture plates and in-cubated for 24 h in DMEM supplemented with 10% FBS at 37 �Cand 5% CO2. Next, 10 ll of the LV solution was added to eachwell and the cells were incubated for 12 h. Cells were next incu-bated for 2 days in DMEM supplemented with 10% FBS.
Illumina high-throughput sequencing. Total RNA was extractedfrom GC-1 cells using TRIzol reagent (Takara, Dalian, China)according to the manufacturer‘s instructions. After purificationof RNA molecules smaller than 45 nt by polyacrylamide gel elec-trophoresis, the 50 and 30-RNA adaptors were ligated to the 50
and 30 ends of the RNAs. The small RNA molecules were thentranscribed to single-stranded cDNA and amplified using theadaptor primers. The PCR products were supplied for the clustergeneration and sequenced using HiSeq 2000 System (TruSeqSBS KIT-HS V3, Illumina) at Beijing Genomics Institute(Shenzhen, China). Image files generated by the sequencer were
2 | TOXICOLOGICAL SCIENCES, 2017
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
processed to produce digital-quality data. The subsequent pro-cedures included summarizing data production, evaluating se-quence quality and depth, calculating small RNA lengthdistribution and collecting clean reads. The final clean readswere obtained by removing short reads (<10 nt) and contami-nant reads (including reads with 50 primer contaminants, readswith 50 connected to poly A and reads without 30 primer). Aftermasking the adaptor sequences, the clean reads were aligned tothe reference Mus musculus genome. The sequences that corre-spond to known piRNAs were determined by perfect sequencematching to the pirnabank database (http://pirnabank.ibab.ac.in/; last accessed October 23, 2017) (Sai Lakshmi and Agrawal,2008). miRNA sequences were determined by matching to themiRNA database (miRBase 19.0), and other small RNAs (rRNA,tRNA, and snoRNA) were annotated by blasting against theRfram database (http://sanger.ac.uk/software/Rfam; lastaccessed October 23, 2017). All data have been uploaded to theGEO database (Accession number: GSE100987).
RNA isolation and qRT-PCR assays. Total RNAs from GC-1 cellsand mouse testis were extracted using TRIzol reagent. For liquidsamples (plasma and serum), RNAs were isolated from all sam-ples by using the miRNeasy kit (Qiagen, Shanghai, China)according to the manufacturer‘s protocol. The qRT-PCR assayswere conducted on the LightCycler 480 RT-PCR system (Roche,Shanghai, China) according to the manufacturer’s instructions.We have provided a detailed description of the method in theSupplementary Material.
Cell viability assay. Purified cells were digested and plated in a96-well culture plate at 1 � 104 cells per well and cultured in theserum-free growth medium. 10 ll of CCK-8 solution was addedto each well and the cells were incubated at 37 �C for a further2 h. Absorbance, expressed as optical density, was measured at450 nm with a multidetection micro plate reader (Versamax,Chester, Pennsylvania).
Western blot. Total protein was isolated from the cells followingvarious treatments. Briefly, protein concentration was deter-mined by the Bradford method. We have provided a detailed de-scription of the method in the Supplementary Material.
Immunohistochemistry. The testes of mice were fixed, embedded,and cut into sections based on previously established protocolswith slight modification (Chen et al., 2017). We have provided adetailed description of the method in the SupplementaryMaterial.
RNA-binding protein immunoprecipitation. RNA was isolated fromGC-1 cells by using the RNA-binding protein immunoprecipita-tion (RIP) Kit (Merck Millipore, Germany) according to the manu-facturer‘s protocol (Supplementary Figure 2). We have provideda detailed description of the method in the SupplementaryMaterial.
Flow cytometric analysis. Cells (1 � 105) were collected and fixedin 75% ethanol at �20 �C overnight. The cells were then har-vested, treated with RNase A (100 ng/ml) for 30 min, and stainedwith propidium iodide (50 ng/ml) for 15 min. After staining,samples were analyzed for cell cycle distribution with a MoFloXDP flow cytometer (Beckman Coulter, Brea, California). Datawere analyzed using Flow Jo. Experiments were performed inde-pendently at least thrice in duplicates.
Dual-luciferase reporter assay. Luciferase assays were carried outto confirm the interaction between mmu_piR_003399 and the30-UTR of CDK6 (Table S5) using the pmirGLO Vector. We haveprovided a detailed description of the method in the SI.
Histopathological evaluation. Testes were fixed by immersion inBouin’s fixative for 24 h and subsequently transferred to 70%ethanol until processing. Fixed tissues were then dehydrated ingraded ethanol, cleared in xylene, and mounted in paraffin, fol-lowed by sectioning at 5 lm and staining with hematoxylin andeosin for light microscopy. Histopathological damage wasassessed by examination of 50 seminiferous tubule cross sec-tions per testis, and evaluated by a board-certified pathologist(W.D.D.) who was blinded to the treatments.
Sperm motility measurement, sperm counts, and sperm morphologyassessment. Sperms were collected as quickly as possible aftersacrifice. We had 6 mice in each treatment group. The testisand epididymis were dissected out, cleaned of adherent fat,and weighed. The caudae epididymis was minced in PBS at37 �C for 20 min and sperm suspensions were prepared. Thesuspensions were filtered through an 80-mm nylon mesh, andthe filtrates were used as described below. The person whocounted the sperms was blinded to treatment. The data areshown in Supplementary Tables 6 and 7. We have provided adetailed description of the method in the SupplementaryMaterial.
Statistical analysis. All calculations and statistical analyses wereperformed using SPSS for windows version 13.0. One-way analy-sis of variance was used to analyze the difference betweengroups, followed by Dunnett’s t-test. p < .05 is regarded as sta-tistically significant.
RESULTS
mmu_piR_003399 Is Most Robustly Increased in GC-1 Cells AfterExposure to MC-LRGC-1 cells were cultured either in the absence or presence of MC-LR at various concentrations. Cell viability was decreased at anMC-LR concentration of 500 nM or above, which became more pro-nounced with time (Supplementary Figure 3). Next, we analyzedthe RNA from GC-1 cells treated with 500 nM MC-LR for 24 h.
After the adapter sequences were trimmed, 10 824 057 and11 566 902 small RNA reads that were smaller than 45 nt wereobtained from control and MC-LR-treated cells, respectively. Weidentified 220 unique reads (from a total of 244 066 reads) pre-sent in control GC-1 cells and 230 unique reads (from a total of359 871 reads) present in MC-LR-treated cells (Figure 1).
All differentially expressed piRNAs are listed in Table S8.piRNAs with changes (> 1.33-fold change or < �1.33-foldchange, reads > 4) were selected for further qRT-PCR analysesand to obtain a deeper insight into their differential expression(Supplementary Figure 4). In this study, mmu_piR_003399 waschosen for further investigation, as the expression ofmmu_piR_003399 was most robustly increased in GC-1 cells af-ter exposure to MC-LR by qRT-PCR.
CDK6 Is Directly Regulated by mmu_piR_003399First, to determine the specificity of the primer set, the no-template control for mmu_piR_003399 was assessed simulta-neously. The piRNA was consistently and efficiently amplified,while the no-template control required a much higher cycle range
ZHANG ET AL. | 3
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
to be amplified, confirming the specificity of the detection ofmmu_piR_003399 by qRT-PCR in GC-1 cells (Supplementary Figure5A). The dissociation curve analysis showed that there was only 1peak in each dissociation curve (Supplementary Figs. 5B–D). Next,TA cloning combined with sequencing was performed to confirmthe qRT-PCR results. The sequences of the products were shown tobe completely identical to the sequences of mmu_piR_003399(Supplementary Figure 5E). These results ruled out the presence ofprimer-dimers or nonspecific products.
The expression of mmu_piR_003399 was up-regulated afterexposure to an MC-LR concentration of 500 nM or above(Figure 2A). The regulation of protein-coding genes by piRNAshas been suggested by recent studies. piRNA induces mRNAdeadenylation and decay through imperfectly base pairing withspecific sites in the 30-UTR of target transcripts (Gou et al., 2014;Watanabe et al., 2015). We hypothesized that altered expressionof piRNAs in GC-1 cells after MC-LR treatment would affect theexpression of downstream target genes. Using the miRanda(John et al., 2004) program, we searched for potential targetingsites of mmu_piR_003399 within the 30-UTRs of mRNAs. CDK6stood out as one of the target mRNAs of mmu_piR_003399. Weobserved down-regulated expression of CDK6 after exposure toMC-LR (Figure 2B). The binding site for mmu_piR_003399 in the30-UTR of CDK6 is shown in Figure 2C.
We next tried to confirm the binding of mmu_piR_003399 tothe 30-UTR of CDK6 in 293 T cells using the dual-luciferase re-porter gene assay system. CDK6 30-UTR and CDK6 30-UTR withmutations in the predicted mmu_piR_003399 binding site werecloned separately into the luciferase reporter system (pmirGLOvector). The pmirGLO vectors were then co-transfected with ei-ther mmu_piR_003399 mimic or its NC. Luciferase activity of the
system containing wild-type, but not mutated, CDK6 30-UTRwas substantially decreased after co-transfection withmmu_piR_003399 (Figure 2D), which supported the predictionthat CDK6 was the target gene of mmu_piR_003399.
It was reported that the piwi complex harbors both piRNAsand mRNAs. The piwi proteins are small RNA-guided nucleasesthat can catalyze cleavage of target nucleic acids identified bybase-pair interactions. In this study, we isolated mili/RNA com-plexes and miwi/RNA complexes in GC-1 cells by RIP with anantimiwi or antimili antibody. The mili and miwi proteins wereexpressed in the GC-1 cells, and the protein bound to beads dur-ing RIP process (Figure 2E). We could confirm the presence ofboth mmu_piR_003399 and the mRNA of CDK6 in mili and miwiprotein complexes (Figure 2F).
mmu_piR_003399 Regulates the Proliferation of GC-1 Cells via theCDK6-Rb PathwayAn increase of mmu_piR_003399 and a decrease of CDK6 in GC-1 cells were found after exposure to MC-LR (Figs. 3A and 3B).CDK6 phosphorylates Rb and inhibition of CDK6 could lead tocell cycle arrest (Knudsen and Knudsen, 2008). We next tried tomeasure the protein expression of CDK6, p-Rb, and Rb, and theproliferation of GC-1 cells. The expression of CDK6 and p-Rbwas consistently decreased in MC-LR-treated GC-1 cells(Figure 3C). The number of cells at G1-phase was increased afterMC-LR exposure (Figure 3D). Furthermore, cell proliferation wasdecreased after MC-LR exposure (Figure 3E).
To examine the regulatory effects of mmu_piR_003399 andCDK6 on cellular functions, we transfected GC-1 cells with apiRNA inhibitor to suppress the expression of mmu_piR_003399or piRNA mimic to enhance the expression of mmu_piR_003399.
Figure 1. The annotation of small RNAs in GC-1 cells. The annotation of piRNAs in the total reads of small RNAs in control (A) and MC-LR-treated cells (B). The annota-
tion of unique reads of small RNAs in control (C) and MC-LR-treated cells (D).
4 | TOXICOLOGICAL SCIENCES, 2017
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
Overexpression of mmu_piR_003399 decreased the level ofCDK6 and p-Rb, whereas down-regulation of mmu_piR_003399increased the level of CDK6 and p-Rb, indicating thatmmu_piR_003399 regulates CDK6 at the level of RNA stabilityand its translation (Figs. 3A–C). Moreover, a decrease of cell pro-liferation was observed by the overexpression ofmmu_piR_003399 in GC-1 cells (Figs. 3D and 3E). We next
confirmed that the expression of CDK6 and p-Rb was up-regulated (Figs. 3A–C) and cell proliferation was increased (Figs.3D and 3E) if intracellular mmu_piR_003399 was suppressed incells treated with MC-LR.
We next determined whether decrease of CDK6 in GC-1 cellssuppressed cell proliferation. Inhibition of CDK6 reduced the ex-pression of p-Rb (Figure 3C). Down-regulation of CDK6 in GC-1
Figure 2. CDK6 is directly regulated by mmu_piR_003399. The expression of mmu_piR_003399 (A) and CDK6 (B) was measured by qRT-PCR in GC-1 cells after exposure
to various concentrations of MC-LR for 24 h. C, Nucleotide sequences of murine mmu_piR_003399 and the 3’-UTR of the CDK6 gene which encompasses the
mmu_piR_003399 seed match region. D, Dual-luciferase reporter assay was carried out in 293 T cells following co-transfection of mmu_piR_003399 mimic (mimic) or its
NC (mimic-NC) together with a wild-type CDK6 vector (WT), mutant CDK6 vector (MUT), or empty vector (empty) as indicated. Firefly luciferase activity was normalized
based on the renilla luciferase activity. E, The presence of piwi-like RNA-mediated gene silencing 2 (mili) and piwi-like RNA-mediated gene silencing 1 (miwi) proteins
in GC-1 cell lysate was confirmed by RNA-binding protein immunoprecipitation (RIP) using antimili or antimiwi antibodies, or normal rabbit IgG as the control anti-
body. The protein was next analyzed by western blot. F, The presence of mmu_piR_003399 and the mRNA of CDK6 was confirmed by RIP using anti-mili or -miwi anti-
bodies, or normal rabbit IgG as the control antibody. The RNA was next analyzed by RT-PCR. mmu_piR_003399 and the mRNA of CDK6 were present in the input and
immunoprecipitated fractions using anti-mili or -miwi antibodies, but not using normal rabbit IgG. Results are expressed as mean 6 SD (n¼3). **p < .01.
ZHANG ET AL. | 5
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
Figure 3. mmu_piR_003399 and CDK6 regulate cellular functions in GC-1 cells. GC-1 cells were incubated without (Con) or with 500 nM MC-LR for 24 h. Alternatively,
GC-1 cells were transfected with piRNA mimic (mimic), mimic-NC, inhibitor (inhibitor) or inhibitor negative control (inhibitor-NC) for 6 h followed by incubation with-
out or with 500 nM MC-LR for 24 h. In some experiments, GC-1 cells were transfected with CDK6 down-regulation lentiviral vector (LV-CDK6) or lentiviral vector nega-
tive control (LV-NC) for 12 h followed by incubation for 2 days. Expression levels of mmu_piR_003399 (A) and CDK6 (B) in GC-1 cells were measured by qRT-PCR. C,
Protein levels of CDK6, retinoblastoma tumor suppressor protein (Rb), p-Rb, and Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in GC-1 cells were measured by
western blot. Protein expression levels were quantified with ImageJ (lower panel). GAPDH was used as a loading control. D, Cell cycle analysis in GC-1 cells was per-
formed by flow cytometry (lower panel). E, Measurement of cell viability was carried out with the CCK-8 assay. Results are expressed as mean 6 SD (n¼3). **p < .01.
6 | TOXICOLOGICAL SCIENCES, 2017
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
cells also increased cell number at G1-phase (Figure 3D) and de-creased cell proliferation of (Figure 3E).
MC-LR Contributes to Reproductive Toxicity In VivoExpression of mmu_piR_003399 and CDK6 was determined intesticular tissues from either control mice or mice that hadbeen exposed to MC-LR. Expression of mmu_piR_003399 was in-creased (Figure 4A) and levels of CDK6 and p-Rb were decreased(Figs. 4B and 4C) when higher concentrations of MC-LR (15 and30 lg/kg) were delivered. Furthermore, the regulated expressionof CDK6 in testicular tissues was confirmed by immunohisto-chemistry. Expression of CDK6 was decreased in testicular tis-sues following MC-LR exposure at 15 and 30 lg/kg(Supplementary Figs. 6A–D).
We further observed decreased paired testis/body weight in-dex of the mice after exposure to MC-LR at 15 lg/kg and 30 lg/kg(Figure 4D). We assessed the number, motility and morphologyof the sperms. Upon exposure to 15 lg/kg and 30 lg/kg MC-LR,the mice produced fewer sperms (Figs. 4E and 4F), and demon-strated decreased numbers of progressively motile sperms(Figure 4G) and increased numbers of sperms with an abnormalmorphology (Figure 4H).
The testes from the exposed mice were assessed by light mi-croscopy. A testis section from the control mouse contained anarray of developing germ cells with spermatids adjacent to thelumen of the seminiferous tubules and an intact seminiferousepithelium (Figure 4I). Tissue morphology after an exposure of7.5 lg/kg MC-LR was indistinguishable from the control, with noevidence of damage or sloughing of germ cells (Figure 4J). Incontrast, the testes after an exposure of 15 and 30 lg/kg MC-LRshowed the following changes in seminiferous tubules: (1) for-mation of multinucleated giant cells, (2) abnormal positioningof germ cells within the epithelium, and (3) loss of germ cells(Figs. 4K and 4L).
Mmu_piR_003399 Regulates MC-LR-Induced Reproductive Toxicityby Targeting CDK6Mice were transfected either with LV-piRNA to down-regulatethe expression of mmu_piR_003399, or with LV-NC as control.Next, mice were exposed to 15 lg/kg MC-LR. Levels of both CDK6and p-Rb were up-regulated when mmu_piR_003399 was down-regulated (Figs. 5A and 5C). Expression of CDK6 was increasedcompared with that in the LV-NC þ MC-LR mouse testis(Supplementary Figs. 6E and 6F). Moreover, the paired testis/body weight index, sperm counts, and sperm motility valueswere increased and the sperm deformity index was lowered inthe LV-piRNA þ MC-LR mice (Figs. 5D–H). Accumulation of mul-tinucleated giant cells and the disordered germ cell arrange-ment were observed in the LV-NC þ MC-LR mouse testis(Figure 5I). The testes of LV-piRNA þ MC-LR mice hardly showedany damage or sloughing of germ cells, which was indistinguish-able from the tissues in mice without MC-LR exposure (Figure 5J).
Furthermore, the mice were transfected with LV-CDK6 todown-regulate the expression of CDK6. Down-regulation ofCDK6 in the mouse testis suppressed the expression of p-Rb(Figs. 5B and 5C). Decreased CDK6 levels were also observed byimmunohistochemistry (Supplementary Figs. 6G and 6H).Down-regulation of CDK6 reduced the paired testis/body weightindex, sperm counts, percentage of progressively motilesperms, and the percentage of sperms with normal morphology(Figs. 5D–H). Furthermore, the multinucleated giant cells andabnormal positioning of germ cells within the epithelium wereobserved in the LV-CDK6 mouse testis compared with control(Figs. 5K and 5L).
Mmu_piR_003399 Is Increased in Mouse Plasma and Serum afterExposure to MC-LRA significant increase of mmu_piR_003399 in the serum andplasma was observed in mice after treatment with either 15 or30 lg/kg MC-LR treatment (Figs. 6A and 6B). Incubation ofplasma or serum at room temperature for up to 24 h or up to8 cycles of freeze-thawing (Figs. 6C and 6D) had minimal effecton levels of mmu_piR_003399.
Given that plasma and serum have been reported to containhigh levels of RNase activity, we sought to determine whetherthe stability of piRNA was intrinsic to its small size or chemicalstructure, or whether the stability came from additional extrin-sic factors. We introduced cel-miR-39, mmu_piR_003399 or wa-ter, into plasma or serum either before or after the addition of adenaturing solution that inhibits RNase activity (Figure 6E). cel-miR-39 and mmu_piR_003399 were rapidly degraded followingdirect addition to plasma or serum, as compared with its addi-tion to plasma or serum that had received a denaturing reagent(Figs. 6F and 6G). These results could confirm the presence ofRNase activity in plasma or serum, and suggested thatmmu_piR_003399 in plasma or serum may be protected fromdegradation by binding to an unknown factor.
DISCUSSION
The World Health Organization published a guideline of 1 lg/l ofMCs in drinking water, and established a tolerable daily intakeguideline of 0.04 lg/kg/day. It is found out that the concentra-tion of MCs in Lake Taihu in China reached 15.6 lg/l (Song et al.,2007). MCs with varying concentrations from 10 to 500 lg/l werealso detected in eutrophic lakes in America (Backer et al., 2010).An important exposure route for humans is through the contami-nation of drinking water with MC-LR. Moreover, MC-LR does notreadily penetrate the cell membrane via simple diffusion butrather requires the presence of multispecific organic anion trans-porting polypeptides (Oatp) for active uptake. We have previouslyshown that Oatp1a5 can mediate the transfer of MC-LR into thegonadotropin-releasing hormone neurons (Ding et al., 2017). Wefound out that Oatp1a5 was expressed in both sertoli cells andspermatogonial cells, and MC-LR could enter into sertoli cells andspermatogonial cells in vitro (Chen, et al., 2017). Another studyalso showed a disruption of the blood-testis barrier integrity byMC-LR and confirmed the accumulation of MC-LR in the testis(Chen et al., 2016). These results demonstrated that MC-LR mayexert direct reproductive toxicity.
Mili and miwi are 2 of the piwi family proteins, which areexpressed in the gonad of mice and regulate the spermatogene-sis. Previous studies have revealed that mili and miwi arepiRNA-guided endonucleases essential for piRNA biogenesisand retrotransposon repression (De Fazio et al., 2011). Also, therole of piRNAs in mediating mRNA cleavage is evidenced inCaenorhabditis elegans (Lee et al., 2012) and Drosophila mela-nogaster (Brower-Toland et al., 2007). In addition, a previousstudy reported miwi/piRNA-mediated mRNA cleavage events inthe mouse testis and demonstrated the functional importanceof the temporal cleavage of piRNA target mRNAs for spermio-genesis (Zhang et al., 2015). Vourekas et al suggest that the sizeranges of piRNAs that can bind to mili and miwi are 23–31 (peakat 26–27) and 25–33 (peak at 29–30) nt, respectively (Vourekaset al., 2012). The length of mmu_piR_003399 is 29 nt, which fallsbetween overlapping lengths. This may explain whymmu_piR_003399 could bind to both mili and miwi.
In this study, we found that CDK6 bound tommu_piR_003399 by a 9-base pair region. However, many other
ZHANG ET AL. | 7
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
Figure 4. Expression of mmu_piR_003399 and its reproductive toxicity are correlated with MC-LR dose. Mice were intraperitoneally injected with various concentrations
of MC-LR. Expression levels of mmu_piR_003399 (A) and CDK6 (B) in mouse testicular tissues were measured by qRT-PCR (n¼6). C, Protein levels of CDK6, retinoblas-
toma tumor suppressor protein (Rb), p-Rb and GAPDH in mouse testicular tissues were measured by western blot. Protein expression levels were quantified with
ImageJ (lower panel n¼ 3). GAPDH was used as a loading control. The paired testis/body weight index (D), sperm counts per epididymis (E), sperm counts per milligram
epididymis (F), sperm motility (G), and sperm malformation (H) were assessed (n¼6). I, The seminiferous tubules in 0.9% saline-treated mice showed normal architec-
ture. J, The seminiferous tubules in mice treated with 7.5 lg/kg MC-LR were indistinguishable from the control. K, L, The seminiferous tubules in mice treated with
15 lg/kg and 30 lg/kg MC-LR demonstrated the formation of multinucleated giant cells (arrows), abnormal positioning of germ cells and the loss of germ cells. Results
are expressed as mean 6 SD. Bar ¼ 100 lm. **p < .01.
8 | TOXICOLOGICAL SCIENCES, 2017
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
Figure 5. The functions of mmu_piR_003399 is validated in mouse testes. Mice received the lentiviral vector that down-regulates mmu_piR_003399 (LV-piRNA) or a len-
tiviral control vector (LV-NC) together with MC-LR. Alternatively, mice received the lentiviral vector that down-regulates CDK6 (LV-CDK6) or LV-NC. A, Expression levels
of mmu_piR_003399 CDK6 in mouse testicular tissues were measured by qRT-PCR (n¼6). B, Expression levels of CDK6 in mouse testicular tissues were measured by
qRT-PCR (n¼6). C, Protein levels of CDK6, retinoblastoma tumor suppressor protein (Rb), p-Rb and GAPDH in the mouse testicular tissues were measured by western
blot. Protein expression levels were quantified with ImageJ (lower panel, n¼ 3). GAPDH was used as a loading control. The paired testis/body weight index (D), sperm
counts per epididymis (E), sperm counts per milligram epididymis (F), sperm motility (G), and sperm malformation (H) were assessed (n¼6). I, The seminiferous
tubules in the LV-NC þ MC-LR-treated mice showed the formation of multinucleated giant cells (arrow) and abnormal positioning of germ cells. J, The seminiferous
tubules in the LV-piRNA þMC-LR-treated mice hardly showed any damage. K, The seminiferous tubules in the LV-NC-treated mice showed the normal architecture. L,
The seminiferous tubules in the LV-NC þ MC-LR-treated mice showed the formation of multinucleated giant cells (arrow) and abnormal positioning of germ cells.
Results are expressed as mean 6 SD. Bar ¼ 100 lm. **p < .01.
ZHANG ET AL. | 9
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
predicated mRNAs can be matched to this 9-base pair region.We detected them in GC-1 cells following the exposure of MC-LR. However, little significant changes of these mRNAs werefound. Moreover, when GC-1 cells were transfected with piRNAmimic to up-regulate the expression of mmu_piR_003399, orpiRNA inhibitor to down-regulated the expression ofmmu_piR_003399, only CDK6 levels were substantially modu-lated consistently. We also confirmed the binding ofmmu_piR_003399 to the 30-UTR of CDK6 using the dual-
luciferase reporter gene assay system. We confirmed that CDK6was the target gene of mmu_piR_003399.
CDK6 is a member of the cyclin-dependent kinase familywhich plays key roles in the cell-cycle transitions and phases ofall eukaryotic organisms (Ekholm and Reed, 2000). CDK6 cancause phosphorylation of the Rb protein. The Rb protein bindswith the transcription factor E2F and lessens the transcriptionactivity of E2F, which can promote cell cycle progression fromG1 to S phase. The p-Rb releases E2F to induce E2F-target gene
Figure 6. MC-LR regulates the levels of mmu-piR-003399 in plasma and serum. Expression levels of mmu-piR-003399 in serum (A) or plasma (B) were measured by qRT-
PCR after the mice were intraperitoneally injected with various concentrations of MC-LR. Expression levels of mmu-piR-003399 in serum (C) or plasma (D) after pro-
longed room temperature incubation or multiple freeze-thaw cycles were measured by qRT-PCR. E, The experimental procedures performed in F is indicated by a flow
chart. cel-miR-39 (from C. elegans), mmu_piR_003399 or water (control) was added to aliquots of fresh serum (F) or plasma (G) followed by incubation at room tempera-
ture for 2 min. Expression of cel-miR-39 and mmu_piR_003399 was measured by qRT-PCR. Results are expressed as mean 6 SD. **p < .01.
10 | TOXICOLOGICAL SCIENCES, 2017
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
expression (Knudsen and Knudsen, 2008). Previously it has beenrevealed that CDK6-null mice have reduced cellularity in its thy-mus (Hu et al., 2009). CDK6-deficient mice also have reducednumbers of red blood cells (Malumbres et al., 2004). The CDK6and p-Rb are decreased and a cell cycle arrest is found afterMC-LR treatment, or overexpression of mmu_piR_003399, ordown-regulation of CDK6 in GC-1 cells. Our data support the re-duction of CDK6 and p-Rb as a plausible mechanism for MC-LR-induced cell proliferation inhibition. Moreover, we observeddown-regulation of CDK6 reduced the paired testis/body weightindex, sperm counts, numbers of progressively motile sperms,and the numbers of sperms with normal morphology. On thecontrary, suppressing mmu_piR_003399 could substantially at-tenuate MC-LR-induced cell cycle arrest through increasing theexpression of CDK6 in vitro. Suppressing the expression ofmmu_piR_003399 also protected MC-LR-induced decrease ofsperm counts and sperm motility, and compromised normalsperm morphology in vivo.
Recent reports have revealed that piRNAs play importantroles in the epigenetic regulation of cancers and other diseases,suggesting that piRNAs can serve as potential diagnostic markersin these disorders (Mei et al., 2013). Interestingly, circulatingmiRNAs as biomarkers have attracted increasing attention in re-cent years (Esteller, 2011). Circulating miRNAs are resistant to ex-ogenous RNase treatment and can withstand long-term storageat room temperature and multiple freeze-thaw cycles (Mitchellet al., 2008). Peng et al. (2016) reported that plasma miRNAs arepromising biomarkers of rat retinal toxicity. Meanwhile, a previ-ous study has shown that piRNAs are stably expressed in bloodserum and plasma (Yang et al., 2015). We were intrigued to inves-tigate whether mmu_piR_003399 was expressed in blood serumand plasma and whether it can serve as a potential biomarker forMC-LR-induced toxicity. Our data show that mmu_piR_003399was stable in blood serum and plasma. Besides, an increased ex-pression of mmu_piR_003399 was observed in MC-LR-treatedmice. We observed a relationship between mmu_piR_003399 inblood and the MC-LR-induced toxicity. However, the evidencewas not strong enough to confirm that this biomarker could beused for screening MC-LR-induced reproductive system toxicity.Some limitations do exist in this study. For example, the exacthalf-life of mmu_piR_003399 in blood is unknown. Additional re-search utilizing different MC-LR toxicity models is needed to fur-ther explore their potential translatability to replace othercommonly used preclinical markers.
In summary, this study revealed an important correlationbetween piRNA and MC-LR-induced reproductive toxicity. Weconfirmed the role of mmu_piR_003399 in regulating MC-LR-induced cellular toxicity. Interfering with the intracellular ex-pression of mmu_piR_003399 may protect germ cells from thetoxic effects of MC-LR. Moreover, our data support the stablepresence of mmu_piR_003399 in blood serum and plasma,which may have diagnostic implications. Our study may pro-vide initial clues enabling further functional characterization ofpiRNAs in male reproductive system toxicity.
SUPPLEMENTARY DATA
Supplementary data are available at Toxicological Sciencesonline.
FUNDING
National Natural Science Foundation of China (31370524),National Natural Science Foundation of China (21377052),
National Natural Science Foundation of China (31670519) andFoundation Research Funds for the Central Universities(021414380332).
ACKNOWLEDGMENTS
We thank Weidong Ding, Xi Chen, Xiaoyan Li, and YetingHong for technical assistance.
REFERENCESBacker, L. C., McNeel, S. V., Barber, T., Kirkpatrick, B., Williams,
C., Irvin, M., Zhou, Y., Johnson, T. B., Nierenberg, K., andAubel, M. (2010). Recreational exposure to microcystins dur-ing algal blooms in two California lakes. Toxicon 55, 909–921.
Brower-Toland, B., Findley, S. D., Jiang, L., Liu, L., Yin, H., Dus, M.,Zhou, P., Elgin, S. C., and Lin, H. (2007). Drosophila PIWI asso-ciates with chromatin and interacts directly with HP1a. GenesDev. 21, 2300–2311.
Chen, L., Zhang, X., Zhou, W., Qiao, Q., Liang, H., Li, G., Wang, J.,Cai, F., and Aldabe, R. (2013). The interactive effects of cyto-skeleton disruption and mitochondria dysfunction lead toreproductive toxicity induced by microcystin-LR. PloS One 8,e53949.
Chen, Y., Wang, J., Zhang, Q., Xiang, Z., Li, D., and Han, X. (2017).Microcystin-leucine arginine exhibits immunomodulatoryroles in testicular cells resulting in orchitis. Environ. Pollut.229, 964–975.
Chen, Y., Zhou, Y., Wang, J., Wang, L., Xiang, Z., Li, D., and Han,X. (2016). Microcystin-leucine arginine causes cytotoxiceffects in sertoli cells resulting in reproductive dysfunctionin male mice. Sci. Rep. 6, 39238.
Clark, S. P., Davis, M. A., Ryan, T. P., Searfoss, G. H., and Hooser,S. B. (2007). Hepatic gene expression changes in mice associ-ated with prolonged sublethal microcystin exposure. Toxicol.Pathol. 35, 594–605.
De Fazio, S., Bartonicek, N., Di Giacomo, M., Abreu-Goodger, C.,Sankar, A., Funaya, C., Antony, C., Moreira, P. N., Enright, A.J., and O’Carroll, D. (2011). The endonuclease activity of Milifuels piRNA amplification that silences LINE1 elements.Nature 480, 259–263.
Deng, W., and Lin, H. (2002). miwi, a murine homolog of piwi,encodes a cytoplasmic protein essential for spermatogene-sis. Dev. Cell 2, 819–830.
Ding, J., Wang, J., Xiang, Z., Diao, W., Su, M., Shi, W., Wan, T., andHan, X. (2017). The organic anion transporting polypeptide1a5 is a pivotal transporter for the uptake of microcystin-LRby gonadotropin-releasing hormone neurons. Aquat. Toxicol.182, 1–10.
Ekholm, S. V., and Reed, S. I. (2000). Regulation of G(1) cyclin-dependent kinases in the mammalian cell cycle. Curr. Opin.Cell Biol. 12, 676–684.
Esteller, M. (2011). Non-coding RNAs in human disease. Nat. Rev.Genet. 12, 861–874.
Gou, L.-T., Dai, P., Yang, J.-H., Xue, Y., Hu, Y.-P., Zhou, Y., Kang, J.-Y., Wang, X., Li, H., Hua, M.-M., et al. (2014). PachytenepiRNAs instruct massive mRNA elimination during late sper-miogenesis. Cell Res. 24, 680–700.
Hirakata, S., and Siomi, M. C. (2016). piRNA biogenesis in thegermline: From transcription of piRNA genomic sources topiRNA maturation. Biochim. Biophys. Acta 1859, 82–92.
Hu, M. G., Deshpande, A., Enos, M., Mao, D., Hinds, E. A., Hu, G-f.,Chang, R., Guo, Z., Dose, M., Mao, C., et al. (2009). A
ZHANG ET AL. | 11
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017
requirement for cyclin-dependent kinase 6 in thymocyte de-velopment and tumorigenesis. Cancer Res. 69, 810–818.
John, B., Enright, A. J., Aravin, A., Tuschl, T., Sander, C., andMarks, D. S. (2004). Human MicroRNA targets. PLoS Biol. 2,e363.
Kist, L. W., Rosemberg, D. B., Pereira, T. C., de Azevedo, M. B.,Richetti, S. K., de Castro Leao, J., Yunes, J. S., Bonan, C. D., andBogo, M. R. (2012). Microcystin-LR acute exposure increasesAChE activity via transcriptional ache activation in zebrafish(Danio rerio) brain. Comp. Biochem. Physiol. Toxicol. Pharmacol.155, 247–252.
Knudsen, E. S., and Knudsen, K. E. (2008). Tailoring to RB: tumoursuppressor status and therapeutic response. Nature Reviews.Cancer 8, 714–724.
Lee, H. C., Gu, W., Shirayama, M., Youngman, E., Conte, D., Jr., andMello, C. C. (2012). C. elegans piRNAs mediate the genome-wide surveillance of germline transcripts. Cell 150, 78–87.
Malumbres, M., Sotillo, R., Santamar�ı;a, D., Gal�an, J., Cerezo, A.,Ortega, S., Dubus, P., and Barbacid, M. (2004). Mammaliancells cycle without the D-type cyclin-dependent kinasesCdk4 and Cdk6. Cell 118, 493–504.
Mei, Y., Clark, D., and Mao, L. (2013). Novel dimensions of piRNAsin cancer. Cancer Lett. 336, 46–52.
Milutinovic, A., Zorc-Pleskovic, R., Petrovic, D., Zorc, M., andSuput, D. (2006). Microcystin-LR induces alterations in heartmuscle. Folia Biol. 52, 116–118.
Mitchell, P. S., Parkin, R. K., Kroh, E. M., Fritz, B. R., Wyman, S. K.,Pogosova-Agadjanyan, E. L., Peterson, A., Noteboom, J.,O’Briant, K. C., Allen, A., et al. (2008). Circulating microRNAsas stable blood-based markers for cancer detection. Proc.Natl. Acad. Sci. U.S.A. 105, 10513–10518.
Ng, K. W., Anderson, C., Marshall, E. A., Minatel, B. C., Enfield, K.S., Saprunoff, H. L., Lam, W. L., and Martinez, V. D. (2016).Piwi-interacting RNAs in cancer: emerging functions andclinical utility. Mol. Cancer 15, 5,
Peng, Q., Collette, W., 3rd, Giddabasappa, A., David, J., Twamley,M., Kalabat, D., Aguirre, S. A., and Huang, W. (2016). Editor’shighlight: Plasma miR-183/96/182 cluster and miR-124 arepromising biomarkers of rat retinal toxicity. Toxicol. Sci. 152,273–283.
Pitois, S., Jackson, M. H., and Wood, B. J. (2001). Sources of the eu-trophication problems associated with toxic algae: An over-view. J. Environ. Health 64, 25–32.
Qin, Y., Xia, Y., Wu, W., Han, X., Lu, C., Ji, G., Chen, D., Wang, H.,Song, L., Wang, S., et al. (2012). Genetic variants in microRNAbiogenesis pathway genes are associated with semen qualityin a Han-Chinese population. Reprod. Biomed. Online 24,454–461.
Sai Lakshmi, S., and Agrawal, S. (2008). piRNABank: a web re-source on classified and clustered Piwi-interacting RNAs.Nucleic Acids Res. 36, D173–D177.
Slotkin, R. K., and Martienssen, R. (2007). Transposable elementsand the epigenetic regulation of the genome. Nat. Rev. Genet.8, 272–285.
Song, L., Chen, W., Peng, L., Wan, N., Gan, N., and Zhang, X.(2007). Distribution and bioaccumulation of microcystins inwater columns: A systematic investigation into the environ-mental fate and the risks associated with microcystins inMeiliang Bay, Lake Taihu. Water Res. 41, 2853–2864.
Su, Y., Li, L., Hou, J., Wu, N., Lin, W., and Li, G. (2016). Life-cycleexposure to microcystin-LR interferes with the reproductiveendocrine system of male zebrafish. Aquat. Toxicol. 175,205–212.
Trinchet, I., Djediat, C., Huet, H., Dao, S. P., and Edery, M. (2011).Pathological modifications following sub-chronic exposureof medaka fish (Oryzias latipes) to microcystin-LR. Reprod.Toxicol. 32, 329–340.
Vourekas, A., Zheng, Q., Alexiou, P., Maragkakis, M., Kirino, Y.,Gregory, B. D., and Mourelatos, Z. (2012). Mili and Miwi targetRNA repertoire reveals piRNA biogenesis and function ofMiwi in spermiogenesis. Nat. Struct. Mol. Biol. 19, 773–781.
Watanabe, T., Cheng, E. C., Zhong, M., and Lin, H. (2015).Retrotransposons and pseudogenes regulate mRNAs andlncRNAs via the piRNA pathway in the germline. Genome Res.25, 368–380.
Yang, X., Cheng, Y., Lu, Q., Wei, J., Yang, H., and Gu, M. (2015).Detection of stably expressed piRNAs in human blood. Int. J.Clin. Exp. Med. 8, 13353–13358.
Zegura, B., Volcic, M., Lah, T. T., and Filipic, M. (2008). Differentsensitivities of human colon adenocarcinoma (CaCo-2), as-trocytoma (IPDDC-A2) and lymphoblastoid (NCNC) cell linesto microcystin-LR induced reactive oxygen species and DNAdamage. Toxicon 52, 518–525.
Zhang, L., Zhang, H., Zhang, H., Benson, M., Han, X., and Li, D.(2017). Roles of piRNAs in microcystin-leucine-arginine (MC-LR) induced reproductive toxicity in testis on male offspring.Food Chem. Toxicol. 105, 177–185.
Zhang, P., Kang, J. Y., Gou, L. T., Wang, J., Xue, Y., Skogerboe, G.,Dai, P., Huang, D. W., Chen, R., Fu, X. D., et al. (2015). MIWIand piRNA-mediated cleavage of messenger RNAs in mousetestes. Cell Res. 25, 193–207.
Zhou, Y., Chen, Y., Yuan, M., Xiang, Z., and Han, X. (2013). In vivostudy on the effects of microcystin-LR on the apoptosis, pro-liferation and differentiation of rat testicular spermatogeniccells of male rats injected i.p. with toxins. J. Toxicol. Sci. 38,661–670.
Zhou, Y., Xiang, Z., Li, D., and Han, X. (2014). Regulation ofmicrocystin-LR-induced toxicity in mouse spermatogonia bymiR-96. Environ. Sci. Technol. 48, 6383–6390.
Zhou, Y., Yuan, J., Wu, J., and Han, X. (2012). The toxic effects ofmicrocystin-LR on rat spermatogonia in vitro. Toxicol. Lett.212, 48–56.
12 | TOXICOLOGICAL SCIENCES, 2017
Downloaded from https://academic.oup.com/toxsci/advance-article-abstract/doi/10.1093/toxsci/kfx209/4466233by Nanjing University useron 23 November 2017