Here, we present an eCLIP protocol to determine major RNA targets of RBP candidates in testis.
Spermatogenesis defines a highly ordered process of male germ cell differentiation in mammals. In testis, transcription and translation are uncoupled, underlining the importance of post-transcriptional regulation of gene expression orchestrated by RBPs. To elucidate mechanistic roles of an RBP, crosslinking immunoprecipitation (CLIP) methodology can be used to capture its endogenous direct RNA targets and define the actual interaction sites. The enhanced CLIP (eCLIP) is a newly-developed method that offers several advantages over the conventional CLIPs. However, the use of eCLIP has so far been limited to cell lines, calling for expanded applications. Here, we employed eCLIP to study MOV10 and MOV10L1, two known RNA-binding helicases, in mouse testis. As expected, we find that MOV10 predominantly binds to 3′ untranslated regions (UTRs) of mRNA and MOV10L1 selectively binds to Piwi-interacting RNA (piRNA) precursor transcripts. Our eCLIP method allows fast determination of major RNA species bound by various RBPs via small-scale sequencing of subclones and thus availability of qualified libraries, as a warrant for proceeding with deep sequencing. This study establishes an applicable basis for eCLIP in mammalian testis.
Mammalian testis represents an excellent developmental model wherein an intricate cell differentiation program runs cyclically to yield numerous spermatozoa. An unique value of this model lies in the emergence of transcriptional inactivation at certain stages of spermatogenesis, typically when meiotic sex chromosome inactivation (MSCI) occurs1,2 and when round spermatids undergo drastic nuclear compaction during spermiogenesis3. These inconsecutive transcriptional events necessitate post-transcriptional gene regulation, in which RNA-binding proteins (RBPs) play a crucial role, shaping transcriptome and maintaining male fertility.
To identify the bona fide RNA targets of an individual RBP in vivo, the crosslinking immunoprecipitation (CLIP) method was developed4,5, based on but beyond the regular RNA immunoprecipitation (RIP)6,7, by incorporation of key steps including ultraviolet (UV) crosslinking, stringent wash and gel transfer to improve signal specificity. The advanced application of the CLIP combined with high-throughput sequencing has provoked large interest in profiling protein-RNA interaction at genome-wide levels8. In addition to genetic studies on RBP function, such biochemical methods that identify the direct interplay of endogenous protein and RNA have been indispensable to accurately elucidate the RNA regulatory roles of RBPs. For example, MOV10L1 is a testis-specific RNA helicase required for male fertility and the Piwi-interacting RNA (piRNA) biogenesis9. Its paralogue MOV10 is known as a ubiquitously expressed and multifunctional RNA helicase with roles in multiple aspects of RNA biology10,11,12,13,14,15,16,17,18. By employing the conventional CLIP-seq, we found that MOV10L1 binds and regulates primary piRNA precursors to initiate early piRNA processing19,20, and that MOV10 binds mRNA 3' UTRs and as well as noncoding RNA species in testicular germ cells (data not shown).
Nevertheless, CLIP is originally a laborious, radioactive procedure followed by sequencing library preparation with a remarkable loss of CLIP tags. In the conventional CLIP, a cDNA library is prepared using adapters ligated at both RNA extremities. After protein digestion, crosslinked short polypeptides remain attached to RNA fragments. This crosslinking mark partially blocks reverse transcriptase (RTase) progression during cDNA synthesis, resulting in truncated cDNAs which represent about 80% of the cDNA library21,22. Thus, only cDNA fragments resulting from RTase bypassing the crosslinking site (read-through) are sequenced. Recently, various CLIP approaches, such as PAR-CLIP, iCLIP, eCLIP and uvCLAP, have been employed to identify crosslink sites of RBPs in living cells. PAR-CLIP involves the application of 365 nm UV radiation and photoactivatable nucleotide analogs and is therefore exclusive to in-culturing living cells, and incorporation of nucleoside analogs into newly synthesized transcripts is prone to producing bias where RNA physically interacts with protein23,24. In iCLIP, only a single adapter is ligated to the 3' extremity of crosslinked RNA fragments. After reverse transcription (RT), both truncated and read-through cDNAs are obtained by intramolecularly circularization and re-linearization followed by polymerase chain reaction (PCR) amplification25,26. However, the efficiency of intramolecular circularization is relatively low. Although older CLIP protocols need labeling of crosslinked RNA with a radioisotope, ultraviolet crosslinking and affinity purification (uvCLAP), with a process of stringent tandem affinity purification, does not rely on radioactivity27. Nevertheless, uvCLAP is limited to cultured cells that must be transfected with the expression vector carrying the 3x FLAG-HBH tag for tandem affinity purification.
In eCLIP, adapters were ligated first at the 3' extremity of RNA followed by RT, and next at the 3' extremity of cDNAs in an intermolecular mode. Hence, eCLIP is able to capture all truncated and read-through cDNA28. Also, it is neither restricted to radioactive labeling, nor to using cell lines based on its principle, while maintaining single-nucleotide resolution.
Here, we provide a step-by-step description of an eCLIP protocol adapted for mouse testis. Briefly, this eCLIP protocol starts with UV crosslinking of testicular tubules, followed by partial RNase digestion and immunoprecipitation using a protein-specific antibody. Next, the protein-bound RNA is dephosphorylated, and adapter is ligated to its 3' end. After protein gel electrophoresis and gel-to-membrane transfer, RNA is isolated by cutting the membrane area of an expected size range. After RT, DNA adapter is ligated to the 3' end of cDNA followed by PCR amplification. Screening of subclones prior to high-throughput sequencing is taken as a library quality control. This protocol is efficient at identifying major species of protein-bound RNA of RBPs, exemplified by the two testis-expressing RNA helicases MOV10L1 and MOV10.
All performed animal experiments have been approved by the Nanjing Medical University committee. Male C57BL/6 mice were kept under controlled photoperiod conditions and were supplied with food and water.
1. Tissue Harvesting and UV Crosslinking
2. Beads Preparation
3. Tissue Lysis and Partial RNA Digestion
4. Immunoprecipitation
5. Dephosphorylation of RNA 3' Ends
6. RNA Adapter Ligation to RNA 3' Ends
7. SDS-PAGE and Membrane Transfer
8. RNA Isolation
9. Dephosphorylation of Input RNA 3' Ends
10. RNA Adapter Ligation to Input RNA 3' Ends
11. Reverse Transcription, DNA Adapter Ligation to cDNA 3' Ends
12. Quantification of cDNA by Real-time Quantitative PCR (qPCR)
13. PCR Amplification of cDNA
14. Gel Purification
15. TOPO Clone of PCR Product
The eCLIP procedure and results are illustrated in Figure 1, Figure 2, Figure 3, Figure 4. Mice were euthanized with carbon dioxide and a small incision was made in the lower abdomen using surgical scissors (Figure 2A,B). Mouse testes were removed, detunicated and then UV-crosslinked after grinding (Figure 2C–I). Representative eCLIP results of using two known RNA-binding helicases in testis tissues are depicted in Figure 3 and 4. We performed MOV10 eCLIP in testes from adult wild-type mice, with common concentration of 40 U/mL of RNase I treating the crosslinked lysate. The top panel of Figure 3A shows that the target protein sized about 114 kDa was successfully enriched. Western blot of the immunoprecipitated MOV10L1 proteins was performed with two concentrations (5 or 40 U/mL) of RNase I during the eCLIP process (Figure 3A). Figure 3B shows qPCR using 1:10 diluted cDNA (already ligated with DNA adapter) from MOV10 and MOV10L1 UV-crosslinked, non-crosslinked, and the paired size matched input (SMInput) sample. Non-crosslinked samples show decreased RNA recovery. We observed that, the Ct values of the non-crosslinked group was generally 5 times more than UV-crosslinked group. Figure 4A displays PCR amplification and size selection via agarose gel electrophoresis (cut 175-350 bp). Primer-dimer product appears at about 140 bp. Figure 4B shows the UCSC genome browser view of two representative subclone sequences. MOV10-bound eCLIP tags are found to be located within the 3' UTR of gene Fto; The approximate rate of 3' UTR targets accounts for 75% (Figure 4C), consistent with the majority of MOV10 targets in HEK293 cells10 and in testes (data not shown). In contrast, MOV10L1-bound eCLIP tags are found to be located within a piRNA cluster indicating MOV10L1 targets piRNA precursors. The approximate rate of piRNA precursor targets accounts for 42% (Figure 4E), which reflects a trend from our previous conventional CLIP experiment20. MOV10L1 eCLIP with 40 U/mL RNase I digestion yields relatively more sequences with less than 20 bp (Figure 4D).
Figure 1: Schematic representation of eCLIP. UV-crosslinked mouse seminiferous tubules (step 1) are lysed in eCLIP lysis buffer and sonicated (step 2). Lysate is treated with RNase I to fragment RNA, after which protein-RNA complexes are immunoprecipitated using the anti-RBP antibody (step 3-4). Dephosphorylation of RNA fragments and ligation of 3′ RNA adapter are performed (step 5-6). Protein-RNA complexes are run on an SDS-PAGE gel and transferred to nitrocellulose membranes (step 7). RNA is recovered from the membrane by digesting the protein with proteinase K and Urea which leaves a short polypeptide remaining at the crosslink site. Dephosphorylation of RNA fragments of input samples and ligation of 3′ RNA adapter is performed (step 8). Perform RT of RNA and ligation of 3′ DNA adapter (step 9-10). Perform PCR amplification of cDNA library, gel extraction, and blunt-end PCR cloning for preliminary library quality control (step 11). Finally, perform high-throughput sequencing (step 12). Please click here to view a larger version of this figure.
Figure 2: Testis tissue harvesting and UV crosslinking. (A) The exposure of the mouse abdomen. (B) A 0.5 cm incision in the abdominal wall exposing the peritoneum. (C) A pair of testes are taken out by pulling out the fat pads. (D) Testicular tissue is removed and placed in a small dish containing ice-cold PBS. (E) Gently remove the tunica albuginea. (F) Press the loose pestle to triturate the tissue in a tissue grinder dounce. (G) Distributed seminiferous tubules in a 10 cm plate. (H) UV crosslinking with 400 mJ/cm2 energy. (I) The crosslinked samples are collected in 1.5 mL centrifuge tubes. Please click here to view a larger version of this figure.
Figure 3: Representative results of MOV10 and MOV10L1 eCLIP. (A) Western blot validation of MOV10 and MOV10L1 immunoprecipitates. (B) qPCR on 1:10 diluted eCLIP libraries of MOV10 and MOV10L1, with replicates for UV, non-UV and the paired SMInput samples. Please click here to view a larger version of this figure.
Figure 4: eCLIP library preparation and quality assessment. (A) The gel images of PCR amplification are shown. Asterisk indicates primer dimer. Red dotted line indicates regions excised for PCR product of cDNA, somewhere between 175 and 350 bp. (B) UCSC genome browser view of two representative subclone sequences31. (C) Small-scale subclone sequencing analysis of MOV10. (D) MOV10L1-bound tags display distinct patterns of length distribution when processed by two different RNase concentrations. (E) Small-scale subclone sequencing analysis of MOV10L1. Please click here to view a larger version of this figure.
Sequence Name | Sequence Information | Description | |||||
RNA adapters | |||||||
RNA X1A | /5Phos/AUAUAGGNNNNNAGAUCGGAAGAGCGUCGUGUAG/3SpC3/ | stock at 200 µM; working at 20 µM | |||||
RNA X1B | /5Phos/AAUAGCANNNNNAGAUCGGAAGAGCGUCGUGUAG/3SpC3/ | stock at 200 µM; working at 20 µM | |||||
RiL19 | /5phos/AGAUCGGAAGAGCGUCGUG/3SpC3/ | stock at 200 µM; working at 40 µM | |||||
DNA adapter | |||||||
Rand103tr3 | /5Phos/NNNNNNNNNNAGATCGGAAGAGCACACGTCTG/3SpC3/ | stock at 200 µM; working at 80 µM | |||||
RT primer | |||||||
AR17 | ACACGACGCTCTTCCGA | stock at 200 µM; working at 20 µM | |||||
PCR primers | |||||||
PCR-F-D 501 |
AATGATACGGCGACCACCGAGATCTACACTATAGCCTACACTCTTTCCCTACACGACGCTCTTCCGATCT | stock at 100 µM; working at 20 µM | |||||
PCR-R-D 701 | CAAGCAGAAGACGGCATACGAGATCGAGTAATGTGACTGGAGTTCAGACGTGTGCTCTTCCGATC | stock at 100 µM; working at 20 µM | |||||
(See Illumina customer service letter for D502-508, D702-712) |
Table 1: Adapter and primer sequences. The adapter contains an in-line random-mer (either N5 or N10) to determine whether two identical sequenced reads indicate two unique RNA fragments or PCR duplicates of the same RNA fragment. "5 Phos" stands for 5' Phosphorylation, which is needed if an oligo is used as a substrate for DNA/RNA ligase. "3SpC3" stands for 3' C3 Spacer, which can prevent ligation between adapters.
With increasing understanding of the universal role of RBPs under both biological and pathological contexts, the CLIP methods have been widely utilized to reveal the molecular function of RBPs20,32,33,34,35. The protocol described here represents an adapted application of the eCLIP method to mouse testis.
One challenge in performing eCLIP in testis is maintaining viability and integrity of fresh testicular cells, which is also important for effective crosslinking. Shearing the testis with mild mechanical force using the loose pestle can prevent cell lysis32,36. Proper digestion of RNA is also critical for successful eCLIP assays. RNA fragments could be more convergent after digestion, but length less than 20 bp can be removed via pre-processing of the library reads. In order to adopt an ideal RNase dosage for an RBP candidate, we suggest a preliminary test based on the results of the subclone sequencing of eCLIP libraries that can be prepared by RNase treatment with concentrations ranging from 0 U to 40 U (per milliliter of lysate). The small-scale subclone sequencing analysis is a recommended step for a reliable examination of library quality in our eCLIP method. First, the percentage of inserts shorter than 20 bp should not be too high, or, the subsequent pre-processing of eCLIP library will cause a costly loss of reads. Secondly, the efficiency of correct ligation of both adapters should be checked. Substandard samples can be eliminated without deep sequencing to ensure successful deep sequencing, the results of which generally take much longer to analyze.
Although the feasibility of eCLIP in mouse testis is still limited by the specificity of the antibody for the step of immunoprecipitation, eCLIP is advantageous over conventional CLIP methods in several aspects. First, it is a non-radioactive method. By eCLIP, RNA targets of RBPs are directly captured in vivo without having to resort to labor-intensive techniques using radioactive materials. Secondly, the method is less time intensive. The whole procedure takes only 4 days through eCLIP library preparation. Third, sequence diversity. Compared with the conventional unified amplification cycle of 25-35 cycles, eCLIP refers to Ct values of qPCR to set the number of PCR cycles specifically. Lastly, it provides stronger signal-to-noise ratio. The size-matched input serves as an appropriate background for authentic targets.
In summary, our eCLIP results consolidate the conclusions that MOV10 and MOV10L1 have a binding preference to mRNA 3' UTR and piRNA precursors, respectively. The protocol we described herein represents the first employment of the eCLIP method in reproduction, an area in which RNA-RBP interaction knowledge is rather insufficient, although genetic studies have provided ample information about the biological roles of RBPs. Visualization of this eCLIP protocol may help guide its widespread applications in broader areas.
The authors have nothing to disclose.
We thank Eric L Van Nostrand and Gene W Yeo for helpful guidance with the original protocol. K.Z. was supported by National Key R&D Program of China (2016YFA0500902, 2018YFC1003500), and National Natural Science Foundation of China (31771653). L.Y. was supported by National Natural Science Foundation of China (81471502, 31871503) and Innovative and Entrepreneurial Program of Jiangsu Province.
Antibodies | |||
Anti-mouse MOV10 antibody | Proteintech, China | 10370-1-AP | |
Anti-mouse MOV10L1 antibody | Zheng et al.20109 | polyclonal antisera UP2175 | provided by P. Jeremy Wang lab(University of Pennsylvania) |
HRP Goat Anti-Rabbit IgG | ABclonal | AS014 | |
Rabbit IgG | Beyotime, China | A7016 | |
Equipment | |||
Centrifuge | Eppendorf, Hamburg, Germany | 5242R | |
Digital sonifier | BRANSON,USA | BBV12081048A | 450 Watts; 50/60 HZ |
DynaMag-2 Magnet | Invitrogen,USA | 12321D | |
Mini Blot Module | Invitrogen,USA | B1000 | |
Mini Gel Tank | Invitrogen,USA | A25977 | |
Shaking incubator | Eppendorf, Hamburg, Germany | Thermomixer comfort | |
Tissue Grinder, Dounce | PYREX, USA | 1234F35 | only the "loose" pestle is used in this protocol |
TProfessional standard 96 Gradient | Biometra, Germany | serial no.: 2604323 | |
Tube Revolver | Crystal, USA | serial no.: 3406051 | |
UV-light cross-linker | UVP, USA | CL-1000 | |
Materials | |||
TC-treated Culture Dish | Corning, USA | 430167 | 100 mm |
Tubes | Corning, USA | 430791 | 15 mL |
Microtubes tubes | AXYGEN , USA | MCT-150-C | 1.5 mL |
Reagents | |||
Acid phenol/chloroform/isoamyl alcohol | Solarbio, China | P1011 | 25:24:01 |
AffinityScript Enzyme | Agilent, USA | 600107 | |
Antioxidant | Invitrogen,USA | NP0005 | |
DH5α competent bacteria | Thermo Scientific, USA | 18265017 | these economical cells yield >1 x 106 transformants/µg control DNA per 50 µL reaction. |
DMSO | Sigma-Aldrich, USA | D8418 | |
DNA Ladder | Invitrogen, USA | 10416014 | |
dNTP | Sigma-Aldrich, USA | DNTP100-1KT | |
Dynabeads Protein A | Invitrogen, USA | 10002D | |
ECL reagent | Vazyme, China | E411-04 | |
EDTA | Invitrogen, USA | AM9260G | |
EDTA free protease inhibitor cocktail | Roche, USA | 04693132001 | add fresh |
Exo-SAP-IT | Affymetrix, USA | 78201 | PCR Product Cleanup Reagent |
FastAP enzyme | Thermo Scientific, USA | EF0652 | |
LDS Sample Buffer | Thermo Scientific, USA | NP0007 | |
MetaPhor Agarose | lonza, Switzerland | 50180 | |
MgCl2 | Invitrogen, USA | AM9530G | |
MiniElute gel Extraction | QIAGEN, Germany | 28604 | column store at 4 ℃; buffer QG=gel dissolving buffer; buffer PE= wash buffer(for step 14) |
MyONE Silane beads | Thermo Scientific, USA | 37002D | nucleic acids extraction magnetic beads |
NaCl | Invitrogen,USA | AM9759 |
|
NP-40 | Amresco, USA | M158-500ML | |
NuPAGE Bis-Tris Protein Gels | Invitrogen, USA | NP0336BOX | 4%–12%,1.5 mm, 15-well |
NuPAGE MOPS SDS Buffer Kit | Invitrogen, USA | NP0050 | |
PBS | Gibco, USA | 10010023 | |
Phase-Locked Gel (PLG) heavy tube | TIANGEN, China | WM5-2302831 | |
PowerUp SYBR Green Master Mix | Applied Biosystems, USA | A25742 | |
proteinase K | NEB, New England | P8107S | |
Q5 PCR master mix | NEB, New England | M0492L | |
RLT buffer | QIAGEN, Germany | 79216 | RNA purification lysis buffer |
RNA Clean & Concentrator-5 columns | ZYMO RESEARCH, USA | R1016 | RNA purification and concentration columns |
RNase I | Invitrogen, USA | AM2295 | |
RNase Inhibitor | Promega, USA | N251B | |
RQ1 DNase | Promega,USA | M610A | |
Sample Reducing Agent | Invitrogen,USA | NP0009 | |
SDS Solution | Invitrogen, USA | 15553027 | 10% |
Sodium deoxycholate | Sigma-Aldrich, USA | 30970 | protect from light |
T4 PNK enzyme | NEB, New England | M0201L | |
T4 RNA ligase 1 high conc | NEB, New England | M0437M | |
TA/Blunt-Zero Cloning Mix | Vazyme, China | C601-01 | |
TBE | Invitrogen,USA | AM9863 | |
Tris-HCI Buffer | Invitrogen, USA | 15567027 | |
Triton X-100 | Sangon Biotech, China | A600198 | |
Tween-20 | Sangon Biotech, China | A600560 | |
Urea | Sigma-Aldrich, USA | U5378 | |
X-ray Films | Caresteam, Canada | 6535876 |