We present a filtration-based protocol to isolate high-quality nuclei from cross-linked mouse skeletal muscle wherein we removed the need for ultracentrifugation, making it easily applicable. We show that chromatin prepared from the nuclei is suitable for chromatin immunoprecipitation and likely chromatin immunoprecipitation sequencing studies.
Chromatin immunoprecipitation (ChIP) is a powerful method to determine protein binding to chromatin DNA. Fiber-rich skeletal muscle, however, has been a challenge for ChIP due to technical difficulty in isolation of high-quality nuclei with minimal contamination of myofibrils. Previous protocols have attempted to purify nuclei before cross-linking, which incurs the risk of altered DNA-protein interaction during the prolonged nuclei preparation process. In the current protocol, we first cross-linked the skeletal muscle tissue collected from mice, and the tissues were minced and sonicated. Since we found that ultracentrifugation was not able to separate nuclei from myofibrils using cross-linked muscle tissue, we devised a sequential filtration procedure to obtain high-quality nuclei devoid of significant myofibril contamination. We subsequently prepared chromatin by using an ultrasonicator, and ChIP assays with anti-BMAL1 antibody revealed robust circadian binding pattern of BMAL1 to target gene promoters. This filtration protocol constitutes an easily applicable method to isolate high-quality nuclei from cross-linked skeletal muscle tissue, allowing consistent sample processing for circadian and other time-sensitive studies. In combination with next-generation sequencing (NGS), our method can be deployed for various mechanistic and genomic studies focusing on skeletal muscle function.
Skeletal muscle plays important roles in physiology and behavior. The multi-nucleated muscle fiber consists of myofibrils where actin and myosin form functional units called sarcomeres to generate contractile force. Skeletal muscle is also the largest metabolic organ in the body, accounting for >80% postprandial glucose intake and regulating insulin response and metabolic homeostasis1,2. Muscle physiology and metabolism are closely regulated by the circadian clock, an intrinsic biological timer3,4,5,6. For example, the skeletal muscle-specific deletion of Bmal1, one of the core circadian clock components, resulted in insulin resistance and diminished glucose uptake in skeletal muscle, and the animals were found to develop type 2 diabetes7. In addition, skeletal muscle is also increasingly being appreciated as an endocrine organ8, secreting myokines to regulate systemic metabolism and physiology. Mechanistic studies are required to fully understand these regulatory functions in skeletal muscle.
ChIP is a powerful approach to delineate promoter recruitment of DNA binding proteins. ChIP was initially developed to identify nucleosome organization on chromatin DNA9,10. A variety of methods have since been reported to cross-link proteins and chromatin DNA using formaldehyde, dimethyl sulfate or ultraviolet irradiation (UV)11,12. Formaldehyde cross-linking is the most commonly used, preserving chromatin structure and DNA-protein interactions9,13,14. Cross-linked chromatin is shredded by sonication and immunoprecipitated with antibody against the particular DNA binding protein of interest15,16. In recent years, ChIP-sequencing (ChIP-seq), a method combining ChIP with NGS, has been developed to interrogate genome-wide transcription factor binding17, and in some cases to monitor dynamic changes over a time course18,19,20. For example, circadian ChIP-seq studies have revealed a highly orchestrated sequence of genomic binding of circadian clock components and histone markers, which drives temporally precise gene expression throughout the ~ 24 h circadian cycle18.
Most available ChIP protocols are designed for soft tissues (e.g. liver, brain, etc.), and very few have been published for hard tissues including skeletal muscle. It is technically challenging to homogenize fiber-rich skeletal muscle and isolate high-quality nuclei21, especially for ChIP experiments which require cross-linking. In a recent muscle ChIP study22, satellite cells were separated from myofibers, and nuclei were prepared from both cell types through a prolonged procedure involving tissue digestion. The entire process took approximately three hours to complete before formaldehyde cross-linking was performed on isolated nuclei. While this procedure avoided cross-linking muscle fiber, which makes muscle tissue even more refractory to efficient homogenization, and was able to produce high-quality nuclei, the significant time-lag from tissue collection to nuclei cross-linking incurs the risk of altered DNA-protein interaction. In contrast, most studies performed cross-linking immediately after experimental treatment or tissue collection in order to preserve the real-time DNA-protein binding12. A second drawback of nuclei isolation before cross-linking is that it precludes time-sensitive applications such as circadian sample collection which typically occurs at 3 – 4 h intervals. Without cross-linking the nuclei, the isolation needs to proceed immediately after dissection, whereas cross-linked samples can be processed together after the entire time course is completed, thus ensuring greater experimental consistency.
Other protocols for nuclei isolation from uncrosslinked skeletal muscle have also been reported. Two studies described the use of gradient ultracentrifugation to separate nuclei from remaining myofibrils and cell debris23,24. While sucrose or colloidal gradient ultracentrifugation is effective with uncrosslinked muscle tissues, our experiments revealed that after crosslinking, gradient ultracentrifugation failed to separate nuclei from cell debris on the gradient.
We therefore developed a procedure to isolate high-quality nuclei using cross-linked mouse skeletal muscle tissues. Rather than gradient ultracentrifugation, we devised a serial filtration method to effectively separate nuclei from debris. Following ultrasonication, the chromatin samples were successfully applied for ChIP studies which showed a circadian pattern of BMAL1 protein binding to target promoters. Our method can be broadly applicable to various mechanistic studies of muscle tissues.
Animal care was performed under Institutional Animal Care and Use Committee (IACUC) guidelines, and the procedures were conducted according to an animal protocol approved by the University of Texas Health Science Center at Houston.
1. Nuclei Isolation from Cross-linked Skeletal Muscle
2. Sonication
3. Evaluation of Sonication and Quantitation
4. ChIP
Here we performed formaldehyde cross-linking immediately after tissue collection to preserve real-time DNA-protein interaction. However, we found that sucrose or colloidal gradient, commonly used for nuclei isolation23,24, was not effective in separating nuclei from myofibrils (data not shown). The reason may be that cross-linking conferred similar gravity for nuclei and myofibrils. Therefore, we developed a serial filtration process to effectively remove large myofibrils and other debris from the nuclei fraction. After 100 µm filtration, there were still large tissue debris and myofibrils (Figure 1A). In comparison, at the end of the sequential filtration, the majority of large tissue debris, intact cells and large myofibrils were successfully removed (Figure 1B).
Those nuclei were used for dish-shaped ultrasonication. Nuclei stored at -80 °C were suspended in SDS lysis buffer and immediately subjected to sonication. Although incubation of nuclei in SDS lysis buffer on ice or room temperature may improve sonication for some tissue such as liver27 or mouse embryonic fibroblast (MEF)28, in our experience it interfered with sonication of skeletal muscle nuclei and resulted in broad smearing, indicating ineffective sonication. Using the current protocol with immediate sonication after suspension in SDS lysis buffer we observed efficient sonication with chromatin gradually shredded in a cycle-dependent manner and eventually producing ~ 500 bp DNA fragments (Figure 2). Pre-incubation in the lysis buffer detectably compromised the sonication efficiency.
To confirm the quality of chromatin preparation for ChIP, we examined DNA binding of the circadian transcription factor BMAL1, which showed binding peak and trough at Zeitgeber time (ZT)6 and ZT18, respectively18,29. Skeletal muscle samples from C57B/6J mice were collected at ZT6 and ZT18, and chromatin samples were prepared as above. Briefly, after sonication, shredded chromatin samples were pre-cleared with IgY beads that have been pre-blocked with BSA followed by incubation with anti-BMAL1 antibody at 4 °C overnight. After elution and purification of chromatin, we performed RT-qPCR to detect BMAL1 binding on E-Box elements of two target genes, Nr1d1 and Dbp30. We detected robust binding of BMAL1 to the E-box elements at ZT6 and minimal binding at ZT18 (Nr1d1: 0.452 ± 0.022 vs. 0.039 ± 0.002, Dbp -0.4: 0.627 ± 0.013 vs. 0.062 ± 0.009, Dbp +0.8: 0.176 ± 0.013 vs. 0.008 ± 0.001, Dbp +2.4: 0.466 ± 0.010 vs. 0.122 ± 0.014; all values are mean ± SEM) (Figure 3), validating the protocol for time-sensitive transcription factor binding analysis in skeletal muscle.
Figure 1: Sequential Filtration Effectively Removed Tissue Debris. (A) Representative images showing samples after 100 µm filtration. Large tissue and fiber debris are observed. (B) Representative images showing samples after serial filtration. Large fiber debris were cleared. Only isolated nuclei and small myofibril fragments are observed. Pictures were taken by using a light microscope at 10X, 20X and 40X magnifications. Scale bars are shown on the right hand side panels. Please click here to view a larger version of this figure.
Figure 2: Progressive Chromatin Shredding Through 10 Cycles of Sonication. Ten cycles of sonication with digested chromatin DNA to ~ 500 bp, as revealed in a 0.8 % agarose gel, run at 150 V for 60 min. The right panel indicates a lower sonication efficiency after pre-incubation in ice-cold SDS lysis buffer for 1 h. Please click here to view a larger version of this figure.
Figure 3: Representative qPCR Results for BMAL1 ChIP with Mouse Skeletal Muscle Samples Collected at ZT6 and ZT18. Data are presented as mean ± SEM. Dbp -0.4, +0.8 and +2.4 indicate locations of the E-Box elements on the Dbp gene. NC: negative control with IgY. The temporal pattern of BMAL1 binding is consistent with previous results showing BMAL1 binding peak at around ZT618. The forward and reverse primers are as follows. Rev-erba: 5'-GTAGACTACAAATCCCAACAATCCTG, and 5'-TGGAGCAGGTACCATGTGATTC; Dbp -0.4: 5'-ACACCCGCATCCGATAGC, and 5'-CCACTTCGGGCCAATGAG; Dbp +0.8: 5'- ATGCTCACACGGTGCAGACA, and 5'- CTGCTCAGGCACATTCCTCAT; Dbp +2.4: 5'- TGGGACGCCTGGGTACAC, and 5'- GGGAATGTGCAGCACTGGTT. Please click here to view a larger version of this figure.
Here we describe a robust method where cross-linked skeletal muscle tissues were used to isolate high-quality nuclei. Sequential filtration was carried out to effectively separate nuclei from debris, and ultrasonic acoustic energy from dish-shaped transducer sheared the chromatin for ChIP analysis. The results showed circadian time-specific binding of BMAL1 to target promoters.
ChIP can be employed to capture real-time protein occupancy on genomic DNA when cross-linking takes place. To take advantage of this potential, we aimed to develop a method to allow cross-linking of skeletal muscle at the time of tissue dissection and to streamline the nuclei isolation without gradient ultracentrifugation. Due to the difficulty of homogenizing fiber-rich skeletal muscle compared with soft tissues such as liver, we minced muscle tissue in ice-cold PBS and then homogenized the sample in a formaldehyde buffer. After quenching, tissue suspension was centrifuged and rinsed with ice-cold base buffer to rinse out any remaining formaldehyde. Nuclei were released by Dounce homogenization, and the homogenates were sequentially filtered to gradually remove cell debris and myofibrils. We devised the series of filtration to minimize filter clogging which could adversely impact yield. Only very short myofibrils remained when the sequential filtration was completed.
The sonication and ChIP procedures were adapted from a previous report12 with modifications including sonication timing and SDS buffer amount. The dish-shaped sonicator allows exposure to centralized ultrasonic wave for samples in glass vials in a cold water bath. Compared with probe sonicators, this sonicator controls the sample temperature to avoid overheating, and also prevents sample cross-contamination. If probe sonicators are used, optimal sonication conditions need to be determined empirically. We also reduced the amount of SDS buffer since the yield of muscle chromatin is lower than that in liver12. Several protocols28,31,32 include incubation on ice or at room temperature prior to sonication. However, in our experience, pre-incubation on ice did not improve the sonication efficiency. In fact, in some instances the sonication was compromised. It is possible that residual myofibrils entangled the chromatin DNA during incubation and attenuated sonication efficacy. With immediate sonication after nuclei suspension in SDS lysis buffer, we succeeded in obtaining progressive chromatin fragmentation with increasing sonication cycles (Figure 2).
We validated the quality of chromatin with ChIP qPCR. As shown in Figure 3, the BMAL1 promoter occupancy was robust at ZT6 and minimal at ZT18, consistent with previously shown BMAL1 circadian promoter binding18. This functional assay confirmed the quality of nuclei and chromatin. In recent years, rapid development in NGS opened up a new horizon for ChIP application where ChIP-seq can quantitatively interrogate genomic binding with high sensitivity17. Particularly for ChIP-seq, high-quality nuclei and chromatin are required to consistently capture protein-DNA interaction. The procedure described herein may constitute a valuable resource for ChIP-seq studies using skeletal muscle. Of note, ChIP-seq library preparation requires additional measures to improve signal resolution, such as PAGE-based size selection18.
In conclusion, we developed a filtration-based protocol to prepare high-quality nuclei from cross-linked skeletal muscle. We remove the need for ultracentrifugation, making it easily applicable. In addition to ChIP for genes of interest, the nuclei and chromatin prepared as described can be broadly applicable to ChIP-seq studies.
The authors have nothing to disclose.
We thank Karyn Esser, Nobuya Koike and Noheon Park for helpful advice. This work was in part supported by NIH/NIGMS (R01GM114424) to S.-H.Y., and the Robert A. Welch Foundation (AU-1731) and NIH/NIA (R01AG045828) to Z.C.
Materials | |||
Formaldehyde solution | Sigma Aldrich | F8775 | |
Glycine | Fisher Scientific | BP381-5 | |
Trypan Blue solution (0.4%) | Fisher Scientific | 15250061 | |
RNase A | Sigma Aldrich | 10109142001 | |
Protease K | Sigma Aldrich | 3115887001 | |
Chicken Anti-BMAL1 antibody | Generated in chicken (Cocalico Biologicals) against antigen aa 318-579, and IgY was affinity purified using the same antigen. | ||
Chicken IgY Precipitating Resin | GenScript | L00405 | |
Equipment | |||
KINEMATICA Polytron PT2100 Benchtop Homogenizer | Fisher Scientific | 08-451-178 | |
15mL Dounce tissue grinder | Whearton | 357544 | |
Falcon Cell Strainers 100µm | Fisher Scientific | 08-771-19 | |
Falcon Cell Strainers 70µm | Fisher Scientific | 08-771-1 | |
Falcon Cell Strainers 40µm | Fisher Scientific | 08-771-2 | |
pluriStrainer 30 µm | PluriSelect | 43-50030-03 | |
pluriStrainer 20 µm | PluriSelect | 43-50020-03 | |
pluriStrainer 10 µm | PluriSelect | 43-50010-03 | |
Covaris S2 Focused-ultrasonicator | Covaris | Model S2 | |
Labquake | Thermo Scientific | C415110 | |
CCD Microscope Camera | Leica Microsystems | DFC3000 G | |
Reagent Kit | |||
DNA extraction kit | Thermo Scientific | K0691 | |
Buffers | |||
All buffer components are discribed in the protocol. Each component was purchased from Sigma Aldrich |