Recently high-throughput sequencing technology has greatly increased sensitivity of Chromatin Immunoprecipitation (ChIP) experiment and prompted its application using purified cells or dissected tissue. Here we delineate a method to use ChIP technique with Drosophila tissue, which can address the endogenous chromatin state in a well-characterized biological system.
Epigenetics remains a rapidly developing field that studies how the chromatin state contributes to differential gene expression in distinct cell types at different developmental stages. Epigenetic regulation contributes to a broad spectrum of biological processes, including cellular differentiation during embryonic development and homeostasis in adulthood. A critical strategy in epigenetic studies is to examine how various histone modifications and chromatin factors regulate gene expression. To address this, Chromatin Immunoprecipitation (ChIP) is used widely to obtain a snapshot of the association of particular factors with DNA in the cells of interest.
ChIP technique commonly uses cultured cells as starting material, which can be obtained in abundance and homogeneity to generate reproducible data. However, there are several caveats: First, the environment to grow cells in Petri dish is different from that in vivo, thus may not reflect the endogenous chromatin state of cells in a living organism. Second, not all types of cells can be cultured ex vivo. There are only a limited number of cell lines, from which people can obtain enough material for ChIP assay.
Here we describe a method to do ChIP experiment using Drosophila tissues. The starting material is dissected tissue from a living animal, thus can accurately reflect the endogenous chromatin state. The adaptability of this method with many different types of tissue will allow researchers to address a lot more biologically relevant questions regarding epigenetic regulation in vivo1, 2. Combining this method with high-throughput sequencing (ChIP-seq) will further allow researchers to obtain an epigenomic landscape.
(The entire ChIP procedure takes approximately two days. Preparation of ChIP libraries for high-throughput sequencing takes another 2-3 days.)
1. Dissect and Prepare Tissue for ChIP Experiment (~1 million cells)
2. Prepare Supernatant with Protein-DNA Conjugation for ChIP Assay
3. Analyze ChIP-ed DNA
3a. Analyze ChIP-ed DNA using qPCR
3b. Amplify ChIP-ed DNA for high throughput sequencing.
3c. Solexa pipeline analysis
4. Representative Results
Examples of ChIP-qPCR results using bam (bag of marbles) mutant testes are shown in Figure 14. In bam testes, there is a failure in the transition from proliferative spermatogonia to differentiating spermatocytes5, 6. We use bam testes as a source for undifferentiated germ cells, which are enriched in this tissue type. Differentiation genes required for sperm differentiation, such as male-specific transcription factor 87 (mst87F), don juan (dj), and fuzzy onion (fzo) are not expressed in bam testes. These genes are highly enriched with the repressive H3K27me3 histone modification7 (Figure 1A), but are devoid of the active H3K4me3 histone modification8 (Figure 1B), a chromatin signature we called ‘monovalent’7. Enrichment of either H3K27me3 or H3K4me3 is determined by normalization to a constitutively expressed Cyclin A (CycA) gene4.
ChIP-seq analysis using the same set of antibodies (i.e. repressive H3K27me3 and active H3K4me3) with bam mutant testes (Figure 2) validated the qPCR results shown in Figure 1. For the three tested terminal differentiation genes mst87F, dj and fzo, their genomic regions are highly enriched with H3K27me3 but not H3K4me3 (Figure 2A-2C). In contrast, the constitutively expressed CycA gene has significant H3K4me3 but little H3K27me3 binding near its transcription start site (TSS) (Figure 2D).
Furthermore, the ChIP profiles of H3K27me3 and H3K4me3 near the TSSs of four classes of differentially expressed genes are consistent with the function of each histone modification. As shown in Figure 3A, enrichment of H3K27me3 downstream of the TSSs is inversely correlated with gene expression level. The silent genes have the highest H3K27me3 whereas the highly expressed genes have no H3K27me3 binding. These data is consistent with the repressive role of H3K27me3 on gene expression. In contrast, enrichment of H3K4me3 around the TSSs showed the opposite correlation with gene expression level (Figure 3B), consistent with the active role of H3K4me3 on gene expression.
Figure 1. qPCR analysis of ChIP-ed DNA using antibody against either the repressive H3K27me3 histone modification (A) or active H3K4me3 histone modification (B) in undifferentiated cell-enriched bam mutant testes. (A) In bam testes, differentiation genes (Mst87F, dj, fzo) are enriched with the H3K27me3 repressive histone modification. (B) Differentiation genes are depleted with H3K4me3 active mark. The level of ChIP DNA (ChIP DNA/input) at the target gene (Mst87F, dj or fzo) was first normalized to the level of ChIP DNA at the control CycA gene. Error bars indicate standard deviation from three independent biological replicates.
Figure 2. UCSC genome browser snapshots of H3K27me3 and H3K4me3 enrichment across the entire genomic regions of (A) Mst87F (B) dj and (C) fzo, (D) CycA genes. The circled H3K27me3-enriched region in (D) reflects the chromatin status of an overlapping gene CG7264, which is lowly expressed in bam testes (RPKM=1) but highly expressed in wild-type testes (RPKM=130)8 (ChIP-seq data from7). Probes used in quantitative PCR analysis of ChIP results in Figure 1 are labeled at the bottom of each plot. Click here to view larger figure.
Figure 3. ChIP-seq profiles using H3K27me3 and H3K4me3 in bam testes7. The four gene groups (7,509 genes) were classified according to their gene expression levels based on RNA-seq results8. The representative classes of genes are plotted for enrichment of a particular histone modification, using sequences from 3kb upstream to 3kb downstream of their transcription start sites (TSSs). This generates a profile of enrichment of (A) H3K27me3 (K27) and (B) H3K4me3 (K4) ChIP-seq analyses in each group. Click here to view larger figure.
The versatility of ChIP analyses discussed in this protocol can be used on different tissues, which provides an opportunity to study the chromatin state in a biologically relevant system. ChIP experiments using cells from cultured systems are convenient to perform because large quantity of cells can be easily obtained. However, cultured cells do not necessarily reflect cells in a multi-cellular environment. By developing this technique using dissected tissue from live animals, we can address many questions that cultured cells cannot.
Despite the usefulness of this protocol, there are several caveats. For example, the dissected tissue is still heterogeneous with several different cell types. While relatively homogeneous cells can be obtained in Drosophila tissues such as imaginal discs, other tissues like guts, testes, and ovaries which all contain limited adult stem cells are heterogeneous. This complication can be partially addressed using powerful Drosophila genetics. For example, although adult stem cells exist in a small population in tissues discussed above and are extremely difficult to be isolated in a sufficient number for ChIP analysis, stem cell population can be enriched using genetically mutated background. The bam gene is necessary for stem cell differentiation in Drosophila germline lineage5,6. Therefore bam mutant gonads are a good source for enriched undifferentiated germ cells. By performing ChIP using bam mutant, we can obtain an epigenomic landscape in undifferentiated germ cells.
The authors have nothing to disclose.
The authors would like to thank Dr. Keji Zhao’s lab (NIH/NHLBI) for their help in providing sequencing results. We would also like to thank the UCSC genome project for the use of Genome Browser to visualize mapped sequencing reads.
This work has been supported by the R00HD055052 NIH Pathway to Independence Award and R01HD065816 from NICHD, the Lucile Parkard Foundation, and the Johns Hopkins University start-up funding to X.C.
Name of the reagent | Company | Catalogue number | Comments |
Complete Mini protease inhibitor cocktail | Roche | 11836153001 | |
Formaldehyde (37%) | Supelco | 47083-U | |
PMSF | Sigma | 78830 | |
Kontes pellet pestle | Fischer Scientific | K749521-1590 | |
PCR Purification Kit | Qiagen | 28104 | |
Linear polyacrylamide | Sigma | 56575-1ML | |
Glycogen | Qiagen | 158930 | |
SYBR green/ROX qPCR Master Mix | Fermentas | K0223 | |
Mini plate spinner | Labnet | Z723533 | |
Real time PCR system | Applied Biosystem | 4351101 | |
Small Volume Ultrasonic Processor | Misonix | HS-XL2000 | Model discontinued |
Dynabeads, Protein A | Invitrogen | 100-01D | |
Dynamag magnet | Invitrogen | 123-21D | |
Phenol:Chlorofrom:IAA | Invitrogen | 15593-049 | |
Epicentre DNA END-Repair Kit | Epicentre Biotechnologies | ER0720 | |
MinElute Reaction Cleanup Kit | Qiagen | 28204 | |
Klenow Fragment (3’→5′ exo–) | New England Biolabs | M0212S | |
T4 DNA ligase | Promega Corporation | M1794 | |
Adaptor oligonucleotides | Illumina | PE-400-1001 | |
Paired-End Primer 1.0 and 2.0 | Illumina | 1001783 1001 784 |
|
E-Gel Electorphoresis system | Invitrogen | G6512ST | |
2X Phusion HF Mastermix | Finnzymes | F-531 |