news and views
The genome shows its sensitive side Anil Raj & Graham McVicker
Many of the traits that make individuals unique are encoded by genetic differences in their genomes. Recent evidence suggests that many of these genetic differences do not affect genes directly but instead alter regulatory
sequences that control when they are switched on and off. Two papers published in this issue of Nature Methods1,2 and one paper published in the December issue3 describe technical advances, and highlight previously
Millions of cells
500–50,000 cells
DNase I
Tn5 transposase
Cleavage of sensitive sites
TF
TF
npg
© 2014 Nature America, Inc. All rights reserved.
New methods for measuring the sensitivity of chromatin to DNase digestion and Tn5 transposition help us map and interpret the genome’s regulatory sequences.
Long nucleosome DNA fragments
Sequence ends of fragments and map back to genome
Short DNA fragments from sensitive region
Infer locations of sensitive sites and nucleosomes
TF footprint?
Figure 1 | Mapping regulatory regions with paired-end DNase-seq or ATAC-seq. Short fragments come from nucleosome-depleted sequences, whereas long fragments originate from flanking nucleosomes. Anil Raj and Graham McVicker are in the Department of Genetics, Stanford University, Stanford, California, USA; and Graham McVicker is in the Department of Biostatistics and Computational Biology, Dana-Farber Cancer Institute, Boston, Massachusetts, USA. e-mail: [email protected]
unknown challenges, for the mapping of regulatory sequences in the genome. Regulatory sequences, when active, are bound by transcription factors (TFs), which are proteins that recognize specific DNA sequences. Once bound, TFs recruit other proteins that transcribe, or ‘turn on’, nearby genes. A complete description of all of the regulatory sequences that are active in a given cell type is therefore fundamentally important for understanding how our genome functions. Direct measurement of TF-bound sequences, such as by chromatin immunoprecipitation, provides information about only one TF at a time, even though hundreds of TFs may be active in a single cell. Another approach is to look for the indirect effects of TFs on chromatin. At its most basic level, chromatin is made up of a repeating series of nucleosomes (complexes of histone proteins) encircled by DNA. When TFs bind to the genome, they displace nucleosomes, thereby exposing the DNA and making it more sensitive to cleavage by enzymes. The methods described in this issue exploit the increased sensitivity of nucleosome-depleted chromatin to identify active regulatory sequences. He et al.1 and Vierstra et al.2 base their work on DNase-seq, a method that has already proven very successful at identifying active regulatory regions in the genome4–7. First, an enzyme known as DNase I is used to preferentially cleave nucleosome-depleted DNA sequences. Pairs of DNase cuts generate short fragments that are then sequenced and mapped back to the genome to identify sensitive ‘open chromatin’ regions. The number of fragments that map to a sequence is a measure of regulatory activity; moreover, sites bound by some TFs show highly specific patterns of DNase I cleavage. These cut patterns, called ‘DNase footprints’, have been used to identify the binding of specific TFs in several studies5,6,8,9. Vierstra et al.2 extend the DNase-seq protocol in an approach they term DNaseFLASH (DNase I–released fragment-length analysis of hypersensitivity). They sequence both ends of the DNA fragments that are released by DNase cleavage1,2, which allows the fragment lengths to be determined after the fragment ends are mapped to
nature methods | VOL.11 NO.1 | JANUARY 2014 | 39
the genome. The authors demonstrate that short fragments, up to 200 base pairs long, are generated when both DNase cuts occur within an open chromatin region between nucleosomes. In contrast, long fragments, approximately 200–500 base pairs in length, are created when one of the cuts occurs in an open chromatin region and the other occurs on the far side of a nucleosome that flanks the open region. Vierstra et al.2 use this knowledge to characterize both regulatory sequences and the nucleosomes that are adjacent to them (Fig. 1). He et al.1 also perform paired-end sequencing of DNase-seq fragments and investigate how DNase I concentration and fragment size affect their results. Similarly to Vierstra et al.2, they find that short fragments predominantly originate from regulatory regions between nucleosomes, whereas longer fragments span nucleosomes. Although the short fragments are more useful for identifying TF binding, the long fragments provide information about nucleosome spacing. In the third paper, Buenrostro et al.3 describe a complementary method called ATAC-seq (assay for transposase-accessible chromatin
using sequencing). ATAC-seq uses an enzyme known as Tn5 transposase to preferentially tag and sequence DNA from open chromatin3. ATAC-seq provides similar information to that given by the new DNase-seq methods, but its protocol is simpler and requires far fewer cells (Fig. 1). This opens up the possibility of applying it to samples for which large numbers of cells cannot be obtained. For example, ATAC-seq could be used to identify regulatory sequences in rare cancer tissues or in cells that are undergoing differentiation from induced pluripotent stem cells. ATAC-seq may also serve as a useful diagnostic tool because it can be applied to small samples on clinically relevant timescales. Despite these technical advances, biases in DNase I and Tn5 cleavage are not well understood and pose serious analytical challenges for identifying TF binding sites. Using ‘naked’ DNA that is free of nucleosomes and bound TFs, He and colleagues demonstrate a strong preference for DNase I to cut some sequences over others1. This suggests that many previously characterized DNase footprints reflect the cutting bias of the enzyme rather than the actual binding of a
npg
© 2014 Nature America, Inc. All rights reserved.
news and views
40 | VOL.11 NO.1 | JANUARY 2014 | nature methods
TF. The authors illustrate this point by showing that the footprints of many TFs mirror the cutting preferences of DNase I and are thus likely to be artifacts. Although the paper by He et al.1 brings to light serious technical challenges in the interpretation of DNase I footprints, methods for measuring chromatin sensitivity will continue to be incredibly useful for mapping active regulatory sequences in the genome. COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. 1. He, H.H. et al. Nat. Methods 11, 72–77 (2014). 2. Vierstra, J., Wang, H., John, S., Sandstrom, R. & Stamatoyannopoulos, J.A. Nat. Methods 11, 65–71 (2014). 3. Buenrostro, J.D., Giresi, P.G., Zaba, L.C., Chang, H.Y. & Greenleaf, W.J. Nat. Methods 10, 1213–1218 (2013). 4. Boyle, A.P. et al. Cell 132, 311–322 (2008). 5. Hesselberth, J.R. et al. Nat. Methods 6, 283–289 (2009). 6. Neph, S. et al. Nature 489, 83–90 (2012). 7. Thurman, R.E. et al. Nature 489, 75–82 (2012). 8. Boyle, A.P. et al. Genome Res. 21, 456–464 (2011). 9. Pique-Regi, R. et al. Genome Res. 21, 447–455 (2011).
The genome shows its sensitive side Anil Raj & Graham McVicker
Many of the traits that make individuals unique are encoded by genetic differences in their genomes. Recent evidence suggests that many of these genetic differences do not affect genes directly but instead alter regulatory
sequences that control when they are switched on and off. Two papers published in this issue of Nature Methods1,2 and one paper published in the December issue3 describe technical advances, and highlight previously
Millions of cells
500–50,000 cells
DNase I
Tn5 transposase
Cleavage of sensitive sites
TF
TF
npg
© 2014 Nature America, Inc. All rights reserved.
New methods for measuring the sensitivity of chromatin to DNase digestion and Tn5 transposition help us map and interpret the genome’s regulatory sequences.
Long nucleosome DNA fragments
Sequence ends of fragments and map back to genome
Short DNA fragments from sensitive region
Infer locations of sensitive sites and nucleosomes
TF footprint?
Figure 1 | Mapping regulatory regions with paired-end DNase-seq or ATAC-seq. Short fragments come from nucleosome-depleted sequences, whereas long fragments originate from flanking nucleosomes. Anil Raj and Graham McVicker are in the Department of Genetics, Stanford University, Stanford, California, USA; and Graham McVicker is in the Department of Biostatistics and Computational Biology, Dana-Farber Cancer Institute, Boston, Massachusetts, USA. e-mail: [email protected]
unknown challenges, for the mapping of regulatory sequences in the genome. Regulatory sequences, when active, are bound by transcription factors (TFs), which are proteins that recognize specific DNA sequences. Once bound, TFs recruit other proteins that transcribe, or ‘turn on’, nearby genes. A complete description of all of the regulatory sequences that are active in a given cell type is therefore fundamentally important for understanding how our genome functions. Direct measurement of TF-bound sequences, such as by chromatin immunoprecipitation, provides information about only one TF at a time, even though hundreds of TFs may be active in a single cell. Another approach is to look for the indirect effects of TFs on chromatin. At its most basic level, chromatin is made up of a repeating series of nucleosomes (complexes of histone proteins) encircled by DNA. When TFs bind to the genome, they displace nucleosomes, thereby exposing the DNA and making it more sensitive to cleavage by enzymes. The methods described in this issue exploit the increased sensitivity of nucleosome-depleted chromatin to identify active regulatory sequences. He et al.1 and Vierstra et al.2 base their work on DNase-seq, a method that has already proven very successful at identifying active regulatory regions in the genome4–7. First, an enzyme known as DNase I is used to preferentially cleave nucleosome-depleted DNA sequences. Pairs of DNase cuts generate short fragments that are then sequenced and mapped back to the genome to identify sensitive ‘open chromatin’ regions. The number of fragments that map to a sequence is a measure of regulatory activity; moreover, sites bound by some TFs show highly specific patterns of DNase I cleavage. These cut patterns, called ‘DNase footprints’, have been used to identify the binding of specific TFs in several studies5,6,8,9. Vierstra et al.2 extend the DNase-seq protocol in an approach they term DNaseFLASH (DNase I–released fragment-length analysis of hypersensitivity). They sequence both ends of the DNA fragments that are released by DNase cleavage1,2, which allows the fragment lengths to be determined after the fragment ends are mapped to
nature methods | VOL.11 NO.1 | JANUARY 2014 | 39
the genome. The authors demonstrate that short fragments, up to 200 base pairs long, are generated when both DNase cuts occur within an open chromatin region between nucleosomes. In contrast, long fragments, approximately 200–500 base pairs in length, are created when one of the cuts occurs in an open chromatin region and the other occurs on the far side of a nucleosome that flanks the open region. Vierstra et al.2 use this knowledge to characterize both regulatory sequences and the nucleosomes that are adjacent to them (Fig. 1). He et al.1 also perform paired-end sequencing of DNase-seq fragments and investigate how DNase I concentration and fragment size affect their results. Similarly to Vierstra et al.2, they find that short fragments predominantly originate from regulatory regions between nucleosomes, whereas longer fragments span nucleosomes. Although the short fragments are more useful for identifying TF binding, the long fragments provide information about nucleosome spacing. In the third paper, Buenrostro et al.3 describe a complementary method called ATAC-seq (assay for transposase-accessible chromatin
using sequencing). ATAC-seq uses an enzyme known as Tn5 transposase to preferentially tag and sequence DNA from open chromatin3. ATAC-seq provides similar information to that given by the new DNase-seq methods, but its protocol is simpler and requires far fewer cells (Fig. 1). This opens up the possibility of applying it to samples for which large numbers of cells cannot be obtained. For example, ATAC-seq could be used to identify regulatory sequences in rare cancer tissues or in cells that are undergoing differentiation from induced pluripotent stem cells. ATAC-seq may also serve as a useful diagnostic tool because it can be applied to small samples on clinically relevant timescales. Despite these technical advances, biases in DNase I and Tn5 cleavage are not well understood and pose serious analytical challenges for identifying TF binding sites. Using ‘naked’ DNA that is free of nucleosomes and bound TFs, He and colleagues demonstrate a strong preference for DNase I to cut some sequences over others1. This suggests that many previously characterized DNase footprints reflect the cutting bias of the enzyme rather than the actual binding of a
npg
© 2014 Nature America, Inc. All rights reserved.
news and views
40 | VOL.11 NO.1 | JANUARY 2014 | nature methods
TF. The authors illustrate this point by showing that the footprints of many TFs mirror the cutting preferences of DNase I and are thus likely to be artifacts. Although the paper by He et al.1 brings to light serious technical challenges in the interpretation of DNase I footprints, methods for measuring chromatin sensitivity will continue to be incredibly useful for mapping active regulatory sequences in the genome. COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. 1. He, H.H. et al. Nat. Methods 11, 72–77 (2014). 2. Vierstra, J., Wang, H., John, S., Sandstrom, R. & Stamatoyannopoulos, J.A. Nat. Methods 11, 65–71 (2014). 3. Buenrostro, J.D., Giresi, P.G., Zaba, L.C., Chang, H.Y. & Greenleaf, W.J. Nat. Methods 10, 1213–1218 (2013). 4. Boyle, A.P. et al. Cell 132, 311–322 (2008). 5. Hesselberth, J.R. et al. Nat. Methods 6, 283–289 (2009). 6. Neph, S. et al. Nature 489, 83–90 (2012). 7. Thurman, R.E. et al. Nature 489, 75–82 (2012). 8. Boyle, A.P. et al. Genome Res. 21, 456–464 (2011). 9. Pique-Regi, R. et al. Genome Res. 21, 447–455 (2011).
