single-chromosome transcriptional proﬁling reveals ...rajlab/pdfs/nature... · reveals...

BRIEF COMMUNICATIONS

246 | VOL.10 NO.3 | MARCH 2013 | NATURE METHODS

with 16 short, fluorescently labeled oligonucleotides4 (Fig. 1a and Supplementary Fig. 1) enabled us to measure whether the gene was actively transcribing5 and, if it was, to identify the three-dimensional coordinates of the gene6–8. Control experiments con-firmed that the intron spot marked the site of transcriptionally active genes (Supplementary Figs. 2–7). Note that even genes considered constitutively active transcribe RNA in ‘bursts’ thought to arise from random aspects of the transcriptional process9–12. The overall transcription rate is proportional to the probability of finding such a spot for each gene (Supplementary Discussion).

To concomitantly measure overall chromosome structure, we designed probes targeting the introns of 20 genes along chromo-some 19 (Supplementary Fig. 1 and Supplementary Table 1). This yielded an average resolution of 3 megabases (Mb) with a minimum of 360 kilobases (kb), but we were also able to distinguish loci sepa-rated by just 30 kb (Supplementary Fig. 6). To measure all 20 genes’ transcriptional statuses simultaneously, we labeled each gene’s introns with a particular ‘pseudocolor’, which is a distinct code for each gene consisting of either two or three (out of five) spectrally distinguishable fluorophores13,14 (Fig. 1a). To assign gene identity, we looked for colocalization of two or three spots in each fluores-cence channel (Supplementary Fig. 8). In human foreskin fibro-blasts, we could discern two clearly separated chromosomes (Fig. 1b) 78% of the time (Supplementary Fig. 9). On average, we found 6 2 expressing genes (out of the 20 labeled) per chromosome. The positions of these genes were more spread out than expected15 (Supplementary Fig. 6 and Supplementary Discussion). We found that using 16 singly colored probes did not change the spot detec-tion efficiency (Online Methods), nor did pseudocoloring incur

Single-chromosome transcriptional profiling reveals chromosomal gene expression regulationMarshall J Levesque & Arjun Raj

We report intron chromosomal expression FISH (iceFISH), a multiplex imaging method for measuring gene expression and chromosome structure simultaneously on single chromosomes. We find substantial differences in transcriptional frequency between genes on a translocated chromosome and the same genes in their normal chromosomal context in the same cell. Correlations between genes on a single chromosome pointed toward a cis chromosome-level transcriptional interaction spanning 14.3 megabases.

The transcription of a gene’s DNA into RNA is thought to be con-trolled largely by the interaction of regulatory proteins with DNA sequences proximal to the gene itself. At the same time, genes are organized by the thousands into chromosomes, thus raising the possibility that the structure or organization of chromosomes influ-ences transcription1,2; however, little is known about how organi-zation at the chromosome-length scale affects gene expression. Here we begin to address this question with a method based on RNA FISH3,4 called iceFISH that enabled us to generate transcriptional profiles of 20 genes simul-taneously along individual copies of human chromosome 19 in single cells.

We took advantage of the fact that introns typically degrade rapidly after being spliced out of nascent RNA. Labeling the intron

Department of Bioengineering, University of Pennsylvania, Philadelphia, Pennsylvania, USA. Correspondence should be addressed to M.J.L. or A.R. ([email protected]).RECEIVED 17 NOVEMBER 2012; ACCEPTED 18 JANUARY 2013; PUBLISHED ONLINE 17 FEBRUARY 2013; DOI:10.1038/NMETH.2372

aSpliced mRNA

Export andTranslation

MaturemRNA

ExonExon Exon

Chromosome 19

Cytoplasm Nucleus

IntronIntron

Introns degraded beforeleaving site of transcription

04

812

1620

1612

84

0

02

4

x (µm)

y (µm)

z (µ

m)

5 base colors

2 colors 3 colors20 pseudocolors

b

RPL18A

UBA52

ZNF91

UBA2EIF3KSUPT5HFBLEGLN2RPS19TOMM40MARK4

SLC1A5

PPP2R1ARPS9ZNF444

PTBP1

EEF2

0 Mb

Chr

omos

ome

19

59.13 Mb

DNMT1ILF3

DHPS

Chr 19

Chr 19

Nucleus

Figure 1 | Unique identification of 20 loci on chromosome (chr) 19 by RNA FISH targeting introns in human foreskin fibroblasts. (a) Pseudocoloring scheme for labeling the site of transcription by targeting gene introns with a series of labeled oligonucleotide probes. (b) Transcriptional activity and location of the 20 genes computationally identified from images from each fluorescence channel. Scale bar, 5 m.

npg

© 2

013

Nat

ure

Am

eric

a, In

c. A

ll rig

hts

rese

rved

.

http://www.nature.com/doifinder/10.1038/nmeth.2372

NATURE METHODS | VOL.10 NO.3 | MARCH 2013 | 247


a significant rate of spot misidentification (Supplementary Fig. 10). We ensured that the cells we analyzed were in the G0/G1 stage of the cell cycle by colabeling cyclin A2 mRNA, which is abundant during the S, G2 and M phases of the cell cycle16 (Supplementary Fig. 11).

By grouping actively transcribing genes into territories corresponding to each chromosome, we constructed tran-scriptional profiles showing which of our 20 genes were on or off per chromo-some. Researchers largely believe that gene transcription depends on both chromosome-extrinsic trans factors (such as transcription factors) and local cis fac-tors on the DNA (typically within 1 Mb of the gene itself). Our method enabled us to examine the possibility that nonlo-cal mechanisms at the chromosome scale may also regulate transcription.

Translocations provide a means to search for such possibilities: although they disrupt the large-scale structure of a chromosome, the cell’s trans envi-ronment and local cis DNA regulatory code remain unchanged for most genes on the translocated chromosome. For example, HeLa cells contain two intact copies of chromosome 19 and one copy that is split into two pieces fused to parts of other chromosomes17: one, denoted t(6;19), consists of the first 17–20 Mb of chromosome 19 fused to part of chromo-some 6; and the other, denoted t(13;19), consists of the remaining 40–43 Mb of chromosome 19 translocated onto a por-tion of chromosome 13 (Fig. 2a and Supplementary Fig. 12). Our iceFISH data recapitulated these genetic rearrange-ments (Fig. 2b). We found that most genes on t(13;19) were up to fivefold more transcriptionally active than those on the normal copies of chromosome 19 (Fig. 2c and Supplementary Fig. 13), consistent with the existence of chromosome-specific transcriptional regulation that the translocation may have dis-rupted. Intron spot intensities were roughly the same on all the chromosomes we examined (Supplementary Fig. 14), suggest-ing that transcriptional hyperactivation results from an increased probability of a gene being active (Supplementary Discussion).

We also found that the transcriptional frequencies of two genes from chromosome 13 (DIAPH3 and MZT1) were roughly twofold higher on t(13;19) than on the normal copies of chromosome 13 (Fig. 2c and Supplementary Fig. 15), suggesting that this trans-location resulted in hyperactivation of all genes on t(13;19) irrespective of location. Meanwhile, transcription of the chro-mosome 19 genes on t(6;19) was similar to that of the normal copies (Fig. 2c), suggesting that translocations do not necessar-ily lead to transcriptional changes. Reports of such effects18 are not widespread because of the averaging effects of most assays. We also found that these transcriptional differences did not

appear to correlate with differences in spatial chromosome conformation (Supplementary Figs. 6, 16 and 17).

We next looked for evidence of interactions governing the transcription of genes within a single chromosome by examin-ing whether the transcriptional status of one gene in our panel affected the transcriptional status of another gene on the same chromosome. Such an interaction would manifest itself as a devia-tion from independence, with positive correlations signifying that the two genes ‘A’ and ‘B’ would have a likelihood greater than chance to be actively transcribing at the same time on the same chromosome, and with anticorrelations indicating that transcrip-tion of genes A and B would be mutually exclusive.

We found that most pairwise interactions on single chromo-somes did not show significant deviations from independence (Fig. 3 and Supplementary Figs. 18 and 19). However, one pair of genes, RPS19 and ZNF444 (separated by 14.3 Mb), showed an anticorrelation (R = −0.40 0.08; P = 3.99 × 10−5, Fisher exact test). One explanation for this anticorrelation is fluctuations in a potential trans-acting factor, such as a transcription factor, that activated RPS19 and inactivated ZNF444 in some cells while acti-vating ZNF444 and inactivating RPS19 in others. Such a trans factor would, however, also affect the copy of the gene on the

Chr 19

Chr 19t(13;19)

Fusionchromosomes

t(6;19)

HeLa karyotype

Chr 19

Chr 19q-arm

Chr 6p-arm

Chr 6q-arm

Chr 19p-arm

t(13;19)

t(6;19)Chr 13q-arm

a b

RPL18AUBA52

ZNF91

UBA2EIF3KSUPT5HFBLEGLN2RPS19TOMM40MARK4SLC1A5PPP2R1ARPS9ZNF444

PTBP1

EEF2

0 Mb

Chr

omos

ome

19

59.13 Mb

DNMT1ILF3DHPS

DIAPH3

MZT1

RPL18A

UBA52

ZNF91

UBA2EIF3K

SUPT5H

FBLEGLN2

RPS19

TOMM

40

MARK4

SLC1A5

PPP2R1A

RPS9ZNF444

*** **

** ******

***

******

***

***

Chromosome 19 gene

Chromosome 19 gene

Chr 13 gene

0

1.0

iceF

ISH

spo

ts p

er c

hrom

osom

e

cCentromereTranslocation

Chr 19Chr 13t(13;19)t(6;19)

PTBP1

EEF2DNM

T1

ILF3DHPS

0

1.0

t(6;19)

n = 50 n = 30

n = 30

Figure 2 | Translocated portions of chromosome (chr) 19 display different expression patterns than intact chromosomes. (a) Schematic showing chr 19 and its derivatives in our HeLa cells. (b) Computational identification of actively transcribing genes revealed the two intact copies and the two translocated pieces of chr 19. Scale bar, 5 m. (c) Comparison of the

transcriptional activity of the genes on the translocated fragments (as measured by frequency of observing a transcription site per chr) to the activity of the intact copies of chr 19. We similarly analyzed two genes on chr 13 (MZT1 and DIAPH3), as described in Supplementary Figure 15. P values for the difference in frequency (using a binomial distribution test) *P < 0.05, **P < 0.01, ***P < 0.001.

npg

© 2

013

Nat

ure

Am

eric

a, In

c. A

ll rig

hts

rese

rved

.

248 | VOL.10 NO.3 | MARCH 2013 | NATURE METHODS


other chromosome 19 (ref. 19). Thus, we looked for an anticorrelation between RPS19 on one chromosome and ZNF444 on the other copy of chromosome 19 in the same cell. We found that the inter-chromosomal interaction between the genes was qualitatively different, having a mild and less statistically significant positive correlation (R = 0.33 0.09; P = 6.90 × 10−4, Fisher exact test), indicating that the interaction between these genes is indeed a cis effect confined to the chromosome itself. The lack of anticorrelation between the chromosome 19 copies (both between the pair of genes and between each gene and its copy; Fig. 3) also precludes the possibility of genetic imprinting. Notably, we also found the exact same pattern of interactions when examining the two intact copies of chromosome 19 in HeLa cells (Fig. 3). The interaction between RPS19 and ZNF444 disappeared, however, on t(13;19) (Supplementary Fig. 20), suggesting that whatever mechanism is responsible requires a fully intact chromosome 19. This interac-tion did not, however, depend on the distance between any pair of genes (Supplementary Fig. 21) or on the distance of the chromo-some from the nuclear periphery (Supplementary Fig. 17).

iceFISH provides a complement to chromosome conforma-tion assays20 that look for interactions at the DNA (rather than transcriptional) level. We believe that iceFISH and similar tools will allow us to determine the prevalence of these chromosome-level regulatory phenomena and uncover their underlying mechanisms.

METHODSMethods and any associated references are available in the online version of the paper.

Note: Supplementary information is available in the online version of the paper.

ACKNOWLEDGMENTSWe thank Biosearch Technologies (especially M. Beal and R. Cook) for providing many of the reagents used in the iceFISH assay; J. Morrissette for her lab’s assistance in performing the cytogenetic analysis; L. Cai, H. Youk, D. Muzzey, J. Gore, H. Maamar and O. Padovan-Merhar for insightful comments; G. Nair for discussions about statistics; and the lab of P. Sharp (Massachusetts Institute of Technology) for providing the HeLa cells. We gratefully acknowledge a University of Pennsylvania University Research Foundation pilot grant, the US National Institutes of Health Director’s New Innovator Award (1DP2OD008514) and a Burroughs-Wellcome Fund Career Award at the Scientific Interface for supporting our work.

AUTHOR CONTRIBUTIONSM.J.L. and A.R. conceived of the project, performed the analyses and wrote the paper. M.J.L. performed the experiments.

COMPETING FINANCIAL INTERESTSThe authors declare competing financial interests: details are available in the online version of the paper.

Published online at http://www.nature.com/doifinder/10.1038/nmeth.2372. Reprints and permissions information is available online at http://www.nature.com/reprints/index.html.

1. Fraser, P. & Bickmore, W. Nature 447, 413–417 (2007).2. Cremer, T. & Cremer, C. Nat. Rev. Genet. 2, 292–301 (2001).3. Femino, A.M., Fay, F.S., Fogarty, K. & Singer, R.H. Science 280, 585–590

(1998).4. Raj, A., van den Bogaard, P., Rifkin, S.A., van Oudenaarden, A. & Tyagi, S.

Nat. Methods 5, 877–879 (2008).5. Fremeau, R.T., Lundblad, J.R., Pritchett, D.B., Wilcox, J.N. & Roberts, J.L.

Science 234, 1265–1269 (1986).6. Gribnau, J. et al. EMBO J. 17, 6020–6027 (1998).7. Xing, Y., Johnson, C.V., Dobner, P.R. & Lawrence, J.B. Science 259,

1326–1330 (1993).8. Vargas, D.Y. et al. Cell 147, 1054–1065 (2011).9. Golding, I., Paulsson, J., Zawilski, S.M. & Cox, E.C. Cell 123, 1025–1036

(2005).10. Chubb, J.R., Trcek, T., Shenoy, S.M. & Singer, R.H. Curr. Biol. 16,

1018–1025 (2006).11. Raj, A., Peskin, C.S., Tranchina, D., Vargas, D.Y. & Tyagi, S. PLoS Biol. 4,

e309 (2006).12. Suter, D.M. et al. Science 332, 472–474 (2011).13. Levsky, J.M., Shenoy, S.M., Pezo, R.C. & Singer, R.H. Science 297,

836–840 (2002).14. Lubeck, E. & Cai, L. Nat. Methods 9, 743–748 (2012).15. Mateos-Langerak, J. et al. Proc. Natl. Acad. Sci. USA 106, 3812–3817 (2009).16. Eward, K.L., Van Ert, M.N., Thornton, M. & Helmstetter, C.E. Cell Cycle 3,

1057–1061 (2004).17. Macville, M. et al. Cancer Res. 59, 141–150 (1999).18. Harewood, L. et al. Genome Res. 20, 554–564 (2010).19. Elowitz, M.B., Levine, A.J., Siggia, E.D. & Swain, P.S. Science 297,

1183–1186 (2002).20. Dekker, J., Rippe, K., Dekker, M. & Kleckner, N. Science 295, 1306–1311

(2002).

–log(P)

RP

L18A

UB

A52

ZN

F91

UB

A2

EIF

3KS

UP

T5H

FB

LE

GLN

2R

PS

19TO

MM

40M

AR

K4

SLC

1A5

PP

P2R

1AR

PS

9Z

NF

444

RP

S19

PTB

P1

EE

F2

DN

MT1

ILF

3D

HP

S

RPL18AUBA52ZNF91UBA2EIF3KSUPT5HFBLEGLN2RPS19

ZNF444

RP

S19

ZNF444

RP

S19

ZNF444

OF

F

ON

RP

S19

ZNF444

P = 2.7 × 10–5

R = –0.42P = 1.0R = 0.008

P = 4.0 × 10–5

R = –0.40P = 6.9 × 10–4

R = 0.33

TOMM40MARK4SLC1A5PPP2R1ARPS9ZNF444

PTBP1EEF2DNMT1ILF3DHPS

RP

L18A

UB

A52

ZN

F91

UB

A2

EIF

3KS

UP

T5H

FB

LE

GLN

2R

PS

19TO

MM

40M

AR

K4

SLC

1A5

PP

P2R

1AR

PS

9Z

NF

444

PTB

P1

EE

F2

DN

MT1

ILF

3D

HP

S

RPL18AUBA52ZNF91UBA2EIF3KSUPT5HFBLEGLN2RPS19TOMM40MARK4SLC1A5PPP2R1ARPS9ZNF444


RP

L18A

UB

A52

ZN

F91

UB

A2

EIF

3KS

UP

T5H

FB

LE

GLN

2R

PS

19TO

MM

40M

AR

K4

SLC

1A5

PP

P2R

1AR

PS

9Z

NF

444

PTB

P1

EE

F2

DN

MT1

ILF

3D

HP

S



RP

L18A

UB

A52

ZN

F91

UB

A2

EIF

3KS

UP

T5H

FB

LE

GLN

2R

PS

19TO

MM

40M

AR

K4

SLC

1A5

PP

P2R

1AR

PS

9Z

NF

444

PTB

P1

EE

F2

DN

MT1

ILF

3D

HP

SRPL18AUBA52ZNF91UBA2EIF3KSUPT5HFBLEGLN2RPS19TOMM40MARK4SLC1A5PPP2R1ARPS9ZNF444


45

31

5

27

27

49

23

9

45

24

6

25

35

34

16

15

1

2

P = 0.05

4

P = 0.015

6

P = 0.001

8

P = 0.0001

High likelihoodof correlation oranticorrelation

InterchromosomalIntrachromosomal

Low likelihoodof correlation oranticorrelation

Hum

an fo

resk

in fi

brob

last

sH

eLa

cells

OFF

ON

Figure 3 | Identification of a pair of genes showing an intrachromosomal but not inter-chromosomal expression relationship. The heat maps show the deviation from independence of the intra- and interchromosomal transcriptional activity of all pairs of genes we measured in human foreskin fibroblasts (n = 54 cells) and the two intact copies of chromosome 19 in HeLa cells (n = 50 cells). A smaller P value (Fisher exact test) indicates a more significant deviation from independence. The grid shows the number of chromosomes in which RPS19 and ZNF444 are transcriptionally active or inactive (designated ON and OFF as marked in lower left; markings apply to all boxes).

npg

© 2

013

Nat

ure

Am

eric

a, In

c. A

ll rig

hts

rese

rved

.






http://www.nature.com/reprints/index.html

http://www.nature.com/reprints/index.html

doi:10.1038/nmeth.2372 NATURE METHODS

ONLINE METHODSCell culture, fixation and fluorescence in situ hybridization. We grew primary human foreskin fibroblasts (ATCC CRL 2097) or HeLa cells in Dulbecco’s modified Eagle’s medium with Glutamax (DMEM, Life Technologies) supplemented with penicillin/streptomycin and 10% FBS. We enriched for G0/G1-phase cells through a double-thymidine block (2 mM thymidine in medium) procedure, which arrested cells at the beginning of S phase. We released the cells and let them go through S, G2, M, G1, S, G2 and M and then fixed them when they were in G1. We let the cells go through one complete cell cycle after the block to minimize any potential transcriptional or structural effects due to the block itself. To fix the cells, we followed the protocol of Raj et al.4 Briefly, we fixed the cells for 10 min at room temperature using 4% formaldehyde/10% formalin in 1× phosphate buffered saline solution (PBS) and followed the fixation by two rinses in 1× PBS, after which we permeabilized the cells with 70% EtOH and stored them at 4 °C at least overnight.

To perform fluorescence in situ hybridization (FISH), we again followed the procedure of ref. 4 with some minor modifications. We prewashed with a wash buffer containing 10% formamide and 2× saline-sodium citrate (SSC) and then hybridized by add-ing the appropriate amount and type of probe (described below) in a buffer containing 10% formamide, 2× SSC and 10% dextran sulfate (w/v). We empirically determined the optimal concentra-tion of each probe, which in most cases was roughly equivalent to the concentrations used in ref. 4. We hybridized our samples overnight in a humidified chamber kept at 37 °C, washed twice for 30 min with wash buffer at 37 °C (adding DAPI at a con-centration of 50 ng/mL in the second wash) and then imaged in 2× SSC as described below.

In the case of the experiments involving actinomycin D, we incubated HeLa cells in 2 g/mL of actinomycin D (Sigma) for 0, 30, 60 and 120 min (as described in Supplementary Fig. 3), after which we fixed the cells and performed FISH. We made sure to thoroughly mix the actinomycin D into the medium before add-ing it to avoid spatial inhomogeneity in the activity of the drug.

For the RNase experiments, we fixed and permeabilized the cells as just outlined, after which we aspirated the 70% EtOH, washed once with 1× PBS and then added 1× PBS with 10 g/mL of RNase A (Sigma). We incubated the fixed cells at 37 °C for 30 min, washed with 1× PBS and then proceeded with FISH as outlined above. As a control, we performed the exact same procedure on cells in a neighboring well but did not add RNase A to the 1× PBS for the incubation (as described in Supplementary Fig. 2).

Imaging. We imaged all our samples on a Nikon Ti-E inverted fluorescence microscope using a 100× Plan-Apo objective (numerical aperture of 1.43) and a cooled CCD camera (Pixis 1024B from Princeton Instruments). We sequentially acquired three-dimensional stacks of fluorescence images in six different fluorescence channels using filter sets for DAPI, Atto 488, Cy3, Alexa 594, Atto 647N and Atto 700. Our exposure times were roughly 2–3 s for most of the dyes except for DAPI (which we exposed for ~100 ms) and Atto 700 (~5 s, owing to somewhat weaker illumination on our apparatus). The spacing between con-secutive planes in our stacks was 0.3 m. The filter sets we used were 31000v2 (Chroma), 41028 (Chroma), SP102v1 (Chroma),

a custom set from Omega as described in ref. 4, SP104v2 (Chroma) and SP105 (Chroma) for DAPI, Atto 488, Cy3, Alexa 594, Atto 647N and Atto 700, respectively.

Image analysis. Once we acquired our images, we put them through an image analysis pipeline made up of custom semiau-tomated spot recognition software we wrote in MATLAB with the following series of steps.

1. We first identified candidate spots in the three-dimensional image by filtering the image with a Laplacian of Gaussian filter and taking the top 300 spots as candidates. In some cases, we also chose cells to analyze on the basis of their phase in the cell cycle. In those cases, we chose cells that had little or no cyclin A2 mRNAs. Our experiments in Supplementary Figure 11 validate this approach.

2. For each candidate, we then fit the candidate to a Laplacian of Gaussian intensity profile, thereby giving us precise estimates of the center, width and intensity of the spot.

3. Using histograms of the intensities and widths, we manually selected a subset of the spots with qualities (uniform width, higher intensity) that were higher than background. This is similar in spirit to the procedure described in ref. 4, in which the experi-menter chose a threshold to separate legitimate RNA spots from background spots. In this case, we erred on the side of including spots that may be background because our multicolor scheme for spot assignment provided us another means by which to discard background spots.

4. Once we had selected the spots, we ran software that found the fiducial markers (in this case, probes in all five RNA colors targeting SUZ12 mRNA, which were present at an abundance of roughly 20–50 clear cytoplasmic spots per cell). In this manner, we could measure the displacements between different fluores-cence channels in each cell individually. We then applied these shifts to align the computationally identified spots between the different fluorescence channels.

5. After alignment, we ran software that looked for colocalized spots corresponding to the particular pseudocoloring scheme we chose for the introns we targeted. We estimate that our software is roughly 75% accurate in assigning colocalized spots to particular genes at this stage.

6. We then went through a manual correction process in which we corrected mistakes the software made in identifying spots. Common issues were failure to detect dim (but clearly present) signals in one of the fluorescence channels and resolution of two spatially close fluorescent spots that the Laplacian of Gaussian filter-ing and candidate identification steps had labeled as a single spot.

7. Once we had correctly annotated the introns of the gene loci we had labeled, we examined cells manually to separate out individual chromosomes. We discarded cells in which the chro-mosomes overlapped because this made it difficult to assign gene spots to particular chromosomes.

To determine the distance of the chromosome from the nuclear periphery, we first determined the average position of the spots of the chromosome in x and y and then found the Euclidean dis-tance between this point and the nuclear periphery as outlined by our DAPI stain.

Characterization of error rate. To gain some sense of the rate of false positives, we performed a hybridization in foreskin

npg

© 2

013

Nat

ure

Am

eric

a, In

c. A

ll rig

hts

rese

rved

.

doi:10.1038/nmeth.2372NATURE METHODS

fibroblasts in which we left out 10 of the 20 genes comprising our iceFISH assay (randomly chosen by another member of the lab) and proceeded with our spot identification procedure as usual (Supplementary Fig. 10). We found that our rate of false identification was very low, with the vast majority (97%) of spots we assigned corresponding to genes that we had targeted in our assay.

We also probed a set of two genes (RPS19, TOMM40) one at a time with oligonucleotides labeled with a single dye rather than the combination of two or three dyes used in our pseudocoloring strategy. Our aim was to determine to what extent our pseudo-coloring strategy would result in false negatives in spot identi-fication. We found that the spot-per-chromosome frequencies measured with a singly colored probe alone were 0.56 and 0.27, whereas the spot frequencies measured by pseudocoloring were 0.57 and 0.25 for RPS19 and TOMM40, respectively, in a total of 30 cells. Although statistical effects preclude a definitive state-ment, our results are consistent with our pseudocoloring strategy correctly identifying virtually all spots detectable by RNA FISH targeting introns.

Probe design. We designed 20 base oligonucleotide probes against introns using custom FISH design software (http://www.biosearchtech.com/stellarisdesigner/). 90% of probes worked without needing optimization. Where possible, we tried to design 16 oligonucleotides targeting the first intron of the gene; for cases in which we were unable to fit all the oligonucleotide probes onto the first intron, we targeted the second intron as well. We ordered the oligonucleotides from Biosearch Technologies, who synthe-sized the oligonucleotides with amine groups attached to the 3 end. We coupled these 3 ends to various organic dyes (including Atto 488 (Atto-Tec), Cy3 (GE), Alexa 594 (Invitrogen), Atto 647N (Atto-Tec) and Atto 700 (Atto-Tec)) as indicated in the text and in Supplementary Table 1. We purified the probes by HPLC4.

Karyotyping of HeLa cells. We performed G-band analysis (karyotyping) on metaphase spreads of our HeLa cells, following standard procedures. The analysis indicated that our cells con-tained two intact copies of chromosome 19 and a full third copy of chromosome 19 split into two fragments and fused to other chromosomes (Supplementary Fig. 12). One fragment included the first half of the chromosome 19 p-arm and was fused to a large portion of chromosome 6. The second fragment was the remaining portion of chromosome 19 (half the p-arm through the centromere and entire q-arm), which was fused to the q-arm of chromosome 13. To conclusively demonstrate that chromo-some 19 was split in this particular way, we performed a DNA FISH analysis on the same metaphase spreads that we performed the G-band analysis on. We used probes targeting loci within the 19p13 and 19q13 cytogenetic bands on chromosome 19, each labeled with a different fluorophore (Abbott Molecular). The results confirmed the results of the G-band analysis. We per-formed this analysis on ten cells, each of which showed the same genetic abnormalities, indicating that the cells did not vary much in this particular characteristic from cell to cell.

Click-iT EdU analysis of cell-cycle progression. To demonstrate that cyclin A2 mRNA was an accurate marker of position in the cell cycle, we used the Click-iT EdU Alexa Fluor 594 imaging kit

(Invitrogen), which incorporates a targetable chemical into newly replicated DNA. In this case, we incubated foreskin fibroblasts with 10 M Click-iT EdU reagent for 5 min before fixing the cells. We performed our FISH protocol on these cells using a cyclin A2 mRNA Cy3 probe and, after the hybridization and wash steps, followed the instructions provided with the kit for fluorescently labeling the incorporated EdU. We ultimately did not elect to use the Click-iT EdU kit directly in most of our experiments (and instead opted to use cyclin A2) because we found that perform-ing the Click-iT EdU procedure interfered with our nascent RNA FISH detection, most likely either because of interference with transcription itself or because it made our spot detection less reliable owing to additional washing steps associated with the Click-iT procedure.

DNA FISH. We performed DNA FISH with BAC probes from Empire Genomics, using their reference hybridization protocol. In the human foreskin fibroblast cells we applied pairs of fluores-cently labeled BAC clones from the human RPCI-11 library tar-geting human chromosome 19 at positions 2.8–4.5 Mb (268O21), 39.0–39.5 Mb (31D10) or 52.5–52.7 Mb (43N16). We denatured the DNA by immersing the cells in a solution containing 70% formamide and 2× SSC buffer at 80 °C for 5 min and then trans-ferred them to a series of ethanol steps increasing to 70%, 85% and then 100% ethanol. We added 10 L of BAC probes to the air-dried sample, applied a coverslip and incubated overnight in a humidified slide chamber. The next day we washed the sample with 0.4× SSC at 73 °C for 2 min, removed the coverslip, trans-ferred it to room temperature 2× SSC for 1 min and then to 10 L of 2× SSC with DAPI at 50 ng/mL, and applied a new coverslip. We performed imaging similar to that for our iceFISH probes with dye pairs red 5-ROX and TAMRA.

Combined DNA-RNA FISH. We performed a sequential DNA-RNA FISH in HeLa cells by first performing DNA FISH using BAC clones and then performing RNA FISH, both by following the protocols outlined above. We found that both the bright exonic transcription sites and the intron spots were con-siderably brighter than single mRNA spots, thus showing that the RNA probes were not simply targeting the DNA directly. We compared the location of SLC1A5 exonic (Alexa 594) and intronic (ATTO647N) RNA to the location of DNA FISH probes using BAC clones RP11-687M15 (TAMRA).

Statistical analysis. In Figure 3, we looked for deviations from independence in the transcriptional frequencies of all pairs of genes we examined. We performed the Fisher exact test on each 2 × 2 table generated by counting the number of chromosomes in which gene A or B was transcriptionally active versus the number that were inactive. We reported the two-sided P value corresponding to the chance of obtaining a similar deviation from independence via random chance, with a smaller P value corresponding to a more significant result. In Figure 3, we show the results we obtained by analyzing a data set consisting of the combination of two independent biological replicates; we also performed the analysis on each individual biological replicate, as shown in Supplementary Figures 18 and 19. Note that we have not applied a multiple hypothesis correction in our presentation of the P values; however, our results would remain statistically

npg

© 2

013

Nat

ure

Am

eric

a, In

c. A

ll rig

hts

rese

rved

.

http://www.biosearchtech.com/stellarisdesigner/

http://www.biosearchtech.com/stellarisdesigner/

doi:10.1038/nmeth.2372 NATURE METHODS

significant if we applied the crude correction of just multiplying our P values by 190, which is the number of pairs of genes we examined. We chose to convey the information in this manner because the number of hypotheses tested depends on the particu-lar question being asked of the data. For instance, if one decides that, from the human foreskin fibroblast data, one wanted to focus on interactions between RPS19 and ZNF444, then the P values for the specific hypothesis comparing these two genes in, say, HeLa cells, would not be subjected to this same correction. We leave such interpretative matters to the reader.

We also report the correlation coefficient between RPS19 and ZNF444; although it is a somewhat imperfect measure of the lack of independence for this sort of data, it has the advantage of being

familiar to many researchers. We obtained standard errors for the correlation coefficient by bootstrapping.

In Figure 2, we obtained P values for the difference in tran-scriptional frequency between the copy of the gene on the t(13;19) (or t(6;19)) chromosome and the copies of the gene on the nor-mal copies of chromosome 19 by rejecting the null hypothesis in which the frequency of transcription was the same for all three copies. We did this by computationally generating the probability density function for the difference in transcriptional frequencies between two sets chosen to match our experimental data in size under the null hypothesis that the frequency is the same for both sets and then directly calculated the probability of finding our observed difference by chance.

npg

© 2

013

Nat

ure

Am

eric

a, In

c. A

ll rig

hts

rese

rved

.

0.79

3.98

10.2

4

10.7

6

12.7

9

17.9

7

18.6

8

23.5

4

34.9

2

39.1

139

.94

40.3

341

.31

42.3

645

.39

45.7

5

47.2

8

52.6

9

54.7

056

.65

RPL1

8A

UBA5

2

ZNF9

1

UBA2

EIF3

KSU

PT5H FB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5

PPP2

R1A

RPS9

ZNF4

44

PTBP

1

EEF2

Chromosome 19p-arm q-arm

DNM

T1

ILF3

DHPS

Supplementary Figure 1

Supplementary Fig. 1. Targeting 20 introns enables chromomal paints. a��'$3,�VWDLQ�RI�WKH�FHOO·V�QXFOHXV��b. RNA FISH targeting mRNA from the EEF2 gene (labeled with Alexa594 dye). c. RNA FISH targeting EEF2 introns (labeled with Cy3). d. Genomic locations of all genes on chromosome 19 whose introns we targeted. :H�VHOHFWHG�PDQ\�SRWHQWLDO�JHQHV�EDVHG�RQ�KLJK�DEXQGDQFH�RI� WKHLU�P51$��GHWHUPLQHG�E\�51$�VHT�DQG�57�3&5��WKHQ�QDUURZLQJ�RXU�IRFXV�WR�D�VXEVHW�RI�WKRVH�ZLWK�KLJK�LQWURQ�VSRW�IUHTXHQFLHV�DV�PHDVXUHG�E\�51$�),6+��8OWLPDWHO\��WKH�VHW�RI�JHQHV�ZH�SLFNHG�KDG�VSRW�IUHTXHQFLHV�UDQJLQJ�EHWZHHQ��DQG��RI�FKURPR-VRPHV�H[KLELWLQJ�D�VSRW��VHH�)LJ��IRU�VRPH�UHSUHVHQWDWLYH�QXPEHUV��7KH�VPDOOHVW�JHQRPLF�GLVWDQFH�EHWZHHQ�loci was 0.36 megabases (between TOMM40 and MARK4). e. 17 gene intron chromosome 19 “paint” (cyan, labeled with Cy3), intron of EEF2 (white, labeled with Atto 647N), EEF2 mRNA (magenta, labeled with Alexa �� DQG� WKH� QXFOHXV� �\HOORZ�� ODEHOHG� ZLWK� '$3,�� $OO� LPDJHV� DUH� PD[LPXP� ]�SURMHFWLRQV� RI� D� WKUHH-GLPHQVLRQDO�]�VWDFN��6FDOH�EDUV�DUH��P�ORQJ�

e

a b cDAPI mRNA intron

Chr. 19EEF2 mRNAEEF2 intron

d

Nature Methods: doi:10.1038/nmeth.2372

DAP

I(n

ucle

us)

TMR

Chr

. 19

pain

tC

y5EE

F2 In

tron

Alex

a 59

4EE

F2 m

RN

A


Supplementary Fig. 2. RNase A eliminates intron spots. We exposed HeLa cells to 10 µg/mL of RNase A for 30 minutes after fixation and before hybridization (right panels), with control cells exposed to the same procedure but without the addition of RNase. The top row contains images of the DAPI nuclear stain. The second row contains images in which we labeled the introns of all genes except EEF2, thus painting active genes in chromosome 19, and in the third row we labeled the introns of EEF2. In the fourth row, we labeled EEF2 mRNA. All images are maximum intensity projections of a z-stack of fluorescence images. The scale bar is 5µm long and applies to all images depicted. These results show that RNase A eliminates all the FISH signals we observed in the cells. This shows that the signals we are detecting are due to binding to RNA and not DNA, showing that our probes are detecting RNA from actively transcribing genes and not the DNA of gene itself.

No RNase A 30 minutes of 10 µg/mL RNase A


0 minutes

DAP

I(n

ucle

us)

TMR

Chr

. 19

pain

tC

y5EE

F2 In

tron

Alex

a 59

4EE

F2 m

RN

A30 minutes 60 minutes

Actinomycin D exposure time (2 µg/mL)120 minutes

Supplementary Fig. 3. Inhibition of transcription by Actinomycin D results in rapid degradation of intronic RNA spots. We exposed HeLa cells to 2 µg/mL of actinomycin D for varying amounts of time as indicated. The top row contains images of the DAPI nuclear stain. The second row contains images in which we labeled the introns of all genes except EEF2, thus painting active genes in chromosome 19, and in the third row we labeled the introns of EEF2. In the fourth row, we labeled EEF2 mRNA. All images are maximum inten-sity projections of a z-stack of fluorescence images. The scale bar is 5µm long and applies to all images depicted. We found that virtually all the intronic RNA disappeared or greatly diminished in intensity after 30 minutes of Actinomycin D exposure. In the case of EEF2 intron, we found absolutely no spots; in the case of the chromosome paint, we did occasionally see dim spots even at later time points, although they were consid-erably dimmer. These may be residual RNA still attached to RNA polymerases stalled by Actinomycin D. We observed mature mRNA at all timepoints, indicating that the cells were still alive and that the treatment did not affect the RNA itself. Altogether, our results show that intronic RNA degrades rapidly; thus, the presence of an intronic RNA spot indicates that the targeted gene is transcriptionally active.




Supplementary Fig. 4. Analysis of exon vs. intron spot intensity. a. We show intron vs. exon spot inten-sity, defined as maximum pixel intensity of the spot minus the nearby background pixel intensity. Blue spots correspond to S/G2 phase cells (as scored by Cyclin A2 mRNA levels); red spots correspond to G1 cells. We analyzed a total of 105 spots. The correlation between intron and exon intensity is 0.378. b. Images depicting all the exon and intron spot pairs we analyzed (exon on left, intron on right). The slight shift between the exon and intron spot results from slight registration shifts in the images between fluorescence channels. We found that only 2 out of 105 intron spots displayed a lack of corresponding exon spot.

These results further establish that the presence of an intron spot corresponds to active transcription of the gene. We almost never observe an intron spot wthout any corresponding exon spot. Moreover, when there is no visible intron spot, we never find a bright transcription site in the exon channel (data not shown). Also, these results provide further evidence that the intron spots are truly located at the site of transcription because the intron spots strongly colocalize with bright exon spots that researchers have shown to represent nascent tran-scripts emanating from the site of transcription (see Levsky and Singer Science 2002, Vargas et al. PNAS 2005, Chubb et al. Current Biology 2006, Raj et al. PLoS Biology 2006).

0 500 1000 1500 2000 2500100

200

300

400

500

600

700

800

G1S/G2D

NM

T1 in

tron

spot

inte

nsity

(arb

itrar

y un

its)

exon

intro

n

a

b

DNMT1 exon spot intensity (arbitrary units)



Supplementary Fig. 5. Comparison of iceFISH spot frequency with GRO-seq data (Core et al. Science 2008). In order to compute the GRO-seq mean intron reads, we measured the average number of reads within the first 3000 intronic bases in the transcript (which was roughly the same region that we probed by iceFISH in our study). We computed the frequency as discussed at length in the main text. We also included data from a further 15 genes on chromosome 19 not studied in the main paper to provide better sampling in the high GRO-seq regime.

Our observation is that above a certain number of GRO-seq reads, the iceFISH spot frequency was always relatively high. We also point out that there is a large set of points that we have not shown that are at the origin of the graph: they have no GRO-seq reads and generally do not show intronic FISH spots. (We have not included these data as we were not interested in intron probes that do not show spots and so did not systemati-cally analyze such probes.) We believe there are many reasons that the correlation we observe is rather noisy: 1. these two sets of data arise from different cell lines and represent completely different treatments and proce-dures; 2. comparing iceFISH signals from gene to gene is not necessarily valid (nor are those from GRO-seq); and 3. our data are solely from G1 cells, whereas GRO-seq is from a mixed population. As such, we believe that the fact that these data show that high GRO-seq signals correspond to high iceFISH spot frequency is a valid means by which to measure transcriptional activity.

0 0.02 0.04 0.06 0.08 0.10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

*52ïVHT�PHDQ�LQWURQ�UHDGV�SHU�QXFOHRWLGH��OXQJ�ILEUREODVWV

iceFI

SH s

pots

per

chr

omos

ome,

fore

skin

fibr

obla

sts

0% (p = 0.0146, n = 35)

25.7%40.0%

34.3%



Supp

lem

enta

ry F

ig. 6

. C

ompa

rison

of g

enom

ic a

nd p

hysi

cal d

ista

nces

bet

wee

n ac

tivel

y tr

ansc

ribin

g pa

irs o

f gen

es.

Each

box

plo

t dep

icts

a s

tatis

tical

ana

lysi

s of

the

dist

ance

s be

twee

n a

parti

cula

r pai

r of a

ctiv

ely

trans

crib

ing

gene

s, a

nd w

e pl

aced

the

box

plot

at a

loca

tion

on th

e x-

axis

repr

esen

ting

the

geno

mic

dis

tanc

e be

twee

n th

e pa

ir of

gen

es.

The

spot

in th

e ce

nter

of t

he b

ox c

orre

spon

ds to

the

med

ian

dist

ance

, the

box

itse

lf co

rresp

onds

to 2

5th

and

75th

per

cent

iles,

and

the

whi

sker

s re

flect

the

rang

e of

the

data

(with

the

open

circ

les

refle

ctin

g da

ta p

oint

s de

emed

out

liers

). T

he o

pen

trian

gles

pro

vide

com

paris

on in

terv

als:

two

med

ians

are

di

ffere

nt a

t a 5

% s

igni

fican

ce le

vel i

f the

inte

rval

s re

pres

ente

d by

thes

e tri

angl

es d

o no

t ove

rlap.

The

red

box

plot

s co

rresp

ond

to d

ista

nces

mea

sure

d by

DN

A FI

SH (t

he e

xper

imen

tal d

etai

ls o

f whi

ch a

re d

escr

ibed

in th

e m

etho

ds).

The

re is

no

parti

cula

r diff

eren

ce b

etw

een

the

DN

A FI

SH re

sults

and

the

dist

ance

s be

twee

n tra

n-sc

riptio

nally

act

ive

loci

at t

he le

ngth

sca

les

we

exam

ined

, alth

ough

our

resu

lts in

dica

te th

at g

enet

ical

ly p

roxi

mal

tra

nscr

iptio

nally

act

ive

loci

(<0.

5 m

egab

ases

) are

mor

e sp

read

out

phy

sica

lly th

an p

ure

DN

A FI

SH m

easu

re-

men

ts th

at d

o no

t dis

tingu

ish

betw

een

trans

crip

tiona

l sta

tus

(Mat

eos-

Lang

erak

et a

l. PN

AS 1

06:3

812-

3817

. 20

09).

How

ever

, a s

impl

e m

odel

dem

onst

rate

s th

at o

ur d

ata

are

com

patib

le w

ith th

ese

prev

ious

resu

lts (s

ee

dicu

ssio

n po

int i

n su

pple

men

t). T

he in

set s

how

s th

e ph

ysic

al d

ista

nce

betw

een

loci

sep

arat

ed b

y 30

kilo

base

s (E

IF3K

and

AC

TN4;

mag

enta

) and

intro

ns a

nd e

xons

of t

he s

ame

gene

(DN

MT1

; gre

en).

The

se d

ata

show

that

in

tron

spot

s ar

e ve

ry c

lose

to th

e si

te o

f tra

nscr

iptio

n as

mea

sure

d by

exo

nic

prob

es (w

e be

lieve

the

mea

sure

-m

ents

are

alm

ost t

o w

ithin

our

spa

tial d

iscr

imin

atio

n lim

it), a

nd th

at g

enet

ical

ly v

ery

prox

imal

gen

es d

o in

deed

co

me

very

clo

se to

eac

h ot

her,

show

ing

that

the

spre

ad o

ut c

hara

cter

istic

s th

at w

e ob

serv

e ar

e no

t jus

t an

artif

act o

f our

met

hod.

Phys

ical

dis

tanc

e be

twee

n ac

tive

gene

s (in

tron

RN

A FI

SH)

Phys

ical

dis

tanc

e be

twee

n tw

o re

gion

s of

DN

A (D

NA

FISH

)Ph

ysic

al d

ista

nce

betw

een

intro

ns a

nd e

xons

(RN

A FI

SH)

Phys

ical

dis

tanc

e be

twee

n tw

o ne

arby

gen

es (i

ntro

n R

NA

FISH

)

01

23

45

67

89

1011

1213

1415

1617

1819

2021

2223

2425

2627

2829

3031

3233

3435

3637

3839

4041

4243

4445

4647

4849

5051

5253

5455

56012345678

Gen

omic

dis

tanc

e (m

egab

ases

)

Physical distance (microns)

0 0.03

0.25

0.5

0

0.2

0.4

0.6

0.81

Gen

omic

dis

tanc

e (m

egab

ases

)

Physical distance (microns)



Supplementary Fig. 7. Colocalization of intron RNA FISH signals, exon RNA FISH signals, and DNA FISH signals. As we describe in the methods, we performed a protocol in which we combined DNA and RNA FISH, here using BAC probes targeting the DNA, RNA FISH probes targeting the exons of SLC1A5, and RNA FISH probes targeting the introns of SLC1A5 (b-d). We found that all the exon and intron signals colocalized with the DNA FISH signals (e) to within our detection limits. These results show that the intron RNA FISH signals we observed indeed remain at the site of transcription. Note that we observe three copies of the gene via DNA FISH, but only two of the three copies of the gene are transcriptionally active. This shows that our RNA FISH probes are not inadvertently binding to DNA. Another proof that our probes are not merely targeting DNA is the fact that both the exon and intron signals are brighter than what one would expect from a single RNA molecule. If the probes ZHUH�ERXQG�WR�WKH�JHQH·V�'1$��RQH�ZRXOG�RQO\�VHH�IOXRUHV-cence intensity equivalent to that of a single RNA molecule. The scale bar is 5µm long.

a eDAPI

DNA

exon

intron

b

c

d


a DAPI cb ATTO488 Cy3

d e fAlexa 594 ATTO647N ATTO700

g

RPL1

8A

UBA5

2

ZNF9

1

UBA2

EIF3

KSU

PT5H FB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5

PPP2

R1A

RPS9

ZNF4

44

PTBP

1

EEF2

0 Mb Chromosome 19 59.13 Mb

DNM

T1IL

F3DH

PS

Chr 19

Chr 19


-

Supplementary Fig. 8. Images in individual fluorescence channels. a-f. Images for each fluorescence channels from the nucleus (labeled iwth DAPI in (a)) that we stained with probes labeled with the pseudocolor-ing scheme detailed in Supplementary Table 1. Along with probes targeting the introns, we also included probes targeting Cyclin A2 mRNA to determine position in cell cycle and SUZ12 mRNA as a fiducial marker. g. Computational identification of chromosome 19 gene positions. All images are maximum z-projections of a three-dimensional z-stack. All scale bars are 5µm long.



Supplementary Fig. 9. Chromosome identification does not bias spot frequency. In our analysis, we had to exclude chromosome in which both copies of chromosome 19 were too close for us to separate. In order to check whether this exclusion introduced any bias into our measurements, we measured iceFISH spot frequency for all chromosomes (including those which were excluded from transcriptional profiles). We found that the frequency was essentially identical to what we obtained from analyzing just the non-overlapping chro-mosomes, showing that our exclusion of overlapping chromosomes did not bias our results.

PTBP1EEF2

DNMT1ILF3

DHPSRPL18A

UBA52ZNF91

UBA2EIF3K

SUPT5HFBL

EGLN2RPS19

TOMM40MARK4

SLC1A5PPP2R1A

RPS9ZNF444

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7ice

FISH

spo

t fre

quen

cy p

er c

hrom

osom

e

Separated chromosomes (n=108)All chromosomes (n=138)


PTBP1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0EEF2 2 2 2 2 1 2 1 2 2 2 2 2 2 2 0 1 2 2 2 2DNMT1 0 0 0 0 0 1 0 2 1 1 2 0 0 2 0 0 1 2 0 0ILF3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0DHPS 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0RPL18A 2 0 0 0 0 1 2 0 1 1 1 1 1 2 0 1 0 1 1 0UBA52 0 0 1 1 1 0 1 0 1 0 0 0 0 0 2 0 0 0 1 1ZNF91 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0UBA2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0EIF3K 0 0 0 0 0 0 1 0 2 0 0 0 0 0 0 0 0 0 1 1SUPT5H 0 1 1 0 0 0 1 0 1 0 1 1 0 0 1 0 1 0 1 0FBL 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0EGLN2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1RPS19 0 0 0 1 0 0 0 0 0 1 0 1 1 0 2 0 1 1 0 2TOMM40 0 0 0 0 1 0 0 0 1 0 1 1 1 1 0 0 0 1 0 1MARK4 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0SLC1A5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0PPP2R1A 1 1 1 1 0 0 1 1 1 1 0 2 2 1 2 2 1 0 2 2RPS9 0 0 1 1 0 0 1 1 1 1 1 2 0 0 2 0 1 0 0 1ZNF444 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

20 cells

Probe added, correct spot(s) identified

Probe left out, missing spot correctly not identified

Probe left out, inappropriate spot identified


Supplementary Fig. 10. False identification rate. In order to assess the degree to which our intron spot color-coding scheme resulted in false positives, we left out 10 of the 20 probes (labeled in green) and ran the resultant images through our normal spot identification pipeline. In blue, we have indicated all the spots that we found that we ended up assigning to genes whose probes had been added to the hybridization (i.e., correct LGHQWLILFDWLRQV��,Q�UHG�DUH�VSRWV�WKDW�ZH�PLVLGHQWLILHG�LQ�WKH�VHQVH�WKDW�WKH\�FRUUHVSRQG�WR�JHQHV�WKDW�ZH�KDGQ·W�added to the hybridization. We found that misidentified spots were relatively rare, with roughly 97% of spots we identified being assigned to genes that we had actually targeted in our hybridization. In order to minimize bias, we had another person in the lab randomly select 10 genes to leave out of the hybridization.


Supplementary Fig. 11. Labeling Cyclin A2 mRNA enables detection of cells in the S, G2, and M phases of the cell cycle. In order to isolate cells in G0/G1 for analysis, we performed RNA FISH targeting Cyclin A2 (magenta), which Eward et al. 2004 have shown to be present during S, G2, and M phases. We observed highly variegated expression, with some cells having having high levels of expression and some having very few mRNA. In order to demonstrate that the high expressing cells were indeed in S-M phase, we incubated the cells with Click-iT EdU 10µM for 5 minutes before fixation (green); the Click-iT EdU reagent incorporates into polymerizing DNA and produces a signal in cells undergoing DNA replication. We found that every cell display-ing Click-iT EdU signal had high levels of Cyclin A2 mRNA, showing that Cyclin A2 mRNA provides a strong marker for cell-cycle. Sometimes a cell would have high levels of Cyclin A2 but would not be undergoing DNA replication; these cells are in G2, as we observed several double intron spots in these cells, indicating that those cells had already duplicated their DNA. In thus study, we were primarily interested in cells that had not undergone replication, so we selected cells with low levels of Cyclin A2. Note that we could not use the Click-iT EdU kit in combination with iceFISH because we found that incubating cells with the Click-iT reagent resulted in an abolition of transcription. We stained the nuclei with DAPI (purple); the scale bar is 5µm long.



a

b

Supplementary Fig. 12. G-band and DNA FISH analysis confirms the translocated copies of chromo-VRPH��LQ�RXU�ODE·V�+H/D�FHOO�OLQH� a. We performed G-band analysis of a metaphase spread of our chromo-somes to identify both the numbers of chromosomes in the cells as well as identify potential translocations. The numbers (and X and Y) correspond to the identities of the intact chromosomes. The specific translocations we were interested in for this study are fusions of portions of chromosome 19 to chromosome 13 and 6, denoted t(13;19) and t(6;19), respectively. b. We wanted to further verify the translocations of chromosome 19 via DNA FISH, which we performed upon the same chromosomes for the G-band analysis in a. We probed the chromo-somes with probes against the p-arm of chromosome 19 (green) and the q-arm of chromosome 19 (red). This confirmed the results of our G-band analysis.



DIAPH3MZT1 RPL18A UBA52 ZNF91 UBA2 EIF3K SUPT5H FBL EGLN2 RPS19 TOMM40 MARK4 SLC1A5 PPP2R1A RPS9 ZNF444

*

**

**

*******

***

*** **

Chromosome 19 gene name

Chromosome 19 gene name

Chr. 13gene name

0

0.4

0.2

0.8

1

0.6

iceFI

SH s

pots

per

chr

omos

ome

Centromere

N=20

N=20

N=53

Translocation

t(13;19) t(13;19)

Chr 19Chr 13t(13;19)t(6;19)

PTBP1 EEF2 DNMT1 ILF3 DHPS 0

0.4

0.2

0.8

1

0.6

t(6;19)


Supplementary Fig. 13. Biological replicate of the data presented in Figure 3. We performed the exact same analysis as described in the legend of that figure. In this case, we analyzed 20 cells with the full set of 20 chromosome 19 probes and 53 cells in which we analyzed expression of MTZ1 and DIAPH3 on chromo-some 13.



Supplementary Fig. 14. Analysis of spot intensity. We computed spot intensity by finding the maximum pixel intensity in the center of a 3x3 pixel box around the center of the spot and subtracting the median local background pixel intensity around the spot. We then normalized this intensity for each color to number of probes on a cell-by-cell basis by calibrating to the intensities of the SuZ12 mRNA signals, which had 13 oligo-nucleotide probes in each fluorescence channel. A value of 1 for spot intensity here indicates and intensity equivalent to 1 complete set of intron probes bound. The box plots show spot intensity for each gene on Chr. 19 in foreskin fibroblasts and HeLa cells, as well as the t(6;19) and t(13;19) derivative chromosomes in HeLa cells. Center white spot represents the median, edges correspond to 25th and 75th percentile, and whiskers extend to extrema, with outliers plotted individually as circles.

We found that the spot intensities varied from gene to gene, with most of the genes showing a median between 1 and 4 full intron equivalents. The foreskin spot intensities appeared to be somewhat lower than in HeLa cells. In HeLa cells, we found that spot intensities were essentially the same when comparing spots on the intact chromosome 19s and the translocated t(6;19) and t(13;19). This result implies that the changes in transcription of the chromosome 19 genes on t(13;19) arise via changes in transcriptional burst frequency but not changes in burst size. It also highlights the fact that the changes in spot frequency that we observe on t(13;19) are not likely to be the result of a relatively small increase in spot intensity leading to an increase in the number of spots above a putative detection threshold. Were that the case, then we would see many more low intensity spots on t(13;19), which would lower the mean intensity, which is not what we observe.

0

1

2

3

4

5

6

7

8

RPL18A UBA52 ZNF91 UBA2 EIF3K SUPT5H FBL EGLN2 RPS19 TOMM40 MARK4 SLC1A5 PPP2R1A RPS9 ZNF444PTBP1 EEF2 DNMT1 ILF3 DHPS

Spot

inte

nsity

(ful

l intro

n eq

uiva

lent

)

Chr. 19 (foreskin)Chr. 19 (HeLa)t(13;19) (HeLa)t(6;19) (HeLa)



Supplementary Fig. 15. Chromosome 13 intron spot frequency. To measure the frequency of transcription of genes on the portion of chromosome 13 fused to chromosome 19 (denoted der(19)t(13;19)) in HeLa cells, we used a strategy in which we painted chromosome 19 with probes in one color (Cy3, panel b) to identify the chromosome territory. We also labeled Cyclin A2 mRNA in Cy3 to ensure analyzed cells were in G0/G1 phase RI� WKH�FHOO�F\FOH�� :H�DOVR� ODEHOHG�SUREHV� WDUJHWLQJ�(()�·V� LQWURQ� LQ�$WWR��1��F��ZKLFK� LV� ORFDWHG�RQ� WKH�SRUWLRQ�RI�FKURPRVRPH��WKDW�LV�SDUW�RI�GHU��W��DQG�LV�QRW�RQ�GHU��W��WKHUHE\�KHOSLQJ�WR�LGHQWLI\�those portions of chromosome 19 that are not fused to chromosome 13. Finally, we labeled the intron of the JHQH�RI�LQWHUHVW�RQ�FKURPRVRPH��HLWKHU�0=7��LQ�SDQHO�G��RU�',$3+��QRW�GHSLFWHG��ZLWK�$OH[D��DQG�LI�the intron spot appeared in the chromosome territory identified by the chromosome 19 paint, we assigned it to der(19)t(13;19). We assigned spots for these chromosome 13 genes that appeared away from chromosome 19 as coming from one of the two normal copies of chromosome 13 in our HeLa cells. The scale bar is 5µm long.

a b

c d

DAPI Chr. 19 intron paint and CyclinA2

MTZ1 intronEEF2 intron

Cyclin A2 mRNA

Chr 19

Chr 19

t(13;19)

t(6;19)

Chr 19

Chr 19

t(13;19)

t(6;19)

Chr 19

Chr 19

t(13;19)

t(6;19)


0 10 20 30 40 50 600

1

2

3

4

Foreskin fibroblast chromosome 19

Genomic vs. physical distance

Genomic distance between loci on Chr. 19 (megabases)

Phys

ical d

istan

ce (m

icron

s)

HeLa normal chromosome 19HeLa t(13;19)

0

4

8RPL18A UBA52 ZNF91 UBA2 EIF3K SUPT5H FBL EGLN2 RPS19 TOMM40 MARK4 SLC1A5 PPP2R1A RPS9 ZNF444EEF2

Inte

rpai

r dist

ance

(micr

ons)

DNMT1 ILF3 DHPS

RPL18AUBA52

ZNF91UBA2

EIF3KSUPT5H

FBLEG

LN2RPS19

TOM

M40

MARK4

SLC1A5PPP2R1A

RPS9PTBP1

EEF2DNM

T1ILF3

DHPS

Foreskin fibroblast chromosome 19HeLa normal chromosome 19HeLa t(13;19)


Supplementary Fig. 16. Pair distance comparison in HeLa cells. In this figure, we compared the physical distances between all pairs of actively tran-scribing genes on chromosome 19 or the t(13;19) derivative chromosome from human foreskin fibroblasts and HeLa cells. Our comparison was between all loci on all chromosomes from human foreskin fibroblasts (blue markers), normal chromosome 19s from HeLa cells (red markers), and the t(13;19) derivative chromosome also present in HeLa cells (green markers; only for those genes that are located on this chromosome). We did not observe any statistically significant differences in average distances for any gene pair when comparing any of these conditions. The inset plot shows a comparison between genomic and physical distance for all active loci, comparing these same classes of chromosomes. The plot contains a rolling average with a ±1 megabase window, with the error bars corresponding to the standard error of the mean. We observed no significant differences in the relationship between genomic distance and physical distance between active loci, suggesting that the increased expression on the t(13;19) derivative chromosome does not correlate with an overall increase in the physical distance between active loci. Nature Methods: doi:10.1038/nmeth.2372


Supplementary Fig. 17. Analysis of distance of chromosomes from nuclear periphery. a. Box plot show-ing distance from nuclear periphery for Chr. 19 in foreskin fibroblasts and HeLa cells, as well as the t(6;19) and t(13;19) derivative chromosomes. Center line represents the median, edges correspond to 25th and 75th percentile, and whiskers extend to extrema. Medians are different with a p < 0.05 if the notched regions do not overlap. We computed p-values using the Kolmogorov-Smirnov test. b. The cumulative density function corre-sponding to A. c. Distance from nuclear periphery for chromosomes in which a given gene is actively transcrib-ing (green) versus inactive (orange). Error bars reflect the standard error of the mean.

2

3

4

5

6

7

Chr. 19 (foreskin)Chr. 19 (HeLa)t(13;19) (HeLa)t(6;19) (HeLa)

Chr. 19(foreskin)

t(6;19) t(13;19)

p = 0.063

p = 1.23x10-5

Chr. 19(HeLa)

t(13;19)(HeLa)

t(6;19)(HeLa)

0.01 2 3 4

Distance from nuclear periphery (µm)

Gene OFFGene ON

Dist

ance

from

nuc

lear

per

iphe

ry (µ

m)

Fore

skin

Chr

. 19

HeLa

Chr

. 19

HeLa

t(6;

19),

t(13;

19)

Mea

n di

stan

ce fr

om n

ucle

ar p

erip

hery

(µm

)

Cum

ulat

ive d

ensit

y fu

nctio

n (p

roba

bility

)

5 6 7

0.2

0.4

0.6

0.8

1.0

0

1

2

3

4

5

0

1

2

3

4

5

0

1

2

3

4

5




a

c

b


-log(p)

RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444

RPS1

9PT

BP1

EEF2

DNM

T1IL

F3DH

PS


OFF

OFFZNF444

ON

ON

RPS1

9

OFF

OFFZNF444

ON

ON

RPS1

9

OFF

OFFZNF444

ON

ON

RPS1

9

OFF

OFFZNF444

ON

ON

Nucleus

Chr. 19Gene A

Gene B

p = 0.00092R = -0.43

p = 0.051R = 0.29

p = 0.0019R = -0.43

p = 0.021R = 0.34TOMM40

MARK4SLC1A5PPP2R1ARPS9ZNF444


RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444

PTBP

1EE

F2DN

MT1

ILF3

DHPS



RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444

PTBP

1EE

F2DN

MT1

ILF3

DHPS



RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444

PTBP

1EE

F2DN

MT1

ILF3

DHPS



Human foreskin fibroblasts

Biol

ogica

l rep

licat

e 1

Biol

ogica

l rep

licat

e 2

2510

514

1520

154

2021

013

1229

85

Nucleus

Chr. 19Gene A

Gene B

1

2

p = 0.05

4

p = 0.01

5

6

p = 0.001


Supplementary Fig. 18. Replicates for foreskin correlations. Characterization of inter and intra-chromosomal interactions between gene pairs in independent biological replicates measured in human fore-eskin fibroblasts. The analysis is exactly the same as that performed in Fig. 3 and as described in the methods

Intra-chromosomal (cis) interactions Inter-chromosomal (trans) interactions

High likelihoodof interaction

Low likelihoodof interaction


-log(p)

RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444

RPS1

9PT

BP1

EEF2

DNM

T1IL

F3DH

PS


OFF

OFFZNF444

ON

ON

RPS1

9

OFF

OFFZNF444

ON

ON

RPS1

9

OFF

OFFZNF444

ON

ON

RPS1

9

OFF

OFFZNF444

ON

ON

Nucleus

Chr. 19Gene A

Gene B

p = 0.0084R = -0.46

p = 1.0R = 0.0

p = 0.0026R = -0.41

p = 1.0R = 0.011TOMM40

MARK4SLC1A5PPP2R1ARPS9ZNF444


RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444

PTBP

1EE

F2DN

MT1

ILF3

DHPS



RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444

PTBP

1EE

F2DN

MT1

ILF3

DHPS



RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444

PTBP

1EE

F2DN

MT1

ILF3

DHPS



HeLa cells, normal copies of chromosome 19

Biol

ogica

l rep

licat

e 1

Biol

ogica

l rep

licat

e 2

2613

516

2019

1110

1911

19

1515

55

Nucleus

Chr. 19Gene A

Gene B

1

2

p = 0.05

4

p = 0.01

5

6

p = 0.001


Supplementary Fig. 19. Replicates for HeLa correlations. Characterization of inter and intra-chromosomal interactions between gene pairs in independent biological replicates measured in HeLa cells. The analysis is exactly the same as that performed in Fig. 3 and as described in the methods.

Intra-chromosomal (cis) interactions Inter-chromosomal (trans) interactions




-log(p)

RPS1

9

OFF

OFFZNF444

ON

ON

RPS1

9

OFF

OFFZNF444

ON

ON

Nucleus

t(13;19)Gene A

Gene B

p = 1.0R = 0.03

p = 0.14R = 0.29

Intra-chromosomal (cis) interactions

HeLa cells, t(13;19)

Biol

ogica

l rep

licat

e 1

Com

bine

d da

tase

tBi

olog

ical r

eplic

ate

2

23

196

74

63

1

2

p = 0.05

4

p = 0.01

5

6

p = 0.001

RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444


RPS1

9

OFF

OFFZNF444

ON

ON

p = 0.33R = 0.33

97

259

RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444


RPL1

8AUB

A52

ZNF9

1UB

A2EI

F3K

SUPT

5HFB

LEG

LN2

RPS1

9TO

MM

40M

ARK4

SLC1

A5PP

P2R1

ARP

S9ZN

F444



Supplementary Fig. 20. Replicates for t(13;19) correlations. Characterization of intra-chromosomal inter-actions between chromosome 19 gene pairs on the der(19)t(13;19) fusion chromosome in HeLa cells. The analysis is exactly the same as that performed in Fig. 3 and as described in the methods. We found no signifi-cant correlation between RPS19 and ZNF444, suggesting that the translocation disrupts the whatever the regu-latory mechanism is that is responsible for the anti-correlation we obseved between this pair of genes on the normal copies of chromosome 19 in both human foreskin fibroblasts and HeLa cells.




Foreskin fibroblast chromosome 19Foreskin, RPS19 ONForeskin, RPS19 OFF

0

4

8RPL18A UBA52 ZNF91 UBA2 EIF3K SUPT5H FBL EGLN2 RPS19 TOMM40 MARK4 SLC1A5 PPP2R1A RPS9 ZNF444EEF2

Inte

rpai

r dist

ance

(micr

ons)

DNMT1 ILF3 DHPS

RPL18AUBA52

ZNF91UBA2

EIF3KSUPT5H

FBLEG

LN2RPS19

TOM

M40

MARK4

SLC1A5PPP2R1A

RPS9PTBP1

EEF2DNM

T1ILF3

DHPS

Foreskin, ZNF444 ONForeskin, ZNF444 OFF


Supplementary Fig. 21. Comparison of pairwise distances for RPS19 and ZNF444. In this figure, we compared the physical distances between all pairs of actively transcribing genes on chromosome 19 in human foreskin fibroblasts. Our comparison was between all loci on all chromosomes (blue markers), chromosomes upon which RPS19 in actively transcribing (red markers), chromosomes upon which RPS19 is transcriptionally inactive (green markers), chromosomes upon which ZNF444 is actively transcribing (cyan) and chromosomes upon which ZNF444 is tran-scriptionally inactive (magenta markers). We did not observe any statistically signifi-cant differences in average distances for any gene pair when comparing any of these conditions (K-S test, p < 0.05), although we note that our statistical power to resolve these differences is in some cases rather low because of a low number of measure-ments. We also did not note any particular distance patterns or relationships, although some pairs did appear to display less distance variability than others.


Supplementary Discussion: Relationship between intron spot measurements and transcriptional activity In our measurements, we obtain both the probability of finding an intron spot as well as the intensity of that spot. Here, we present a simple model of intron dynamics that relates transcriptional dynamics to these two measurements. We assume that transcription occurs in bursts, which is supported by several studies in higher eukaryotes as noted in the main text. We assume that the transcription of a gene as a function of time is given by the function µ(t) = µ0f(t), where µ0 is a constant and f(t) is a stochastic process that randomly fluctuates between having value 0 and value 1 (corresponding to the gene being active or inactive, respectively). We do not assume any form for f(t) other than that the time in the active state or the inactive state is on average considerably longer than the time to degrade introns, although several groups model the dwell times in the active or inactive state as being exponentially distributed, and there is experimental support using time-lapse imaging for this view (Golding et al. Cell 2005; Chubb et al. Curr. Biol. 2006; Suter et al. Science 2011). We assume that the fraction of time the gene is in the active state is given by a. The (continuous) equation governing the intron dynamics is: dI/dt = µ(t) - δI where I is the number of intron molecules and δ is the rate of intron degradation. The steady state of this equation when the µ = 0 or µ = 1 is 0 or µ0/δ, respectively. The degradation rate, δ, is what determines how rapidly I heads to steady state. Based on our Actinomycin D experiments (Supplementary fig. 3), we believe the intron half-lives of the genes we examined to be less than 5 minutes. In this case, where δ is considerably larger than the rates of the gene switching on or off, then I(t) ~= µ(t)/δ i.e., I(t) is non-zero only when the gene is actively transcribing, and zero when the gene is inactive. The time average of I(t) is then <I(t)> = aµ0/δ while the time averaged rate of transcription is given by <txn> = aµ0 By measuring the percentage of the time we observe the gene actively transcribing, we can estimate a, the probability of the gene being active, in absolute terms. When the gene is active, the rate of transcription is µ0, but we can only measure the intron spot intensity, which is proportional to the rate of transcription. Thus, we cannot measure the rate of transcription when


the gene is active up to a constant of proportionality that is 1/δ, which in principle may vary from one gene to another. Nevertheless, we can compare the relative changes in the rate of transcription of the same gene from one chromosome to another by comparing our measurements of both a and µ0/δ. In our experiments, we found that in virtually all situations, the spot intensity (µ0/δ) did not change (Supplementary fig. 14), but we did observe changes in the probability of finding an intron spot (a), which implies a proportional change in the overall time-averaged rate of transcription. We interpret this to mean that whatever causes the changes in transcription on the hyperactivated t(13;19) chromosome in HeLa cells (as compared with the intact chromosome 19s in HeLa cells), it is most likely not something that is changing the rate of transcription when the gene is active, but rather is changing the probability that the gene is active itself. We note that this is not necessarily the same as saying that the transcriptional burst frequency has changed while the transcriptional burst size remains the same: if transcriptional bursts lasted for longer, then both the burst size and the probability of finding a spot would increase, even if the burst frequency remained constant. Comparison of the distance between active DNA loci to previous experiments We were somewhat surprised to find that the distance we observed between transcriptionally active loci was quite large even for relatively short genomic separations; for instance, we observed a mean physical displacement of 1.7µm for genes separated by only 0.36 kilobases. We suspected that these large distances were due to the relatively decondensed chromatin thought to accompany actively transcribed genes. To check whether such a hypothesis was consistent with the published literature, we examined the data from the excellent study by Mateos-Langerak et al. PNAS 2009, in which the authors measured the relationship between the physical and genetic separation of DNA loci. In particular, they examined the distances genes in transcriptionally active regions of DNA (ridges) and transcriptionally inert regions of DNA (anti-ridges), finding that the transcriptionally active regions were considerably more physically spread out than the transcriptionally inert regions. We posit, given that transcription is fundamentally pulsatile, that the mean physical separation between two loci in a transcriptionally active region fluctuates between a short distance when the genes are inactive and a long distance distance when the genes are active (consistent with the findings of Tumbar et al. JCB 1999). From this perspective, the observations by Mateos-Langerak et al. correspond to measuring the mean inactive gene separation (DNA FISH between transcriptionally inert regions) and the weighted average of the mean inactive and active separation (DNA FISH between transcriptionally active regions), weighted by (1-a) and a, respectively, where a is the probability of the gene being active. Our measurements of interpair separations correspond to the mean active gene separation. We checked for consistency between these different sorts of measurements when comparing our data to that of Fig. 2B, left panel from Mateos-Langerak et al. At a genetic distance scale of roughly 490 kilobases, Mateos-Langerak report an mean square distance of around 0.23µm2 for inactive loci and 0.84µm2 for “ridges”, the latter of which we believe corresponds to the weighted average of active and inactive loci as described above. Our measurements of a mean square distance of 3.57µm2 between active loci at this genetic distance scale would imply a weighting


factor of 0.18, which falls squarely within our observed variation in probabilities of genes transcribing. Thus, we conclude that our data are at least consistent with the previous DNA FISH observations of Mateos-Langerak et al. with this simple model for the distance between active and inactive loci. Further studies may elucidate whether such a model is indeed an accurate description of conformational dynamics. !


single-chromosome transcriptional proﬁling reveals ...rajlab/pdfs/nature... · reveals...

Documents