© 2013 Nature America, Inc. All rights reserved. A.R. ( A.R. Department of UniversityBioengineering, of Pennsylvania, Philadelphia, Pennsylvania, USA. Correspondence should be addressed to M.J.L. or from each fluorescence channel. Scale bar, 5 20 computationally identified from images ( a series of labeled oligonucleotide probes. of transcription by targeting introns with ( introns in human foreskin fibroblasts. (chr) 19 by RNA FISH targeting F intron the Labeling RNA. nascent of out typically degrade rapidly after being spliced of copies cells. single in 19 chromosome human individual along taneously of simul 20 genes profiles transcriptional FISH RNA us to that enabled generate iceFISH called on based method a question with this address to begin we Here gene affects scale chromosome-length the at zation ences transcription possibility that the structure or organizationthe raising thus of chromosomeschromosomes, into thousands influ the by organized proximal to the Atgene sequences itself. the same genes aretime, largely trolled by the of interaction regulatory with DNA of The transcription a gene’s DNA into RNA is thought to be con spanning toward a C genes in their normal chromosomal context in the same cell. between genes on a translocated chromosome and the same We findsubstantial differences in transcriptional frequency chromosome structure on simultaneously single . a multiplex imaging method for measuring and We report intron chromosomal expression Marshall & J Arjun Levesque Raj expression regulation reveals chromosomal gene transcriptional profiling S Received Received 17 Novembe b a igure orrelations orrelations between genes on a single chromosome pointed ) Pseudocoloring scheme for labeling the site ) Transcriptional activity and location of the We took advantage of the fact that introns ingle-chromosome ingle-chromosome [email protected] 1 | Unique identification of 20 loci on cis 1 4.3 4.3 megabases. chromosome-level transcriptional interaction r 2012; accepted 18 Ja 1 , 2 ; ; however, little is known about how organi ).

n µ ua m.

r

3 y y 2013; published o , 4 -

FISH (ice a Cytoplasm leaving siteoftranscription Introns degradedbefore FISH

expression. expression. Intron ), ), n Nucleus Spliced mRNA li

n e e 17 feb Exon - - - Intron r Mature ua mRNA of 360 kilobases (kb), but we were also able to distinguish loci sepa yielded an average resolution of 3 megabases (Mb) with a minimum some 19 ( chromo along genes 20 intronsof the targeting probes designed ( gene for each a spot such finding The rate overall transcription is to proportional the probability of process transcriptional the of aspects random from arise to considered constitutively active transcribe RNA in ‘bursts’ thought ( genes active transcriptionally of site the marked spot intron the that firmed dimensional coordinates of the gene gene was actively transcribing and oligonucleotides labeled fluorescently short, 16 with tion efficiency (Online Methods), nor did pseudocoloring incur incur pseudocoloring did nor Methods), (Online efficiency tion that using 16 singly colored probes did not change the spot detec ( The positions of these genes were more spread out than expected 6 ( time the of 78% blasts, we could discern two clearly separated ( channelcence chromosomes ( weforlooked colocalization of twoor fluoresspotsthreeeach in distinguishable fluorophores spectrally (out offive) or three two ofconsistingeither gene each gene’sintrons with a particular ‘pseudocolor’, each which is labeled a distinct code for we simultaneously, statuses transcriptional rated by just 30 kb ( Supplementary 20 pseudocolors 2 colors3 r ± To concomitantly measure overall chromosome structure, we structure, chromosome overall measure To concomitantly y y 2013; 2 expressing genes (out of the 20 labeled) per chromosome. per labeled) 20 the of (out genes expressing 2 Exon Supplementary Export and d Supplementary o i : 1

0

. 5 base colors base 5 Supplementary 1 0 Supplementary nature methods

3 Fig. 6 8 Supplementary / b Supplementary n z (µm) m Chr 19 0 2 4

16 Fig. 1 e and t h .

2 brief communications Fig. 12 3 Supplementary y 7 ) enabled us to measure whether the the whether measure to us enabled )

2 ( 13 µ

m) 1 5 ,

1 and, if it was, to identify the three- and Chr 19 8 Figs.

4

Fig. ( | ADVANCE Fig.

Supplementary Fig. Fig. Supplementary 6– 4

8 2–7 8

). In humanforeskinInfibro ). 1 . Control experiments con

6 a 9 ). To measure all 20 genes’ ). To assign gene identity, ). On average, we found we average, On ). 0 ). Note that even genes genes even that Note ).

Nucleus 0 Discussion ONLINE 4

8 12 16 x

20 ( PUBLICATION µ

m)

Discussion Table 1 59.13 Mb ). We found 0 Mb

Chromosome 19 4 ( Fig. 1 Fig. 1 ZNF444 RPS9 PPP2R1A SLC1A MARK TOMM4 RPS19 EGLN2 FB SUPT5 EIF3 UBA2 ZNF9 UBA52 RPL18A DHPS ILF3 DNMT1 EEF2 PTBP1 ). This L 9– K 1 4 | 5 H 1 0

b ). 1 2  a ­ 5 - - - - - ) .

© 2013 Nature America, Inc. All rights reserved. not widespread because of the averaging effects of most assays. assays. most of effects averaging the of because widespread not to lead ily normal the ( copies of that to similar was t(6;19) on genes 19 chro mosome the t(13;19) of on transcription genes Meanwhile, location. all of of irrespective hyperactivation in resulted location ( 13 of chromosome copies normal on the than on t(13;19) higher from ( active being a gene of probability results ing from hyperactivation that an transcriptional increased ( examined we chromosomes the all on same the roughly were intensities spot Intron rupted. dis have may translocation the that regulation transcriptional Fig. 13 ( 19 the chromosome on of copies those normal than rearrange active transcriptionally genetic more fivefold to up these recapitulated ( ments data iceFISH Our ( 13 chromosome of tion por a onto of translocated 19 Mb chromosome 40–43 remaining the of consists t(13;19), denoted other, the and 6; some chromo of part to fused 19 chromosome of Mb 17–20 first the of consists t(6;19), chromosomes other of parts to fused pieces two into split is that copy one and 19 chromosome of intact copies two contain For cells HeLa chromosome. example, translocated the on genes most for unchanged remain code local and ronment cell’s structure the chromosome, large-scale a of the although disrupt they possibilities: such for search transcription. regulate also may scale chromosome the at mechanisms cal nonlo that possibility the examine to us enabled method Our itself). gene the of Mb 1 within (typically DNA the on tors local and factors) both transcription as on depends chromosome-extrinsic transcription that believe gene largely Researchers some. of chromo per off or which on were genes 20 our showing profiles scriptional tran constructed we to chromosome, each corresponding territories into genes ( during cycle cell the of phases abundant M and G2 S, the is which mRNA, A2 colabeling by cycle cell the of stage G0/G1 the in were analyzed we cells the ( misidentification spot of rate significant a  not did differences transcriptional these that found also We Supplementary Supplementary Fig.

brief communications | Translocations provide a means to to means a provide Translocations transcribing actively grouping By We also found of that frequencies two the genes transcriptional ADVANCE

chromosome 13 ( 2 c ), consistent with the existence of chromosome-specific chromosome-specific of existence the with consistent ), and and Fig. Fig.

transcriptional changes. Reports of such effects such of Reports changes. transcriptional

ONLINE 2 Supplementary 2 c b ), suggesting that translocations do not necessar not do translocations that suggesting ), ). We found that most genes on t(13;19) were were t(13;19) on genes most that found We ).

Fig. 10 Fig.

PUBLICATION cis

11 DIAPH3 trans DNA regulatory regulatory DNA ). ). We ensured that 1 7 Fig. : one, denoted denoted one, : factors (such factors

Fig. Supplementary

trans 2 and | a

nature methods and and

15 Supplementary cis

envi MZT1 cyclin cyclin Fig. 2 ), suggesting that this trans this that suggesting ), fac Supplementary 1 6 - - - - ­ - -

c ) ) were roughly twofold and and for the difference in frequency (using a binomial distribution test) * analyzed two genes on chr 13 ( observing a transcription site per chr) to the activity of the intact copies of chr 19. We similarly transcriptional activity of the genes on the translocated fragments (as measured by frequency of

Fig. c a

Supplementary iceFISH spots per chromosome 1.0 1.0 0 0

Discussion 14 q-arm Chr 19 p-arm Chr 19 ), suggest ),

PTBP1 Chr 13gene

Fig. 12 ** n Chromosome 19gene MZT1 =50

Translocation

1

DIAPH3 Chr 19 8 EEF2 * are are q-arm Chr 13 HeLa karyotype n

=30 ). ). ). ------DNMT1

t(13;19)

RPL18A

factor would, however, also affect the copy of the gene on the the on gene the of copy the affect also however, would, ­factor vating activated potential in a is fluctuations anticorrelation for this explanation One test). ( anticorrelation of genes, ( independence from deviations significant show not did somes exclusive. mutually be would B and A genes of tion chromosome, and with anticorrelations indicating that transcrip same the on time same the at transcribing actively be than to chance greater likelihood a have would ‘A’‘B’ and genes two the tion from independence, with positive correlations signifying that chromosome. same Such an the interaction wouldon manifest gene itself as a another devia of status transcriptional the affected panel our in gene one of status transcriptional the whether ing examin by chromosome single a within genes of transcription ( conformation to appear UBA52 ILF3 Fig. t(6;19) We found that most pairwise interactions on single chromo single on interactions pairwise most that Wefound the governing interactions of evidence for looked next We t(6;19) MZT1 Centromere

p-arm Chr q-arm Chr DHPS *** ZNF91

3 and and 6 6 ZNF444 and

*** UBA2 chromosomes Fusion RPS19 RPS19 trans b

correlate with differences in spatial chromosome chromosome spatial in differences with correlate DIAPH3 Supplementary EIF3K of chr 19. Scale bar, 5 the two intact copies and the two translocated pieces identification of actively transcribing genes revealed its derivatives in our HeLa cells. ( chromosomes. ( display different expression patterns than intact F Chromosome 19gen -acting factor, such as a transcription factor, that that factor, transcription a as such factor, -acting igure

and inactivating inactivating and Supplementary

and and and inactivated and inactivated R SUPT5H n ** t(6;19) =30 ), as described in t(13;19) = −0.40 −0.40 = 2

ZNF444

*** | FBL Translocated portions of chromosome (chr) 19

*** EGLN2 a ±

) ) Schematic showing chr 19 and Figs.

0.08; 0.08; (separated by 14.3 Mb), showed an showed Mb), by 14.3 (separated

e RPS19

Figs. ZNF444 S

µ

RPS19 TOMM40 upplementary 18 m. ( P Chr 19 P < 0.05, ** = 3.99 × 10 × 3.99 =

Chr 19

and

c 6 ) Comparison of the MARK4 ** ,

16 in others. Such a a Such others. in in some acti cells while

19 SLC1A5 *** b

and and t(6;19) t(13;19) Chr 13 Chr 19 ) ) Co ). However, one pair pair However, one ). P

< 0.01, *** PPP2R1A *** F m 17 igure putational −5 ).

, Fisher exact exact Fisher , RPS9 59.13 Mb 1 0 Mb 5 Chromosome 19

. *** P ZNF444 P < 0.001. values RPS9 PPP2R1A SLC1A TOMM40 EGLN2 SUPT5H UBA ZNF9 DHP DNMT1 ZNF444 MARK RPS1 FB EIF3 UBA52 RPL18A ILF3 EEF2 PTBP L

trans K 2 S 1 9 1 4 5

- - - - -

© 2013 Nature America, Inc. All rights reserved. National Institutes of Health Director’s New Innovator Award (1DP2OD008514) Award (1DP2OD008514) New Innovator of Director’s Health National Institutes grant, the USpilot Foundation Research of University Pennsylvania University of a Technology) for acknowledge providing the WeHeLa cells. gratefully and the of lab P.statistics; Institute about discussions (Massachusetts Sharp J. Gore, H. Maamar and O. G. Nair for for comments; insightful Padovan-Merhar H. Youk,L. Cai, D. Muzzey, analysis; the cytogenetic in performing assistance for her many lab’s of J. Morrissette the in the reagents assay; used iceFISH for and providing M. R. Beal Cook) Technologies We(especially Biosearch thank A Note: information Supplementary is available in the v and are Methods references any in available the associated M these of prevalence their the uncover mechanisms. and underlying phenomena determine to regulatory similar us chromosome-level and allow iceFISH will that believe tools We level. transcriptional) assays tion ( periphery nuclear the from some genes ( tion did not, however, depend on the distance between any pair of interac This 19. chromosome intact fully a requires responsible ( between ( cells HeLa in 19 chromosome of copies exact same pattern of interactions when examining the two intact the found also Notably, we imprinting. genetic of possibility the of copy; its and gene each pair between and genes the between (both copies 19 chromosome the between anticorrelation of lack The itself. chromosome a the to indeed is confined genes these between interaction the that 10 × 6.90 significant ( correlation positive statistically less and mild a having different, the qualitatively was genes between interaction inter chromosomal the that found We cell. same the in 19 chromosome of copy other the on between RPS19 anticorrelation an for looked (ref. 19 chromosome other to all boxes). and OFF as marked in lower left; markings apply transcriptionally active or inactive (designated ON of chromosomes in which from independence. The grid shows the number exact test) indicates a more significant deviation cells ( the two intact copies of chromosome 19 in HeLa human foreskin fibroblasts ( activity of all pairs of genes we measured in the intra- and interchromosomal transcriptional maps show the deviation from indepe chromosomal expression relationship. The heat showing an intrachromosomal but not inter F supporting our work. our supporting Award for Fund Career at the Interface and Scientific a Burroughs-Wellcome Supplementary igure 3 ckno e ethods iceFISH provides a complement to chromosome conforma chromosome to complement a provides iceFISH r s i o n w = 50 cells). A smaller n on one chromosome and and chromosome one on Supplementary | ledgments

o Identification of a pair of genes RPS19 f −4

t 2 h , Fisher exact test), indicating indicating test), exact Fisher , 0 e that look for interactions at the DNA (rather than than (rather DNA the at interactions for look that

p a and and p

Fig. e r . ZNF444

R RPS19 20

= 0.33 0.33 = n Fig. ), ), suggesting that whatever mechanism is = 54 cells) and P value (Fisher and

21 1 disappeared, however, on t(13;19) t(13;19) on however, disappeared, 9 ) ) or on the distance of the chromo ). Thus, we we Thus, ). ZNF444 ± n dence of 0.09; 0.09; ZNF444 Supplementary ­ are P = o - Fig. n

l i Fig. n e

3 v e ). The interaction interaction The ). r 3 OF

s HeLa cells Human foreskin broblasts ON i ) also precludes precludes ) also o F n

Fig. R P

PTBP1 R P PTBP1 o =2.7 =–0.42 f =4.0 EEF2 =–0.40 EEF2

RPS19 RPS19 ZNF444 ZNF444 t

24 45 DNMT1 31 45 DNMT1 h

cis e ILF3 ILF3

17 OFF

p × 25 27 DHPS × DHPS 6 5 Intrachromosomal 10 a 10 effect effect o

). RPL18A RPL18A p ON n

e UBA52 UBA52 –5 –5 r l

. ZNF91 ZNF91 i

n UBA2 UBA2 e - - - EIF3K EIF3K SUPT5H SUPT5H FBL FBL EGLN2 EGLN2 R Published online at o The authors declare competing financial interests: details are available in the COM paper. M.J.L. performed the experiments. M.J.L. and A.R. conceived of the project, performed the analyses and wrote the AUTHOR 20. 19. 18. 17. 16. 15. 14. 13. 11. 10. 9. 8. 7. 6. 5. 4. 3. 2. 1. c 12. o n eprints and permissions information is available online at RPS19 RPS19

m l

i

TOMM40 TOMM40 n / P (2002). (2002). 1183–1186 (2004). 1057–1061 (2002). 836–840 (2006). 1018–1025 (2005). (1993). 1326–1330 Science Methods Nat. (1998). e309 (2006). e309 Dekker, J., Rippe, K., Dekker, M. & Kleckner,N. & Dekker,M. K., Rippe, Dekker,J., P.S.Swain, & E.D. Siggia, A.J., Levine,M.B., Elowitz, L. Harewood, M. Macville, Helmstetter,C.E. & M. Thornton, M.N., Ert, Van K.L., Eward, Mateos-Langerak,J. L. Cai, & E. Lubeck, Singer,R.H. & R.C. Pezo, Shenoy,S.M., J.M., Levsky, S. Tyagi, & D.Y.Vargas, Tranchina,D., C.S., Peskin, A., Raj, Singer,R.H. & Shenoy,T.,S.M. Trcek, J.R., Chubb, E.C. Cox, & S.M. Zawilski, J., Paulsson, I., Golding, D.Y.Vargas, J.B.Lawrence, C.V.,Dobner,P.R.& Johnson,Y., Xing, J. Gribnau, J.L. Roberts, & J.N. Wilcox, D.B., Pritchett, J.R., Lundblad,R.T., Fremeau, S. Tyagi, & Oudenaarden,A. van S.A., P.,Rifkin, Bogaard, den van A., Raj, Singer,R.H. & K. Fogarty,F.S.,Fay, A.M., Femino, Cremer,C. Cremer,& T. Bickmore,W.Fraser,P.& Suter, D.M. Suter, MARK4 MARK4 e r ETING e SLC1A5 SLC1A5 v p e

PPP2R1A PPP2R1A r CONTRIBUTIONS r i

s RPS9 RPS9 n i o t

ZNF444 ZNF444 PPP2R1A SLC1A5 MARK4 TOMM40 RPS19 EGLN2 FBL SUPT5H EIF3K UBA2 ZNF91 UBA52 RPL18A DHPS ILF3 DNMT1 EEF2 PTBP1 RPS9 PPP2R1A TOMM40 UBA2 DHPS DNMT1 ZNF444 RPS9 ZNF444 SLC1A5 MARK4 RPS19 EGLN2 FBL SUPT5H ZNF91 UBA52 RPL18A ILF3 EEF2 PTBP1 EIF3K FINANCIAL s 2 n /

34 o i n f

et al. et et al. et , 1265–1269 (1986). 1265–1269 , d t et al. et et al. et

e h 5 et al. et x e nature methods

, 877–879 (2008). 877–879 , PTBP1 PTBP1 .

R P R P h h p

=6.9

=1.0 EEF2 EEF2 =0.33 =0.008

t RPS19 RPS19 a t

EMBO J. EMBO Science ZNF444 ZNF444 t m p Cell 34 35 49 27 Cancer Res. Cancer DNMT1 DNMT1

p INTERESTS et al. et e Genome Res. Genome l

: ILF3 ILF3 Nat. Methods Nat. r . / .

15 16 DHPS 23 DHPS / × 9 1 Interchromosomal w 10 Nat. Rev. Genet. Rev. Nat. RPL18A RPL18A 47

w

Nature UBA52 UBA52 Proc.Natl.Acad.USASci. 33

–4 brief communications w 1 , 1054–1065 (2011). 1054–1065 , ZNF91 ZNF91 . 7 2

n UBA2 UBA2 , 6020–6027 (1998). 6020–6027 ,

, 472–474 (2011). 472–474 , a

59 EIF3K EIF3K t

u SUPT5H SUPT5H 447 2 , 141–150 (1999). 141–150 , r

FBL FBL e 9 0 | .

, 743–748 (2012). 743–748 , EGLN2 EGLN2 , 554–564 (2010). 554–564 , ADVANCE c , 413–417 (2007). 413–417 , o RPS19 RPS19

m TOMM40 TOMM40

/ 2 MARK4 MARK4 d , 292–301 (2001). 292–301 , SLC1A5 SLC1A5 o

i PPP2R1A PPP2R1A f

i RPS9 RPS9 n

d ZNF444 ZNF444 ONLINE PPP2R1A SLC1A5 MARK4 TOMM4 RPS19 EGLN2 FBL SUPT5H EIF3K UBA ZNF91 UBA52 RPL18A DHP ILF3 DNMT1 EEF2 PTBP1 ZNF444 RPS9 PPP2R1A SLC1A5 MARK4 TOMM4 RPS19 EGLN2 FBL SUPT5H EIF3K UBA ZNF91 UBA52 RPL18A DHP ILF3 DNMT1 EEF2 PTBP1 ZNF444 RPS9 Science e

r Science / 1 Curr. Biol. Curr. 2 S 2 S 1 Cell 06 0 Science 0 0 Science Science . , 3812–3817,(2009).

h 1 –log( 2 PUBLICATION 12 t 0 95 t

3 p 2 3 8 1 2 4 5 6 8 : of correlationor Low likelihood P P P P anticorrelation of correlationor High likelihood anticorrelation , 1306–1311 , P 80 , 1025–1036 , / PLoS Biol. PLoS

/ Cell Cycle Cell

=0.01 =0.001 =0.0001 ) 2 /

1 = 0.05 n 2 2 w 97 , 585–590 , 6 m 97 59 w , e , w

, t ,

. h n

. a 2

t

3 3 u 4 | 7 ,

r , e 2  . . .

© 2013 Nature America, Inc. All rights reserved. secutive planes in our stacks was 0.3 0.3 was stacks our in planes secutive somewhat to weaker illumination on owing our apparatus).s, The spacing (~5 between con 700 Atto and ms) ~100 for we exposed (which DAPI for except dyes the of most for s were 2–3 times roughly exposure Our 700. Atto and 647N Atto 594, Alexa Cy3, 488, Atto DAPI, for sets filter using channels fluorescence different six in images fluorescence of stacks three-dimensional acquired sequentially We Instruments). objective Princeton (Pixis from camera 1024B CCD Plan-Apo cooled a and 100× 1.43) of a aperture (numerical using microscope fluorescence Imaging. in described (as incubation the for Supplementary PBS 1× the to A add RNase not did but well neighboring a in cells exact on the procedure same performed we control, a As above. outlined as FISH with proceeded then and PBS 1× with washed min, 30 for °C 37 70% the with 10 PBS 1× aspirated added we then and PBS which 1× with once after washed EtOH, outlined, just as cells the drug. the of activity the in inhomogeneity spatial avoid to it ing add before medium D the into actinomycin the mix thoroughly to sure We made FISH. performed and cells the fixed we which in 30, 60 and 120 min (as described in 2 HeLa cells incubated below. described as SSC 2× con in a imaged then and at wash) second DAPI the in ng/mL 50 of (adding centration °C 37 at buffer wash with min 30 for twice washed °C, 37 at kept chamber humidified a in overnight samples our We hybridized 4. ref. in used concentrations the to equivalent roughly was cases most in which probe, each of tion concentra the optimal We (w/v). determined sulfate empirically dextran 10% and SSC 2× formamide, 10% containing a buffer in below) (described of probe type and amount appropriate the ing by add hybridized then and (SSC) citrate saline-sodium formamide 2× and 10% containing buffer wash a with prewashed We modifications. minor some 4 with of ref. procedure the followed overnight. at least at °C 4 them stored and EtOH 70% with cells the permeabilized we which after PBS, 1× in rinses two by fixation the followed and (PBS) solution saline buffered phosphate 1× in formalin formaldehyde/10% 4% using al. et To to itself. the block of we the protocol Raj followed the cells, fix due effects structural or transcriptional potential any minimize to block the after cycle cell complete one through go cells the let We G1. in were they when them fixed then and M and G2 S, G1, M, G2, S, through go them let and cells the We phase. S released of beginning the at cells arrested which procedure, medium) in thymidine mM (2 block a double-thymidine through cells phase G0/G1- for Weenriched FBS. 10% and penicillin/streptomycin medium Eagle’s with supplemented modified Technologies) Life CRL (DMEM, Dulbecco’s Glutamax with (ATCC in cells fibroblasts HeLa or foreskin 2097) human primary grew We Cell ONLINE nature methods (Chroma), SP102v1 (Chroma), 41028 (Chroma), 31000v2 were For the RNase experiments, we fixed and permeabilized permeabilized and fixed we experiments, RNase the For we D, actinomycin involving experiments the of case the In To fluorescence perform µ

g/mL of RNase A (Sigma). We incubated the fixed cells at cells fixed the Weincubated (Sigma). A RNase of g/mL 4 culture, Briefly, we fixed the cells for 10 min at room temperature temperature room at min 10 for cells the fixed we Briefly,

METHODS We imaged all our samples on a Nikon Ti-E inverted inverted Ti-E Nikon a on samples our all Weimaged

fixation

Fig.

2

). and µ g/mL of actinomycin D (Sigma) for D 0, (Sigma) of g/mL actinomycin

in in situ fluorescence hybridization (FISH), we (FISH), again hybridization Supplementary µ m. The filter sets we used used we sets filter The m.

in situ in

hybridization.

Fig.

3 ), ), after - - - - -

spirit to the procedure described in ref. 4, in which the experi the which in 4, ref. in described procedure the to spirit in similar is This background. than higher were that intensity) selected a subset of the spots with qualities (uniform width, higher spot. the of intensity and center, width the of estimates precise us giving thereby profile, intensity Gaussian of approach. this in experiments mRNAs. Our A2 or that had little no cyclin cells we chose cases, In those cycle. cell the in phase their of basis we the on analyze cases, to cells some chose In also candidates. as spots 300 top the taking and filter Gaussian of Laplacian a with image the filtering by image steps. of series following tomated spot recognition software we wrote in MATLAB with the semiau custom of up made pipeline analysis image an through Image respectively. Atto and 700, 647N Atto 594, Alexa Cy3, 488, Atto DAPI, for (Chroma) SP105 and a custom set from Omega as described in ref. of false positives, we performed a hybridization in foreskin foreskin in hybridization a performed we positives, false of Characterization DAPI stain. by our outlined as periphery nuclear the and point this between tance in chromosome of the spots the of position average the determined first we periphery, chromosomes. particular to spots to this made gene it assign because overlapped mosomes difficult chro the which in cells We discarded chromosomes. individual out separate to manually cells examined we labeled, had we loci ing and candidate identification steps had labeled as a single spot. spatially close fluorescent spots that the Laplacian of two resolutionof Gaussianand channels fluorescence the of one in filter signals present) clearly (but dim detect to failure spots. were issues identifying Common in made software the mistakes corrected we stage. at this genes to spots particular colocalized in 75% roughly accurate assigning chose for the introns we targeted. We estimate that our software is we scheme pseudocoloring to the particular corresponding spots channels. fluorescence different the between spots identified computationally the align to shifts these applied Wethen individually. cell each in channels cence fluores different between displacements the measure could we manner, this In cell). per spots cytoplasmic clear 20–50 roughly targeting colors RNA five all in probes case, this (in markers fiducial the spots. background discard to by which means another us provided assignment spot spots that may our forbecause scheme multicolor be background including of side the on erred we case, In this spots. background from RNA spots legitimate separate to threshold a chose menter 3. Using histograms of the intensities and widths, we manually we manually and widths, of intensities 3. the Using histograms Laplacian a to candidate the fit then we candidate, each For 2. three-dimensional the in spots candidate We1. identified first To the determine distance of the chromosome from the nuclear gene the of introns the annotated correctly had we Once 7. We 6. went then manual througha correction process which in for colocalized that looked we ran software alignment, 5. After found that software ran we spots, the selected had we Once 4.

analysis. SUZ12 mRNA, which were present at an abundance of abundance an at present were which mRNA, Once we acquired our images, we put them them put we images, our acquired we Once

of

error x and and

rate. y Supplementary and then found the Euclidean dis Euclidean the found then and To gain some sense of the rate rate the of sense some Togain doi:10.1038/nmeth.2372 4 , SP104v2 (Chroma)

Figure

11 validate validate ------© 2013 Nature America, Inc. All rights reserved. that cyclin A2 mRNA was an accurate marker of position in the the in position of marker accurate an was mRNA A2 cyclin that Click-iT cell. to cell from characteristic particular this in that indicating the did cells not much abnormalities, genetic vary per We same the showed of which each analysis. on cells, ten analysis this formed G-band the of results the The confirmed Molecular). results (Abbott fluorophore different a each with 19, labeled chromosome on bands cytogenetic 19q13 and 19p13 the on. Wewithin loci analysis G-band the targeting probes used that we spreads DNA performed on metaphase the same FISH analysis a performed we way, particular this in split was 19 some chromo that demonstrate conclusively To 13. chromosome of q-arm to the fused was which q-arm), entire and centromere the through p-arm the (half 19 chromosome of the portion was remaining fragment second The 6. a chromosome to of fused portion was large and p-arm 19 chromosome the of half first the ( chromosomes other to fused and fragments two into split 19 chromosome of copy third full a and 19 chromosome of copies intact two tained con cells our that indicated analysis The procedures. standard of following spreads our cells, HeLa on metaphase (karyotyping) Karyotyping Supplementary (Atto-Tec) and Atto 700 (Atto-Tec)) as indicated in the text and in Atto 488 (Atto-Tec), Cy3 (GE), Alexa 594 (Invitrogen), Atto 647N the to attached 3 groups amine with oligonucleotides the sized synthe who Technologies, Biosearch from oligonucleotides the the first intron, we targeted the intron second as well. We ordered in we which were unable to probes fit ontoall the oligonucleotide 16 targeting the oligonucleotides first intron of the gene; for cases without needing optimization. Where possible, we to tried design b ( software design FISH custom using introns against Probe introns. targeting FISH RNA by detectable spots all virtually identifying correctly state ment, our results are strategy consistent with our pseudocoloring definitive a preclude effects statistical Although cells. 30 for 0.25 and 0.57 were pseudocoloring by measured frequencies spot the whereas 0.27, and 0.56 were alone probe colored singly a with measured frequencies spot-per-chromosome the that found We fication. identi spot in negatives false in result would strategy coloring pseudo our extent what to determine to was aim Our strategy. the of combination two or dyes in three used our pseudocoloring than rather dye single a with labeled oligonucleotides with time assay. our in targeted had we that genes to corresponding assigned we spots of (97%) majority vast low, the with very was identification false ( usual procedure as identification of spot our member with proceeded another and by lab) the chosen (randomly assay comprising iceFISH genes 20 our the of 10 out left we which in fibroblasts doi:10.1038/nmeth.2372 kit imaging 594 Fluor Alexa EdU Click-iT the we used cycle, cell ′ i end. We coupled these 3 o We also probed a set of two genes ( genes two of set a We probed also s e a

r c design. h

EdU t Supplementary e c h

. analysis of c We designed 20 base oligonucleotide probes probes oligonucleotide base 20 designed We Supplementary o

Table

m RPS19 HeLa / s t e

l 1 of

l and and . We purified the probes by HPLC probes the We. purified a cells.

r ′ cell-cycle ends to various organic dyes (including i

s Fig. d TOMM40 e We performed G-band analysis analysis G-band performed We s

i

Fig. g 10 n e ). We found that our rate of of rate our that found We ).

progression. r

12 / RPS19 ). 90% of probes worked worked probes of 90% ). , respectively, in a total of total a in respectively, , ). One fragment included included ). One fragment , , TOMM40 To demonstrate h t t p ) one at a at one ) : / / 4 w . w w ------.

the DNA by immersing the cells in a solution containing 70% 70% containing solution a in cells the immersing by DNA the We denatured (43N16). Mb 52.5–52.7 or (31D10) Mb 39.0–39.5 geting human chromosome 19 at positions Mb 2.8–4.5 (268O21), tar library RPCI-11 human the from BAC clones labeled cently In the we human cells fibroblast applied foreskin pairs of fluores protocol. hybridization reference their using Genomics, Empire DNA the procedure. with Click-iT associated steps washing less additional to detection owing spot reliable our made it because or itself transcription with interference of because either likely most detection, FISH ing EdU the procedure Click-iT interfered with our nascent RNA perform that found we because A2) cyclin use to (and opted instead experiments our of most in directly kit EdU Click-iT the EdU. the incorporated Welabeling did to not use ultimately elect fluorescently for kit the with provided instructions the followed steps, wash and hybridization the after and, probe Cy3 mRNA We our A2 a using FISH on cells protocol cyclin these performed with 10 fibroblasts foreskin incubated we case, this In DNA. replicated (Invitrogen), which incorporates a targetable chemical into newly of the the of presentation our in correction hypothesis multiple a applied not in shown as replicate, biological on individual the each analysis also performed we replicates; biological independent two of combination the of consisting set data a analyzing by obtained we results the In result. significant a to more corresponding smaller a with chance, deviation random via similar a independence from obtaining of chance the to corresponding the two-sided versus the reported We active inactive. were that transcriptionally number was B or A gene which in chromosomes of number the counting by generated table 2 × 2 each on test exact Fisher of the We performed pairs examined. we genes all of frequencies transcriptional the in independence Statistical (TAMRA). RP11-687M15 BAC clones using probes FISH DNA of location the to RNA (ATTO647N) compared the location of DNAWe directly. the targeting that simply not were RNA probes the showing thus spots, mRNA single than brighter siderably con were spots intron the and sites transcription exonic bright by both the both that FISH, We found above. RNA outlined protocols the following performing then and clones BAC using FISH DNA performing first by cells HeLa in FISH DNA-RNA Combined TAMRA. and 5-ROX red pairs dye We performed imaging similar to that for our iceFISH coverslip. probes new a with applied and ng/mL, 50 at DAPI with SSC 2× of ferred it to room temperature 2× SSC for 1 min and then to 10 trans coverslip, the removed min, 2 for °C 73 at SSC 0.4× with sample the washed we day next The chamber. slide humidified a in overnight incubated and coverslip a applied sample, air-dried 10 Weadded ethanol. 100% then and 85% 70%, to increasing steps ethanol of series a to them ferred trans then and 5 min for °C at 80 buffer SSC 2× and formamide

FISH. P µ values; however, our results would remain statistically statistically remain would results our however, values; M M Click-iT EdU reagent for 5 min before fixing the cells. Supplementary

analysis.

We performed DNA FISH with BAC probes from from probes BAC with FISH DNA performed We DNA-RNA In In Figure

SLC1A5 FISH.

Figures

3 We performed a sequential sequential a performed We , we looked for deviations from from deviations for looked we , exonic (Alexa 594) and intronic

18

and µ L of BAC probes to the the to probes BAC of L

19 . Note that we have have we that Note . Figure nature methods

3 , we , show we P P value value value value µ L ------

© 2013 Nature America, Inc. All rights reserved. of independence of for independence this sort of data, it has the advantage of being ZNF444 reader. the to matters interpretative such Weleave correction. same this to subjected be not would cells, say, HeLa in, genes two these comparing hypothesis specific the on interactions between that, from the human foreskin fibroblast data, one wanted to focus decides one if instance, For data. the of asked being question lar the because number of tested hypotheses on depends the particu manner this in information the convey to Wechose examined. our multiplying just of correction crude the applied we if significant nature methods We also report the correlation coefficient between between coefficient correlation the report We also P values by 190, which is the number of pairs of genes we we genes of pairs of number the is which 190, by values ; although it is a somewhat imperfect measure of measure it the lack ; is although imperfect a somewhat RPS19 and ZNF444 , , then the P RPS19 values for and and - observed difference by chance. difference our observed finding of probability the calculated directly then and sets is for same the both frequency that the hypothesis null the under size in data experimental our match to chosen sets two between frequencies in transcriptional for difference the function density copies. We did this by computationally generating the probability three all for same the was transcription of frequency the which in hypothesis null the by 19 rejecting of chromosome copies mal nor the on gene the of copies the and chromosome t(6;19)) (or scriptional frequency between the copy of the gene on the t(13;19) by bootstrapping. coefficient correlation familiar to many researchers. We obtained standard errors for the In In Figure

2 , we obtained obtained we , P values for the difference in tran in difference the for values doi:10.1038/nmeth.2372 - -