fundamental principles of in situ hybridization

7/28/2019 Fundamental Principles of in Situ Hybridization

1/9

0022-1554/931$3.30Th e J ou dof Histochemistry and CytochemistryCopyright 0 1993by Th e Histochemical h i ct y, Inc.

WorkshopIVol. 41, No. 12, pp. 1725-1733, 1993Printed n UXA.

I Fundamental Principlesof In Situ Hybridization'J OSIAH N. WILCOX~Department of Medicine, Division of Hematology/Oncology, Emory University, AtZanta, Georgia.Received for publication September 17, 1993; accepted September

Insituhybridization provides invaluable nformation regard-ing the localization of gene expression i n heterogeneous tis-sues. The techniqueisextremely sensitive andcandetect theamount of mRNA contained in a single cell. This review pro-vides a starting point for those who wish tobegin using insituhybridization in their own laboratories. The procedureoutlined here is based on 35S-labeled riboprobes and hasbeen used with many probes andtissueswith a greater than9O0/o success rate on the first hybridization. The importanceof appropriate controls is stressed. Clusters of silver grains

IntroductionI n situ hybridization is an invaluable tool for the examination ofgene expression. I n situ hybridization is very similar to Northernblots and depends on the hybridization of a labeled nucleic acidprobe (RNA orDNA) to a complementary sequence of "A.Thetwo techniques differ in that the starting material for a North-ern blot isa tissue digest, whereas the primary material for in situhybridization is a histological tissue section. A ll of the cell rela-tionships are lost with Northern blots and mRNA levels are aver-aged from all of the cells contained in the original sample. How-ever, in situ hybridization is exquisitely sensitive and can detectthe amount of mRNA contained in a single cell. Furthermore, sincein situ hybridization is a histological technique, cell relationshipsare maintained and it is possible to precisely identify cell types ex-pressing the gene of interest. Thus, potentially important interac-tionsbetween cells that express different proteins may be uncovered.A major advantage of in situ hybridization s that it allows themaximal use of tissues that may be in short supply (i.e., clinicalbiopsies, embryos, cultured cells). A tissue digest from a surgicalbiopsy might yield sufficient RNA for one or twoNorthern blots.These Northern blots would only yield information regarding thepresence or absence of mRNA s without any other information asto the cellular source of that RNA. A lthough Northern blots canbe probed multiple times for different "As in fact this maybe limited to four or five different probes at most, as the transfer

Supported by NI H Grant HL 47838 and by a grant from the EmoryCorrespondenceto: BoxAJ , Hematology,Rm. 1115WM B, 1639PierceUniversity Department of Medicine Development Fund.Dr., Emory U,, Atlanta, GA 30322.

1 1993 (3W 3159).

af ter hybridizationdo not n e d y ndicatespecificmRNAlocalization. Regions of the tissue rich in nudei often ap-pear to cause spurious binding of probes and have high back-grounds often mistaken as positive signals. The mosta-cultaspect of insitu hybridizationisnot to get dustersofsilver grains on the slide but rather to do the appropriatecontrolled experiments to ensure that the signal is real andis not due to some artifactual binding of the probe to thetissue. (J Hisrochem Cyrochem 41:1725-1733, 1993)KEYWORDS: n situ hybridization; mRNA ; Riboprobes.

membranes break down and there can be some RNA loss withrepeated stripping of the blots. However, literally hundredsof differ-ent hybridizations can be performed on the same piece of tissuewith in situ hybridization. A single surgical biopsy of a humanatherosclerotic plaque obtained at carotid endarterectomy, for ex-ample, might yield up to 1000tissue sections, each of which canbe used for a separate hybridization. Kept at -70C with desic-cant, these tissue sections can be stored for more than 6 years with-out significant loss of the hybridization signal (unpublished ob-servations). It is therefore possible to make libraries of tissues, storedassections in the freezer, that can be probed in the future as newprobes become available or newquestions arise. The repeated hy-bridization of these tissues with multiple probes over a long periodof timealso allows the eventual serial reconstitution of a tissue withall of the information concerningmRNA content, cell identity, andregional localization available for study.There are many different ways to do in situ hybridization in-cluding probing with cDNAs, cRNAs, or synthetic oligonucleo-tides. There are also multiple ways to label these probes using ra-dioactive (3H, 32P, 33P, 3%) or non-radioactive (biotin, alkalinephosphatase, digoxigenin, fluorescence) nucleotides. Over the past10years wehave compared most of these methods in my labora-tory and have come to the conclusion that 35S iboprobes repre-sent the most sensitive method for the detection of mRNA in tis-sue sections. A comparison of different labeling strategies for insitu hybridization ranked according to their relative sensitivity isshown below.

Riboprobes>cDNAs>Synthetic oligomers35S>32P>3H>>BiotinldigoxigeninFrozen >Paraffin1725


2/9

1726 WILCOX

This review presents an overviewof in situ hybridization, firstintroducing the various uses of the procedure in the analysis of lo-cal gene expression in tissues and then focusingon some of themore technical aspects of the procedure dealing with the choiceofprobe and the question of appropriate controls. I t concludes withan overviewof an in situ hybridization protocol used in my labora-tory, which is based on3%labeled riboprobes that we have usedwith great success and reliability on a variety of different tissuesranging from surgical biopsy specimens to animal tissues. To date,wehave used this protocol with more than 1000different tissuesand more than 100different probes, without modification. Thesuccess rate of this method with a new probe and tissue is greaterthan 90% on the first series of hybridizations, thus providing rapidresults without the necessity of a series of experiments establishingoptimal conditions individualized for a specific tissue or probe.It ishoped that this protocol may provideastarting point for thosewho wish to begin in situ hybridization in their own laboratories.The method we employ for in situ hybridization is basically amodification of the original protocol for radiolabeled riboprobespresented by Cox and colleagues(2), and has been used by ourlab for many different projects (7,9,12,14,15,17-19). This methodcan be adapted for synthetic oligomers (40-50-mers), cultured cells,or paraffin-embedded tissues.

1.2.

3.4.

5.

6.

Tissue PreparationRemove tissue and rinse in PBS or saline.Immerse n 4% paraformaldehyde-0.1M sodium phosphate buffer, pH7.4, at 4C for 1-3 hr. Try to avoid overnight fixation if possible, as thiscauses problems with keeping the section on the slide during the hy-bridization procedure.Immerse in sterile 15% sucrose-PBS solution 3 hr-overnight at 4C.Embed tissue in OCT (Miles; Elkhart, IN), M1 (Lipshaw; Pittsburgh,PA), or any other convenient embedding matrix for frozen sectioning.Tissue should be oriented n the block appropriately or sectioning. Notethe tissue number on the block directly and indicate which face of theblock should be sectioned.Freeze block with tissue in liquid nitrogen. Place the bottom third ofthe block in the liquid nitrogen, freeze until all but the center of theblock is frozen, and continue freezing on dry ice.Store at -70C in a sealed containeror wrapped in foil. Ship on dryice if necessary.Paraformaldehyde is prepared by dissolving the appropriate amountof powdered paraformaldehyde (Polysciences; Warrington, PA) in 0.1M phos-phate buffer and heating to 80C. The subsequent solution s filtered with

Whatman #1 paper after cooling and is stored at 4C until use. The parafor-maldehyde should be freshly prepared at least weekly to prevent the for-mation of degraded aldehyde byproducts which may cause an increase inthe autoradiography background.The optimal period for fixation is 1-3 hr for tissue of 1-10 mm3. Thisprocess does not necessarily fix the entire tissue block. Large tissue blockswill have a core of relatively unfixed tissue, and small blocks will be fixedto a much greater extent. However, the purpose of the fixation step is notto completely fix the tissue but to preserve and harden the tissue suffi-ciently to improve the tissue morphology from that of unfixed snap-frozenmaterial. Complete fixation is not critical at this point in the protocol,asa second fixation step on the tissue sections is performed just before hy-bridization and will fix all sections equally.A lthough tissues can be fixed for periods longer than 3 hr, extendedfixation times create additional problems. Overfiition often causes a greater

loss of tissue from the slides during hybridization. Excessive cross-linkingof the tissue by paraformaldehydemay reduce potential binding to chargedmolecules in the slide coating which are responsible for tissue adhesion.Overfixation also reduces the in situ hybridization signal, presumably bylimiting probe access to cellular As. This can sometimes be recoveredby increasing the concentrationor the time in proteinaseK (1).Fresh frozen tissue without fixation can be used for in situ hybridiza-tion. T issue should be rinsed in saline or PBS and frozen n liquid nitrogenin OCT blocks as outlined above. Although not optimal, tissue can alsobe snap-frozen without an embedding matrix and used for hybridization.The main determinant of the fixation method is based on the tissue mor-phology after hybridization. Some tissues, such as lung, are not suitablefor frozen sections and must be preparedas paraffin blocks for the bestcell morphology. Paraffin-embedded tissues can be used for in situ hybrid-ization with an approximate25% mRNA signal loss. This loss of signalintensity is acceptable, especially for tissues in which frozen sections arenot practical. A dditional details on using paraffin-embedded sections arecovered below.

Sectioning of Frozen TissueThe tissues are sectioned on a cryostat to 5-10-pm thickness, thaw-mountedon coated slides(0.2 N NaOH-PLL, Vectabond, or Fisher SuperfrostlPlus),and immediately re-frozen by placing the slides in plastic slide boxes keptin cryostator on dry ice. A fter completion of the sectioning the slides arestored at -70C with Humi-Cap desiccant capsules (United Desiccants;Pennsauken, NJ ) in the slide boxes. The mRNA on the slides s stable whenstored in this fashion, and we have used tissue sections for up to 8 yearsafter sectioning with excellent hybridization results. This can be very valu-able for working with human tissues orsmall tissue biopsy specimens andenables the researcher to conduct multiple studies on related adjacent sec-tions over the yearsasnew probes become available. n addition, it is possi-ble to store hundreds of sections typical of a particular animal model andto continue to examine the expression of additional mRNAs without hav-ing to prepare new tissues. Given that the major effort of performing insitu hybridization lies in the preparation of the tissue sections, it is rela-tively easy for a researcher to hybridize tissue sections from a wide rangeof animal models, human biopsy specimens, and normal tissues when prob-ing with new cDNAs. This type of random screening does not take verymuch additional time but may yield a wealth of information and suggestnew directions for research.

SLide Preparation/T issue AdberenceOne of the biggest problems with in situ hybridization is keeping the tis-su e sectionson the slide during the hybridization and washing procedures.A variety of slide coatings have been tested (Tables1and2). and althoughsome seem to be better than others, none is absolutely perfect. We havefound that the best coatings for tissue retention include poly-L-lysine andVectabond (Vector Laboratories; Burlingame, CA). Vectabond slide coat-ing is a proprietary modification of the amino-acyl silane procedure andhas given excellent in situ hybridization results with good tissue morphol-ogy after the hybridization procedure. We have recently begun to useSu-perfrostlPlus microscope slides (Fisher Scientif ic; Pittsburgh, PA) for allof our frozen tissue sectioning. All three methods of sl ide coating appearto have similar properties regarding tissue morphology/retention on theslide after the hybridization procedure (l ible2). The advantage of Super-frost/Plus slides is that they require no preparation time in the laboratoryand are competitive n termsof cost when technician time and reagent ex-penses are considered.Sections placed on poly-L -lysine, Vectabond, or Superfrost/Plus slidesdo not require coverslips during hybridization. With gelatin-coated slides


3/9

PRINCIPLES OF IN SITU HYBRIDIZATION 1727

Table 1. Tissue adherence t o glass microscope sliderprepared with different coatings (see text)"PLLAcid/PLLNaOH /PLL0.5% AAS1.0% AAS2.0% AASMaple'sGelatin-chromalumVectabond

1.81.82.71.81.61.32.003.0

a At least si x slides from two tissues in each experiment were evaluated for tis-sue retention and morphology after a mock in situ hybridization. The results weregradedas complete tissue loss (O), ew fragments of tissue left ( I ) , fair morphologywith moderatefoldingor tissuelass(2), good tissue morphology withoutti ssue loss(3).

the hybridization buffers tend to roll off to one side of the tissue becausethe gelatin layer wets unevenly. Poly-L-lysine and silane coatings, on theother hand, do not wet, and the high surface tension in the hybridizationbubble keeps the buffer in place. As long as the hybridizations are con-ducted in a humid chamber containing a buffer (4 x SSC, 50% forma-mide) with a similar vapor pressure as the hybridization buffer, the sec-tions will not dry out during the overnight incubation. Water alone shouldnot be used to maintain the humid atmosphere,as the hybridizationsol u-tions are hygroscopic and will take up water during the incubation and di-lute the probe.To compare different slide coatings, wehave prepared tissue sectionsfrom different tissues that in particular tended not to adhere well duringthe in situ hybridization procedure. These tissue sections were then carriedthrough a mock hybridization procedure (no labeled probe added) andthe ultimate tissue morphology evaluated after the final washing and dry-ing steps. The fol lowing slide coatings were evaluated.

Maple's. Slides were acid-washed by boiling at 130C for 10 min i nH2S04:HNOs (9:1), then allowed to cool to room temperature and washedthoroughly in'dH2O. The washed slides were then soaked in 2% aqueoussolution of triethoxy-3-aminopropyl ilane, pH 3, overnight at 65'C, washedin dH20, immersed in 10% glutaraldehyde, pH 6.5, for 60 min. andwashed in dH20. Before use the slides were activated by treatment with0.15 M sodium metaperiodate for 30 min. followed by washing in dH20and drying. Some slides were coated as above except for substitution ofdH 20 and ethanol washes similar to that used in the PLL method belowfor the acid-wash step, with similar results.Amino-acyl Silane (AAS). The slides were washed in dH20, dried inan oven, and dipped for 10-15 sec into asolution of 0.5, 1, or2% triethoxy-3-aminopropyl silane (Sigma; St L ouis, MO) in acetone. The slides were

then dried and rinsed in at least three changes of dH20, followed by dry-ing at 50C and storage.

Table 2. Companion of Vectabond and Supeq+ost/Plusslides for tissue adherence after in situ hybn2izationaSuperfrost Plus 2.9Vectabond-coated Superfrost Plus 2.2Sigma Silane-Prep 0NaOH/PLL 2.9Neoprene 1.9Vectabond 1.9

a Mock hybridizations were conducted as indicated in Table 1 and scored from0-3 for the amount of tissue left after hybridization.

Poly-L -lysine (PU). Slides were washed for 2 min each in dH20 and100% ethanol twice, followed by drying in an oven at 42'C. The slides weredipped in 100 pglml poly-clpine (M W 47,000; Sigma) in sterile dH2Ofor 20 sec. The slides were allowed to dry for 30 min at roomtemperature,dipped again in the PLL solution, and dried overnight at 42'C. The slidescan be stored at room temperature for up to 6 months.NaOHIPLL. This method included initial immersion of the slides in0.2 N NaOH for 30 min. followed by two dH20 washes, two ethanolwashes, and PLL coatingas above. Some slides were etched in 1N NaOHor by boil ing in H2S04:HNO3 as outlined in the Maple's procedure(acidlPLL) before PLL coating, but this did not appear to improve tissueretention on the slides.

Gelatin-Chrodum Subbing. Thi s wasa standard coating techniqueused in many histology laboratories, consisting of 2.5 g/liter gelatin and0.25 g/li ter chromalum dissolvedindH20 and used to coat ethanol-washedslides.Vectabond. Thiswasused as directed by the manufacturer. Slides werewashed for 5 min in acetone, immersed in a Vectabond/acetone solution

(1 bottle V ectabond with 350 ml acetone) for 5 min, rinsed for 1 min indH 20 with some agitation, drained, and dried overnight at 42'C.Neoprene. Neoprene (1.5 g) was dissolved in 300 ml isopentyl acetatein a fume hood over a 6-hr period (do not apply heat, as the isopentyl ace-tate is highly flammable). T he slides were then treated for 5 min in isopen-tyl acetate, 5 min in0.5%neoprene-isopentyl acetate, followed by a washfor 30 sec in dH20, and allowed to dry overnight at 37'C before use.

35S-RiboprobeSynthesis (8)1. Pipet 12.5p1[3'S]-UTP 1200 Ci/mmol) into Eppendorf tube. Finalconcentration should be 12 pM. Dry in speed vac. Do not use nucleo-tide that has been previously thawed after receipt from manufacturer.Order 250-pCivialsandusethemal l up at once orthrowtheexcessaway.To tube with the dried down probe add

2 p l 5 x Transcription buffer1pl IYlT, 100 mM1pl FWAsinpl DNA (linearized plasmid 1 @pl)2 p1GT P+CTP+ATP mix (stock solution containing 2.5 mM each)2 pl sterile dH2OMix thoroughly and centrifuge.Add 1 plRNA Polymerase (SP6, T7, or T3 as required)Mix gently by pipetting, do not vortex, incubate 1-2 hr, 37%. (Allof the above buffers are ready-madestocks from Promega Biotech, Madi-son, WI).VortexTake out 1plof the transcription mixture to determine total incorpo-ration (Step 3b)Add 1plRQ1 DNAse to the transcription reaction above Incubate 15 min,37'C3b. To monitor incorporation:

2.

3a. To stop the reaction:

Eke 1pl from reaction and place in microfuge tubeAdd 99 pl 1 x TEPipet 1pl of 1:lOO dilution onto a small piece of DE81 paperWash three times for 5 min in 0.5M NaP04, pH 7.4Wash 10sec in dH20Rinse briefly (


4/9

1728 -ox

5 .

OR

6.

7.

VortexAdd 15 p l phenol, vortexAdd 15 p1CHC13, vortex thoroughly, centrifuge 3 minRemove top (aqueous) fraction carefully with pipette, place in cleanmicrofuge tubeAdd 15 p l 1 x TE to the first tube, vortex, centrifuge 3 minRemove top fraction, adding it to the first aqueous fractionAdd to the aqueous extracts:55 PI 1 x TE10 p1 3 M NaAc, pH 7.5250 p cold 100% ETOHKeep on dry ice for at least 30 min or store at -70C up to 1 weekCentrifuge 5 min, pour off supernatant (HOT waste)Add 500 pl cold 70% ETOH, vortexCentrifuge 5 min, pour off supernatant (HOT waste)Dry in speed-vacDissolve pellet in 1 x TE to a final concentration of 300,000 cpm/pl(total determined at Step 3b)Use immediately for in situ hybridizationStore 100-pl aliquots at -7OC up to 1week as ethanol precipitatewith 10 pl 3M NaAc, pH 7.5, 250 pl cold 100% ETOHJust before starting the hybridization step:Centrifuge5 min. pour off supernatant (HOT waste)Add 500 p1 cold 70% ETOH, vortexCentrifuge 5 min. pour off supernatant (HOT waste)Dry in speed-vacDissolve pellet in 100p l 1 x TE (to a final concentration of 300,000Make up hybridization mix as described in the in situ methodscpm/d)

Probes are labeled by transcription (8), using 35S-labeledUTP (specificactivity >o00 Ci/mmol) (Amersham; Arlington Heights,E). n aliquotof the 3S-labeled nucleotide is pipetted into a microcentrifuge tube anddried under vacuum with centrifugation n a speed-vac (Savant; Farming-dale,NY) . he transcription reaction is begun by adding in sequential or-der: 2 PI 5 x transcription buffer (containing 200 mM Tris-HC1. pH 7.5,30 mM MgC12, 10 mM spermidine,50mM NaCl (Promega;Madison,WI).1p1100 mM DTT, 1plRNAsin (20-40 U) (Promega), 1pl linearized plas-mid DNA as a template (1 pg/pl), 2 pIof a GCA solution containing2 . 5mM each of GTP, CTP, and ATP in H20, and 2 p1sterile dH2O. The com-ponents are thoroughly mixed to ensure that the labeled nucleotide is insolution and then briefly centrifuged to bring the mixture to the bottomof the tube. The reaction is then started by the addition of 1pl of the ap-propriate RNA polymerase (SP6, T7,orT3). depending on the orientationof the cDNA insertorthe vector used. The polymerase is added and mixedgently by pipetting without vortexingor centrifugation and the reactionincubated for 60min at 37C. To increase the amount of probe produced,additional enzyme can be added after 60 min and the reaction allowedto continue for another 60min. The size of the probes produced shouldbe checked on a 5.2% acrylamide, 7 M urea gel and autoradiography. Al-though there will be many small bands on the gel, the major band of tran-scribed material should be equal to the size of the cDNA insert.The protocol presented here does not call for base hydrolysis of theriboprobe before use in hybridization studies. We use full- length (whichis really a mixture of sizes) transcripts from probes up to 1.3KB in length.If the probe is longer than that, it would be preferable to find internalrestriction sites for linearizationsothe resulting template will be less than1.3K B. Alternatively, short non-overlapping >and3fragments ofthe probecould be subcloned into riboprobe vectors. These can be linearized andused individuallyoras a cocktail for transcription and hybridization. Basehydrolysis produces fragments withawide range of sizes; the smaller frag

menu produced may lack specificity and increase the spurious binding ofthe probe in the tissue, thus increasing the background.The concentration of 3S-labeled UTP in the transcription reactionshould exceed 10 pM without adding any unlabeled material. It is usuallymore diff icult to get full-length transcripts using nucleotide concentrationsless than 10 pm, as the labeled nucleotide becomes rate-limiting and thereaction stutters. It is important that the 35S-labeledUTP is used immedi-ately after thawing and that any remaining label shouldbediscarded ratherthan re-frozen or use at a later time. We have found that the single greatestcontribution to high backgrounds using 3S-labeled probes comes fromrepeated freezing and thawing of the isotope before use. A ccording to Amer-sham, thereisabout a 5% degradation of the nucleotide with freezing andthawing that might contribute to background problems. It istherefore con-venient to use only 250-pCi vials of nucleotide, dividing each vial into twotranscription reactions with 125 pCi each. If the specific activity of the iso-tope is 1200 Ci/mmol and the final volume of the transcription reaction10 PI , the final concentration will exceed 10 pM.After the transcription reaction, 1plof the reaction mixture s removed,diluted 1:1000, and an aliquot pipetted onto a small piece of WhatmanDE81 paper to monitor incorporation of the 3S label. The unincorporatednucleotide is washed off the paper with three 5-min washes in 0.5M so-dium phosphate buffer, pH 7.4. The fi lter paper is rinsed briefly in H20and ethanol, dried, and counted in the scintillation counter.After removing an aliquot to monitor incorporation of nucleotide, 1pl of RQ1 DNAse (1 U) (Promega) is added and the reaction allowed toproceed for an additional 15 min at 37C to degrade the template. Theprobe is then cleaned up by a single phenol-chloroform extraction afterthe addition of 20 p11 x TE (to increase the volume) and 1 pl of tRNA( 50mg/ml in dH2O) to act as a carrier, RNA se sink, and to provide a visi-ble pellet in future ethanol precipitations. After the phenol chloroformextraction, the aqueous extract is brought up to 100pl with the additionof 1 x TE and ethanol-sodium acetate added to precipitate the RNA. Af-ter centrifugation, washing in 70% ethanol, and drying, the pellet is re-suspended in 1 x TE to a final concentration of 300,000 cpm/pI, usingthe total incorporation determined aboveasa guide. A liquots of the probe(100111)can then be re-precipitated with ethanol-sodium acetate and theprobe stored this way for up to 1 week at -70C. It is convenient to makemultiple aliquots of the probe for a series of hybridizationsoverthe follow-ing week rather than to continuously thaw and refreeze a single aliquotorto prepare probe daily. We have examined probe stabilityovertime whenstored under ethanol-sodium acetate and have found that there is detect-able degradation of the probe size after a week, which could affect the back-ground or hybridization result. Given the relative ease of probe prepara-tion relative to the amount of effort required in the hybridizationprocedureand the time one invests in long autoradiographyexposures, t is preferableto use a freshly transcribed probe for these studies.

Procedure for In Situ Hybridizationof Frozen Sections (17-19)1.2.3.4.

5 .6.

Remove slides from freezer, thaw for 5 min at 55CFix 10 min in 4% paraformaldehyde-0.1 M NaP04, pH 7.4, 4CWash 5 min in 0.5 x SSC at room temperatureImmerse slides in proteinaseK solution, 1-5 pg/ml in RNAse bufferfor 10 min at room temperature. The amount of proteinaseK mustbe optimized with each new preparation. Once optimized, aliquotscan be frozen and used for some timeWash for 10 min in 0.5 x SSC at room temperaturePre-hybridization: Dry around sections with K imwipe, lay slides flatin an air-tight box with a piece of filter paper which has been satu-rated with Box Buffer (4 x SSC, 50% formamide) on the bottom.


5/9

PRINCIPLES OF IN SI TU HYBRIDIZATION 1729

Cover each section wi th 100pl of rHB2 without probe (can use 50 plif the tissue is small ). Incubate at 42C for 1-3 hHybridizationM k For 35S-labeled iboprobc. Assuming that you haveused 100p l of pre-hybridization buffer, combine the following: 2 p1probe per slide (stock solution 300,000 cpm/pl in 1 x "E);14 RNAper slide(50 mglml stock in sdH2O). Heat 3min,95'C. im"Ltei'yadd 17pl ice-cold rHB2 per slide, vortex, place on ice (adjust volumesif you have used less than 100p1for pre-hybridization)Hybridization: Add 20 p1of above hybridizationmix toeach 100p lofpre-hybridization solution directly nto the bubble covering the sec-tion. Incubate overnight at 55'C (adjust volume f you haveused lessthan 100pl for pre-hybridization)9. Wash twice for 10 min in 2 x SSC with 10mMPME-1 mM EDTAat room temperature (HOT waste)10. Immerse in RNAse A solution (20 p g / d in RNAse bu&r) 30 minat room temperature (HOT waste)11. Wash twice for 10 min in 2 x SSC with 10 mM PME-1 mMEDTAat room temperature (HOTwaste)

12. Wash2hrin4litersofO.l x SSCw t hl OmMPME- l mMEDTA. 55' C13. Wash twice for 10 min in 0.5 x SSC without PME or EDTA at roomtemperature14. Dehydrate2 min each in 50%, 70%. and90%ethanol, eachcontain-ing 0.3M N&Ac15. Dry in vacuum desiccator (3-4 hr), store with desiccantuntil autora-diography16. Dip in Kodak NTB2 nudear emulsion diluted 1:lwthwater at 42'C,dry for 2 hr in the dark, expose in the dark at 4'C with desiccant for2-12 weeksa. 3 min Kodak Dl9 diluted 1:l with waterb. 20 sec inwater stopc. 3 min K odak Fixer, full strengthd. Three times for 5 min inwatere. Counterstain with hematoxylin and eosin

7.

8.

17. Develop at 15'C

Once the probe and slideshave been prepared it is relatively easy tohybridizeas many as 100-200 slides per day. The frozen sections are re-moved from the freezer and collected on a metal slide tray. The slides arethen thawed and dried by incubating them for 5min exactly at 50C. Thedrying at this step has been found to improve tissue retention on the slidethroughout the hybridization procedure. After drying, the slides are putinto plastic slide carriers (each holding 25 slides) and immersed in 4%paraformaldehyde for 10 min at 4'C to ensure that the tissue is equallyfixed. This fixation also tends to improve tissue retention onthe slide. Af-ter fixation the slides are washed in 0.5 x SSC (1 x SSC = 150mM NaCI,15mM sodium citrate, pH 7) and permeabilized in 1 pg/ml proteinaseK (Sigma) n500mM NaCI, 10 mM Tris, pH 8,for 10 min at room temper-ature. This is important, as permeabilization with proteinase K, pronase,orHCI has been shown to improve the signal intensity after in situ hybrid-ization, presumably because they either loosen up the membranes and al-l owgreater penetration of the probes or because they partly digest proteinsassociated with the RNA, allowing longer regions access to the probes (lJ6).The amount of proteinase K must be optimized for each lot used. Itis suggested that this should be tested by doing a mock hybridization witha range of proteinase K concentrations before starting actual experiments.A good starting point is that concentration of proteinase K which providesthe best final tissue morphology and tissue retention on the slides at theend of the mock hybridization procedure. Once a hybridization signal isobtained, it would be appropriate to go back and then optimize the signalby again testing various concentrationsof proteinase K. One of the biggestproblems with the technique, as mentioned earlier, is keeping the sectionson the slide throughout thehybridization procedure. If this becomes a se-rious problem, the first thing to do is eliminate the proteinase K step. It

is important to remember that although proteinase K improves the signalintensity, it is not necessary to get a signal in the first place.After the proteinase K treatment the slides are washed for 10 min in0.5 x SSC, dried carefully around the section with a lint-free tissue, andplaced in air-tight hybridization boxes (Nalgene utility boxes, Baxter HealthCare Prodbcts UL.1995-4) containing filter paper saturated with Box Buffer(4 x SSC,50% formamide). The pre-hybridization s begun by pipettinga small amount of hybridization buffer (100 pl ) onto the sections (forriboprobes rHB2: 10 mM DTT, 0.3M NaCI. 20 mM Tris, pH 8 , 5mMEDTA,1 x Denhardt's, 10% dextran sulfate,50%formamide; or oligomersHB8:10 mM DM, 1 x Denhardt's, 5 x SSC, 100 pg/ml ssDNA, 100 pg/mltRNA, 10% dextran sulfate, 20% formamide). Theti ssucSare pre-hybridizedfor 1-3 hr at 42C and then the hybridization is begun by adding 20 pof a probe mixture which is pipetted carefully into the bubble of pre-hybridization solution covering the tissue section. The hybridization mix-ture contains2 p l 3sS-labeled iboprobe (300,000 cpm/pI in 1 x E) ,p1 tRNA (50 mg/ml stock in dH2O), and 17pl rHB2 for each 100p1ofpre-hybridizationbufferonthe slides. This is prepared by adding the probeand tRNA to a microfuge tube and heating this at 95'C for 3 min in aheat block, followed by the immediate addition of ice-cold rHB2. Theso-lution is then vortexed and kept on ice until used for the hybridization.Once the probe mix is added to the slides, the boxes are covered again andplaced in a 55'C incubator overnight.Coverslippingof the slides s not necessary, nor is it recommended dur-ing the hybridization ncubation. T he surface tension of the hybridizationbuffer is sufficient to keep the probe mixture in contact with the tissueovernight, and as long as the hybridization is carried out in sealed boxeswith fi lter paper saturated with 4 x SSC, 50% formamide, the tissue sec-tions will not dry out. Coverslippinghas its own problems with air bubblesand potential damage to the tissueorradioactive contamination of the labon removal of the coverslips after the hybridization is complete.Extreme care should be taken to ensure that all buffers, reagents, slideholders are ribonuclease (RNAse) free up to and including the hybridiza-tion at Step 8 . RNAse contamination is not a great problem in the post-hybridization steps with riboprobes, since the tissues will be treated withRNAseas part of the stringency wash. For this reason, care should be takenthat all of the slide holders and beakers used after Step8are kept for thatpurpose alone and never used during the pre-hybridization procedure, asthese materials will probably remain contaminated with RNA se. It i s suggested that all equipment and containers used for the pre-hybridizationsteps be periodically treated with diethyl pyrocarbonate (DEPC; Sigma).This is done by soaking the equipment in fresh 0.1% solution of DEPCovernight in the fume hood. The DEPC should be diluted in sterile dH20and used immediately, as it degrades quickly.The hybridization reaction is essentially complete after 4 hr but it isconvenient to let the incubation go overnight before the post-hybridizationwashes are performed. The slides are removed from the hybridization boxesand rinsed twice for 10 min each in 2 x SSC containing 10 mM p-mer.captoethanol and 1 mM EDTA (PME/EDTA) at room temperature to washoff the majority of radioactive probe associated with the slides. This is fol-lowed by RNAse A treatment (Sigma) (20 pg/ml in 500mM NaCI, 10 mMTris, pH 8) for 30min at room temperature. RNA se A will attack single-stranded RNA but will spare RNA-RNA duplexes, thus ensuring that theprobe is properly complexed to the mRNA by complementary base pair-ing. RNAse treatment will significantly reduce the specific signal but willhelp to provide the extremely owbackgrounds that enable the detectionof very low copy number mRNAs in tissue using this technique.Greatcaremust be taken with the handling of the RNase at this stageto prevent any cross-contaminationof pipettes or equipment, which willcause problems with later hybridizations. For this reason, all equipmentused in the post-hybridization washing steps is kept completely separatefrom material used for the pre-hybridization steps. The RNAse treatmentis followed by t womore2 x SSC PMEIEDTA washes atroomtemperature


6/9

1730 m o x

and then a high-stringency wash in 4 li ters of 0.1 x SSC BMElEDTA at55'C for 2 hr. This can be done in a large 4-l iter beaker on a stir plateheated with an immersion heater. This is followed by two washes with 0.5x SSC without BMElEDTA at room temperature. The PMEIEDTA is addedto keep the 3'S-labeled probe in a reduced state throughout the washingsteps. which may reduce backgrounds. I t is important. however, to removethe PMElEDTA before coating the sections with photographic emulsion,as reducing agents would cause chemical reduction of the silver salts andhigh background.Finally, the tissue sections are dehydrated in graded alcohols contain-ing0.3M NH4Ac. dried in a vacuum desiccator, and can be stored desic-cated at room temperature until coating with photographic emulsion. Theslides are exposed for various lengths of time in the dark in sealed slideboxes with desiccant at 4'C. The exposure time may vary depending onthe amount of mRNA in the tissue. This can range from 3 days for an abun-dant mRNA (pro-opiomelanocortin in the rat pituitary) (16) to 8 weeksfor platelet-derived growth factor expression in human vascular tissue (19)or even 12 weeks for rare mRNAs such as tumor necrosis factor or tissueplasminogen activator ( 5 ) . Because it is not always possible to know theoptimal exposure time for each tissue or probe, it is a good idea to performall hybridizations in triplicate and develop one slide every 4 weeks untilthe optimal exposure is determined.

Modifcutions or Purufin Sectionsor Cultured Cell'sThe same basic protocol described in the preceding sections can also beused for paraffin sections or cultured cells, with slight modifications. Par-affin embedding after paraformaldehyde fixation does provide good pres-ervation of the cellular mRNA for hybridization. Direct comparison ofparaffin to frozen blocks from the same tissue indicates that there isapproximately a 25% loss in hybridization signal due to paraffin (Smithand Wilcox. unpubli shed observations). Therefore, when paraff in embed-ding is necessary for appropriate tissue morphology (i.e.. lung), then it canbe used. Sections are deparaffinized by washing twice for 2 min in xylene,twice in 100% ethanol for 1min. for 1min each in 95% ethanol, 70%

t

ethanol. 50% ethanol, and 0.5 x SSC. T he sections are then ready forhybridization in the above protocol, beginning with the paraformaldehydefixation at Step 2. In the same manner, cultured cells can be fixed andprepared for hybridization. Cells are either grownon Lab-Tek cul ture slidesor centrifuged onto V ectabond-coated or SuperfrostlPlus slides and thenfixed by immersion in 4% paraformaldehyde at 4'C for 10 min. T he slidesare then stored at 4'C for up to3 months in 70% ethanol. Before hybrid-ization the slides are taken from the ethanol and the hybridization proce-dure isbegun by immersing them in 0.5 x SSC. starting with Step3directly.

Control'sThe choice of an appropriate control is the most challenging aspect to usingin situ hybridization properly. A number of controls have been used byinvestigators to establish the validity of their results (see below): some arebetter than others. T he best control is common sense. Does the signal makesense to you? Is i t localized over the cytoplasm of the cell rather than overthe nucleus? Are there negative cells in the tissueaswell aspositive cells?Is the gene expressed in appropriate sites in control tissues? These are im-portant questions that must be addressed each time the results of an insitu experiment are evaluated. I t is easy to assume that any concentrationof silver grains represents a positive signal but careful examination and re-hybridization may yield very different results (Figure 1).

Figure 1. Examples o a good (A ) and bad (B ) n situ hybridization result. Sec-tions o rat pituitary were hybridized with 35S-labeled pro-optomelanocortin(POMC)riboprobes using twodifferent methods. (A ) Section hybridized accord-ing to theprotocol outlined in this review reveals strong hybridization to the in-termediate lobe (central band of positive cells), scattered hybridization to an-terior lobe corticotrophs (tothe left),with no hybridization to the posterior lobe(tothe right of the intermediate obe). (B ) The same pituitary hybridized in par-allel with the section in A using a riboprobe procedure outlined in a recentse-riesof publications(6,11,13). Note the high background in the posterior lobein B and the lacko any significant hybridization signal in the anterior lobe. Inthe absence o any additional nformation, the background n B may have beenassumed to represent positive hybridization and il lustratesthe problem as-sociated with accepting all hybridization results at face value. Exposure1week;photographed with polarized light epiluminesence (Leitz). Original magnifica-tion x 50. Bar = 100pm

Controls for In Situ HybridizationSenseIAntisense CO-localization with proteinMultiple non-overlapping probes Northern blotsMultiple probes RNAselDNAseCommon SenseSense and anti-sense hybridizations are widely used as controls for insitu hybridizations with riboprobes (Figure2). Riboprobes are synthesizedby transcription utilizing the cDNA asa template for probe synthesis. Mes-


7/9

PRINClPLES OF I N SITU HYBRlDIZATION 1731

One word of caution is warranted regarding the use of sense and anti-sense riboprobes. It is often possible to see positive hybridization or evehhigh backgrounds with sense riboprobes yet to find a very strong, specificand reproducible signal inthe parallel anti-sense hybridization.Howshouldthis be interpreted? Is there anti-sense mRNA in the tissue? s the enurehybridization flawed? The first assumption s that the researcher has donesomething wrong and the experiment isusually repeated until the sensehybridizations are appropriatelynegative and the anti-sense hybridizationsappropriately positive. However, the question remains:wasthe first or sec-ond experiment flawed? There is no easy answer to these questions. andthe assumption is that sense hybridizations should be negative.A common approach used in my laboratory s hybridization with mul-tiple probes in a single experiment. This can ensure specificity and pro-vides the maximal amount of information per experiment. Typically, fouranti-sense probes are hybridized ina given experiment and the results ofeach hybridization compared for specificity. The probes should show morethan one hybridization pattern for the experiment to succeed and shouldinclude at least one probe that will hybridize to a known subset of cellsin the sections being examined. A good control for human tissue is vonWil lebrands Factor(VWF),which is synthesized only by endothelial cells.Other cell types n the tissue will be, or shouldbe, negative, and the researchercan evaluate tissue background wi th thisapproach. The use of a positivecontrol probe with a known cellular distribution of hybridization in thetissues also provides a control for positive mRNA hybridization and theefficiency of the experiment. The specificity of the hybridization is indi-cated when one probe, e.g., VWF, hybridizes to endothelial cells and an-other probe hybridizes to other cells in serial sections(19).This is a goodcontrol and is generally accepted by journal reviewers.The mistake people commonly make with this approach is the choiceof the control probe. Often tissues are hybridized with poly d(T) to detectpoly(A ) mRNA tails or with housekeeping genes suchas actin, known tobe expressed by a l l cells. These are poor choices for controls, as the resultwill be positive hybridization to all cells with no negative cells present todetermine the amount of background hybridization. The biggest problemwith in situ hybridization s background and determining when accumula-tion of silver grains is real ar not. For this reason it is important to haveclear positive and negative hybridizations ncluded in an experiment o thatit is clear what each will look like.Hybridization of cDNA or riboprobes to tissue sections obviously de-pends on the presence of mRNA in the tissues. For ths reason, anothercontrol that has been used is predigestion ofthe tissue sectionswith RNAseand DNAse. Since RNAse, but not DNAse, should degrade the signal,this is assumed to confirm the necessity of cellular mRNA for the hybrid-ization reaction. One problem with thisapproach is its appli cation to hy-bridization with riboprobes and the possible degradation of the probe it-self rather than the tissue A.Additional controls that have been used include many that fall i n thecommon sense category. M ultiple non-overlappingprobes encoding themPNA should each give similar hybridization signals, whereas an irrele-vant nucleotide sequence should not. This isan especially useful controlwhen synthetic oligomers are used for hybridization and it is possible tomake a number of different probes. Alternatively, non-overlapping5 and3 ends of the cDNA can be subcloned and used for hybridization. De-pendingon the leve ofgene expression in the tissue it should be possibleto detect the mRNA by Northern blots. This would at least confirm thepresence of mRNA in the tissues under study. Of course, t h i s assumes thatthere is enough tissue for a Northern blot or enough cells expressing thegeneso that the mRNA can be detected with this technique. Alternatively,co-localizationof the protein and mRNA using mmunohistochemistry andin situ hybridizationon serial 01 same sections can provide another levelof assurance that the m situ hybridization signal is real.When exon-specific probes are used, the in situ hybridization signal

5 DN AGTCCACCAGGT

33

E C o R l BamHlu3,enseG G A C Ciboprobe3CAGGU

mRNA5

5 G G ~ C C AI I I I I ICCAGGUmRNA

3

Antisense Riboprobe

5 G G ~ C C AI I I I I ICCAGGU 39 probe

Figure2. Transcription of sense and anti-sense riboprobes or use as controlsfor in situ hyb ridization. Cellular mRNA is transcribed by RNA po lymerase inthe 3to 5direction using one strand of the chrom osomal DNA as a template.Depending on the orientation of the cDNA insert in a ribop robe vector relativeto the SP6 or T7 RNA polym erase initiation sites, either sense or anti-senseRNA pro bes can b e synthesized. Transcription from the CDNA template usingSP6 RNA polymerase in this example prod uces a nucleotide sequence denti-cal to the mRNA, otherwise known as the sense ranscript. Since the senseprobe is the same sequence as the cellular mRNA , it will not bind during thehybridization reaction and can beused as a n egative control. Transcription ofthe cDNA emplate rom the opposite end, using theT7RNA polym erase nitia-tion site, will prod uce an anti-sense RNA prob e which will hybridize tothe mRNAin tissues. Anti-sense transc riptswil l be generatedwhen the 3end of the cDNAis closest to the site of transcription initiation.

senger RNA is normally synthesized from chromosomal DNA in the3 to5 direction, producing sense mRNA. Tomake an anti-senseriboprobe thecDNA insert is subcloned in a transcription vector in the 3 to 5 directionrelative to an RNA polymerase initiation site. Transcription takes place inthe presenceof a labeled nucleotideso that the resulting anti-sense probeis complementary to the mRNA and will bind to i t during the hybridiza-tion reaction. Insertion of the cDNA insert in the opposite orientation rel-ative to the RNA polymerase nitiation site will transcribe the control senseriboprobe. a sequence dentical to the mRNA in the tissues which will nothybridize. Since the sense and anti-sense probes both contain the samerelative proportion of all of the nucleotides (A /Us or GIG) n the samesequence, this has been considered an excellent control or riboprobe hybrid-izations (2).


8/9

1732 m o x

I?

6

- . as t o use the number of grains over the nucleus to quantify the amountof mRNA associated with the cell (11.13). The mRNA signal with exon-specific probes should be over the cytoplasm (Figure3). There will be somespread of grains over the nucleus, but 90%or more of the signal normallyappearsas a cluster of silver grains around the nucleus of the cell. Depend-ingon the cell morphology, it is possible tosee some nuclear silver grains.For example, monocytesorsmall macrophages with a very thin cytoplasmoften show some grains over the nucleus (10). This is probably because themRNA is concentrated in a restricted cytoplasm of which the major partappearing in a tissue cross-section overlies the nucleus.On the other hand,one of the most common background problems is nuclear silver grains. Whenthe hybridization is pushed by increasing the probe concentration too highor by dropping the stringency of the final wash. the first thing that hap-pens is that all of the cells in the section show a nuclear signal. This doesnot mean that the mRNA is in every cell but rather that it is time t o beginto trouble-shoot the experiment todetermine the source of background.

Figure 3. Localization of the in situ hybrid ization signal in the c ytoplasm withexon-specific probes. Rat pituitary section s were hybridized with (A ) 3H-or (B)biotin-labeled ribopro bes directed agains t exon 3 of the POMC cDNA. POMCriboprobes were transcribed using 3H-labeled UTP and CTP (10 pM each) inthe transcription reaction , hybridized accord ing to the protocol outlined in thisreview, and exposed to p hotog raphic emulsio n for 5 weeks. Biotin-labeledriboprobes were transcribed using b iotinylated UTP and the B RL non-radioactiveRNA labeling system kit (Gibco BRL; Gaithersburg, MD) and hybridized usingthe B RL in situ hybridization and detection system kit exactly as d escribed bythe m anufactur er. Although hybr idizations with 3H-lab eled or biot in-labeledribopr obes are not as sensitive as 35S-labeledprobes , they can prov ide excel-lent resolution of the mR NA signal. Note the specific cyto plasmic signal usingboth metho ds and the lack of any n ucl ear bindin g of th e probes. Original mag-nifications:A x 200; B x 100. Bars = 100 pm .

should be found primarily over the cytoplasm of the cell (3.4). This is incontrast to recently published reports examining local gene expression inbiopsy specimens of human coronary arteries (6.11.13). These authors ac-cepted nuclear grains as indicating a positive hybridization and go so far

SummaryIn situ hybridization provides the researcher with invaluable infor-mation regarding the localization of gene expression in heteroge-neous tissues. The technique is extremely sensitive and can detectthe amount of mRNA contained in a single cell. However, be care-ful how you apply this technique. In situ hybridization looks goodand is often used not because of scientific necessity but rather forthe fl ash it adds to a presentation. I t is a difficult procedure,and one must ask whether it is worthwhile todevelop i t for a singleexperiment or whether a Northern blotorsome other assay mightprovide the desired information with a lot less effort.I have tried to provide a starting point for those who wish tobegin using in situ hybridization in their own laboratories. T heprocedure outlined here has been used with many probes and tis-sues with a greater than 90% success rate on the first hybridiza-tion. This protocol was designed to be compatible with tissues com-ing from many different sources ncluding surgical biopsy, autopsy,and animal experimentation. The procedure is streamlined withthe goal of being able to hybridizeasmany as100-200 slides perday withat least four different probes. Many steps, suchasacetyla-tion, defatting in xylene or alcohols, post-fixation, or multiple pro-teinase steps, have been eliminated. These are often used to re-duce background in the tissue. However, if this procedure is appliedcorrectly, background is not a problem. A good hybridization willhave 25-50 cytoplasmic (not nuclear) grains per positive cell and


9/9

PRINCIFES OF IN SITU HYBRlDIZATION 1733

section after hybridization andcall it positive. For example, regionsof the tissue rich in nuclei often appear to cause spurious bindingof the probe and have high backgrounds. The trick isnot to getclustersof silver grainsonthe slide but rather to do the appropri-ate controlled experiments to ensure that the signal isreal andnotdue to some artifactual binding of the probe to the tissue.AcknowledgmentsI am deep/y indebted to a number ofpeople who have worked with meover the past 10 years, each of whom bas contributed to the developmentof this in situ protocol in some way, including Kathy Smith, udy Hasko,Andrew Augustine, and Romesh Subramanian. I am especial/y indebtedto ose Rodriguez or providing the in situ hybridizationspresented here.

Literature Cited1. Armstrong E, Partanen ,CannizzaroL , Huebner K, A litalo K Local-ization of the fibroblast growth factor receptor-4 gene to chromosome

region 5q33-qter. Genes Chromosom Cancer 494, 19922. Cox K H, DeLeon DV, AngererLM, Angerer RC Detection of Asin sea urchin embryos by in situ hybridization using asymmetric RNAprobes. Dev Biol 101:485, 19843. Fremeau RT, Auteli tano DJ, BlumM, WilcoxJ ,Roberts L Interven-ing sequence-specific in situ hybridization: detection of the pro-opiomelanocortingene primary transcriptin individual neurons. BrainRes Mol Brain Res 6:197, 19894. Fremeau RT, Lundblad R, Pritchett DB, WilcoxJN, Roberts L Regu-lation of pro-opiomelanocortin gene transcription in individual cellnuclei. Science 234:1265, 19865. Gordon D, Augustine AJ, Smith KM, Schwartz SM, Wilcox N Local-ization of cells expressing tPA, PAIL and urokinase by in situ hybrid-ization n human atherosclerotic plaques and in the normal rhesus mon-

key. Thromb Haemost 62:131, 19896. LeclercG,IsnerJM, Kearney M, SimonsM, SafianRD,Baim DS, WeirL Evidence implicating nonmuscle myosin in restenosis. Use of in situhybridizationCO analyze human d a r esions obtained by directionalatherectomy. Circulation 85:543, 19927. MajeskyMW,ReidyMA, BowenFbpeDF, HartCE, WilcoxJN, Schwartz

SM: PDGF ligand and receptor gene expression during repair of ar-terial injury. J Cell Biol 111:2149, 19908. Melton DA, K rieg PA, RebagliatiMR, Maniatis T, Zinn K, GreenM REfficient in vitro synthesis of biologically active RNA and RNA hy-bridization probes from plasmids containing a bacteriophage SP6 pro-

moter. Nucleic Acids Res 1237035, 19849. Nelkin NA. Coughlin SR, Gordon D, WilcoxJN: M onocyte chemoat-tractant protein-1 inhuman atheromatousplaques. Clin Invest 881121,199110. Neken NA, Soifer SJ, Okeefe ,VuTK, Char0IF, CoughlinSR Throm-bin receptor expression in normal and atherosclerotic human arteries.J Clin Invest 901614, 199211. Nikol S,IsnerJM, Pickering G, KearneyM, Leclerc G, WeirL: Expres-sion of transforming growth factor-beta 1 s increased n humanvascu-lar restenosis lesions.J Clin Invest 90:1582, 199212. Rosenthal A. ChanSY, Henzel W, Haskell C, KuangWJ, ChenE, WilcoxJN, Ullrich A, Goeddel DV, Routtenberg A: Primary structure andmRNA localization of protein F1, a growth-related protein kinase Csubstrate associated with synaptic plasticity. EMBO J 6:3641, 198713. Simons M, Leclerc G, SafianRD, IsnerJM, Weir L , Baim DS: Relationbetween activated smooth-muscle cells in coronary-artery esions andrestenosis after atherectomy. N Engl J Med 328:608, 199314. WilcoxJ N: Analysis of local gene expression in human atheroscleroticplaques. J Vasc Surg 15:913, 199215. Wilcox JN, Derynck R Developmental expression of transforminggrowth factors alpha and beta in m o w etus. Mol Cell Biol8:3415, 198816. WilcoxJN, Gee CE, Roberts L In situ cDNA :mRNA hybridization:development of a technique to measure mRNA levels in individualcells. Methods Enzymol 124510, 198617. WilcoxJN , Pollard H, MoreauJ ,Schwartz C, Malfroy B: Localizationof enkephalinasemRNA in rat brain by in situ hybridization: compar-isonwith immunohistochemical ocalization of the protein. Neuropeptides 1477, 198918. Wilcox N, SmithKM, Schwartz SM, Gordon D: Localization of tissuefactor in the normal vessel wall and in the atheroscleroticplaque. ProcNatl Acad Sci USA 86:2839, 198919. WilcoxJN, SmithKM,WilliamsIT,SchwartzSM, Gordon D: Platelet-derived growth factor mRNA detectioninhuman atherosclerotic plaquesby in situ hybridization. J Clin Invest 82:1134, 1988

fundamental principles of in situ hybridization

Documents