the neurons produce a three-dimensional character that differs slgmficantly
from cell lines in culture. Therefore the detection of transcripts using ISH meth-
ods with radiolabeled probes and emulsion provide poor spatial localization.
Consequently, we have switched to using only nonradioactive detection of tran-
scripts with digoxlgenm-labeled riboprobes.
A second problem 1sthe direct result of the extensive cytoskeletal structure
present within the neurons. A more aggressive permeabilization protocol of
neurons m culture 1s required for penetration of macromolecules (such as
riboprobes), yet such protocols must not disrupt the attachment of the neurons
to the cover slips. In our laboratories, ISH techniques have been developed that
have proven to be both reliable and reproducible. These methods assume
knowledge of the neuronal culture system,which has been described elsewhere
(8), and a general famlllarlty with the establishment and maintenance of
RNAse-free conditions (see ref. 9) for general methods for working with RNA).
1 NTP labeling mix 10 mMATP, 10 mA4CTP, 10 mMGTP, 6 5 mMUTP, 3 5 mM
dig-UTP. Boehringer-Mannhelm (Mannhelm, Germany) #1277 073.
2 2 mMCaCi2
3. 50X Denhardts
5. DEPC-treated phosphate-buffered saline (PBS).
6. 100 mM dlthlothretol (DTT)
7. 20% Dextran sulfate m formamlde.
8. 5% Dimethyldlchlorosilane in chloroform
9. DNA-Dependent RNA polymerase.
11 Geneclean II Kit (BIO 101, La Jolla, CA)
12. 1.5 mM levamisole.
14 1 MMgC12
16 60 mM Na2C03/40 nu!4 NaHC03, pH 10.2
17. 75 mg NBT/mL of 70% v/v DMF
18 4% Phosphate-buffered paraformaldehyde (prepared fresh)
19. 10 mg/mL Protemase K (Boehrmger-Mannheim).
21 50 Clg/mL RNAse A (Qiagen)
22. 2 U/mL Tl RNase
Wilcox and Smith
23 20% Sodium dodecyl sulfate (SDS)
24 0 2M Sodium acetate, 1% (v/v) acetlc acid, pH 6.0
25 20X SSC.
26. I OX Transcription buffer
27. 1M Tris-HCl, 9.5.
28. 20 mMTris-HCl, pH 7 5
29. 10 mg/mL yeast tRNA.
30. 50 mg X-phosphate/mL m DMF.
RNase buffer. 20 mM Tris-HCl, pH 8 0, 0.5M NaCl, 1 mM EDTA
32. Buffer 1: 100 mA4 Tns-HCl, pH 7 5, 150 mA4NaCl.
33 Buffer 2 Prepare only the volume that is needed, as preclpltation occurs with
storage 100 mM Tns-HCl, pH 9 5, 100 mM NaCl
34. 4 mL Buffer 2 + 18 pL of NBT stock (75 mg NBT/mL of 70% v/v DMF) +I4 pL
of X-phosphate stock (50 mg X-phosphate/mL m DMF) + 1 5 mM levamlsole.
The molecular biology reagentshave been used interchangeably from Promega
(Madison, WI) and Boehringer-Mannheim, with the exception of the nucle-
otide mixture, which has only been purchased from Boehrmger-Mannhelm. Other
chemicals are purchased from the supplier indicated or Sigma (St. Lours, MO).
3.1. Fixation and Storage of Cultures Prior to Hybridization
1 Grow neuronal cultures on plastic, collagen-coated cover slips in 24-well
2 To begin fixation, rinse cultures with DEPC-treated PBS
3. Fix with 4% phosphate-buffered paraformaldehyde for 12 h at 4Â°C
4. Rmsewith DEPC-treatedPBS,and dehydrate through grades of ethanol (70,95,
100%) and an-dry.
5. At this point fixed cultures are removed from the cluster dishes with a sterile
disposable hypodermic needle that has been bent at the tip to faclhtate removal of
cover slips from the culture dish
6. Store the cover slips at -20Â°C m the presence of a desiccant Strong hybrldizatlon
signal has been obtamed from cultures stored for 6 mo under these condltlons
3.2. Silanization of Glass Cover Slips
to Be Placed Over Neuronal Cultures
Glass 15-mm round cover slips are used to cover the samples during the
hybridization to prevent drying. To prevent nonspecific sticking of probe, glass
cover slips are silamzed at least 1 d m advance.
1. In a fumehood, mdlvidual glass cover slips are dipped m 5% dimethyldichloro-
sllane m chloroform and placed m a 100-mm glass Petri dish and air-dried.
HSV Latency In Vitro
2. The cover slips are then washed, in mass, by repeatedly filling the Petri dishes
with distilled deionized H20, replacing the cover, gently swirling, and decanting
3 The cover shps are then baked m an oven at 250Â°C for at least 4 h to destroy
RNAse activity These can be stored at room temperature for at least 6 mo
3.3. Preparation of Template DNA
To obtain template DNA, we have used a plasmid preparation method based
on the differential precipitation of DNA (9). We have avoided methods that
require addition of exogenous RNAse to facihtate the purification. Considerable
RNA contaminatton remains in the preparation without presentmg a problem.
The plasmid is linearized with the appropriate restriction enzyme and template 1s
purified by electrophoresis in an agarose gel followed by purification of DNA
using the Geneclean II Kit (BIO 101).
2 The recovery of DNA is then estimated by running on a gel against a set of DNA
standards. All efforts are made to use RNAse-free reagents m preparmg the tem-
plate. In our hands, provided reagents are handled to avoid Introduction of
RNAse, no further purification steps are required Templates subjected to phenol
extraction have produced lower yields in transcription reactions m our hands.
3.4. Transcription of Riboprobes
(The methods described below were derived from those provided by
Boehrmger-Mannheim with the Genms Non-Radioactive Detection Kit).
1. Transcribe by adding to a microfuge tube: 1 pg DNA template; 0 5 pL 100 mA4
DTT; 2 pL NTP labeling mix (10 mM ATP, 10 mM CTP, 10 mM GTP, 6 5 mM
UTP, 3.5 mM dig-UTP), 2 uL 10 X transcription buffer; 2 U of DNA-dependent
RNA polymerase, 2 pL RNasin Adjust final vol to 20 uL with DEPC-dH20
2. Centrifuge briefly.
3. Incubate at 37Â°C for 2 h.
4. Remove template by adding 2 U of RNase-free DNase I, incubate 15 mm, 37Â°C
5. Add 1 pL of 0.2MEDTA and precipitate the riboprobe with 2 5 pL 4M L1C1, and
75 @ 100% ethanol, overnight at -20Â°C.
6. Centrifuge, wash pellet with cold 80% ethanol, dry, and resuspend m 100 pL
DEPC-H,O + 1 pL RNasm + 2 pL 100 mMDTT Incubate at 37Â°C for 15 mm to
dissolve the riboprobe
7. Generally, IO-20 pg of riboprobe will be produced per each reaction, depending
upon template. Methods for confirming the yield and analyzing the size of the
probe are well described (9).
3.5. Fragmentation of the Riboprobe by Alkaline Hydrolysis
1. Hydrolyze drgoxigenm-labeled RNA by addmg an equal volume of DEPC-H,O and
2 vol of 60 mMNazCO,, 40 miVfNaHC03, pH 10.2 Incubate at 60Â°C for 10 min.
2. Stop hydrolyses by adding an equal volume of 0 2M sodium acetate/l% (v/v)
acetic acid, pH 6.0
322 Wilcox and Smith
3 Aliquot 1 pg of rrboprobe per microfuge tube, add 5 pL of 10 mg/mL tRNA per
tube, and 3 vol of ethanol Store mdefimtely at -70Â°C Each tube contains a suf-
ficient amount of rtboprobe to hybridize 1O-l 5 cultures on 13-mm cover slips
3.6. Pretreatment of Samples
1 The cultures on cover shps are hydrated in 20 mMTrts-HCl, pH 7 5 at 37â€™C, 2
rnMCaC1, with 5 mg/mL proteinase K for 30 min at 37â€™C This can conveniently
be performed by placing up to eight cover slips, culture side up, m a 60-mm
sterile plastic Petri dish. Some care 1s required to be certain that trapped an
bubbles have been removed to prevent the cover slips from floatmg
2 Thereafter, the samples are washed once m DEPC-treated PBS, dehydrated
through graded ethanol and allowed to an-dry
3.7. Addition of the Riboprobe-Hybridization #ix to the Samples
1 For one ahquot of an ethanol precipitated rtboprobe, spin, rinse pellet with 70%
ethanol, and resuspend m 20 pL DEPC-dH,O
2 Heat to 75Â°C for 3 mm, cool to room temperature, and then add m order.
100 mMDTT 2ccL
20% Dextran sulfate in formamide
20x ssc 4OcIL
50X Denhardtâ€™s 4lJ-
10 mg/mL yeast tRNA 5&
20% SDS 2c1L
DEPC-dH,O 23 cl˜
02M EDTA 2&
Final volume 200 pL
3. Vortex vigorously to mix completely and then centrifuge to remove trapped
1. To the samples on the cover shps that have been pretreated with proteinase K and an
drted, place culture surface up m a Petri dish, and add 20 pL of the probe-hybndizatton
nuxture to each cover shp Great care must be taken to avoid mtroductton of an bubbles
2. The hybrtdtzatton mixture 1s then covered with a stlanated glass cover slip We
have not found it necessary to seal the edges of the cover slip tf a humidified
environment 1sprovided for the hybridization
3 The Petri dish containing the cover slips must be leveled carefully If excessive
problems occur with the stlanated cover slips sliding off of the samples, either
the volume of hybridtzatton mix can be reduced or an edge of the cover slip can
be tacked in place with a bead of rubber cement delivered via a syringe
HSV Latency In Vitro
4. Place Petri dash contammg samples covered with a lid and place m a sealable
container (1 e., Tupperware) containing water-saturated paper towels Carefully
place the sealed contamer in an oven at 52Â°C overmght. Temperatures for
hybridizatton and washes are for HSV-specific probes, however, we have found
them to be equally successful for probes for many cellular genes
3.9. Posthybridization Washes
1 Remove cover slips and wash samples at 50Â°C in 5 mL/Petri dish of 2X SSC +
50% formamide for 45 mm.
2. Rinse 2X m RNase buffer and then incubate 30 mm at 37Â°C with 50 pg/mL RNase
A + 2 U/mL Tl RNase.
3 Wash at 50Â°C in 2X SSC + 50% formamide for 40 mm
4 Wash with buffer 1 for 10 mm
5 Block nonspecific antibody binding with buffer 1 + 0.3% Triton-X100 and 2%
normal sheep serum for 30 mm.
6 Incubate with buffer 1 + 0.3% Triton, 2% normal sheep serum and 1*5000
antidigoxigenm antibody for 4 h at room temperature
7 Wash with 2X with buffer 1 for 10 mm each
8. Wash with buffer 2 for 10 min
9. Incubation of the cover slips m a vertical posmon during the calorimetric devel-
opment reduces deposition of precipitate that forms. This results m reduced back-
ground stammg and greatly improves the final appearance of the cultures. Cover
slip holders are commercially avatlable from Ltpshaw or Thomas. Incubation with
buffer 2 + NBT + X-phosphate Incubation m the dark for either 2-4 h at room
temperature or 6-12 h at 4Â°C produce similar color development reactions.
10. Stop in 20 MTrts-HCl, pH 8.0, 1 rnM EDTA
11 Several approaches can be taken to mountmg the slides A simple method is
to place the cover slip on a standard microscope slide with a small drop of
Aquamount below the slide. A second drop is placed on the slide and a
glass cover slip is then mounted over the sample Slides prepared by this
method are best photographed immediately Some fadmg and loss of cell
morphology appears over time.
1 Currently, 13-mm plastic tissue culture cover slips (Sarstedt) are used for grow
the neuronal cultures. This has been an area for problems, with plastic cover slips
from some suppliers proving to be toxic to neurons. Unfortunately, although the
cover slips from Sarstedt have not been toxic, some of the cover shps curl during
processing for ISH. Use of acid-washed, glass cover slips is likely to be feasible,
however, we have not yet used them for ISH. Retention of cultures and nonspe-
cific sticking of probe to glass may require some modificattons.
2 During the development of zn situ methods, it became clear that the signal-
obtained ISH protocol was greatly altered by the degree of fixation, In initial
experiments brief fixation protocols were used, however, the results were highly
Wilcox and Smith
variable. Once the importance of permeabhzatton was appreciated, it became clear
that variations could be ehmmated with relatively long fixation times.
3 Two approaches for the generation of appropriate sized probes can be used. Either
small (less than 300 bp) probes can be used, or longer probes can be reduced m
length by limited alkaline hydrolysis, as described
4 As noted in the mtroduction, permeabihzation of neurons IS of critical impor-
tance for success of ISHs. With use and familiarity, adjustments of the concen-
tration of the protemase may be required with different lots of protemase K. To
facilitate this adjustment process, the proteinase K IS made up as a 10 mg/mL
stock and immediately frozen m many small ahquots These ahquots are only
thawed once and used immediately In our hand, the best ISHs have been obtained
when some degradation of the neuronal processes is detectable. Extensive loss of
neurons from the cover slips IS a sign of excessive protemase treatment In addi-
tion, mclusion of SDS has been necessary for neurons m culture
5 We have included a figure to illustrate some of the patterns of transcript distribu-
tion that have been encountered by our laboratories m using the described system
(Fig 1) In (B) the commonly observed nuclear LAT signal is demonstrated dur-
mg the latent infection with a probe specific for the major LAT (For review, see
ref. 10) In (C) IS shown another relatively common pattern of stammg observed
during the latent mfection with a probe homologous to the leader region of the
minor LAT. This consists of discrete punctate stammg m the nuclei. We have
obtained similar results usmg probes for regions 3â€™ of the major LAT Arthur et
al have reported similar results m DRG neurons m the mouse model (11) The
sigmficance of the punctate stammg remams unknown, but may indicate RNA
processmg This punctate nuclear zn sztu signal can easily be overlooked because
the staining is very limited and it IS usually m more than one focal plane Finally,
shown in (E) is a mostly cytoplasmic pattern of stammg of neurons during reac-
tivation with a probe specific for the IE 1 transcript The almost exclusively cyto-
plasmic pattern of stainmg is more difficult to see m this micrograph because the
3D character of the cells produces a slightly diffuse signal. Stammg of the nucleo-
lus has been observed. Probably this IS related to nonspecific staining and
suggests that inadequate RNase treatment was used
6 The methods described here can be directly applied to sections of ganglia. How-
ever, several modifications can be made that improve the histology. First, pro-
temase K pretreatment can be performed under much milder conditions, such as
with 1 mg protemase WmL Second, SDS is not required m the hybridtzation
mixture. With these modifications LAT has been detected readily m para-
formaldyhde or Boumâ€™s fixed tissue If one of the recently mtroduced noncross
linking fixatives (e g., Amresco HC fixatrve) are used, protemase K pretreatment
should absolutely not be done and the use of SDS m the hybridization mixture
can cause a loss in signal
This work was supported by Public Health Service grant NS29046.
HSV Latency In Vitro 325
Fig. 1. Patterns of ISH signals in the in vitro neuronal HSV latency model. DRG
neurons were hybridized for specific transcripts using the methods described in this
review. Shown are (A) mock-infected neurons and (B) neurons during the latent infec-