THE ANATOMICAL RECORD 231:559-562 (1991)

Quantitative Non-Radioactive In Situ Hybridization of Preproenkephalin mRNA With Digoxigenin-Labeled cRNA Probes

ELAINE ROBBINS, FRANK BALDINO, JR., JILL M. ROBERTS-LEWIS, SHERYL L. MEYER, DEBRA GREGA, AND MICHAEL E. LEWIS

Cephalon, Inc., West Chester, Pennsylvania (E.R., F.B., J.M.R.-L., S.L.M., M.E.L.); and Boehringer Mannheim, Inc. Biochemicals, Indianapolis, Indiana (D.G.)

ABSTRACT Non-radioactive detection of mRNA with in situ hybridization histochemistry has emerged as an important new technology for the study of gene expression. Quantitative in situ hybridization studies have generally relied upon counting of autoradiographic grains in the emulsion overlying cells containing hybridized, radioactively labeled probe. However, such high resolution studies require tedious grain counting over individual cells, frequently in addition to weeks of exposure to nuclear emulsion. The present report describes a quantita- tive, non-radioactive approach to the detection of a specific mRNA in the brain with the advantages of comparatively rapid tissue processing and computerized image analysis. The validity of this approach was tested by measuring the halo- peridol-induced increase in the level of preproenkephalin mRNA in striatal sec- tions of the rat brain using an RNA probe labeled with digoxigenin-11-UTP. De- tection of probe hybridized to tissue sections was carried out enzymatically following complex formation with an antidigoxigenin-alkaline phosphatase conju- gate. Using computerized image analysis, it was found that chronic treatment of rats with haloperidol resulted in a 50 5 6% increase in striatal neuronal optical density, a value in good agreement with previous studies using low-resolution radioactive methods, showing a 30-80% increase in striatal preproenkephalin mRNA hybridization signal.

The utility of in situ hybridization histochemistry lies in the capability to analyze the cellular locus as well as the level of specific mRNAs. Traditional meth- ods utilize radioactively labeled probes, with detection and quantification of specific hybrids in cells carried out by grain counting with liquid emulsion autoradiog- raphy (e.g., Baldino et al., 1989; Lewis et al., 1989). However, because of a variety of difficulties associated with the use and detection of radioactive probes (e.g., the need for tedious grain counting as well as pro- longed exposure to emulsions), several investigators have developed alternative methods using non-radio- active labeling of probes (see Lewis and Baldino, 1990 for review). One problem that stands out with the use of these probes is quantification of the hybridization signal, which is crucial to realizing the full value of the in situ hybridization technique. In this study we de- scribe a simple approach to this problem, based on in- corporating the use of computerized image analysis into the digoxigenidalkaline phosphatase probe tech- nology, which we (Baldino and Lewis, 1989; Lewis et al., 1990; Springer et al., 1991) and others (Young, 1989) have employed for the non-radioactive detection of mRNA in tissue sections. As a test of the validity of this approach, we have examined its utility in measur- ing a well-established regulatory event, i.e., the induc- tion of striatal preproenkephalin mRNA by chronic treatment with the dopamine receptor antagonist halo- peridol (Sabol et al., 1983; Tang et al., 1983; Hong et 0 1991 WILEY-LISS, INC.

al., 1985; Angulo et al., 1986; Sivan et al., 1986; Ro- man0 et al., 1987).

MATERIALS AND METHODS Treatment of Animals

Six male Sprague-Dawley rats from Hilltop Labora- tories (150-200 g) received approximately 2 mg/kg daily of haloperidol for a period of 3 weeks by means of a 7.5 mg slow-release pellet implanted subcutaneously (Innovative Research of America). Five control animals received no drug treatment. Both groups were sacri- ficed 22 days after the treated group received the pellet implantation.

Brain Tissue Preparation All animals were sacrificed by decapitation, and each

brain was quickly frozen in powdered dry ice and stored at -80°C until sectioning. Corona1 sections through the striatum (12 p,m) were collected onto gelatin-coated slides, thaw mounted, and stored at -80°C for 3 days prior to hybridization.

Received January 22, 1991; accepted March 13, 1991. Address reprint requests to Elaine Robbins, Cephalon, Inc., 145

Brandywine Parkway, West Chester, PA 19380.

560 E. ROBBINS ET AL

Preparation and Hybridization of the Rat Preproenkephalin cRNA Probe

A digoxigenin-labeled cRNA probe was synthesized using digoxigenin-11-UTP (Boehringer Mannheim Biochemicals) and the linearized plasmid pYSEA1, which terminates at the end of a 935 bp preproen- kephalin cDNA insert. Brain sections were postfixed, acetylated, washed in Tris-glycine (pH 7.0), and dehy- drated (Baldino et al., 1989). Sections were then cov- ered with hybridization buffer containing digoxigenin- labeled probe, 40% formamide and ~ x S S C , and incubated under a parafilm coverslip for 3.5 hours at 50°C in humid boxes. Posthybridization treatments in- cluded washes in 50% formamidel2 x SSC at 52°C and treatment in RNase A (Baldino et al., 1989). Detection of probe hybridized to striatal tissue sections was car- ried out enzymatically following complex formation with an antidigoxigenin-alkaline phosphatase conju- gate (Springer et al., 1991).

Quantitative Analysis A computerized image-analysis system (R & M Bio-

metrics) was used to quantitate the relative amounts of preproenkephalin mRNA hybridization signal in stri- atal neurons of control and haloperidol-treated rats. First, a standard curve was constructed by measuring grey levels on a precalibrated neutral density filter step wedge (Kodak). Grey level values obtained from cell populations in tissue sections could then be auto- matically converted to optical density units. Optical density readings, as determined by NBT reaction prod- uct intensity in enkephalin mRNA positive neurons, were taken from two different fields of approximately 20 cells each from successive sections obtained from the same animal. Readings from a total of four sections for each animal were used to calculate the mean optical density (OD) for each animal in the treated group and six animals in the untreated group. Student’s t-test was used to evaluate the difference in striatal neuronal OD between treated and untreated groups.

RESULTS A digoxigenin-labeled cRNA probe complementary

to proenkephalin mRNA hybridized to forebrain neu- rons (Fig. 1) in an anatomical pattern indistinguish- able from that seen previously with 35S-labeled probes (Harlan et al., 1987). For example, densely labeled perikarya were detected throughout striatum (Fig. lA,B) and in lateral septum (Fig. lB), whereas less densely labeled neurons were scattered in deep layers of parietal cortex (Fig. 1A) and cingulate cortex (Fig. 1C). Following 3 weeks of treatment with haloperidol, there was a visible increase in striatal neuronal pre- proenkephalin mRNA hybridization signal (Fig. 2). Computerized image analysis of striatal neurons dem- onstrated a haloperidol-induced increase in hybridiza- tion signal of 50 ? 6% (Fig. 31, reflecting a statistically significant difference (t = 4.67, df = 9, P = 0.001, two-tailed) between the control and treatment groups.

DISCUSSION Digoxigenin-labeled cRNA probes were found to be

useful for the detection of preproenkephalin mRNA in

Fig. 1. Distribution of proenkephalin mRNA-containing neurons. A Note relatively low expression of mRNA in parietal cortical layer VI (VI) compared to striatum (STR). B: Striatal and lateral septal (LS) neurons have similar mRNA levels. C: Cingulate cortical layer VI neurons express intermediate levels of proenkephalin mRNA. The corpus callosum (CC) is unlabeled.

the brain by in situ hybridization histochemistry, con- sistent with our previous report of their utility for the detection of nerve growth factor receptor (NGFR) mRNA in brain tissue sections (Springer et al., 1991). The use of these probes provides sensitive histochemi- cal detection of mRNA with very low tissue back- ground, apparently because the steroid hapten does not

NON-RADIOACTIVE IN SITU HYBRIDIZATION 561

Fig. 2. Striatal proenkephalin mRNA-containing neurons in control (left) and haloperidol-treated (right) rats. Note increased reaction product in striatal neurons from treated animal.

0.30

0.25

5 2

0.20 - m M wl E

- 0.15 -

d 0.10

’ 0.05

.- - ‘CI

P L

0.M) Control Haloperidol

Group

Fig. 3. Preproenkephalin mRNA hybridization signal in striatal neurons from control and haloperidol-treated rats, expressed as mean optical density values ? S.E.M.

have a detectable counterpart in tissue. The results obtained using this procedure are quantitatively con- sistent with the results reported by other investigators

using radioactive methods (Sabol et al., 1983; Tang et al., 1983; Hong et al., 1985; Angulo et al., 1986; Sivan et al., 1986, Romano et al., 1987). Whereas these re- sults validate the quantitative utility of the present method, i t is also clear that the method, like autorad- iographic methods (e.g., Lewis et al., 1989), does not provide absolute quantitation of cellular mRNA. To be- gin extending this method to absolute quantitation, it is necessary to convert optical density to units of alka- line phosphatase activity in order to ultimately express the results as hybriddcell. In preliminary studies, fol- lowing the method of Biegon and Wolff (1985) for car- rying out quantitative acetylcholinesterase histochem- istry, we have prepared tissue pastes containing different amounts of exogenous alkaline phosphatase. We have detected enzyme concentration-dependent in- creases in optical density in these standards, indicating their potential utility for the further quantitative anal- ysis of digoxigenin-labeled probe hybridization. Al- though the quantitation of hybridslcell may therefore be possible, relating this value to the absolute amount of mRNA/cell is more difficult. In particular, such a conversion would depend on having lost no mRNA dur- ing tissue processing as well as labeling all appropriate mRNA, conditions that are virtually impossible to achieve. Nevertheless, for most experiments, knowl- edge of the relative levels of mRNA (e.g., during dif-

562 E. ROBBINS ET AL.

ferent developmental stages or following different treatments) provides sufficient information to answer the question under investigation. Thus because of the rapidity and ease of carrying out quantitative, non- radioactive in situ hybridization, the methods de- scribed here should be broadly useful for the analysis of gene expression in brain and other tissues.

ACKNOWLEDGMENTS We gratefully acknowledge the technical assistance

of Ms. Ellen Riley and the generosity of Dr. Steven Sabol (N.I.H.) in contributing the pYSEAl plasmid.

LITERATURE CITED Angulo, J.A., L.G. Davis, B.A. Burkhart, and G.R. Christoph 1986

Reduction of striatal dopaminergic neurotransmission elevates striatal proenkephalin mRNA. Eur. J . Pharmacol., 130t341-343.

Baldino, F. Jr., and M.E. Lewis 1989 Non-radioactive in situ hybrid- ization histochemistry with a digoxigenin-dUTP labeled oligonu- cleotides. In: Methods in Neuroscience. P.M. Conn, ed. Academic Press, New York, Vol. 1, pp. 282-292.

Baldino, F. Jr . , M.-F. Chesselet, and M.E. Lewis 1989 High resolution in situ hybridization histochemistry. In: Methods in Enzymology. P.M. Conn, ed. Academic Press, New York, Vol. 168, pp. 761-777.

Biegon, A., and M. Wolff 1986 Quantitative histochemistry of acetyl- cholinesterase in rat and human brain postmortem. J . Neurosci. Methods, 16:39-45.

Harlan, R.E., B.D. Shivers, G.J. Romano, R.D. Howells, and D.W. Pfaff 1987 Localization of preproenkephalin mRNA in the rat brain and spinal cord by in situ hybridization. J. Comp. Neurol., 258;159-184.

Hong, J.S., K. Yoshikawa, T. Kanamatsu, and S.L. Sabol 1985 Mod- ulation of striatal enkephalinergic neurons by antipsychotic drugs. Fed. Proc., 44t2535-2539.

Lewis, M.E., and F. Baldino, J r . 1990 Probes for in situ hybridization

histochemistry. In: In Situ Hybridization Histochemistry. M.-F. Chesselet, ed. CRC Press, Boca Raton, pp. 1-21.

Lewis, M.E., W.T. Rogers, R.G. Krause 11, and J.S. Schwaber 1989 Quantitation anddigital representation of in situ hybridization histochemistry. In: Methods in Enzymology. P.M. Conn, ed. Aca- demic Pres, New York, Vol. 168, pp 808-821.

Lewis, M.E., E. Robbins, D. Grega, and F. Baldino, Jr. 1990 Nonra- dioactive detection of vasopressin and somatostatin mRNA with digoxigenin-labeled oligonucleotide probes. Ann. N.Y. Acad. Sci., 579t246-253.

Romano, G.J., B.D. Shivers, R.E. Harlan, R.D. Howells, and D.W. Ptaff 1987 Haloperidol increases preproenkephalin mRNA levels in the caudate-putamen of the rat: A quantitative study at the cellular level using in situ hybridization. Mol. Brain Res., 2:33- 41.

Sabol, S.L., K. Yoshikawa, and J . 3 . Hong 1983 Regulation of methio- nine-enkephalin precursor messenger RNA in the rat striatum by haloperidol and lithium. Biochem. Biophys. Res. Commun., 113: 391-399.

Sivan, S.P., C. Strunk, D.R. Smith, and J.3. Hong 1986 Proenkepha- lin-A gene regulation in the rat striatum: Influence of lithium and haloperidol. Mol. Pharm., 30t186-191.

Springer, J.E., E. Robbins, B.J. Gwag, M.E. Lewis, and F. Baldino, J r . 1991 Non-radioactive detection of nerve growth factor receptor (NGRF) mRNA in the rat brain using in situ hybridization his- tochemistry. J . Histochem. Cytochem., 392t231-324.

Tang, F., E. Costa, and J.P. Schwartz 1983 Increase in proenkephalin mRNA and enkephalin content of rat striatum after daily injec- tion of haloperidol for 2 to 3 weeks. Proc. Natl. Acad. Sci. U.S.A., 80;3841-3844.

Yoshikawa, K., C. Williams, and S.L. Sabol 1984 Rat brain preproen- kephalin messenger RNA: cDNA cloning, primary structure, and distribution in the central nervous system. J . Biol. Chem., 259: 14301-14308.

Young, W.S. 111 1989 Simultaneous use of digoxigenin- and radiola- beled oligodeoxyribonucleotide probes for hybridization his- tochemistry. Neuropeptides, 13:271-275.