Immunofluorescence analysis of antisense oligodeoxynucleotide-mediated ‘knock-down’ of the mouse δ opioid receptor in vitro and in vivo

ELSEVIER

Neuroscience Letters 213 (1996) 205-208

NiUHOSI[NC[ LETiIIIS

Immunofluorescence analysis of antisense oligodeoxynucleotide-mediated 'knock-down' of the mouse t5 opioid receptor in vitro and in vivo J o s e p h i n e L a i a'*, M a u r e e n R i e d l b, L a u r a S. S t o n e c, U l f A r v i d s s o n b, E d w a r d J. B i l s k y a, G e o r g e L. W i l c o x c, R o b e r t E l d e b , F r a n k P o r r e c a a aDepartment qf Pharmacology, University of Arizona Health Sciences Center, Tucson, AZ 85724, USA hDepartment of Cell Biology and Neuroanatomy, University of Minnesota, Minneapolis, MN 55455, USA CDepartment qf Pharmacology, University of Minnesota, Minneapolis, MN 55455, USA Received 26 March 1996; revised version received 19 June 1996; accepted 27 June 1996

Abstract

We have previously used antisense oligodeoxynucleotides (ODN) to the cloned t5 opioid receptor (DOR) to inhibit the antinociceptive response to spinally administered ~ opioid receptor selective agonists in mice. Here we have examined the effect of DOR antisense ODN treatment on the level of DOR expressed in NG 108-15 cells and the spinal cord, through immuno-fluorescence microscopy, to determine the efficiency and selectivity of the antisense ODN-mediated 'knock-down' of the DOR in these tissues. Antisense ODN, but not mismatch control, treatment resulted in a significant reduction in DOR immunoreactivity (-ir) in NG 108-15 cells and spinal cord. Thus, the inhibition of antinociceptive response to intrathecal t5 selective agonists by DOR antisense ODN correlates with the loss of DOR-ir in the superficial layers of the dorsal horn of the spinal cord.

Keywords: Cloned delta opioid receptor; Antisense oligodeoxynucleotides; Antinociception; Mice; Spinal cord; Immunofluorescence microscopy

We have previously explored the use of antisense oligodeoxynucleotide (ODIN) to transiently and reversibly inhibit the function of the cloned t5 opioid receptor (DOR) in mouse brain [4,5,9]. Antisense ODN are short, synthetic, single-stranded DNA whose mode of action is through hybridization to complementary sequences in the target gene or its messenger RNA; the latter results in the formation of an RNA/DNA duplex which disrupts the normal translation of that gene, and/or leads to the degradation of the messenger I~NA by RNaseH (for review, see [6,13]). Consequently, these molecular events result in a reduction in the level, or 'knock-down', of the protein product. In our studies, when an antisense ODN that was specific to the cloned DOR [7,8] was administered intrathecally, it inhibited the spinal antinociceptive response of the opioid ~i receptor selective agonists [DAla2,Glua]deltorphin and [~PenZ,D-PenS]enkephalin [4]. * Corresponding author. Tel.: +1 520 6262147; fax: +1 520 6264182; e-mail: lai @aruba.ccit.adzona.~du

Treatment with a mismatch control ODN, on the other hand, had no significant effect on the antinociceptive response to these drugs, suggesting that the effect of the antisense ODN was sequence-specific. The DOR antisense ODN also had no effect on the antinociceptive response to /x or K selective agonists. The data suggest that the t5 receptor subtype defined by DOR may play a major role in eliciting spinal tit-mediated antinociception. On the other hand, the efficiency with which the DOR antisense ODN suppresses the synthesis of DOR in vivo had not been addressed. The present study sought to evaluate the level oI'DOR in cultured NG 108-15 cells and in spinal cord after antisense or mismatch control ODN treatment to determine the efficiency and selectivity of the antisense ODN-mediated 'knock-down' of the DOR in these tissues. Three ODN were used in this study. (i) A Texas redconjugated ODN (5'-CTG TGG CCC CTT GCC GCT GC3'), which is complementary to the mismatch control sequence for the cloned DOR from NG 108-15 cells (see

0304-3940/96/$12.00 © 1996 Elsevier Science Ireland Ltd. All rights reserved PII S 0 3 0 4 - 3 9 4 0 ( 9 6 ) 1 2 8 8 3 - 4

206

j. Lai et aL /Neuroscience Letters 213 (1996) 205-208

below). The ODN was synthesized by solid phase procedure, conjugated with Texas-red at the 5' end of the ODN and purified by reverse phase HPLC (Midland Certified Reagent Co., Midland, TX). This Texas-red ODN was reconstituted in nuclease-free water and stored in the dark at 4°C. (ii) DOR antisense ODN (5'-GCA CGG GCA GAG GGC ACC AG-3'; complementary to nucleotide 7 - 2 6 of the DOR coding region). (iii) Mismatch control ODN (5'-GCA GCG GCA AGG GGC CAC AG-3'). All three ODN sequences were screened through the GenBank database to ensure that these sequences were not likely to cross-react with other gene sequences found in the NG 108-15 cells. Intrathecal (i.th.) injections of ODN were made by direct lumbar puncture into the subarachnoid space between L5 and L6 of unanesthetized ICR mice using a Hamilton microliter syringe fitted with a 30 gauge needle with a volume of 5 /~1 containing 12.5 /~g of ODN as previously described [4]. Injections were made twice daily for 3 days, after which the animals were perfusion fixed as described previously [2]. The spinal cords were dissected from animals, cut at 14 #m thickness using a cryostat, and mounted onto gel coated slides for immunohistochemical analysis. All studies involving animals were carried out in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health. NG 108-15 cells were cultured in 5% fetal calf serum/ 5 % newborn calf serum/45 % Ham' s F- 12/45% Dulbecco' s modified Eagle's medium/100 U ml -z penicillin/100 /~g ml-I streptomycin. For experiments, the cells were initially seeded onto cover slips at 100 000 ceUs/50 mm 2 Petri dish in normal media 24 h prior to treatment. On the day of the experiments, the cells were washed twice with serumfree medium and treated with the antisense or the mismatch ODN (in serum-free medium) at a final concentration of 5/~M for a total of 4 days to mimic the time course in vivo [5], during which the ODN-containing medium

was refreshed daily. Our initial experiments showed that NG 108-15 cells remained viable in the absence of serum over a period of several days if they were initially established in relatively high density (50% confluency) in serum-containing medium. These culturing conditions were designed to circumvent the lability of ODN in serum during incubation [1]. Cells cultured under these conditions did not actively divide; overnight loading of the cells with eosin-dextran showed that the cells were metabolically viable and expressed a substantial amount of DOR-ir [10]. On day 5, the cells were rinsed briefly and incubated for 24 h with 5/~M of Texas-red ODN (in a total volume of 40/xl with the cover slip inverted). At the end of this incubation, the cells were processed for immunocytochemical analysis. For immunostaining of NG 108-15 cells, the media were aspirated and the cells washed twice with serum-free medium. Immunolabeling was carried out as described previously [11] using an antiserum raised against a fusion protein containing the C-terminal 35 amino acid peptide from the DOR [10]. Immunostaining of mouse spinal cord sections was carried out as previously described using an antiserum raised against peptide corresponding to amino acids 3-17 of DOR (see [2]; 1:1000) and secondary antisera (donkey anti-rabbit) conjugated to cyanine 3.18 (Jackson ImmunoResearch). Confocal laser microscopy was performed using a BioRad 1000 confocal microscope equipped with a krypton/argon laser as described previously [3]. For NG108-15 cells, confocal images were captured with a Leica TCS-4D laser scanning confocal microscope and SCANWARE software. Laser optics for fluorescein included excitation centered at 488 nm and a 30 nm band pass emission filter centered at 530 nm. For Texas-red, the excitation was centered at 568 nm and emission was captured with a 590 nm long pass filter. Image analysis was performed using customized software on a Silicon Graphics IRIS 10/900. Fluorescence intensity of a cell was corrected for background flugrescence. Spe-

Fig. 1. Fluorescenceimaging of NG 108-15 cells pretreatedwith DOR antisense ODN showing (A) accumulationof Texas-redODN and (B) immunofluorescence of DOR. Cells that show particularly intense Texas-redstaining are indicated by arrows. Magnification, 100 x.

J. Lai et al. / Neuroscience Letters 213 (1996) 205-208

207

Fig. 2. Fluorescence imaging of NG 108-15 cells pretreated with mismatch control ODN showing (A) accumulation of Texas-red ODN and (B) immunofluorescence of DOR. Cells that show a high level of Texas-red staining are indicated by arrows. Magnification, 100 × .

cific labeling was defined as a level of in-cell fluorescence which was significantly greater than background. For Texas-red ODN loading, because the NG 108-15 cells do not exhibit significant autofluorescence at the specified wavelength, specific labeling was expressed as arbitrary units of fluorescence (AUF) per unit area (tzm2) above that of out-of-cell background. For immunofluorescence, non-specific labeling was expressed as AUF//zm 2 measured from cells stained with the secondary fluorophorelabeled antibody alone. When the NG 108-15 cells were incubated with the Texas-red ODN to mon:itor the uptake of ODN by these cells, the cells exhibited variable degrees of accumulation of the tagged ODN within a population. The uptake char-

acteristics of the Texas-red ODN were similar in cells which had been incubated with the tagged ODN only (data not shown), or which had been preincubated with the antisense or mismatch ODN (non-tagged) for up to 4 days prior to loading with the tagged ODN (Fig. 1A and Fig. 2A), indicating that the cells were not adversely affected by prolonged exposure to these ODNs. Fluorescence of the ODN was found both in the cytoplasm and the nucleus. Fig. 1A points out several cells which exhibited a robust accumulation of Texas-red ODN consistently had a lower level of DOR antibody staining when compared with at least eight surrounding cells which accumulated much less ODN (Fig. I B). In contrast, in cells that had been pretreated with the mismatch ODN control, significant

Fig. 3. Immunofluorescent conl'ocai images of DOR-ir in the dorsal horn of mouse lumbar spinal cord from untreated mice (naive) or mice that had received i.th. injections of nuclease-free water (vehicle), mismatch control ODN (mismatch), or DOR antisense ODN (antisense). Note the marked reduction in DOR-ir in the superficial dorsal horn after antisense treatment that was not evident after i.th. injections of nuclease-free water or mismatch control ODN. Scale bar = 200 ,~m.

208

J. Lai et al. / Neuroscience Letters 213 (1996) 2 0 5 - 2 0 8

amount of DOR immunostaining could be seen despite the high levels of Texas-red ODN accumulated (Fig. 2). Digitization analysis shows that the mean fluorescence intensity of DOR-ir was 771 + 187 AUF//~m 2 (n = 27) in DOR antisense ODN-pretreated cells and 2950 + 380 AUF/#m 2 (n = 13) in mismatch ODN-treated cells. The mean fluorescence intensity of Texas-red in the same cells was 5250 + 684 AUF//~m2 and 6130 + 1330 AUF//zm 2, respectively. DOR-ir was located primarily in the superficial layers of the dorsal horn of the spinal cord, within lamina I and II (Fig. 3). Following i.th. treatment with DOR antisense ODN, DOR-ir was greatly reduced in lumbar spinal cord (Fig. 3, antisense). No reduction in DOR-ir was seen after i.th. injections of vehicle or mismatch control ODN. No obvious abnormalities were present in these superficial lamina as a result of this treatment and furthermore, serotonin-, [LeuS]enkephalin-, calcitonin gene related peptide-, and # opioid receptor-ir remained at levels similar to that of normal animals (data not shown). The fundamental mode of action of antisense targeting is an inhibition of the synthesis of the target protein, thereby blocking the function of that protein by reducing its abundance in the target tissue. Thus, the most direct means to evaluating antisense-mediated 'knock-down' of the target protein is by measuring the changes in the level of the target protein after antisense treatment. Our data support this mode of action of DOR antisense ODN by demonstrating a significant reduction in the immunoreactivity of the DOR in both the spinal cord and in NG 108-15 cells after antisense but not mismatch ODN treatment. Interestingly, in NG 108-15 cells, the density of DOR was reduced only in the cells which had accumulated a significant amount of tagged ODN, suggesting further that the effect of the antisense ODN depends critically on the intracellular localization of ODN. The DOR immunostaining observed in a population of DOR antisense ODN-treated cells was 26% that of mismatch control ODN-treated. The reduction in the level of the DOR in the NG 108-15 cells is consistent with previous findings that DOR antisense ODN treatment resulted in a reduction in the binding sites for the t5 receptor-selective ligand, [3H][D-Pen2,DPenS]enkephalin in NG 108-15 cells [12] and [3H]naltrindole in mouse brain [5]. That mismatch ODN treatment had no effect on the DOR-ir in either NG 108-15 cells or the spinal cord argues against non-specific effects of ODNs. Furthermore, the dosage of ODN used did not compromise cell viability in vitro or precipitate behavioral toxicity in vivo [4]. The DOR antisense ODN did not result in any changes in the level of a number of endogenous peptides or the tt opioid receptor in the spinal cord. Together, these data indicate that the DOR antisense ODN targets selectively the DOR in the NG 108-15 cells and in the spinal cord, and this selectivity is based on the sequence of the ODN. The significant reduction in the level of DOR-ir in the spinal cord correlates with the inhi-

bition of receptor-mediated antinociception after DOR antisense ODN, but not mismatch control, treatment [4]. These data support further the selectivity of DOR antisense ODN targeting in vivo, that the inhibition of antinociceptive response to i.th. ~ selective agonists is a result of a loss of the DOR in the superficial layers of the dorsal horn of the spinal cord. These data provide further evidence in support of the utility of antisense targeting as an effective tool in neuropharmacological studies. This work is supported by USPHS grants RO1-DA04274, KO2-DA-00145 (GLW), PO1-DA-06284, KO200185 (FP), and R37-DA-06299 (RE). [1] Akhtar, S., Kole, R. and Juliano, R.L., Stability of antisense DNA oligodeoxynucleotide analogs in cellular extracts and sera, Life Sci., 49 (1991) 1793-1801. [2] Arvidsson, U., Dado, R.J., Riedl, M., Lee, J.H., Law, P.Y., Loh, H.H., Elde, R. and Wessendorf, M.W., Delta-opioid receptor immunoreactivity: distribution in brainstem and spinal cord, and relationship to biogenic amines and enkephalin, J. Neurosci., 15 (1995) 1215-1235. [3] Arvidsson, U., Riedl, M., Chakrabarti, S., Lee, J.H., Nakano, A.H., Dado, R.J., Lob, H.H., Law, P.Y., Wessendorf, M.W. and Elde, R., Distribution and targeting of a V-opioid receptor (MOR 1) in brain and spinal cord, J. Neurosci., 15 (1995) 3328-3341. [4] Bilsky, E.J., Berstein, R.N., Pasternak, G.W., Hruby, V.J., Patel, D., Porreca, F. and Lai, J., Selective inhibition of [D-AlaZ,Glu4]del torphin antinociception by supraspinal, but not spinal, administration of an antisense oligodeoxynucleotide to an opioid delta receptor, Life Sci., 55 (1994) PL37-PL43. [5] Bilsky, E.L, Bernstein, R.N., Hruby, V.J., Rothman, R.B., Lai, J. and Porreca, F., Characterization of antinociception to opioid receptor selective agonists following antisense oligodeoxynucleotide-mediated 'knock-down' of opioid receptors in vivo, J. Pharmacol. Exp. Ther., 277 (1996) 491-501. [6] Crooke, S.T., Therapeutic applications of oligonucleotides, Annu. Rev. Pharmacol. Toxicol., 32 (1992) 329-376. [7] Evans, C.J., Keith, D.E. Jr., Mordson, H., Magendzo, K. and Edwards, R.H., Cloning of a delta opioid receptor by functional expression, Science, 258 (1992) 1952-1955. [8] Kieffer, B.L., Befort, K., Gaveriaux-Ruff, C. and Hirth, C.G., The 6 opioid receptor: isolation of a eDNA by expression cloning and pharmacological characterization, Proc. Natl. Acad. Sci. USA, 89 (1992) 12048-12052. [9] Lai, J., Bilsky, E.J., Rothman, R.B. and Porreca, F., Treatment with antisense oligodeoxy-nucleotide to the opioid 5 receptor selectively inhibits 52-agonist antinociception, NeuroReport, 5 (1994) 10491052. [10] Lai, J., Crook, T.J., Payne, A., Lynch, R.M. and Porreca, F., Antisense targeting of delta opioid receptors in NGI08-15 cells: direct correlation between oligodeoxynucleotide uptake and receptor density, J. Pharmacol. Exp. Ther., (1996) submitted. [11] Lynch, R.M., Fogarty, K.E. and Fay, F.S., Modulation of hexokinase association with mitochondria analyzed with quantitative three-dimensional confocal microscopy, J. Cell Biol., 112 (1991) 385-395. [12] Standifer, K.M., Chien, C.C., Wahlestedt, C. and Pasternak, G.W., Selective loss of delta opioid analgesia and binding by antisense oligodeoxynucleotides to a delta opioid receptor, Neuron, 12 (1994) 805-810. [13] Wahlestedt, C., Antisense oligonucleotide strategies in neuropharamacology, Trends Pharmacol. Sci., 15 (1994) 42-46.