Enzymatic amplification of latent pseudorabies virus nucleic acid sequences

Journal of Virological Methods, 34 (1991) 4555 0 1991 Elsevier Science Publishers B.V. / All rights reserved / 0166-0934/91/$03.50 ADONIS 016609349100413F

45

VIRMET 01208

Enzymatic amplification of latent pseudorabies virus nucleic acid sequences J.R. Lokensgard,

D.G. Thawley and T.W. Molitor

Department of Clinical and Population Sciences, College of Veterinary Medicine, University of Minnesota, Saint Paul, Minnesota, U.S.A. (Accepted 9 April 1991)

summary To investigate various aspects of the latency of pseudorabies virus in swine (PRV, suid herpesvirus 1) we developed in vitro nucleic acid amplification methods based upon the polymerase chain reaction. Primers flanking a 156-bp region of the pseudorabies virus gp II gene were annealed to purified PRV DNA as well as DNA isolated from the trigeminal ganglia of swine latently infected with PRV and subjected to PCR amplification. Following amplification, 100 fg of PRV DNA was visualizable on stained gels and 1 fg (equivalent to 6 viral genome copies) was detectable when amplification was combined with blot hybridization. PRV-specific DNA sequences which remained undetectable by direct blot hybridization assays were amplified to levels visualizable on ethidium-bromide-stained gels in 5 of 5 experimental latently infected animals. In addition, oligonucleotide primers specific for a 223-bp region of the PRV immediate-early gene (IE 180) were capable of amplifying overlapping latency associated transcripts (LATs), via a cDNA intermediate, in 6 of 6 latently infected swine. These nucleic acid amplification methods should be applicable to the investigation of PRV latency, and gene expression during latency and reactivation, in which few cells harbor latent virus. Pseudorabies virus; Latency; Polymerase chain reaction

Correspondence to: Dr. T.W. Molitor, Dept. of Clinical and Population Teaching Hospitals, 1365 Gortner Avenue, Saint Paul, MN 55108, U.S.A.

Sciences, 225 Veterinary

46

Introduction A characteristic of alphaherpesviruses, including pseudorabies virus, is the ability of their genome to persist in nervous tissue in a latent form following primary infection. This latent state may reactivate later to produce virus which can subsequently infect susceptible animals. Perpetuation of PRV in swine herds is insured through transitions between latency and reactivation. Latent carriers have long been suspected as an important factor in the spread of pseudorabies and their presence could complicate attempts to control and eradicate (McFerran et al., 1984). A fundamental understanding of the mechanisms involved in the establishment, maintenance, and reactivation of viral latency may be required in order to control effectively the spread of pseudorabies. Analyses of latent PRV infections have been difficult because of the frequent failure of conventional methods to reactivate latent infections in swine and obstacles encountered with early DNA hybridization studies, e.g. cross reactivity (McFarlane et al., 1986). Viruses such as PRV which establish latent infections are particularly attractive targets for the polymerase chain reaction in which the superb sensitivity of PCR is employed to detect the extremely small quantity of viral nucleic acid present in latently infected tissues. PCR amplification techniques for pseudorabies virus DNA have been published (Belak et al., 1989; Jestin et al., 1990; Maes et al., 1990); however, these reports have not concentrated on the latent state nor have they described the amplification of latency-associated transcripts, also known as ‘LATs’. Pseudorabies virus has been shown to be transcriptionally active during latency and the presence of these LATs should allow PCR RNA amplification via a cDNA intermediate. PRV LATs are encoded from a region which overlaps the immediate-early gene but in an antisense orientation (Rock et al., 1988; Cheung, 1989a; Lokensgard et al., 1990; Priola et al., 1990). At present, the significance of this latency-associated transcription remains obscure. No sequence data is available for PRV LATs; however, primers specific for the 3’ end of the IE gene should also amplify regions of the LATs. This approach has been utilized to amplify the latency-associated transcripts of HSV (Lynas et al., 1989). The objectives of the present study were twofold: first, to develop a PCR DNA amplification technique specifically applicable to studies of PRV latency in swine and, secondly, to develop an RNA PCR technique capable of detecting This RNA PCR technique should be latency-associated transcription. appropriate for the investigation of viral gene expression, and transcript quantitation during latency and reactivation in which few cells harbor latent virus.

47

Materials and Methods Tissue and nucleic acid preparation

Mixed breed pigs were inoculated intranasally with 1 x lo5 PFU of PRV, strain S-62 (USDA challenge strain). At various times representing the acute (4-7 days p.i.) and latent (30, 60, or 90 days p.i.) phases of infection animals were euthanized and their trigeminal ganglia were immediately excised and flash frozen in liquid nitrogen. Virus was not recovered from tonsillar swabs or tissues collected from pigs sacrificed at 30, 60, or 90 days p.i. by standard virus isolation techniques. DNA was extracted from the frozen ganglia by treating the homogenized tissue with 0.1 mg/ml proteinase K and 0.5% SDS in 100 mM NaCl, 25 mM EDTA, and 100 mM Tris, pH 8.0. The digestion was followed by phenol, phenol-chloroform, and chloroform extraction and ethanol precipitation. RNA was extracted from frozen ganglia by the guanidine thiocyanate method (Chirgwin et al., 1977). After isolation, the RNA was treated with 1 U of RNase-free DNase l/100 ,ul RNA in 10 mM MgCl*, 50 mM Tris, pH 7.4, (37°C 1 h), extracted with phenol-chloroform, and precipitated with ethanol to remove any contaminating DNA. Nucleic acid concentrations were determined spectrophotometrically prior to PCR assay. Primer design

Oligonucleotides were synthesized by the standard phosphoramidite procedure on an automated synthesizer (Applied Biosystems, Foster City, CA). The DNA sequence data necessary for primer construction was obtained from the published sequences of the PRV gp II (Robbins et al., 1987) and immediate-early (IE) genes (Cheung, 1989b). The genomic positions of the gp II and IE genes as well as sequences of the primers and probes synthesized are shown in Fig. 1A. Primer 1 extended from nucleotides 613 to 632, numbered with reference to the translational start site (ATG) on the coding strand of PRV gp II, primer 1.1 extended from nucleotides 608 to 627 on the same strand, and primer 2 extended from nucleotides 803 to 784 on the template strand; both primers are 55% G + C and flank a region which is 67.3% G + C. Primer 4 (same polarity as the IE coding strand) and primer 5 (same polarity as the IE template strand) flanked a 223-bp region of the PRV IE gene and extended from nucleotides 37143733 and 3957-3976 of BamHl fragment I, respectively (Fig. 1B). Oligonucleotides 3 and 6 were 30-mers synthesized for use as probes. The application of oligonucleotide probes circumvents the potential problems associated with using probes which contain primer sequences (i.e., hybridization to nonspecifically incorporated primers). Oligonucleotide 3 extended from nucleotides 686 to 716 on the coding strand of PRV gp II and is 66.6% G+ C. Oligonucleotide 6 extended from nucleotides 3769-3798 of BamHl fragment I and has the same polarity as IE mRNA. Probes were made by T4 kinase

48 A. Map Units:

Oi”

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.6

0.9

1.0 1

PRV Genome:

B.

gp II amplification:

primers and probe

627

606

*I 716

696

632

613

I

primer Xl

S-CAGGAGATCACGGACCTGAT-3

primerXl.1

S-CCGTGCAGGAGATCACGGAC3

primer #Z

S-CCGATCTTGGTGTAGGTGTC-3

oligo#3

SAGGTGACCGCCTTCGACCGCGACGAGAACG3

LAT amplification: primers and probe

IE

180 mRNA

primer#4

S-TCATCGTGCTGGACACCATC-3

wimer#S

S-CGTGTAGCGCACGTTGTCCT-3’

oligo#6 S-CTACCACGTCTACGTCCGCGCCCGCCTGGAB

Fig. 1. Schematic representation of the PRV genome. A: the genome of PRV consists of approximately 144 kb of double-stranded DNA which contains two unique segments (Ul and Us), the smaller of which is bounded by inverted repeat sequences (Ir and Tr). The locations of the gp II and IE genes are indicated. An exact map position of the LATs has not yet been determined; however, it is known that they overlap the IE gene. B: primer and probe locations and sequences. The gp-II-specific primers were used for DNA PCR and IE-specific primers were used for RNA PCR.

49

labeling of the internal oligonucleotide specific activity of 2 x 10’ cpm/pg.

at the 5’ end using [Y-~*P]ATP to a

Amplification of viral genomic sequences

The gp II amplification reaction was conventional and contained 0.2 PM each 20 base primer (i.e., primer 1 or 1.1 and 2), 200 PM each dNTP, 10 mM Tris, pH 7.6, 50 mM KCl, 0.001% gelatin, 1.5 mM MgC12 (optimal Mg*+ concentration between 1 and 4 mM was determined empirically) and 2.5 U of Amplitaq polymerase. A 5 min initial denaturation of target DNA (1 pg) at 95°C was performed. The gp II reaction profile consisted of a 94°C denaturation step for 1 min, annealing at 57°C for 1 min, and extension at 72°C for 1 min. In testing 32,35 and 40 cycles, only 40 cycles provided enough amplification to detect latent PRV gp II sequences on ethidium-bromidestained gels. Reactions were performed in an automated thermocycler (Coy Laboratories, Ann Arbor, MI). Reverse transcription and RNA amplification

Synthesis of first strand cDNA was carried out on 1 pg total ganglionic RNA from latently infected animals extracted by the guanidine thiocyanate procedure. After heating to 90°C for 5 min to denature secondary structure, the RNA and random hexamers (100 pmol) were allowed to anneal during cooling to room temperature. A mixture consisting of 1 mM each dNTP, 50 mM KCl, 10 mM Tris, 5 mM MgC12, 1 mM DTT, 1 U/p1 RNasin, and 8 U of AMV reverse transcriptase was added and the reaction was incubated at room temperature for 10 min, to extend the hexameric primers, before raising the temperature to 42°C for 1 h. Before being subjected to PCR, the cDNA mix was heated at 99°C for 5 min and quickly chilled on ice to denature the RNAcDNA hybrids and inactivate reverse transcriptase. The PCR components consisting of 0.2 ,uM each 20-base primer, 50 mM KCI, 10 mM Tris, 0.01% gelatin, and 2.5 U Amplitaq were added (final Mg*+ in the PCR mix was 2.0 mM). In some instances, 1 U of Perfect Match polymerase enhancer (Stratagene, La Jolla, CA) was employed to accentuate amplification of the PRV-specific target and eliminate spurious products, although its use was not absolutely required. The mixture was amplified by 40 cycles of denaturation (1 min), annealing at 56°C (1 min), and polymerization at 72°C (1 min). To insure complete denaturation of the long products, due to the high G + C content of the region (76%; Cheung, 1989b), the first three cycles were denatured at 97”C, followed by melting at 94°C for the next 37 cycles. It was also mandatory to use a 3:l mixture of 7-deaza dGTP (c’dGTP):dGTP to allow for amplification (McConlogue et al., 1988).

50

Confirmation of product specificity Ethidium-bromide-stained DNA bands at 196 bp and 263 bp were presumptively PRV-specific and confirmation of this specificity was obtained by Southern blotting. Briefly, gels (2% NuSieve:l% GTG agarose) were soaked in 1.0 M NaCl/O.S M NaOH (2 x 15 min) and 0.5 M Tris, pH 7.4/1.5 M NaCl(2 x 15 min) followed by capillary transfer in 20 x SSC (1 x SSC is 0.15 M NaCl, 0.015 M Na citrate, pH 7.0) to Hybond-N nylon membranes. Membranes were prehybridized for 2 h at 50°C in a solution containing 50% formamide, 5 x SSC, 5 x Denhardt solution (1 x is 0.02% BSA, 0.02% ticoll, 0.02% polyvinyl-pyrrolidone), 0.5% SDS, 0.5 mg/ml denatured ssDNA, and 25 mM sodium phosphate. Hybridization was performed overnight at T,- 10°C in 28% formamide, 2 x SSC, 1 x Denhardt solution, 0.5% SDS, 0.2 mg/ml denatured ssDNA, and 20 mM sodium phosphate with 2 x lo7 cpm/ membrane of oligonucleotide 3 or 6. Washings were performed 3 times for 15 min in 2 x SSC, 0.1% SDS at 45”C, followed by one wash for 5 min at T,- 10°C (i.e., 68.3”C for oligonucleotide 3, 69.7”C for oligonucleotide 6, in 2 x SSC, 0.1% SDS), exposure time was 3 h.

gp II DNA detection limits Initially, two 20-base primers, 1 and 2, which flanked a 151-bp region of PRV gp II, were synthesized (Fig. 1B). This particular primer set amplified a 191-bp product specific for PRV but also amplified an approximately 3 lo-bp product from normal swine cellular DNA. The swine-specific band was present under high stringencies and every reaction condition tested. Primer 1.1, a 20mer identical to primer 1 but moved 5 bases upstream, was synthesized to eliminate the swine-specific band. Using primers 1.1 and 2 a single 196-bp fragment of PRV gp II could be amplified in the presence of swine DNA. Sensitivity was quantified as the minimum amount of PRV DNA which could be detected following PCR amplification in the presence of 1 pg of uninfected swine cellular DNA. The gp-II-specific product migrated at 196 bp and was easily detectable out to 100 fg (Fig. 2A). The negative control was 1 ,ug of uninfected swine DNA and the positive control was PRV grown in Vero cells. A buffer control (i.e., all reagents except target) was always included to check for carryover from previous amplifications. Slot-blot hybridization with the internal probe following PCR further increased sensitivity to 1 fg of PRV DNA, in 1 pg of total DNA (Fig. 2B). Detection of latent PRV DNA Regardless of how sensitive the assay, the important question is whether viral DNA sequences from latently infected tissues can be detected. Following

51

PRV DNA1

1

100 PS

196bp-

neg. B.C.

Fig. 2. Sensitivity titration of PRV DNA amplification. A: ethidium bromide stained gel. The lanes contain 100,10, and 1 pg, and 100, 10, and 1 fg of purified PRV DNA in a background of 1 pg of uninfected swine cellular DNA. The PRV-specific product migrates at 196 bp and is apparent up to 100 fg. The neg. lane is uninfected swine cellular DNA and the pos. is PRV grown in Vero cells. A buffer control is always included to control for carryover. B: slot-blot hybridization with the internal probe further increased sensitivity to 1 fg of PRV DNA.

amplification using primers 1.1 and 2, a 196-bp PCR product was demonstrated in all 5 latently infected animals tested (Fig. 3A). PRV DNA was not detected in any of these samples by direct blot hybridization without prior PCR amplification (data not shown). The corresponding Southern blot with the internal oligonucleotide probe hybridized under conditions of high stringency (T,- 1O’C) proved the specificity of the gp II band (Fig. 3B), and the use of blotting to increase sensitivity was not required. Ampilfxation

of PRV-latency-associated

transcripts

The results of the RNA PCR technique applied to tissues infected latently with PRV are presented in Fig. 4. The animals were the same ones known to .be latently infected by gp II DNA PCR. Following cDNA synthesis and PCR amplification a 263-bp product was visualized for 6 of 6 latently infected animals (30-90 days p.i.; Fig. 4A). The positive control was ganglionic RNA obtained from an animal during the acute phase of infection (i.e., when IE was being synthesized). To amplify this region, it was also necessary to use a 3:l mixture of 7-deaza dGTP (c dGTP):dGTP. PCR using primers 4 and 5 was unsuccessful when standard dNTPs were employed (data not shown). Note that the bands seen on

52

Fig. 3. Detection of latent PRV DNA. A: ethidium bromide stained gel. Lanes 242, 255 and 260 are I-pg ganglionic DNA samples taken from latently infected animals at 30 days pi. and subjected to PCR amplification. B60 and C90 are samples from animals sacrificed 60 and 90 days p.i., respectively. Pos. and neg. controls are as described in the legend to Fig. 2. B: the corresponding Southern blot probed with the internal oligonucleotide 3 proves the band is PRV gp-II-specific.

the gel (Fig. 4A) do not appear as intense as those of the gp II region. It is known that PCR products containing c7dGTP simply do not stain as efficiently with ethidium bromide as those containing standard dNTPs (Innis, 1990). When the gel was probed in a Southern blot assay with oligonucleotide 6, the band was determined to be PRV-specific (Fig. 4B). To confirm that any signal was derived from RNA and not contaminating DNA, control reactions were

Fig. 4. Amplification of a PRV LAT region. A: ethidium bromide stained gel. Lanes 242, 255 and 260 are I-pg ganglionic RNA samples taken from latently infected animals at 30 days p.i. and subjected to PCR amplilication with primers 4 and 5. A30, B60 and C90 are samples from animals sacrificed 30, 60 and 90 days p.i., respectively. The pos. control lane is 1 pg of ganglionic RNA extracted from an animal during the acute phase of infection (i.e., a time when IE synthesis is known to occur). The neg. control is 1 pg of ganglionic RNA taken from an uninfected animal. B: the corresponding Southern blot with the internal oligonucleotide 6 proves that the 263-bp band is PRV IE(LAT)-specific.

53

carried out without the addition of reverse transcriptase and no evidence of bands was detected (data not shown), even though Taq DNA polymerase has been reported to have minimal RT activity (Tse and Forget, 1990).

Discussion Until recently, in situ hybridization was the only viable method to study the PRV latent state; it is, -however, technically demanding and labor intensive. There are now enzymatic ~pli~~ation methods for very precise molecular studies of gene expression during latency and reactivation of PRV. We have now demonstrated that latent PRV genomic sequences, as well as viral gene expression during latency, can be detected by enzymatic amplification using the polymerase chain reaction. Previous reports of PCR. amplification techniques for the detection of pseudorabies virus (Belak et al., 1989; Jestin et al., 1990; Maes et al., 1990) have not described the amplification of LATs. Separate enzymatic amplification methods for both PRV-specific DNA and RNA are communicated here. The PCR DNA assay employs primer sequences derived from an essential glycoprotein gene (gp II) and encompasses much greater sensitivity than is required to simply detect the latent viral genome. The level of sensitivity is at least 1 fg when DNA PCR is combined with nucleic acid hybridization techniques. With a genome size of 144 kbp, 1 fg is equivalent to 6 genomic copies of PRV. In addition, PCR methods described here were able to identify latent viral nucleic acid sequences in tissues which were undetectable by other methods. The development of this technique to amplify latent PRV nucleic acids will find an extensive range of research applications, for example, the dete~nati~n of the infection status of single serologic reactors in otherwise clean herds (Annelli et al., 1991). The high G-i-C content of our IE (LAT)-specific PCR template (76%; Cheung, 1989b) posed some difficulties for nucleic acid amplification, presumably due to the presence of secondary structures which may hinder annealing and/or extension of the primers. The use of 7-deal-~-deoxy~anosine (c7dGTP) in the dNTP mix was found to be essential for ampli~cation of this region. PCR with c7dGTP incorporates this structure-destabilizing base analog into the amplified DNA, precludes secondary structure formation and amplifies difficult templates more efficiently than PCR with dGTP alone (McConlogue et al., 1988). The main downfall of incorporating c7dGTP into PCR products is that they do not stain as efficiently with ethidium bromide because adjacent base stacking is diminished (Innis, 1990). Previous studies have indicated that PRV expresses one immediate-early (IE) poly (A)+ mRNA of 5.6 kb (Campbell and Preston, 1987; Cheung, 1988). PRV latency-associated transcripts (LATs) which are encoded from an approximately 11-kb region overlapping the IE gene in an antisense orientation have been reported, no other region of the viral genome is transc~ptionally active at this time (Cheung, 1989a; Lokensgard et al., 1990; Priola et al., 1990). The

54

region encoding LATs has also recently been shown to be transcriptionally active late in the replicative cycle during productive infection (Cheung, 1990). Still, the transcripts present in latently infected tissues can be considered as markers for latency by PCR if no other viral transcripts (e.g. gp II) are present. In this report, we have shown the feasibility of amplifying PRV LATs by using primers specific for the IE gene. Assay by PCR should also be useful for quantifying the amount of PRV DNA in latently infected tissues and the level of expression of specific genes (e.g. thymidine kinase or major glycoproteins) during the establishment; maintenance, and reactivation of the latent state. The assays can be made quantitative by including an appropriate synthetic RNA as an internal standard (Wang et al., 1989); this work is now in progress. Acknowledgement

This work was funded in part by a grant from the United States Department of Agriculture (86~CRSR-2-2929).

References Annelli, J.F., Morrison, R.B., Goyal, S.M., Bergland, M.E., Mackey, W.J. and Thawley, D.J. (1991) Pig herds having a single reactor to serum antibody tests to Aujeszky’s disease virus. Vet. Rec. 128,49-53. Belak, S., Ballagi-Pordany, A., Flensburg, J. and Virtanen, A. (1989) Detection of pseudorabies virus DNA sequences by polymerase chain reaction. Arch. Virol. 108, 279-286. Campbell, M.M. and Preston, C.M. (1987) DNA sequences which regulate the expression of the pseudorabies virus immediate early gene. Virology 157, 307-316. Cheung, A.K. (1988) Fine mapping of the immediate-early gene of the Indiana-Funkhauser strain of pseudorabies virus. J. Virol. 62, 4763-4766. Cheung, A.K. (1989a) Detection of pseudorabies virus transcripts in trigeminal ganglia of latently infected swine. J. Virol. 63, 2908-2913. Cheung, A.K. (1989b) DNA nucleotide sequence analysis of the immediate-early gene of pseudorabies virus. Nucleic Acids Res. 17, 46374646. Cheung, A.K. (1990) The BamHl J Fragment (0.706 to 0.737 map units) of pseudorabies virus is transcriptionally active during viral replication. J. Virol. 64, 977-983. Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter, W.J. (1979) Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18, 52945299. Innis, M.A. (1990) PCR with 7-deaza-2’-deoxyguanosine triphosphate. In: M.A. Innis, D.H. Gelfand, J.J. Sninsky and T.J. White (Eds), PCR Protocols, Academic Press, San Diego, pp. 54 59. Jestin, A., Foulon, Th., Pertuiset, B., Blanchard, Ph. and Labourdet, M. (1990) Rapid detection of pseudorabies virus genomic sequences in biological samples from infected pigs using polymerase chain reaction DNA amplification. Vet. Microbial. 23, 317-328. Lokensgard, J.R., Thawley, D.G. and Molitor, T.W. (1990) Pseudorabies virus latency: restricted transcription. Arch. Virol. 110, 129-136. Lynas, C., Laycock, K.A., Cook, S.D., Hill, T.J., Blyth, W.A. and Maitland, N.J. (1989) Detection of herpes simplex virus type 1 gene expression in latently and productively infected mouse

55 ganglia using the polymerase chain reaction. J. Gen. Virol. 70, 2345-2355. Maes, R., Beisel, C., Spatz, S. and Thacker, B. (1990) Polymerase chain reaction amplification of pseudorabies virus DNA from acutely and latently infected cells. Vet. Microbial. 24, 281-295. McConlogue, L., Brow, M.A.D. and Innis, M.A. (1988) Structure-independent DNA amplification by PCR using 7-deaza-2’-deoxyguanosine. Nucleic Acids Res. 16, 9869. McFarlane, R.G., Thawley, D.G. and Solarzano, S.F. (1986) Detection of latent pseudorabies virus in porcine tissues using a DNA hybridization dot-blot assay. Am. J. Vet. Res. 47, 23292336. McFerran, J.B., McCracken, R.M. and Dow, C. (1984) The role of the carrier pig in the epidemiology of Aujeszky’s disease. Curr. Top. Vet. Med. Animal Sci. 27, 403-415. Priola, S.A., Gustafson, D.P., Wagner, E.K. and Stevens, J.G. (1990) A major portion of the pseudorabies virus genome is transcribed in the trigeminal ganglia of latently infected pigs. J. Virol. 64, 4755-4760. Robbins, A.K., Darney, D.J., Wathen, M.W., Whealy, M.E., Gold, C., Watson, R.J., Holland, L.E., Weed, S.D., Levine, M., Glorioso, J.C. and Enquist, L.W. (1987) The pseudorabies virus gI1 gene is closely related to the gB glycoprotein gene of herpes simplex virus. J. Virol. 61,269 l2701. Rock, D.L., Hagemoser, W.A., Osorio, F.A. and McAllister, H.A. (1988) Transcription from the pseudorabies virus genome during latent infection. Arch. Virol. 98, 99-106. Tse, W.T. and Forget, B.G. (1990) Reverse transcription and direct amplification of cellular RNA transcripts by Taq polymerase. Gene 88, 293-296. Wang, A.M., Doyle, M.V. and Mark, D.F. (1989) Quantitation of mRNA by the polymerase chain reaction. Proc. Natl. Acad. Sci. USA 86, 9717-9721.