acute infection.3,7,11,16 Recrudescence of clinical signs and shedding of infectious virus can occur at later times after reactivation by a variety of stimuli or by.
J Vet Diagn Invest 11:122–126 (1999)
Comparison of PCR, virus isolation, and indirect fluorescent antibody staining in the detection of naturally occurring feline herpesvirus infections Kent M. Burgesser, Stephanie Hotaling, Anita Schiebel, Scott E. Ashbaugh, Steven M. Roberts, James K. Collins Abstract. Cats with clinical signs suggestive of ocular infection with feline herpesvirus type 1 (FHV 1) and cats without such signs were assayed by 3 methods to detect FHV. Comparison of polymerase chain reaction (PCR), virus isolation, and indirect fluorescent antibody staining techniques for the detection of FHV demonstrated higher sensitivity of PCR in detecting this common infectious agent of cats. Compared with PCR, sensitivity and specificity for virus isolation was 49% and 100%, respectively, and those of indirect immunofluorescence were 29% and 96%, respectively. FHV was detected in 13.7% of client-owned cats with conjunctivitis and in 31% of shelter cats with no ocular signs. The use of FHV PCR as a diagnostic test for FHVassociated disease is limited because of the occurrence of healthy carriers.
Feline herpesvirus (FHV) type 1 is a ubiquitous alphaherpesvirus that infects cats. In adult cats, FHV infection has been associated with serious upper respiratory tract disease, including rhinitis, conjunctivitis, keratoconjunctivitis, and corneal disease, with high morbidity and low mortality.16 In kittens, FHV-associated mortality can exceed 60%.9 During the acute stages of infection, large amounts of infectious virus are present in nasal and ocular discharges.13,17 As with many alphaherpesviruses, latency in sensory nerve ganglia occurs commonly after recovery from the acute infection.3,7,11,16 Recrudescence of clinical signs and shedding of infectious virus can occur at later times after reactivation by a variety of stimuli or by corticosteroids. 4,5 During periods of reactivation, smaller amounts of infectious virus are shed for a shorter period of time than during acute primary infection. The techniques used most commonly to detect FHV have included viral isolation in cell culture and indirect fluorescent antibody staining of tissue samples. These methods, although adequate in detecting virus during the acute stages of disease, may not have the required sensitivity to identify FHV in chronic or recrudescent infections where viral shedding is low. Characterization of the FHV genome15,19 has resulted in identification of the thymidine kinase gene as a suitable target for polymerase chain reaction (PCR).6,15,18 PCR techFrom the Departments of Clinical Sciences (Burgesser, Roberts) and Microbiology (Collins) and The Diagnostic Laboratory (Hotaling, Schiebel, Ashbaugh, Collins), College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Ft. Collins, CO 80523. Received for publication March 13, 1998.
niques have been used consistently to detect latent FHV infections in tissues of the head (trigeminal ganglia, optic nerves, olfactory bulbs, and corneas) and inconsistently in oral fauces, salivary gland, lacrimal gland, cerebellum, and conjunctiva.18 PCR has also been useful for detecting FHV from ocular swabs in outbreaks of FHV-associated respiratory disease6 and from tissue specimens from naturally occurring ocular disease.20 The objective of this study was to compare PCR, virus isolation, and indirect fluorescent antibody staining in the detection of naturally occurring FHV infection. The hypothesis was that nested PCR is a more sensitive test than either of the other methods and should be the test of choice for detecting FHV in suspected clinical cases. Materials and methods Sample collection. Conjunctival swabs from 211 cats showing clinical signs of conjunctivitis (ocular discharge, blepharospasm, chemosis, conjunctival hyperemia) or upper respiratory tract disease (nasal discharge) and 84 clinically normal cats were collected using sterile Dacron swabs. One swab was placed in viral transport medium (minimum essential medium with 3% fetal calf serum [FCS]) for virus isolation (VI), and a second swab was placed in Hanks’ balanced salt solution (HBSS) for PCR. Conjunctival epithelial cells were scraped from the palpebral conjunctiva after topical anesthesia and placed on glass slides for indirect fluorescent antibody staining (IFA). Cats were client-owned animals or recent arrivals at a local animal shelter. Samples were stored at 270 C until assayed by VI or PCR, usually within 3–5 days. Virus and cells. FHV-1 (ATCC No. VR-636)a was propagated on Crandell feline kidney cells (CrFK) using Eagle’s minimum essential medium containing 5% FCS and antibiotics (50 U/ml penicillin, 50 mg/ml streptomycin, 1.25 mg/
122
PCR detection of feline herpesvirus infection
ml amphotericin B) at 37 C. Feline herpesvirus isolates obtained from clinical cases of conjunctivitis were propagated similarly. Cultures were inspected every 48 hr postinoculation for a minimum of 10 days, and isolates were identified with a commercial FHV conjugateb using immunofluorescence on suspected infected CrFK cells. Viral DNA was prepared as described for bovine herpesvirus-1 (BHV-1).1 IFA. Slides with conjunctival cells were fixed in acetone for 10 min at room temperature and stained with a mixture of 2 mouse monoclonal antibodies specific for FHVc using an IFA procedure. The pooled primary antibody in ascitic fluid was used at an optimal dilution of 1:2,000, and the secondary rabbit anti-mouse antibodyd was used at a dilution of 1:160. Samples were graded for cellularity, and those samples with inadequate cellular material were excluded from further analysis. PCR. The thymidine kinase (TK) gene was amplified using published primer sequences18 and yielded a 383-bp product. Samples consisted of 1 ml of HBSS into which a conjunctival swab had been placed. Final concentrations in the 50 ml reaction mixture were 0.8 mM MgCl2, 6% glycerol, 0.01% gelatin, 160 mM dNTP, 1.25 U Taq polymerase,e and 15 pmoles of each primer. The mixture was overlaid with 50 ml mineral oil and denatured for 5 min at 94 C followed by 40 cycles of 1 min at 94 C, 1 min at 55 C, and 1.5 min at 72 C. Primers for the nested reaction consisted of upstream primer 59-TGG TCA GAG CGG ATG AAA ATC G-39 and downstream primers 59-CGT CGG GGT GTT CCT CAC ATA C-39. A 1-ml sample of the flanking reaction product was amplified in a 50-ml reaction volume containing 0.8 mM MgCl2, 5% glycerol, 0.1% gelatin, 160 mM dNTP, and 15 pmoles of each primer. After initial denaturation at 94 C for 4 min, 1.25 U Taq polymerase was added, and 40 cycles of 1 min at 94 C, 1 min at 59 C, and 1.5 min at 72 C were completed, resulting in a 267-bp PCR product. Twenty-five microliters of the nested PCR product was visualized by electrophoresis on 1.4% agarose gels with ethidium bromide staining. Positive controls consisting of plasmid DNA containing a 282-bp insert into the expected PCR product and negative controls consisting of sterile water were included in each reaction. PCR sensitivity was determined with purified FHV DNA, and specificity was determined by Southern blot hybridization. The probe for Southern blot analysis was constructed by substituting dTTP with fluorescene dUTP (1:5 ratio of U:T) in the reaction mixture. Additional specificity analysis was performed by amplification of extracted genomic nucleic acid from equine herpesvirus-1 (Kentucky D strain), BHV-1,1 pseudorabies virus (Aujesky strain), canine herpesvirus (ATCC No. VR-552),a feline calici virus (ATCC No. VR-782),a and feline infectious peritonitis virus (TN402 strain). Construction of positive control plasmid. Primers for EcoRI cloning of the flanking PCR amplification product were made by the addition of GAATTC and GAATTC to the 59 ends of the FHV TK upstream and downstream primer, respectively. FHV DNA was amplified using these primers, the amplification product was electrophoresed through 1.4% agarose, and the 396-bp fragment was gel purified using Quiaex II.f The DNA was digested with EcoRIg and elec-
123
Figure 1. Agarose gel electrophoresis of flanking (a) and nested (b) PCR products. Lane 150.25 fg DNA; lane 252.5 fg DNA; lane 3525.0 fg DNA; lane 45250 fg DNA; lane 552.5 pg DNA; lane 6525.0 pg DNA; lane 75250 pg DNA; lane 852.5 ng DNA; lane 9525 ng DNA; lane 105negative control; lane 115kb ladder.
trophoresed, and the 390-bp insert fragment was gel purified. SP64 plasmid DNAh was digested with EcoRI,g treated with calf intestinal phosphatase (CIP),g and electrophoresed, and the 3,388-bp vector fragment was gel purified. Following overnight ligation of the vector and insert using T4 DNA ligase,h the ligants were used to transform Escherichia coli TOP 10 (F9R) cells according to standard protocols.8 Recombinants were selected on LB/Amp (75 mg/ml) plates, and single colonies were grown in liquid LB/Amp cultures for miniprep and PCR analysis.8 A clone was identified that generated the correct EcoRI digestion products and amplified the correct FHV flanking and nesting PCR products. A DNA fragment to be used for insertion into the FHV TK of the identified clone at a site internal to the nesting PCR primers (the EcoRI site) was generated from BHV-1 DNA. Plasmid pGC12 containing the glycoprotein C gene (gC) from BHV-1 was digested with NaeI and electrophoresed, and a 260-bp gC fragment was gel purified. The identified clone miniprep DNA was digested with EcoRV,h treated with CIP, and electrophoresed through 1.4% agarose, and the 3,388-bp vector fragment was gel purified. These 2 fragments were ligated and transformed, and the recombinants were selected as described above. A clone was selected that generated the correct 260-bp EcoRI digestion product and the correct 666-bp flanking and 550-bp nesting PCR products. A large LB/Amp culture of this clone (renamed pFHV1) was grown from which was prepared a large quantity of purified plasmid DNA for use as a positive control in FHV PCR reactions.
Results Sensitivity and specificity of FHV PCR. The FHV PCR flanking reaction detected 25 fg of DNA (170 genome equivalents) showing the expected 383-bp band (Fig. 1a). The nesting reaction provided a 10fold increase in sensitivity, detecting 2.5 fg of DNA (17 genome equivalents) showing the expected 267-bp band (Fig. 1b). Nucleic acid from 6 unrelated viruses were tested
124
Burgesser et al.
Table 1. Results of PCR, VI, and IFA on 295 clinical ocular samples from cats. Clinical signs of FHV Test result
Figure 2. Agarose gel electrophoresis of positive control plasmid (lanes 6, 7) and isolated FHV DNA (lanes 9, 10).
using the FHV flanking and nesting primers. DNA from these viruses was not amplified (data not shown). A total of 37 FHV strains that had been previously isolated at the Colorado State University Veterinary Diagnostic Laboratory were recultured and tested for detectability by PCR, and all generated the expected bands (data not shown). Positive control plasmid. Figure 2 shows a representative gel demonstrating the difference between the positive control plasmid DNA following nested PCR and isolated viral DNA. The positive control plasmid DNA (1 pg) and 2 negative control specimens were included in every PCR test to assure sensitivity and lack of contamination Detection of FHV from clinical cases. The results of PCR, VI, and IFA from the 295 cats tested are summarized in Tables 1 and 2. Overall, PCR was positive in 18.6% of the samples, and VI and IFA were positive in 9.2% and 5.1% of the samples, respectively. PCR and VI were in agreement on 49% (27/55) of the positive results and 100% (240/240) of the negative results. PCR and IFA were in agreement on 29% (16/ 55) of the positive results and 96.25% (231/240) of the negative results. Of the samples from cats with clinical signs consistent with FHV infection, 13.7% were positive by PCR, 8.5% were positive by VI, and 7.6% were positive by IFA. Additionally, 8 samples that were negative for FHV by VI were positive for feline calici virus. PCR was positive for FHV in 7 of these 8 samples. Samples from cats that were clinically normal showed that 30.9% were PCR positive for FHV. Of these same samples, 10.7% were positive for FHV by both VI and IFA. Southern blot hybridization was used to verify 9 of the 55 positive PCR samples (data not shown). Discussion This study demonstrates the superior capability of PCR, compared to the most commonly utilized techniques, in the detection of FHV in naturally acquired infection. PCR was almost twice as likely to detect FHV as was VI, previously the most sensitive test available. The improved detection rate would suggest that PCR is the test of choice for the detection of FHV. Recent reports have described the utility of PCR in the
Positive PCR Positive VI Positive IFA Negative Total samples
Present
29 18 16 182 211
Absent
(13.7%) (8.5%) (7.5%) (86.3%)
26 9 9 57 84
(30.9%) (10.7%) (10.7%) (67.8%)
detection of FHV from ocular tissues,14,20 but direct comparison between methods in the detection of clinical disease has not yet been made. Proposed explanations for the low sensitivity of VI include envelope destruction during virus transport, inactivation of viral infectivity by enzymes or antibodies present in saliva or tears, and shedding of immature virus that is not fully infective.18 Detection of FHVspecific immunofluorescence may be complicated by low amounts of viral antigen and interference by hostderived antibody and mediators of inflammation, increasing the potential for false-negative results. Additionally, IFA is a very subjective test, and results may vary depending on the individual interpreting the slides.14 False-positive results are likely due to high levels of background fluorescence. Because PCR does not require infectious virus, or accessible viral antigen for antibody binding, greater sensitivity would be expected. The percentage of positive samples from clinically affected cats was lower in this study than has been previously reported (13.7% vs. 54%).20 Possible explanations include regional differences in the prevalence of FHV and differences in sampling technique. In the present study, swabs of the conjunctival surface were used as samples for PCR, whereas in a previous study20 conjunctival snip biopsies were used. Because a larger and deeper tissue sample is obtained via biopsy, the probability of collecting infected cells would Table 2. Comparison of PCR results with VI and IFA for infection in cats. PCR Test results
Positive
Negative
Total
VI Positive Negative Total samples
27 28 55
0 240 240
27 268 295
16 39 55
9 231 240
25 270 295
IFA Positive Negative Total samples
PCR detection of feline herpesvirus infection
be greater, which may be important in diagnosing cases of chronic or recrudescent disease when levels of virus shedding are small. The proportion of clinically normal individuals that would be positive for FHV using this sampling technique may also be high. The clinical signs of conjunctivitis in cats include conjunctival hyperemia, intermittent blepharospasm, and serous to mucopurulent ocular discharge. Depending on the etiologic agent, upper respiratory signs such as sneezing, nasal discharge, and possibly coughing may also be present. Pathogens that have been associated with feline conjunctivitis in addition to FHV include Chlamydia psittaci, Mycoplasma spp., and other less common bacteria and viruses.10 The clinical signs of conjunctivitis are not pathognomotic for any etiologic agent but represent a reaction pattern of the conjunctiva to insult. Noninfectious causes of conjunctivitis may also result in similar clinical signs. There was no attempt made to identify other causes of conjunctivitis. The large number of PCR-positive results from samples without clinical signs suggestive of FHV infection likely represents detection of latent or reactivated virus or an early stage of disease before clinical signs are apparent. Like most alphaherpesviruses, FHV becomes latent in sensory ganglia3,7,11,16 and other tissues.18 FHV replicates in the corneas of experimentally infected cats,12,13 and latency has been demonstrated by PCR in the cornea and conjunctiva.18 FHV DNA was present in the corneas of 46% of clinically normal cats in a recent study.20 These samples were from cats obtained from an animal shelter, as were the 84 clinically normal cats in this study. The environment in which the cat lives may therefore be an important in the etiology of FHV infection. However, many cat owners complain of periodic ‘‘cold’’ symptoms in their pets, and never present the cat for evaluation. Such anecdotal reports may represent FHV infection, and the increased sensitivity of PCR could help provide evidence of a possible etiologic agent. The finding of samples positive by PCR that were also positive for infectious feline calici virus demonstrates the ability of PCR to detect FHV in cases of coinfection with other viruses. In this study, increased detection of FHV by PCR was demonstrated through direct comparison with the more common laboratory identification methods. This increased sensitivity would make PCR ideal for epidemiologic studies, where population prevalence data could be ascertained. Clinically, the same increased sensitivity may sacrifice diagnostic specificity. The significance of a positive PCR result may be of question in diagnosis of the underlying cause of conjunctivitis, especially acute disease. In light of the large number of apparently normal cats that were positive for FHV by PCR and the fact that viral recrudescence
125
occurs, a positive PCR result may represent inapparent shedding or reactivation of latent virus secondary to a different primary cause. Additionally, it is not known if modified-live virus vaccine strains of FHV will result in a positive PCR result, and if so, how long after vaccination positive results can be anticipated. Treatment with topical or systemic antivirals therefore may not be indicated. When other causes of conjunctivitis or keratitis have been eliminated, especially in chronic or recurrent disease, a positive PCR result may be more meaningful. Because latent virus cannot be completely eliminated, a positive PCR test for FHV may give the clinician insight as to the possibility of recurring clinical signs at some time in the future if conditions are right to induce viral activation from sites of latency. Acknowledgements This work was supported by grants from the American College of Veterinary Ophthalmologists Research Fund and the Colorado State University Foundation Ophthalmology Research Fund.
Sources and manufacturers a. American Type Culture Collection, Rockville, MD. b. Veterinary Medical Research and Development, Pullman, WA. c. Dr. Jim Guy, School of Veterinary Medicine, North Carolina State University, Raleigh, NC. d. Sigma Chemical Co., St. Louis, MO. e. Perkin Elmer Cetus Corp., Norwalk, CT. f. Quiagen, Santa Clarita, CA. g. Life Technologies, Gaithersburg, MD. h. Promega Corp., Madison, WI.
References 1. Ashbaugh SE, Thompson KE, Belknap EB, et al.: 1997, Specific detection of shedding and latency of bovine herpesvirus 1 and 5 using a nested polymerase chain reaction. J Vet Diagn Invest 9:387–394. 2. Chowdhury SI: 1995, Molecular basis of antigenic variation between the glycoprotein C of respiratory bovine herpesvirus 1 (BHV-1) and neurovirulent BHV-5. Virology 213:558–568. 3. Gaskel R, Moore PE, Goddard LE, et al.: 1985, Isolation of felid herpesvirus I from the trigeminal ganglia of latently infected cats. J Gen Virol 66:391–394. 4. Gaskel RM, Povey RC: 1973, Re-excretion of feline viral rhinotracheitis virus following corticosteroid treatment. Vet Rec 93:204–205. 5. Gaskel RM, Povey RC: 1977, Experimental induction of feline rhinotracheitis virus re-excretion in FVR recovered cats. Vet Rec 100:128–133. 6. Hara M, Fukuyama M, Suzuki Y, et al.: 1996, Detection of feline herpesvirus 1 DNA by the nested polymerase chain reaction. Vet Microbiol 48:345–352. 7. MacEwen EG, Hess PW: 1995, Evaluation of the effect of immunomodulation of the feline eosinophilic granuloma complex. Am J Vet Res 46:263–265. 8. Maniatis T, Fritsch EF, Sambrook J: 1982, Molecular cloning, a laboratory manual. Cold Springs Harbor Laboratory Press, Cold Springs Harbor, NY. 9. Mickman MA, Reubel GH, Hoffman DE, et al.: 1994, An epi-
126
10.
11.
12.
13.
14. 15.
Burgesser et al.
zootic of feline herpesvirus type 1 in a large specific-pathogenfree cat colony and attempts to eradicate the infection by identification and culling of carriers. Lab Anim 28:320–329. Nasisse MP: 1991, Feline ophthalmology. In: Veterinary ophthalmology, ed. Gelatt KN, 2nd ed., pp. 534–538. Lea & Febiger, Philadelphia, PA. Nasisse MP, Cavis BJ, Guy JS, et al.: 1992, Isolation of feline herpesvirus 1 from the trigeminal ganglia of acutely and chronically infected cats. J Vet Intern Med 6:102–103. Nasisse MP, English RV, Tompkins MB, et al.: 1995, Immunologic, histologic and virologic features of herpesvirus-induced stromal keratitis in cats. Am J Vet Res 56:51–55. Nasisse MP, Guy JS, Davidson MG, et al.: 1989, Experimental ocular herpesvirus infection in cats: sites of virus replication, clinical features and effect of corticosteroid administration. Invest Ophthalmol Visual Sci 30:1758–1768. Nasisse MP, Weigler BJ: 1997, The diagnosis of ocular feline herpesvirus infection. Vet Comp Ophthalmol 7:44–51. Nunberg JH, Wright DK, Cole GE, et al.: 1989, Identification
16.
17.
18.
19. 20.
of the thymidine kinase gene of feline herpesvirus: use of degenerate oligonucleotides in the polymerase chain reaction to isolate herpesvirus gene homologs. J Virol 63:3240–3249. Povey RC: 1979, A review of feline rhinotracheitis (feline herpesvirus 1 infection). Comp Immunol Microbiol Infect Dis 2: 373–387. Reubel GH, George JW, Barlough JE, et al.: 1992, Interaction of acute feline herpesvirus-1 and chronic feline immunodeficiency virus infection in experimentally infected specific-pathogen free cats. Vet Immunol Immunopathol 35:95–119. Reubel GH, Ramos RA, Hickman MA, et al.: 1993, Detection of active and latent feline herpesvirus 1 infections using polymerase chain reaction. Arch Virol 132:409–420. Rota PA, Maes RK, Ruyechan WT: 1986, Physical characterization of feline herpesvirus-1. Virology 154:168–179. Stiles J, McDermott M, Bigsby D, et al.: 1997, Use of nested polymerase chain reaction to identify feline herpesvirus in ocular tissue from clinically normal cats and cats with corneal sequestra or conjunctivitis. Am J Vet Res 58:338–342.
