- Email: [email protected]
Mixed ocular infections identify strains of herpes simplex virus for use in genetic studies
Journal of Viroiogical Methods, 35 (1991) 127-135 0 1991 Elsevier Science Publishers B.V. All rights reserved. / 0166-0934/91/$03.50
127
VIRMET 01242
Mixed ocular infections identify strains of herpes simplex virus for use in genetic studies Curtis R. Brandt Departments of Ophthalmology and Medical ~~crobioiogy and Imrn~o~ogy, University of Wisconsin-Mad~on, Madison, Wisconsin U.S.A. (Accepted
14 June 1991)
Summary
Studies on the genetic mechanisms involved in the ocular virulence of herpes simplex virus (HSV) require the careful selection of parental strains. We used the technique of mixed ocular infection in vivo to identify strains of HSV for use in genetic studies. A pair of viruses (OD4 and 994) were identified that cause significantly more severe ocular disease when mixed together and used to infect the eyes of Balb/c mice compared to each strain when used alone. The mixed infection with 0D4 and 994 did not result in increased neurovirulence. The technique of mixed ocular infections provides a sensitive screen to identify strains of virus that can act s~er~sticaily to cause more severe disease. Marker transfer can then be used to map the genes involved. Herpes simplex virus; Ocular virulence; Neurovirulence; Mixed infections; Viral genetics
Temperature-sensitive (ts), host range (hr), drug resistance, and plaque morphology mutants of herpes simplex virus (HSV) have proven very useful in determining the function of many HSV genes (reviewed by Subak-Sharpe and Timbury, 1977). These types of mutants are generally not useful in studying the role of viral genes in virulence, since their effects are unpredictable. M~ntaining the ts mutant is difficult in an animal host and drug-resistant Correspondence to: CR. Brandt, Dept. of ~hthaImology, University of Wisconsin, Medical Sciences Center, 1300 University Avenue, Madison, WI 53706, U.S.A.
128
mutants vary in their virulence properties, making it difficult to absolutely define the role of the genes in virulence (Field and Darby, 1980; Klein et al., 1981; Larder and Darby, 1984, 1985). Host range mutants, by definition, do not grow in certain cell types, and it is difficult to convincingly show that lack of virulence is not due to the inability to grow in certain host cells. Changes in plaque morphology are not necessarily correlated with altered virulence, and changes in cell culture properties are not predictive of virulence changes (Wheeler, 1964; Dix et al., 1983). Null mutants (viruses that do not express specific genes) have been constructed and used in studies of virulence (Meignier et al., 1988; Jacobson et al., 1989; Turk et al., 1989; Gordon et al., 1990; Brandt et al., 1991), but the genes that can be studied are limited to those that are dispensable for growth of the virus in cell culture systems. Some of these mutants are essentially hr mutants. Mutants that are avirulent have been isolated fortuitously (Thompson and Stevens, 1983; Becker et al., 1986; Taha et al., 1989) but selection for avirulence is difficult and requires plaque purification of a mutant in the absence of selection and subsequent testing in an animal. Another strategy is to construct recombinant viruses. However, individual recombinants must be isolated and characterized prior to testing for virulence properties, mapping of the recombination site is laborious, the location of the recombination event(s) cannot be controlled, and may be biased in location on the HSV genome (Centifanto-Fitzgerald et al., 1982; Thompson and Stevens, 1983; Oakes et al., 1986; Pogue-Geile and Spear, 1986; Halliburton et al., 1987; Javier et al., 1987; Batra and Brown, 1989). Recently, a new strategy has been devised that makes the use of recombinant viruses more precise and less time consuming. Two viruses differing in virulence properties are identified. The genome of one, usually the more virulent virus, is cloned. Purified intact viral DNA from the less virulent parent is mixed with a restriction fragment from the cloned DNA of the virulent parent, transfected into cells and the resulting virus stock containing wild type and recombinant viruses is inoculated directly into animals. Any alteration in virulence suggests that the cloned fragment carries a virulence determinant and can be studied further (Thompson et al., 1983). This variation of marker transfer has the added advantage of also providing the map location of the gene involved. Javier et al. (1986) have shown that a mixed in vivo infection with two avirulent viruses can result in increased virulence. Mixed in vivo infection thus provides a simple method for identifying parental strains for genetic studies of virulence. The use of mixed ocular infections to isolate viruses with both increased ocular and neurovirulence was described recently (Brandt and Grau, 1990). In this communication, we further describe the usefulness of mixed in vivo ocular infections to identify a pair of viruses which are avirulent when used alone but together act synergistically to cause significantly increased ocular disease. In contrast to our previous communication (Brandt and Grau, 1990), the viruses described below do not cause death from encephalitis in a mixed infection.
129
Materials and Methods Cell cultures
High-titer stocks of virus were prepared and titered in Vero cells as described previously (Grau et al., 1989; Brandt and Grau, 1990). Virus
The virulence properties and origins of the viruses used in this study have been described previously (Grau et al., 1989). Table 1 summarizes the relevant properties of the viruses. Animal inoculation
Procedures for animal inoculation have been described previously (Grau et al., 1989; Brandt and Grau, 1990). Briefly, 34week-old female Balb/c mice (Harlan Sprague Dawley, Indianapolis, IN) were anesthetized by 2.5% halothane inhalation. The right cornea of each animal was scratched three times vertically and three times horizontally with a 30-gauge needle and a 5-,ul drop of Dulbecco’s Modified Eagles Medium (DME) with 2% fetal bovine serum (FBS) containing 1 x lo5 PFU of each virus to be tested (2 x lo5 total in mixing experiments), was placed on the cornea. After 30 s the inoculum was removed using a sterile cotton swab and the animals were allowed to recover from the anesthesia. The scratches extend through the cornea1 epithelium but do not penetrate the stroma (unpublished results). All animal procedures were carried out according to NIH guidelines. Disease scoring
Mice were examined microscopically for disease on the days indicated. The severity of blepharitis, vascularization of the cornea, stromal keratitis, and death due to encephalitis were scored as we described previously (Grau et al., 1989; Brandt and Grau, 1990). Epithelial disease (corneal) ulceration is difficult to distinguish from the eye scratches. This results in unreliable scores and we TABLE 1 Properties Strain
of viral strain&!
HSV type
Mean peak disease scores Stromal disease
0D4 994
1 1
Vascularization
0 0
“Data from Grau et al., 1989; bmean + standard deviation.
Blepharitis
Neurovirulence
0.44*0.s3b 1.2SkO.86
-
130
therefore do not report epithelial disease. We also scored the animals for extraocular disease (ulcers extending beyond the eyelids) but none was observed (data not shown). Statistical
analysis
Methods for Analysis of Variance (ANOVA) (Grau et al., 1989; Brandt and Grau, 1990).
were described
previously
Results The virulence properties of a number of HSV-1 isolates were characterized previously (Grau et al., 1989). To determine if these isolates could interact synergistically, we carried out mixed ocular infections with several pairwise combinations. One pair of viruses (OD4/994) appeared to result in more severe ocular disease than either parent alone. We therefore quantitated the ocular disease caused by the OD4/994 mixture. The results are shown in Fig. 1. As described previously, 0D4 and 994 alone caused very little if any ocular disease. Blepharitis was first detected on day 3 in all 3 groups but rapidly cleared in animals infected with either OD4 or 994. In animals infected with the OD4/994 mixture, blepharitis continued to increase, peaking at a mean score of 2.5 on day 7. Blepharitis was still severe on day 14 post-infection (p.i.) in the OD4/994-infected animals (mean score 1.85) but had healed in the animals that were separately infected with either OD4 or 994. Vascularization of the cornea was never seen in ODbinfected animals. A transient vascularization of the cornea developed in the mice infected with strain 994 beginning on day 3, peaking on day 7 (mean score 0.8), and then clearing by day 11. This transient vascularization was due to swelling of the blood vessels at the cornea1 limbus and penetration of new vessels only a short distance from the corneal-limbal boundary. By day 11, the vascular swelling had subsided and the vessels that had penetrated a short distance into the cornea were no longer visible. The vessels may have been present but no longer contained blood and were therefore not scored. In our previous studies with strain 994, the mice did not develop even a transient vascularization (Grau et al., 1989). The reason for this difference is not clear, but the mice in the mixed infection group (OD4/994) clearly developed permanent vascularization and blindness. The blood vessels in the OD4/994 infected animals penetrated further into the cornea than the transient vascularization seen in the 994 infected animals and these vessels remained filled with blood. Neither OD4 nor 994 induced any stromal keratitis. Animals infected with the OD4/994 mixture developed stromal disease which was first visible on day 7 and continued to increase until day 14 (mean score 2.0) when scoring was halted. The increased severity of ocular disease in the animals infected with the
131
OD4/994 mixture was also reflected in the mean peak disease scores (MPDS, Grau et al., 1989) shown in Fig. 2. Statistical analysis of the MPDS indicated that the mixed OD4/994 infection resulted in significantly more severe BLEPHARITIS 4-iXM
0
2
-I3
, 2
0
14
Diys ‘Pos!infd?konf 2
.
.
1.1 4
6
Days
I 8
.
I IO
*
, 12
.
5 14
I 12
.
1 14
Postinfection
4 STROYAL
-1)
. 0
I 2
.
I 4
KERATITIS
.
I 6
.
I 8
.
9 10
.
Days Postinfection Fig. 1. Time course for developing ocular disease. Mice were infected with 1 x 10’ PFU of 0134 or 994, or a mixture of 1 x IO5PFU of each virus (2 x lo5 PFU total) and the severity of blepharitis, vascularization of the cornea, and stromal keratitis, was scored on the days indicated.
132
Fig. 2. Mean peak disease scores. The peak disease scores for each animal used in the experiment shown in Fig. 1 were averaged and analyzed statistically by ANOVA.
blepharitis (P < 0.05) and stromal keratitis (P -=c0.05) than either virus alone. The MPDS for vascularization induced by the OD4/994 mixture was significantly more severe (P -=z0.05) than strain 0D4 but was not significantly higher than the vascularization induced by strain 994 (P > 0.05). The increased virulence of the OD4/994 mixture was also reflected in the percentage of animals that showed some evidence of ocular disease (Table 2). None of the animals infected with strains OD4 and 994 showed evidence of TABLE 2 Percentage
of animals with ocular disease”
Group
No. animals
Stromal keratitis (%)
Vascularization (%)
Blepharitis (%)
0D4 994 OD4/994 mixture
10 11 6
0 0 50
0 55 67
20% 82 100
“No. of animals showing disease at any time during the experiment/animals
infected.
133
stromal keratitis compared to 50% of the animals infected with the OD4/994 mixture. None of the animals infected with strain OD4 developed vascularization while about 55% and 67% of the animals infected with strain 994 and the OD4/994 mixture, respectively, developed vascularization. Only 20% of the ODCinfected animals developed blepharitis while 82% of the 994 and 100% of the OD4/994_infected animals developed blepharitis.
Discussion
The use of mixed infections provides a relatively quick, easy, and powerful tool to identify parental virus strains for use in genetic studies of virulence. An additional advantage is that several strains can be tested at the same time. As many as six strains of viruses were tested in a single experiment (C. Brandt, unpublished results). Once the initial screening has been completed and potentially interesting pairs of viruses identified, there are two strategies that can be used to map the gene or genes involved. One strategy is to isolate recombinants and the other is to use marker transfer methods. The isolation of recombinants is difficult as these viruses are naturally occurring isolates that do not necessarily have phenotypically distinct traits useful in selecting recombinants. Brute force screening of hundreds of isolated plaques might be required. The analysis of recombinants presents additional problems. Many of the isolates we have tested, including OD4 and 9W (unpublished results), and others (Brandt and Grau, 1990), are HSV type 1 and have very similar restriction maps. This results in imprecision in mapping recombination boundaries, making it difficult to localize the gene or genes involved, and large numbers of recombinants. must be analyzed. Because of the similarity in the restriction maps, a region of the genome could be scored as parental even though it is recombinant. The analysis of recombinants is time consuming. An additional consideration is that marker transfer must still be used to confirm the genes involved in the altered virulence. It is therefore more productive to use mixed infections to screen and identify isolates, then clone the genomes of the parental strains, and use marker transfer and subsequent in vivo screening to identify and map the genes involved in virulence. These studies are in progress with OD4 and 994. In the original description of mixed infections for virulence studies, the viruses were inoculated via the footpad (Javier et al., 1986). Footpad inoculation is useful for studying changes in neurovirulence or neuroinvasiveness, but is less useful in studying changes in the severity of disease at the site of inoculation. Mixed ocular infections are useful for studying changes in neurovirulence (Brandt and Grau, 1990), but they have the added advantage that changes in the severity of ocular disease can be identified. In our ocular infection model, the severity of stromal keratitis, vascularization of the cornea, blepharitis, and extraocular disease can be scored. Ocular infection thus provides more information concerning virulence properties and is potentially a
134
more sensitive indicator of changes in virulence. Our results show that the ocular model is useful for identifying viruses that interact synergistically to increase the severity of both localized disease (Fig. 1) as well as neurovirulence (Brandt and Grau, 1990). Both complementation and recombination have been shown to occur following mixed infections (Javier et al., 1986; Sederati et al., 1988; Brandt and Grau, 1990). Although we have not determined whether complementation, recombination, or both occurred in the OD4/994 infections, it would not be productive to do so since the mixed infection is only being used to screen parental HSV strains. As mentioned above, the next step is to use marker transfer (i.e., construct recombinants) to map the gene or genes involved.
Acknowledgements This work was supported by a grant from the United Service (EY07736) and by an unrestricted grant from Blindness, Inc. to the Department of Ophthalmology. I Dorene R. Grau for expert technical assistance and preparing the manuscript.
States Public Health Research to Prevent would like to thank Bernadette Bull for
References Batra, SK. and Brown, SM. (1989) Analysis of unselected HSV-I McRae/HSV-2 HG52 recombinants demonstrates preferential recombination between intact genomes and restriction endonuclease fragments containing an origin of replication. Arch. Virol. 105, l-1 3. Becker, Y., Hadar, J., Tabor, E., Ben-Hur, T., Raibstein, I., Rosen, A. and Darai, G. (1986) A sequence in HpaI-P fragment of herpes simplex virus-l DNA determines intraperitoneal virulence in mice. Virology 149, 255-259. Brandt, C.R. and Grau, D.R. (1990) Mixed infection with HSV-I generates recombinants with increased ocular and neurovirulence. Invest. Ophthalmol. Vis. Sci. 31, 2214-2223. Brandt, C.R., Kintner, R.L., Pumfery, A.M:, Visalli, R.J. and Grau, D.R. (1991) The herpes simplex virus ribonucleotide reductase is required for ocular virulence. J. Gen. Virol. (in press). Centifanto-Fitzgerald, Y.M., Yamaguchi, T., Kaufman, H.E., Tognon, M. and Roizman, B. (1982) Ocular disease pattern induced by herpes simplex virus is genetically determined by a specific region of viral DNA. J. Exp. Med. 155, 475489. Dix, R.D., McKendall, R.R. and Baringer, J.R. (1983) Comparative neurovirulence of herpes simplex virus type 1 strains after peripheral or intracerebral inoculation of BALB/c mice. Infect. Immun. 40, 103-l 12. Field, H.J. and Darby, G. (1980) Pathogenicity in mice of strains of herpes simplex virus which are resistant to acyclovir in vitro and in vivo. Antimicrob. Agents Chemother. 17, 209-216. Gordon, Y.J., Romanowski, E., Balouris, C. and Araullo-Cruz, T. (1990) A herpes simplex type 1 ICPO deletion mutant demonstrates diminished pathogenicity during acute ocular infection in different host animals. Invest. Ophthalmol. Vis. Sci. 31, 681688. Grau, D.R., Visalli, R.J. and Brandt, C.R. (1989) Herpes simplex virus stromal keratitis is not titerdependent and does not correlate with neurovirulence. Invest. Ophthalmol. Vis. Sci. 30, 2474 2480. Halliburton, I.W., Honess, R.W. and Killington, R.A. (1987) Virulence is not conserved in recombinants between herpes simplex virus types 1 and 2. J. Gen. Virol. 68, 1435-1440.
I35 Jacobson, J.G., Leib, D.A., Goldstein, D.J., Bogard, CL., Schaffer, P.A., Weller, SK. and Coen, D.M. (1989) A herpes simplex virus ribonucleotide reductase deletion mutant is defective for productive acute and reactivatable latent infections of mice and for replication in mouse cells. Virology 173, 276283. Javier, R.T., Sedarati, F. and Stevens, J.G. (1986) Two avirulent herpes simplex viruses generate lethal recombinants in vivo. Science 234, 746748. Javier, R.T., Thompson, R.L. and Stevens, J.G. (1987) Genetic and biological analyses of a herpes simplex virus intertypic recombinant reduced specifically for neurovirulence. J. Virol. 61, l9781984. Klein, R.J., Friedman-Kein, A.J. and DeStefano, E. (1981) Pathogenesis of experimental skin infections induced by drug-resistant herpes simplex virus mutants. Infect. Immun. 34, 693-701. Larder, B.A. and Darby, G. (1984) Virus drug resistance: mechanisms and consequences. Antiviral Res. 4, 142. Larder, B.A. and Darby, G. (1985) Selection and characterization of acyclovir-resistant herpes simplex virus type I mutants inducing altered DNA polymerase activities. Virology 130, 262271. Meignier, B., Longnecker, R., Mavromara-Nazos, P., Sears, A.E. and Roizman, B. (1988) Virulence of and establishment of latency by genetically engineered deletion mutants of Herpes simplex virus I. Virology 162, 251-254. Oakes, J.E., Gray, W.L. and Lausch, R.N. (1986) Herpes simplex virus type I DNA sequences which direct spread of virus from cornea to central nervous system. Virology 150, 513- 517. Pogue-Geile, K.L. and Spear, P.G. (1986) Enhanced rate of conversion or recombination of markers within a region of unique sequence in the herpes simplex virus genome. J. Virol. 58, 704708. Sederati, F., Javier, R.T. and Stevens, J.G. (1988) Pathogenesis of a lethal mixed infection in mice with two non-neuroinvasive herpes simplex virus strains. J. Virol. 62, 3037-3039. Subak-Sharpe, J.H. and Timbury, M.C. (1977) Genetics of Herpes viruses. In: H. Fraenkel-Conrat and R.R. Wagner (Eds), Comprehensive Virology. Academic Press, New York, pp. 89-131. Taha, M.Y., Clements, G.B. and Brown, S.M. (1989) A variant of herpes simplex virus type 2 strain HG52 with a 1.5 Kb deletion in RL between 0 to 0.02 and 0.81 to 0.83 map units is nonneurovirulent for mice. J. Gen. Virol. 70, 705716. Thompson, R.L. and Stevens, J.G. (1983) Biological characterization of a herpes simplex virus intertypic recombinant which is completely and specifically non-neurovirulent. Virology 131, 171-179. Thompson, R.L., Wagner, E.K. and Stevens, J.G. (1983) Physical location of a herpes simplex virus type 1 gene function(s) specifically associated with a IO million-fold increase in HSV neurovirulence. Virology I3 I, 180-l 92. Turk, S.R., Kik, M.A., Birch, G.M., Chiego, D.J. Jr. and Shipman, C. Jr. (1989) Herpes simplex virus type I ribonucleotide reductase null mutants induce lesions in guinea pigs. Virology 173, 733-735. Wheeler, C.E. Jr. (1964) Biologic comparison of a syncytial and a small giant cell-forming strain of herpes simplex. J. Immunol. 93, 749-756.
127
VIRMET 01242
Mixed ocular infections identify strains of herpes simplex virus for use in genetic studies Curtis R. Brandt Departments of Ophthalmology and Medical ~~crobioiogy and Imrn~o~ogy, University of Wisconsin-Mad~on, Madison, Wisconsin U.S.A. (Accepted
14 June 1991)
Summary
Studies on the genetic mechanisms involved in the ocular virulence of herpes simplex virus (HSV) require the careful selection of parental strains. We used the technique of mixed ocular infection in vivo to identify strains of HSV for use in genetic studies. A pair of viruses (OD4 and 994) were identified that cause significantly more severe ocular disease when mixed together and used to infect the eyes of Balb/c mice compared to each strain when used alone. The mixed infection with 0D4 and 994 did not result in increased neurovirulence. The technique of mixed ocular infections provides a sensitive screen to identify strains of virus that can act s~er~sticaily to cause more severe disease. Marker transfer can then be used to map the genes involved. Herpes simplex virus; Ocular virulence; Neurovirulence; Mixed infections; Viral genetics
Temperature-sensitive (ts), host range (hr), drug resistance, and plaque morphology mutants of herpes simplex virus (HSV) have proven very useful in determining the function of many HSV genes (reviewed by Subak-Sharpe and Timbury, 1977). These types of mutants are generally not useful in studying the role of viral genes in virulence, since their effects are unpredictable. M~ntaining the ts mutant is difficult in an animal host and drug-resistant Correspondence to: CR. Brandt, Dept. of ~hthaImology, University of Wisconsin, Medical Sciences Center, 1300 University Avenue, Madison, WI 53706, U.S.A.
128
mutants vary in their virulence properties, making it difficult to absolutely define the role of the genes in virulence (Field and Darby, 1980; Klein et al., 1981; Larder and Darby, 1984, 1985). Host range mutants, by definition, do not grow in certain cell types, and it is difficult to convincingly show that lack of virulence is not due to the inability to grow in certain host cells. Changes in plaque morphology are not necessarily correlated with altered virulence, and changes in cell culture properties are not predictive of virulence changes (Wheeler, 1964; Dix et al., 1983). Null mutants (viruses that do not express specific genes) have been constructed and used in studies of virulence (Meignier et al., 1988; Jacobson et al., 1989; Turk et al., 1989; Gordon et al., 1990; Brandt et al., 1991), but the genes that can be studied are limited to those that are dispensable for growth of the virus in cell culture systems. Some of these mutants are essentially hr mutants. Mutants that are avirulent have been isolated fortuitously (Thompson and Stevens, 1983; Becker et al., 1986; Taha et al., 1989) but selection for avirulence is difficult and requires plaque purification of a mutant in the absence of selection and subsequent testing in an animal. Another strategy is to construct recombinant viruses. However, individual recombinants must be isolated and characterized prior to testing for virulence properties, mapping of the recombination site is laborious, the location of the recombination event(s) cannot be controlled, and may be biased in location on the HSV genome (Centifanto-Fitzgerald et al., 1982; Thompson and Stevens, 1983; Oakes et al., 1986; Pogue-Geile and Spear, 1986; Halliburton et al., 1987; Javier et al., 1987; Batra and Brown, 1989). Recently, a new strategy has been devised that makes the use of recombinant viruses more precise and less time consuming. Two viruses differing in virulence properties are identified. The genome of one, usually the more virulent virus, is cloned. Purified intact viral DNA from the less virulent parent is mixed with a restriction fragment from the cloned DNA of the virulent parent, transfected into cells and the resulting virus stock containing wild type and recombinant viruses is inoculated directly into animals. Any alteration in virulence suggests that the cloned fragment carries a virulence determinant and can be studied further (Thompson et al., 1983). This variation of marker transfer has the added advantage of also providing the map location of the gene involved. Javier et al. (1986) have shown that a mixed in vivo infection with two avirulent viruses can result in increased virulence. Mixed in vivo infection thus provides a simple method for identifying parental strains for genetic studies of virulence. The use of mixed ocular infections to isolate viruses with both increased ocular and neurovirulence was described recently (Brandt and Grau, 1990). In this communication, we further describe the usefulness of mixed in vivo ocular infections to identify a pair of viruses which are avirulent when used alone but together act synergistically to cause significantly increased ocular disease. In contrast to our previous communication (Brandt and Grau, 1990), the viruses described below do not cause death from encephalitis in a mixed infection.
129
Materials and Methods Cell cultures
High-titer stocks of virus were prepared and titered in Vero cells as described previously (Grau et al., 1989; Brandt and Grau, 1990). Virus
The virulence properties and origins of the viruses used in this study have been described previously (Grau et al., 1989). Table 1 summarizes the relevant properties of the viruses. Animal inoculation
Procedures for animal inoculation have been described previously (Grau et al., 1989; Brandt and Grau, 1990). Briefly, 34week-old female Balb/c mice (Harlan Sprague Dawley, Indianapolis, IN) were anesthetized by 2.5% halothane inhalation. The right cornea of each animal was scratched three times vertically and three times horizontally with a 30-gauge needle and a 5-,ul drop of Dulbecco’s Modified Eagles Medium (DME) with 2% fetal bovine serum (FBS) containing 1 x lo5 PFU of each virus to be tested (2 x lo5 total in mixing experiments), was placed on the cornea. After 30 s the inoculum was removed using a sterile cotton swab and the animals were allowed to recover from the anesthesia. The scratches extend through the cornea1 epithelium but do not penetrate the stroma (unpublished results). All animal procedures were carried out according to NIH guidelines. Disease scoring
Mice were examined microscopically for disease on the days indicated. The severity of blepharitis, vascularization of the cornea, stromal keratitis, and death due to encephalitis were scored as we described previously (Grau et al., 1989; Brandt and Grau, 1990). Epithelial disease (corneal) ulceration is difficult to distinguish from the eye scratches. This results in unreliable scores and we TABLE 1 Properties Strain
of viral strain&!
HSV type
Mean peak disease scores Stromal disease
0D4 994
1 1
Vascularization
0 0
“Data from Grau et al., 1989; bmean + standard deviation.
Blepharitis
Neurovirulence
0.44*0.s3b 1.2SkO.86
-
130
therefore do not report epithelial disease. We also scored the animals for extraocular disease (ulcers extending beyond the eyelids) but none was observed (data not shown). Statistical
analysis
Methods for Analysis of Variance (ANOVA) (Grau et al., 1989; Brandt and Grau, 1990).
were described
previously
Results The virulence properties of a number of HSV-1 isolates were characterized previously (Grau et al., 1989). To determine if these isolates could interact synergistically, we carried out mixed ocular infections with several pairwise combinations. One pair of viruses (OD4/994) appeared to result in more severe ocular disease than either parent alone. We therefore quantitated the ocular disease caused by the OD4/994 mixture. The results are shown in Fig. 1. As described previously, 0D4 and 994 alone caused very little if any ocular disease. Blepharitis was first detected on day 3 in all 3 groups but rapidly cleared in animals infected with either OD4 or 994. In animals infected with the OD4/994 mixture, blepharitis continued to increase, peaking at a mean score of 2.5 on day 7. Blepharitis was still severe on day 14 post-infection (p.i.) in the OD4/994-infected animals (mean score 1.85) but had healed in the animals that were separately infected with either OD4 or 994. Vascularization of the cornea was never seen in ODbinfected animals. A transient vascularization of the cornea developed in the mice infected with strain 994 beginning on day 3, peaking on day 7 (mean score 0.8), and then clearing by day 11. This transient vascularization was due to swelling of the blood vessels at the cornea1 limbus and penetration of new vessels only a short distance from the corneal-limbal boundary. By day 11, the vascular swelling had subsided and the vessels that had penetrated a short distance into the cornea were no longer visible. The vessels may have been present but no longer contained blood and were therefore not scored. In our previous studies with strain 994, the mice did not develop even a transient vascularization (Grau et al., 1989). The reason for this difference is not clear, but the mice in the mixed infection group (OD4/994) clearly developed permanent vascularization and blindness. The blood vessels in the OD4/994 infected animals penetrated further into the cornea than the transient vascularization seen in the 994 infected animals and these vessels remained filled with blood. Neither OD4 nor 994 induced any stromal keratitis. Animals infected with the OD4/994 mixture developed stromal disease which was first visible on day 7 and continued to increase until day 14 (mean score 2.0) when scoring was halted. The increased severity of ocular disease in the animals infected with the
131
OD4/994 mixture was also reflected in the mean peak disease scores (MPDS, Grau et al., 1989) shown in Fig. 2. Statistical analysis of the MPDS indicated that the mixed OD4/994 infection resulted in significantly more severe BLEPHARITIS 4-iXM
0
2
-I3
, 2
0
14
Diys ‘Pos!infd?konf 2
.
.
1.1 4
6
Days
I 8
.
I IO
*
, 12
.
5 14
I 12
.
1 14
Postinfection
4 STROYAL
-1)
. 0
I 2
.
I 4
KERATITIS
.
I 6
.
I 8
.
9 10
.
Days Postinfection Fig. 1. Time course for developing ocular disease. Mice were infected with 1 x 10’ PFU of 0134 or 994, or a mixture of 1 x IO5PFU of each virus (2 x lo5 PFU total) and the severity of blepharitis, vascularization of the cornea, and stromal keratitis, was scored on the days indicated.
132
Fig. 2. Mean peak disease scores. The peak disease scores for each animal used in the experiment shown in Fig. 1 were averaged and analyzed statistically by ANOVA.
blepharitis (P < 0.05) and stromal keratitis (P -=c0.05) than either virus alone. The MPDS for vascularization induced by the OD4/994 mixture was significantly more severe (P -=z0.05) than strain 0D4 but was not significantly higher than the vascularization induced by strain 994 (P > 0.05). The increased virulence of the OD4/994 mixture was also reflected in the percentage of animals that showed some evidence of ocular disease (Table 2). None of the animals infected with strains OD4 and 994 showed evidence of TABLE 2 Percentage
of animals with ocular disease”
Group
No. animals
Stromal keratitis (%)
Vascularization (%)
Blepharitis (%)
0D4 994 OD4/994 mixture
10 11 6
0 0 50
0 55 67
20% 82 100
“No. of animals showing disease at any time during the experiment/animals
infected.
133
stromal keratitis compared to 50% of the animals infected with the OD4/994 mixture. None of the animals infected with strain OD4 developed vascularization while about 55% and 67% of the animals infected with strain 994 and the OD4/994 mixture, respectively, developed vascularization. Only 20% of the ODCinfected animals developed blepharitis while 82% of the 994 and 100% of the OD4/994_infected animals developed blepharitis.
Discussion
The use of mixed infections provides a relatively quick, easy, and powerful tool to identify parental virus strains for use in genetic studies of virulence. An additional advantage is that several strains can be tested at the same time. As many as six strains of viruses were tested in a single experiment (C. Brandt, unpublished results). Once the initial screening has been completed and potentially interesting pairs of viruses identified, there are two strategies that can be used to map the gene or genes involved. One strategy is to isolate recombinants and the other is to use marker transfer methods. The isolation of recombinants is difficult as these viruses are naturally occurring isolates that do not necessarily have phenotypically distinct traits useful in selecting recombinants. Brute force screening of hundreds of isolated plaques might be required. The analysis of recombinants presents additional problems. Many of the isolates we have tested, including OD4 and 9W (unpublished results), and others (Brandt and Grau, 1990), are HSV type 1 and have very similar restriction maps. This results in imprecision in mapping recombination boundaries, making it difficult to localize the gene or genes involved, and large numbers of recombinants. must be analyzed. Because of the similarity in the restriction maps, a region of the genome could be scored as parental even though it is recombinant. The analysis of recombinants is time consuming. An additional consideration is that marker transfer must still be used to confirm the genes involved in the altered virulence. It is therefore more productive to use mixed infections to screen and identify isolates, then clone the genomes of the parental strains, and use marker transfer and subsequent in vivo screening to identify and map the genes involved in virulence. These studies are in progress with OD4 and 994. In the original description of mixed infections for virulence studies, the viruses were inoculated via the footpad (Javier et al., 1986). Footpad inoculation is useful for studying changes in neurovirulence or neuroinvasiveness, but is less useful in studying changes in the severity of disease at the site of inoculation. Mixed ocular infections are useful for studying changes in neurovirulence (Brandt and Grau, 1990), but they have the added advantage that changes in the severity of ocular disease can be identified. In our ocular infection model, the severity of stromal keratitis, vascularization of the cornea, blepharitis, and extraocular disease can be scored. Ocular infection thus provides more information concerning virulence properties and is potentially a
134
more sensitive indicator of changes in virulence. Our results show that the ocular model is useful for identifying viruses that interact synergistically to increase the severity of both localized disease (Fig. 1) as well as neurovirulence (Brandt and Grau, 1990). Both complementation and recombination have been shown to occur following mixed infections (Javier et al., 1986; Sederati et al., 1988; Brandt and Grau, 1990). Although we have not determined whether complementation, recombination, or both occurred in the OD4/994 infections, it would not be productive to do so since the mixed infection is only being used to screen parental HSV strains. As mentioned above, the next step is to use marker transfer (i.e., construct recombinants) to map the gene or genes involved.
Acknowledgements This work was supported by a grant from the United Service (EY07736) and by an unrestricted grant from Blindness, Inc. to the Department of Ophthalmology. I Dorene R. Grau for expert technical assistance and preparing the manuscript.
States Public Health Research to Prevent would like to thank Bernadette Bull for
References Batra, SK. and Brown, SM. (1989) Analysis of unselected HSV-I McRae/HSV-2 HG52 recombinants demonstrates preferential recombination between intact genomes and restriction endonuclease fragments containing an origin of replication. Arch. Virol. 105, l-1 3. Becker, Y., Hadar, J., Tabor, E., Ben-Hur, T., Raibstein, I., Rosen, A. and Darai, G. (1986) A sequence in HpaI-P fragment of herpes simplex virus-l DNA determines intraperitoneal virulence in mice. Virology 149, 255-259. Brandt, C.R. and Grau, D.R. (1990) Mixed infection with HSV-I generates recombinants with increased ocular and neurovirulence. Invest. Ophthalmol. Vis. Sci. 31, 2214-2223. Brandt, C.R., Kintner, R.L., Pumfery, A.M:, Visalli, R.J. and Grau, D.R. (1991) The herpes simplex virus ribonucleotide reductase is required for ocular virulence. J. Gen. Virol. (in press). Centifanto-Fitzgerald, Y.M., Yamaguchi, T., Kaufman, H.E., Tognon, M. and Roizman, B. (1982) Ocular disease pattern induced by herpes simplex virus is genetically determined by a specific region of viral DNA. J. Exp. Med. 155, 475489. Dix, R.D., McKendall, R.R. and Baringer, J.R. (1983) Comparative neurovirulence of herpes simplex virus type 1 strains after peripheral or intracerebral inoculation of BALB/c mice. Infect. Immun. 40, 103-l 12. Field, H.J. and Darby, G. (1980) Pathogenicity in mice of strains of herpes simplex virus which are resistant to acyclovir in vitro and in vivo. Antimicrob. Agents Chemother. 17, 209-216. Gordon, Y.J., Romanowski, E., Balouris, C. and Araullo-Cruz, T. (1990) A herpes simplex type 1 ICPO deletion mutant demonstrates diminished pathogenicity during acute ocular infection in different host animals. Invest. Ophthalmol. Vis. Sci. 31, 681688. Grau, D.R., Visalli, R.J. and Brandt, C.R. (1989) Herpes simplex virus stromal keratitis is not titerdependent and does not correlate with neurovirulence. Invest. Ophthalmol. Vis. Sci. 30, 2474 2480. Halliburton, I.W., Honess, R.W. and Killington, R.A. (1987) Virulence is not conserved in recombinants between herpes simplex virus types 1 and 2. J. Gen. Virol. 68, 1435-1440.
I35 Jacobson, J.G., Leib, D.A., Goldstein, D.J., Bogard, CL., Schaffer, P.A., Weller, SK. and Coen, D.M. (1989) A herpes simplex virus ribonucleotide reductase deletion mutant is defective for productive acute and reactivatable latent infections of mice and for replication in mouse cells. Virology 173, 276283. Javier, R.T., Sedarati, F. and Stevens, J.G. (1986) Two avirulent herpes simplex viruses generate lethal recombinants in vivo. Science 234, 746748. Javier, R.T., Thompson, R.L. and Stevens, J.G. (1987) Genetic and biological analyses of a herpes simplex virus intertypic recombinant reduced specifically for neurovirulence. J. Virol. 61, l9781984. Klein, R.J., Friedman-Kein, A.J. and DeStefano, E. (1981) Pathogenesis of experimental skin infections induced by drug-resistant herpes simplex virus mutants. Infect. Immun. 34, 693-701. Larder, B.A. and Darby, G. (1984) Virus drug resistance: mechanisms and consequences. Antiviral Res. 4, 142. Larder, B.A. and Darby, G. (1985) Selection and characterization of acyclovir-resistant herpes simplex virus type I mutants inducing altered DNA polymerase activities. Virology 130, 262271. Meignier, B., Longnecker, R., Mavromara-Nazos, P., Sears, A.E. and Roizman, B. (1988) Virulence of and establishment of latency by genetically engineered deletion mutants of Herpes simplex virus I. Virology 162, 251-254. Oakes, J.E., Gray, W.L. and Lausch, R.N. (1986) Herpes simplex virus type I DNA sequences which direct spread of virus from cornea to central nervous system. Virology 150, 513- 517. Pogue-Geile, K.L. and Spear, P.G. (1986) Enhanced rate of conversion or recombination of markers within a region of unique sequence in the herpes simplex virus genome. J. Virol. 58, 704708. Sederati, F., Javier, R.T. and Stevens, J.G. (1988) Pathogenesis of a lethal mixed infection in mice with two non-neuroinvasive herpes simplex virus strains. J. Virol. 62, 3037-3039. Subak-Sharpe, J.H. and Timbury, M.C. (1977) Genetics of Herpes viruses. In: H. Fraenkel-Conrat and R.R. Wagner (Eds), Comprehensive Virology. Academic Press, New York, pp. 89-131. Taha, M.Y., Clements, G.B. and Brown, S.M. (1989) A variant of herpes simplex virus type 2 strain HG52 with a 1.5 Kb deletion in RL between 0 to 0.02 and 0.81 to 0.83 map units is nonneurovirulent for mice. J. Gen. Virol. 70, 705716. Thompson, R.L. and Stevens, J.G. (1983) Biological characterization of a herpes simplex virus intertypic recombinant which is completely and specifically non-neurovirulent. Virology 131, 171-179. Thompson, R.L., Wagner, E.K. and Stevens, J.G. (1983) Physical location of a herpes simplex virus type 1 gene function(s) specifically associated with a IO million-fold increase in HSV neurovirulence. Virology I3 I, 180-l 92. Turk, S.R., Kik, M.A., Birch, G.M., Chiego, D.J. Jr. and Shipman, C. Jr. (1989) Herpes simplex virus type I ribonucleotide reductase null mutants induce lesions in guinea pigs. Virology 173, 733-735. Wheeler, C.E. Jr. (1964) Biologic comparison of a syncytial and a small giant cell-forming strain of herpes simplex. J. Immunol. 93, 749-756.