Arch Virol (1989) 104:77-86 Archives

Vi rology © by Springer-Verlag 1989

Physical mapping of a genome of equine herpesvirus 2 (equine cytomegalovirus)

G. F. Browning and M. J. Studdert

School of Veterinary Science, The University of Melbourne, Parkville, Victoria, Australia

Accepted October 6, 1988

Summary. The genome of a low equine cell passage equine herpesvirus 2 was partially cloned and physical maps for the restriction endonucleases BamHI, EcoRI, HindIII, and SalI determined. The genome length was estimated to be 192 kilobase pairs (kbp) and no evidence of isomerization was found. Two separate repeat structures were detected: 18 kbp direct terminal repeats; and an unrelated second pair of short internal, indirect repeats at 0.20 and 0.75 map units. Such a genomic structure does not appear to have been reported amongst the herpesviruses--all the genomes that do not isomerize either have repeat structures only at the termini, or if present internally, have only direct repeats.

Introduction

Equine herpesvirus 2 (EHV2), sometimes called equine cytomegalovirus, is provisionally grouped among the betaherpesviruses [21]. It has been shown to be one of two distinct, but related, types of slowly cytopathic herpesviruses which infect the horse [4]. Although definitive evidence is lacking, EHV 2 has been implicated in immunosuppression in foals, upper respiratory tract disease, conjunctivitis, general malaise and poor performance [-1, 2, 12, 17, 20, 23, 24, 26, 27]. Although appearing to have much in common with human cytomeg- alovirus (HCMV), including antigenic, biologic and genomic intratypic hetero- geneity [4, 9, 11, 16, 27, 31] and the ability to establish latent infections in circulating leukocytes [-5, 13], they have several major epidemiological and pathogenetic differences. In contrast to HCMV where persistent infection by a single virus seems to be the rule [11], horses appear to be frequently reinfected by different strains of EHV 2, and may be infected by two or more strains concurrently [3]. Also EHV 2, although unable to cross the placenta of the horse, was shown, when inoculated directly into the amniotic sac to be unable to cause the developmental disorders ascribed to other betaherpesviruses [8].

Previous studies of the genome of a high non equine cell passage EHV2 strain LK indicated that the genome was approximately 186kbp pairs and

78 G.F. Browning and M. J. Studdert

shared less than 3% homology with two o f the three equine alphaherpesvirus genomes (EHV 1 and 3) [22, 30]. This paper records restriction endonuclease maps o f a low equine cell passage E H V 2 as a prel iminary to examining the molecular biology and heterogenei ty of these viruses in more detail.

Materials and methods

Virus and cells

A plaque purified isolate of EHV2 strain 86/67 [4], passaged 17 times on equine foetal kidney (EFK) cells, was used for all studies. Only EFK cells were used for growing the virus.

Preparation and cloning of viral DNA: DNA was extracted from infected and uninfected EFK cells using the method of Pignatti et al. [18]. Whole viral DNA digested with either EcoRI, or EcoRI and SalI, or alternatively single restriction endonuclease generated frag- ments purified from low melting point agarose gels [15, 29], were ligated into the plasmid pUC 19 [ 15]. This mixture was used to transform E. coli JM 83 and transformants selected on LB agar plates impregnated with 50 gg/ml ampicillin and 30 gg/ml 5-bromo-4-chloro- 3-indolyl-[3-D-galactoside [15].

Plasmid DNA was obtained by heat lysis of an overnight broth culture [15].

Restriction endonuclease digestion and Southern blotting

Viral DNA was digested with BamHI, EcoRI, HindIII, and SalI, as well as all possible pairs of these four restriction endonucleases. The fragments generated were separated by electrophoresis through 0.5% agarose submersion gels, stained with ethidium bromide and photographed [25]. After photography DNA was transferred and fixed to nylon membranes [t5].

EcoRI generated EHV2 genome fragments cut from low melting point agarose gels, or plasmid DNA, were radiolabelled by DNA synthesis primed by random hexanucleotides in the presence of dCTPa[35S] [6]. These radiolabelled fragments were hybridized to the Southern blots (1 M Na +, 68 °C), then washed under highly stringent conditions (16.5 mM Na +, 68 °C), dried and autoradiographed a t - -70 °C [15].

Recleavage of isolated DNA fragments

Individual restriction endonuclease generated fragments of the EHV 2 genome were re- covered from low melting point agarose gets [29] and subjected to a second digestion with a different restriction endonuclease. Similarly all cloned fragments of the genome were digested with BamHI, HindIII, and SalI, as well as all possible paired combinations of these enzymes.

2 Exonuclease digestion

The termini of the EHV 2 genome were tentatively identified by digestion with 1 U Z exonuclease/lag viral DNA for 30 minutes. The reaction was stopped by treatment with phenol:chloroform:isomylalcohol [15]. The viral DNA was then digested with either BamHI, EcoRI, HindIII, or SalI, electrophoresed and fragments with lowered molarities identified.

Results

The sizes of the E H V 2 genomic f ragments by each o f the tbur restriction endonucleases, BamHI, EcoRI, HindIII, and SalI, are shown in Table 1. Al- though several bands were supramolar , no submolar bands were detected,

Genome of equine herpesvirus 2

Table 1. EHV 2 strain 86/67 restriction endonuclease fiagments

79

Designation BamHI EcoRI HindIII SalI

A 26 a B 22 C 17 D 16" E 15.5 F 13.0" G 8.8 H 8.1 I 7.2 J 6.9 K 6.7 L 6.2 M 4.85 N 4.85 O 4.40 P 3.65 Q 3.30* R 3.30* S 2.70 T 2.00* U 2.00 V 1.80 W 1.80 X 1.60 Y 1.40 Z 1.20 a 0.80

40* 66 30 26* 27* 21 22 15.5 14.5 14.0 14.0 13.0 12.0 12.0" 8.7 6.5 7.1 3.50* 6.0 3.50* 5.4 3.25 2.70 3.00" 2.45 3.00 1.85 2.65 1.25 2.00

2.00

90 43 26 18" 18"

* Fragments tentatively identified as terminal by )~ exonuclease digestion a Kilobase pairs

suggesting that the genome did not isomerize. Several bands (Fig. 1) were present in both mock and virus infected preparations of DNA, and corresponded to equine mitochondrial D N A [7-1, which would be expected to be present in samples prepared by the method of Pignatti et al. [18]. The fragment or frag- ments at or near each terminus of the genome for each of the four enzymes were tentatively identified by )~ exonuclease digestion (Fig. 1, Table 1) and con- firmed by subsequent comprehensive double digestion and hybridization ex- periments as detailed below.

Southern blots of single and all possible paired digests of the four endo- nucleases were probed with each of the EcoRI fragments (Fig. 2, Table 2). Although background hybridization was high when genomic fragments cut from gels were used as probes, resolution was sufficient for tentative linkage maps to be constructed. These maps were confirmed by recleavage experiments. Spe- cifically, the ten cloned fragments, that is EcoRI fragments D, G, H, J, K, L,

80 G.F. Browning and M. J. Studdert: Genome of equine herpesvirus 2

Fig. 1. Terminal fragments of EHV 2 strain 86/67 as determined by exonculease digestion. Control DNA digests are shown in the left lane of each pair, with mitochondrial DNA fragments arrowed. The patterns obtained following )~ exonculease digestion are shown in the right lane of each pair, with those fragments which showed decreased molarity arrowed and labelled. Note that the two EcoRI digests were not run concurrently, and thus do not

coincide perfectly in the lower molecular size regions

M, and O, and EcoRI/SalI fragments I and J, were each digested with BamHI, HindIII, and SalI, as well as all possible paired combinations of these three enzymes. Additionally EcoRI fragments, A, B, C, and F were digested with BamHI, HindIII, and SalI, EcoRI fragments E and I were digested with BamHI, and EcoRI fragment N was digested with HindlII. BamHI fragments A to L were digested with EcoRI and HindIII fragments A to J were digested with BamHI. SalI fragments A and B were digested with BarnHI and EcoRI, and SaII fragments D and E with BarnHI. In all instances the recleavage experiments confirmed the tentative linkage maps constructed from hybridization data (Fig. 3). Table 2 was compiled from the results of both the hybridization and recleavage experiments.

Hybridization was detected between the regions at either end of the genome suggesting the presence of terminal repeat structures. The BarnHI digestion products of the SalI D and E fragments were indistinguishable, suggesting that these two SalI terminal fragments were probably identical. The cleavage maps deduced for the genome from the hybridization and cleavage experiments sug- gested the orientation of these repeated sequences was direct. This was confirmed by additional hybridization experiments using the cloned EcoRI/SalI J fragment

Fig. 2. Hybridization of EcoRI fragmens (A to O) of the EHV 2 strain 86/67 genome with digests of whole viral genomic DNA with BamHI (B), EcoRI (E), HindIII (H) or SalI (S) and all possible pairs of these four enzymes. Z HindIII digest )~ phage DNA, a size marker

Barn HI F

E_~ RI

'i i Hind III . I

II

Sal I D

kilobase 0 pa~rs

j,, I 0

map units

'~K Ji " ° °

L J C TX,~ ~ B t1 H,O,a l ~ H II I ' l I II

B KM F 0 C P A N D I t l II II I I

B C A

20 40 60 80 1 O0 120 140 I 1 I I I I I I I I I I j l I I I I I I

0.1 0.2 0.3 0.4 0.5 0.6 0.7

~////~/..-.-.~

I I I I I N WS D RT

I ,tT I C

160 180 192

I l l l i I I I 0.8 0,9 1.0

, 4 - -

pE86J pE86! pES86J

I i I El I,,, i i l i l l l I pE86G pES861 pE86M pE86D pE86H pE860 pE86K

Fig. 3. Physical maps of the EHV 2 strain 86/67 genome for BamHI, EcoRI, HindIII, and SalI. The terminal and internal repeat structures and their orientation (arrows) are depicted below by boxes. The fragments of the genome cloned into pUC 19, and the designation of

these plasmids, are also shown (bottom line)

82 G.F. Browning and M. J. Studdert

Table 2. Cross-hybridization analyzis of EHV 2 strain 86/87 restriction endonuclease fragments

EcoRI BamHI EcoRI HindIII S a l I BamHI/EcoRI fragment

A (26) *,a, (16), 13.0, 40, (27), (6.0) 26, (15.5), 13.0, (90), 43, 18, (18) 6.9, 6.7, 6.2, 4.85, (12.0), (6.5), (3.5), 3.3, (3.3), (2.0) 3.5, 3.25, 3.0, 3.0

(16), 13.0, 6.9, 6.7, 6.2, (6.0), 3.3, (3.3), 3.0, 2.0, (2.0)

B 22, 17, 8.8, 2.0, 30 66, 21, 13.0, 2.0, 43, 18 8.8, 7.2, 6.5, 2.0, 1.8, 1.6, 1.4, 1.2 2.0 1.8, 1.6, 1.4, 1.2

C 16, (13.0), (6.7), (40), 27 (26), 14.0, 12.0, 90, 18, (18) t6, (t3.0), (6.7), 4.85, 3.3, (3.3), 6.5, 3.5, (3.5), 3.3, (3.3), 2.7, 2.7, 2.0, 1.8 (3.0) 2.3, 2.0, 1.8

D 15.5, 8.1, 7.2, 4.4, 22 66 90 8.1, 6.1, 4.4, 2.45, 0.8 0.8

E 26, 15.5, 3.65 14.5 66 90 10.5, 3.65, 1.1 F 22, 7.2 14.0 66 90, 26 8.4, 5.0 a (26), 17, 4.85 12.0, (8.7) (66), 13.0 (90), 43 9.6, (8.7), 1.9 H 26, (17, 4.85) (12.0), 8.7 66, 15.5, (t3.0), 90, (43) (9.6), 8.7

2.65 I 26 7.1 15.5, 14.0 90 7.1 J 26, (4.85) (40), 6.0 15.5 (13.0) 90, (43) 6.0, (3.0) K 26, 4.85 5.4 14.0 90 3.4, 2.35 L 22 2.7 66 26 2.7 M 22 2.45 66 26 2.45 N 22 1.85 66 26 1.85 O 26 1.25 15.5 90 1.25

* Bracketed fragments are those cross hybridizing due to repeat structure a Kilobase pairs

and HindIII fragments B, G, and H probes. The EcoRI/SalI J fragment hy- bridized to HindIII fragments E and H, while HindIII H hybridized to HindIII H and L. The HindIII B fragment hybridized to HindIII fragments B and G, while HindIII G also hybridized only to HindIII B and G and not to HindIII L or I. Thus those HindIII fragments containing some of the terminal repeated sequences, but also some of the unique internal sequence, hybridized to the terminal fragment at the opposite end of the genome, confirming that the terminal 18 repeats were direct (Fig. 3).

A second, more complex set of repeat structures was also identified by cross- hybridization between regions located at 0.20 and 0.75 map units (Fig. 3). Spe- cifically, the cloned EcoRI H hybridized to EcoRI G, BamHI M and C, HindIII F and SalI B, while EcoRI G, also a cloned fragment, hybridized to EcoRI H, BamHI A, HindIII A and D, and SalI A. Addit ional ly EcoR J hybridized to EcoRI A, BamHI M, and SalI B. Al though this cross hybridization is faint in

Genome of equine herpesvirus 2 83

BamHI/HindlII BamHI/Satl EcoRI/HindIII EcoRI/SalI HindIII/SalI

(15.5), 7.2, (7.2), (15.5), 13.0, (13.0), 6.7, 6.2, (6.0), 3.5, 6.9, 6.2, 4.85, 4.6, (3.5), 3.3, (3.3), 3.25, 3.0, 2.75, 2.75, 2.0, 2.0, 1.0 20, 8.8, 6.6, 2.0, 2.0, 17, 15, 8.8, 2.0, 1.6, 21, 5.4, 2.0, 2.0

26, (12.0), (6.5), (6.0), 3.5, (3.5),

3.3, (3.3), 2.0, (2.0), 3.25, 3.0, 3.0, 1.1

2.0, 1.8, 1.6, 1.4, 1.2 1.6, 1.4, 1.2 7.2, (7.2), (6.7), 6.0, 14.0, (13.0), 4.85, 4.85, 3.5, (3.5), 3.3, 3.3, (3.3), 3.0, 2.7, (3.3), (3.0), 2.1, 2.0, 2.0, (2.0), 1.8 1.6, 1.1 15.5, 8.1, 7.2, 4.4, 16.5, 8.1, 7.2, 4.4, 0.8 0.8 15.5, 3.65, 3.2 26, 15.5, 3.65 20, 7.2 t5, 7.2, 7.0 10.5, (3.4), 2.75 (26), 17, 4.85 15.5, (10.5), 3.4, 26, (17) 2.75, (2.75) 15.5, 5.0 26 15.5, (2.75) 26, (4.85) 5.2, 5.0 26, 5.2 20 16.5 20 16.5 20 16.5 15.5 26

20, 18, (18), (6.0) 15, 13.0, 12.0, (12.0), 3.5, (3.5), 3.25, 3.0, 3.0, (3.0)

20, 10.5

(26), 12.0, 6.5, 5.5, 18, (18), 9.5 3.5, (3.5), 3.25, (3.0)

14.0, 13.0, 13.0, 7.4, 3.5, 3.0, 2.0, 2.0 14.0, 12.0, (12.0), 3.5, 3.5, (3.5), 3.0, 3.0, (3.o)

22 22 55

14.5 14.5 55 14.0 12.5, 1.4 55, 14.0 12.0 (3.6) (2.55) 12.0, (8.7) (55), 13.0 (12.0), 3.6, 2.75, (12.0), 8.7 55, 15.5, (13.0), 2.75 2.55 4.4, 2.8 7.1 15.5, 14.0 6.0, (1.1) (20), 6.0 15.5, (13.0) 5.4 5.4 14.0 2.70 2.70 14.0 2.45 2.45 14.0 1.85 1.85 14.0 1.25 1.25 15.5

Fig. 2, it was highly reproducible and with cloned fragments as probes, the background seen in other blots when fragments cut from genomic digests were used, was not evident. As the more external regions of these repeats hybridize to each other, and the internal regions also hybridize only to each other, their orientation appears to be indirect. More precise mapping of these cross-hy- bridizing regions revealed a complex relationship.

Each of the regions could be further resolved into two smaller regions interrupted by non repeating segments as shown in Fig. 4. Within the EcoRI G fragment two repeat regions were identified, "a" , a 1.30 kbp region bounded by EcoRI and BstEII sites, and "b" a 2.80 kbp region bounded by BstEII and BstXI sites. Within the EcoRI H fragment "c", a 2.55 kbp region bounded by EcoRI and HindIII sites, and "d", a 3.50 kbp region bounded by HindIII and EcoRI sites, were identified. Regions "a" and "b" hybridized to both "c" and "d" , but not to each other. In contrast, "c" and "d" hybridized to all four

84 G . F . Browning and M. J. Studdert

regions. The degree of sequence similarity between "a" and "c", "b" and "d", and "c" and "d" was not as great as that between "a" and "d", and "b" and "c", as reflected by both the amount of probe bound during hybridization, and the stability of the hybrids formed during washing at high stringency (330 mM Na + , 68 °C compared to 16.5 mM Na + , 68 °C). Neither the HindIII N fragment (the region between "c" and "d") nor the region between "a" and "b", cross hybridized with any other region on the genome.

Discussion

The genome size of EHV 2 deduced from these restriction endonuclease mapping studies was 192 kbp. This agrees well with studies by Wharton et al. [30], who estimated the genome of EHV 2 strain LK to be about t 86 kbp in sedimentation analyzes as well as by summation of restriction endonuclease fragments.

The structure of the EHV 2 genome deduced from these studies has not previously been reported for a betaherpesvirus, or indeed any other herpesvirus [10]. This structure, comprising a non-isomerising genome with 18 kbp direct terminal repeats and an unrelated, complex set of internal indirect repeats at 0.20 and 0.75 map units, contrasts with that suggested by the study of Wharton et al. [30]. In their study of restriction endonuclease digests of EHV 2 strain LK, 0.5 molar fragments were identified, suggesting an isomerising genome. This conflict may be due to changes to the genomic structure induced by passage in heterologous host cells (in the case of LK, rabbit cells), similar to those recorded for pseudorabies virus passaged in chick embryo cells [14].

The orientation and position of both the terminal and internal repeat struc- tures has been further confirmed by the restriction mapping of a further 14 low equine cell passage EHV 2 [4 a] previously shown to vary considerably in re- striction endonuclease patterns [4]. In all 14 of these isolates there was cross- hybridization between the termini and between the regions at 0.20 and 0.75 map units, and the apparent orientation was the same as reported here.

EcoRt BstEIl BstEII BstXI EcoRI

I I 1 I 1 a b

EcoRI G (12 kb)

EcoRI Hindlll Hindttl EcoRI

I I I I c d

Eco RI H (8.7 kb)

Fig. 4. The internal repeat regions of EHV 2. EcoRI fragments G and H are shown. The cross-hybridizing regions within these two fragments (a, b, c, and d), and the restriction endonuclease sites which bound them are depicted. Detail of the relationships between

these four regions is described in the text

Genome of equine herpesvirus 2 85

The size o f the internal repeat regions shown in Figs. 3 and 4 represents the maximal extent of these regions. Al though complex, the general orientat ion appears to be indirect, with the strongest cross-hybridizations detected between regions "a" and " d" and "b" and '%", as well as between EcoRI fragments A and J.

The complexity of these repeats, highlighted by the differential stability of the hybrids when washed under high stringency, would probably be best resolved by sequencing both regions.

Acknowledgements

These studies were funded by The University of Melbourne and Australian Equine Research Funds. We are grateful to Nino Ficorilli for technical assistance. G.F.B. was in receipt of a Commonwealth Postgraduate Research Award.

References

1. Belak S, Palfi V, Tuboly S, Bartha L (1980) Passive immunization of foals to prevent respiratory disease caused by equine herpesvirus type 2. Zentralbl Veterinarmed [B] 27: 826-830

2. Blakeslee JR Jr, Olsen RG, McAUister ES, Fassbender J, Dennis R (t975) Evidence of respiratory tract infection induced by equine herpesvirus, type 2, in the horse. Can J Microbiol 21:1940-1946

3. Browning GF, Studdert MJ (1987a) Epidemiology of equine herpesvirus 2 (equine cytomegalovirus). J Clin Microbiol 25:13-16

4. Browning GF, Studdert MJ (1987b) Genomic heterogeneity of equine betaherpesvi- ruses. J Gen Virol 68:1441-1447

4a. Browning GF, Studdert MJ (1989) Physical mapping of the genomic heterogeneity of isolates of equine herpesvirus 2 (equine cytomegalovirus). Arch Virol 104:87-94

5. Diosi P, Moldovan E, Tomescu N (1969) Latent cytomegalovirus infection in blood donors. Br Med J 4:660-662

6. Feinberg AP, Vogelstein B (1983) A technique for radiolabelling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132:6 13

7. George M Jr, Ryder OA (1986) Mitochondrial DNA evolution in the genus Equus. Mol Biol Evol 3:535-546

8. Gleeson L J, Studdert MJ (1977) Equine herpesvirus: experimental infection of a foetus with type 2. Aust Vet J 53:360-362

9. Harden TJ, Bagust TJ, Pascoe RR, Spradbrow PB (1974) Studies on equine herpesvirus 5. Isolation and characterization of slowly cytopathic equine herpesviruses in Queens- land. Aust Vet J 50:483488

10. Honess RW (1984) Herpes simplex and "the herpes complex": diverse observations and a unifying hypothesis. J Gen Virol 65:2077-2107

11. Huang E-A, Huong S-M, Tegtmeier GE, Alford C (1980) Cytomegalovirus: genetic variation of viral genomes. Ann NY Acad Sci 354: 332-346

12. Jolly PD, Fu ZF, Robinson AJ (1986) Viruses associated with respiratory disease of horses in New Zealand: an update. NZ Vet J 34:46-50

13. Kemeny L, Pearson JE (1970) Isolation of herpesvirus from equine leukocytes: com- parison with equine rhinopneumonitis virus. Canad J Comp Med 34:59-65

14. Lominiczi B, Gielkens A, Csobai I, Ben-Porat T (1987) Evolution ofpseudorabies virus virions containing genomes with an invertible long component after repeated passage in chicken embryo fibroblasts. J Virol 61:1772-1780

86 G.F. Browning and M. J. Studdert

15. Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cloning. A laboratory manual. Cold Spring Harbor Laboratory, New York

16. Mumford JA, Thomson GR (1978) Serological methods for identification of slowly growing herpesviruses isolated from the respiratory tract of horses. In: Bryans JT, Gerber H (eds) Equine infectious diseases IV. Proc Fourth Int Conf Equine Infect Dis. Veterinary Publications Inc., Princeton, New Jersey, pp49-55

17. Patti V, Belak S, Molnar T (1978) Isolation of equine herpesvirus type2 from foals, showing respiratory symptoms. Zentralbl Veterinarmed [B] 25:165-167

18. Pignatti PF, Cassai E, Meneguzzi G, Chenciner N, Milanesi G (1979) Herpes simplex virus DNA isolation from infected cells with a novel procedure. Virology 93:260-264

19. Plummer G, Goodheart CR, Studdert MJ (1973) Equine herpesviruses: antigenic re- lationship and deoxyribonucleic acid densities. Infect Immun 8:621-627

20. Plummer G, Waterson AP (1963) Equine herpesviruses. Virology 19:412-416 21. Roizman B (1982) The family Herpesviridae: general description, taxonomy and clas-

sification. In: Roizman B (ed) The Herpesviruses, vol 1. Plenum Press, New York, pp 1-23

22. Staczek J, Atherton SS, O'Callaghan DJ (1983) Genetic relatedness of the genomes of equine herpesvirus types 1, 2, and 3. J Virol 45:855-858

23. Studdert MJ (197 t) Equine herpesvirus 4: concurrent infection in horses with strangles and conjunctivitis. Aust Vet J 47:434-436

24. Studdet MJ (1974) Comparative aspects of equine herpesviruses. Cornell Vet 64: 94- 122

25. Studdert MJ (1983) Restriction endonuclease DNA fingerprinting of respiratory, foetal and perinatal foal isolates of equine herpesvirus type 1. Arch Virol 77:249-258

26. Sugiura T, Fukuzawa Y, Kamada M, Ando Y, Hirasawa K (1983) Isolation of equine herpesvirus type 2 from foals with pneumonitis. Bull Equine Res Inst 20:148-153

27. Turner A J, Studdert MJ (1970) Equine herpesviruses. 3. Isolation and epizootiology of slowly cytopathic viruses and the serological incidence of equine rhinopneumonitis. Aust Vet J 46:581-586

28. Turner AJ, Studdert M J, Peterson JE (1970) Equine herpesviruses. 2. Persistence of equine herpesviruses in experimentally infected horses and the experimental induction of abortion. Aust Vet J 46: 90-98

29. Vogelstein B, Giltespie D (1979) Preparative and analytical purification of DNA from agarose. Proc Natl Acad Sci USA 76:615-619

30. Wharton JH, Henry BE, O'Callaghan DJ (1981) Equine cytomegalovirus: cultural characteristics and properties of viral DNA. Virology 109:106-119

31. Witks CR, Studdert MJ (1974) Equine herpesviruses. 5. Epizootiology of slowly cy- topathic viruses in foals. Aust Vet J 50:438-442

Authors' address: Dr. M. J. Studdert, School of Veterinary Science, The University of Melbourne, Parkvitle, Victoria 3052, Australia.

Received June 2, 1988