In vivo and in vitro characterization of two camelpoxvirus isolates with decreased virulence

Today almost 20 million camels (Camelus dromedarius and Camelus bactrianus) are kept in Africa, Asia, India and the Middle East for milk, meat and wool production, transport or racing (18). As the interest in husbandry and diseases of camels has increased, camelpox in particular has received attention (20). It is caused by the so-called camelpoxvirus (CPV) which represents aseparate species within the genus Orthopoxvirus (19). Mainly came1 calves are affected and develop a proliferative skin disease often complicated by bacterial infections (12, 23). The mortality cari be as high as 30 % and recovering animals may develop lifelong immunity (12). With no therapy at hand, prophylactic vaccination seems to be a promising approach to protect camels. Higgins (8) reported about early attempts of nomads to protect their animals by exposing them to scabs from recent outbreaks ground in milk. In 1992, camels were vaccinated with vacciniavirus (Orthopoxvirus commune) (S), and in the same year homologous vaccines derived from CPV isolates serially passaged in ce11 culture were described by two groups (7, 10). According to both studies homologous vaccination proved to be effective after subsequent challenge in a limited number of animals. However, no data are available that would allow discrimination of the vaccine strains from wildtype isolates.

are kept in Africa, Asia, India and the Middle East for milk, meat and wool production, transport or racing (18). As the interest in husbandry and diseases of camels has increased, camelpox in particular has received attention (20). It is caused by the so-called camelpoxvirus (CPV) which represents a separate species within the genus Orthopoxvirus (19). Mainly came1 calves are affected and develop a proliferative skin disease often complicated by bacterial infections (12,23). The mortality cari be as high as 30 % and recovering animals may develop lifelong immunity (12). With no therapy at hand, prophylactic vaccination seems to be a promising approach to protect camels. Higgins (8) reported about early attempts of nomads to protect their animals by exposing them to scabs from recent outbreaks ground in milk. In 1992, camels were vaccinated with vacciniavirus (Orthopoxvirus commune) (S), and in the same year homologous vaccines -derived from CPV isolates serially passaged in ce11 culture -were described by two groups (7, 10). According to both studies homologous vaccination proved to be effective after subsequent challenge in a limited number of animals. However, no data are available that would allow discrimination of the vaccine strains from wildtype isolates.
In this study data on biological characteristics of two serially passaged CPV isolates and their progenies are presented. In a mouse model, passagation is shown to have led to a significant decrease in virulence.

II MATERIALS AND METHODS
Viruses and cells CPV strains variole de dromadaire 49 (VD49) and variole de dromadaire de Maurétanie (VDM) have been isolated from camels with generalized skin lesions in Niger (VD49) and Mauretania (VDM) (kindly provided by Dr. N. Ba-Vy, CIRAD-EMVT). Both strains were subcultured five times on MA-104 cells (African green monkey kidney ce11 line). In the following, these 5th passages Will be referred to as wildtype virus (VDM wt and VD49 wt, respectively). After 107 passages on Vero cells the strains were plaque purified three times. Subsequently, the 114th passages (referred to as VDM 114 and VD49 114, respectively) were investigated.

Infection of embryonated eggs
The pock morphology of VDM and VD49 wt and 114, respectively, was examined after infection of the chorioallantoic membrane (CAM) of 11 day old chicken eggs according to the method of Mayr et al. (16). Briefly, 0.1 ml of diluted (100 to 10-4) freeze-thawed infected ce11 culture material was applied to four eggs. The titer of the virus material used was determined on Vero cells (11). Eggs inoculated with 0.1 ml phosphate buffered saline served as a control. Six days post inoculation the CAMs were collected and examined.

Animal experiments
Suckling F81 POP mice (27) were infected with lO-fold dilutions (100 to 10-e) of purified virions of VDM and VD49 wt and 114, respectively, 36 1: 8 h after parturition. Six animals received 0.1 ml intraperitoneally (i.p.) and 7 to 8 animals received 0.02 ml intracerebrally (i.cer.). The titer of each virus preparation was determined on Vero cells (11). Untreated animals or animals inoculated with dilution buffer (2.5 mM Tris -1 mM EDTA, pH 8.0) served as controls. Infected animals were inspected twice a day. Animal experiments were approved by the government of Oberbayern (reference nbr. 21 l-2531-28/93).

Restriction enzyme analysis
Viral DNA of both CPV strains was isolated from purified virions from passages 0 (wildtype), 30, 80 or 82, 110 (not plaque purified) and 114, respectively. DNA was cleaved with restriction endonucleases HindIII and XhoI according to manufacturer's instructions (Boehringer, Mannheim, Germany). Resulting fragments were analyzed in 0.4 % and 1.2 % agarose gels and their molecular weights were estimated by comparison with molecular weight standards (1 KB ladder and HMW marker; Pharmacia, Freiburg, Germany). The total size of each genome was obtained by summation. Southern blots were prepared on Hybond Nt membranes (Amersham/Buchler, Braunschweig, Germany) using 0.4 N NaOH as transfer buffer. For hybridization restriction enzyme generated DNA-fragments of CPV: vacciniavirus strain Elstree and cowpoxvirus strain Brighton were gel-isolated and labeled according to manufacturer's instructions (DIG-DNA-Labeling and Detection Kit Nonradioactive; Boehringer, Mannheim, Germany).

Replication of VDM and VD49 in vitro
Replication of VDM and VD49 wt and 114, respectively, was observed in ce11 lines derived from various hosts. Generally, 12 to 24 h p.i. rounding of cells could be seen. The formation of visible plaques became evident after 24 h p.i. On the second day plaque diameters averaged from 0.2 mm to 0.5 mm on most ce11 lines and up to 1.5 mm on Dubca cells. A characteristic feature of CPV is the formation of multinucleated giant cells. This was most distinct in African green monkey kidney ce11 lines Vero and MA-104 with 20 to 200 nuclei involved. Some differences in virus yield and plaque size were observed between wt and passage 114. The formation of cornets was only seen with the highly passaged strains. Most obvious was the failure of both virus strains from passage 114 to replicate in MDCK cells and in Dubca cells. Fortyeight hours after infection the titer of wildtype viruses had increased at least 400-fold (VD49 wt) and 6000-fold (VDM wt), respectively, whereas no increase was observed for VDM 114 and VD49 114 infected cells as demonstrated for Dubca cells in figure 1. Furthermore, agglomerations of rounded cells was observed in BHK, E.derm (except VDM) and L929 cells after infection with wildtype strains and only in E.derm cells after infection with passages 114. Table 1 summarizes the differences in cytopathic effects of both strains in different ce11 lines (rounding up of cells, plaque formation and plaque diameter up to 0.5 mm are not noted).

Replication of VDM and VD49 in vivo
The pock morphology on the CAM was determined after six days of incubation. By then, small white-opaque pocks had developed with 0.2 to 1.0 mm in diameter. Comparison of either wt or passage 114 infected CAMs displayed no differences in pock morphology, and only a slight shift in the ratio of plaque forming units to pockforming units was observed. By day 6,25 % of VDM and VD49 wt infected embryos and 20 % of VD49 114 infected embryos had died, whereas no losses were seen after infection with VDM 114.
A marked attenuation was seen after application of the highly passaged viruses VDM 114 and VD49 114. The LD,, p.m. was calculated to be in the range of 1 to 7 x 10s pfu for both strains and both modes of application (figure 2,3 and table II).

Genome analysis
The restriction fragments obtained after digestion of different passages of VDM and VD49 with either XhoI or HindIII are shown in figure 5. The majority of fragments were comigrating. However, some fragments present in the wt strain (i.e. 11, 14.5 and 18 kilobasepair (kbp) HindIII-fragment of VDM) were absent in subsequent passages. In contrast, some new fragments could be

HDISCUSSION
The pheno-and genotype of two CPV isolates (VDM and VD49) were examined before and after serial in vitro passages. Both wildtype viruses displayed simi1a.r characteristics as compared to data previously described (1,2,4,13,14,23,24,26). However, after passagation differences were noted in vivo and in vitro. It seems remarkable that both strains have lost the ability to replicate in Dubca cells that originally were derived from came1 skin cells. The second difference noted is the formation of cornets that could be seen with the passaged viruses in MA-104 cells. Formation of cornets is caused by a unidirectionally spread of extracellular enveloped virions (EEV). It is well accepted that EEV plays an important role in the pathogenesis in vivo and might contribute to a protective immune response (25). This is of course essential for an effective vaccine.
However, the essential part of this study was the comparison of virulence of VDM and VD49 before and after ce11 culture passagation in a mouse model. According to data described in references 2, 4, 13, 14, 26, the susceptibility of mice seems to be age dependent. In this study, 36 f 8 h old mice proved to be highly susceptible for both VDM wt and VD49 wt. After intracerebral application, the pathogenicity of VD49 wt was slightly lower than the pathogenicity of VDM wt. Comparing the modes of application, intraperitoneal inoculation leads to a higher mortality and results cari be obtained within six days, which should be of help in future experiments.
However, a striking low pathogenicity of ce11 culture passaged virus could be demonstrated. Comparini the LD,, p.m.-values, the amount of pfu increased up to 260,000-fold. Only a lO,OOO-fold increase was described for recombinabt vacciniaviruses with a thymidin kinase negative phenotype (3). It cari only be assumed that during passagation mutations and/or deletions have occurred. It is tempting to speculate that failure to replicate in Dubca cells (the reasons are unknown) point to a highly attenuated phenotype in camels too.
Of interest are two animals which received a dilution (10'3 and 10-5 fold) of VDM 114 i.p. and were eaten by their dams. Therefore, an examination was not possible. However, in both cases only one animal from a group of seven was missing and because no animal of the other groups died (10-1, 10-2 and 10-4 fold virus dilutions) these two animals were not included in the survey on the LD,, value.
Investigation of the viral genome of VDM and VD49 indicates two different M'ays of genomic evolution during ce11 culture passages: VDM wt displays migration patterns and an overall genomic size similar to other CPV strains (5, 17, 24). Continuous passagation leads to a loss of 22 kbp in the terminal regions and thereby probably to a loss of genes which are not essential for replication in vitro. Many authors (6,9,15,22) assume that in the variable-sized termini proteins are encoded which interfere with the host's immune response.
In contrast, the mouse pathogenic VD49 wt displays a significantly shorter genome than VDM wt and other CPV strains (165 versus 190 kbp) (5, 17, 24) due to a truncated right terminus, thus resembling the VDM 114 strain with a highly attenuated phenotype. Passagation of VD49 wt leads only to a loss of approximately 1 kbp but is associated with a marked attenuation.
The mechanisms leading to attenuation by ce11 culture passagation are not yet understood. Maybe minute deletions or even point mutations in genes interfering with the host's immune response are responsible for the loss of pathogenicity (15,22). Thus, further detailed investigations like marker rescue experiments would be helpful to identify the gene(s) involved in virulence. However, at present molecular investigations cannot replace animal experiments for the proof of apathogenicity for animals or man.
The results show that VDM 114 and VD49 114 might be candidates for live vaccines against CPV infection. However, this cari only be evaluated in the natural host and must include subsequent challenge experiments. Thereby, two characteristics (failure to replicate in Dubca cells and formation of cornets) could serve as important in vitro markers to distinguish the attenuated strains from field strains.  I  I  III  ,  I  II  I  I I I  I  I  I  I  I  I  I  I  H  D  SRP  E  C  TNKQOI  J  M  A  1  c  B  F   VD Mauritanien  114.Pass.  Hindlll   I  I-III  ,  I  I I  I  I   F  KRQO  c  B  SM'  I 'PIN'  G  H  L'  A  'J'E'D' ' K VD 49 (Niger) wt Hindlll  l  I  III  I  I  II  I  I 1 1  I  I  ,  I  l