dilution of virus which caused degeneration of 25% or more of the cells in the cultures after 7 to 10 days incubation. In the neutralization tests, serial 2-fold dilutions of serum inactivated at 56°C for 30 minutes were incubated for one hour at 37°C with an equal volume of virus diluted to contain about 200 TCD30 per 0.2 ml. The inoculum was 0.2 ml of the virus-serum mixture and the end point was the highest initial dilution of serum which effected complete suppression of cytopathic change in cultures observed 7 to 10 days following inoculation.

Preparation of antiserum against the vacuolating agent. Young adult rabbits were injected intravenously with 1.0 or 2.0 ml of infected grivet monkey kidney culture fluid plus 1.0 ml of infected fluid emulsified in Arlacel-Drakeol adjuvant administered intramuscularly in a single dose or intracutaneously in divided doses. A final 1.0 ml dose of infected fluid was administered intravenously 3 weeks later and the animals were bled one week thereafter.

Results. Recovery of strains of vacuolating virus. Vacuolating virus was recovered from "normal" cell cultures of rhesus and cynomolgus monkey kidney and from the seed stocks of a variety of viruses which had become contaminated with the vacuolating virus during passage in cultures of rhesus and cynomolgus monkey kidney.

The vacuolating virus was recovered from contaminated virus seeds by passage of the virus seed, in the presence of its homologous antiserum, in cell cultures of Cercopithecus aethiops or grivet monkey. Under these conditions, the vacuolating virus "broke through" and was identified in serum neutralization tests using vacuolating virus antiserum prepared in rabbits. Recovery of the virus from "normal" rhesus or cynomolgus cell cultures was accomplished by simple passage, in grivet cell cultures, of the fluid from the rhesus or cynomolgus cultures which had been held for 7 to 10 days with maintenance medium. Vacuolating virus isolates were identified in serum neutralization tests using rabbit antisera prepared against prototype strains.

421

More than 20 strains of vacuolating virus have been recovered in our laboratory. Table I shows the origin of 8 of the strains examined most extensively. The vacuolating virus has appeared as a contaminant in the seed stocks of all of types 1 through 7 adenovirus. These seeds had been prepared in cultures of rhesus or cynomolgus monkey kidney and were obtained from Dr. R. J. Huebner or were brought to this laboratory by one of us (M.R.H.) from Walter Reed Army Inst. of Research. Additionally, vacuolating virus was recovered from seed stocks of Myxovirus parainfluenza 1 and 3, from the SA virus, and from the respiratory syncitial agent received from Dr. R. Chanock and Dr. K. Habel.

Vacuolating virus strains S207, S211 and S215 were derived from types 1, 2 and 3, respectively, of Dr. Albert Sabin's live attenuated poliomyelitis vaccine. For recovery of vacuolating agent, the poliovirus was neutralized by homologous polio immune serum prepared in rabbits immunized with Syverton's HeLa cell lines of poliovirus propagated in human stable amnion cell culture. The infectivity titer of the vacuolating virus in each of the Sabin vaccines ranged from 10-3.7 to 104.5. These isolates of vacuolating virus were identified in serum neutralization tests with rabbit antisera against prototype strain

776. At the time of the initial isolations of vacuolating virus, the Sabin vaccines were also passaged in presence of a mixture of homologous poliovirus and vacuolating agent antiserum. The antiserum mixtures suppressed both the poliovirus and vacuolating agent and showed the absence of other detectable viruses which might have been present in the Sabin materials.

The essentially ubiquitous occurrence of vacuolating virus in the various virus seed stocks and vaccines suggested a high infection rate among normal rhesus and cynomolgus monkey kidney cell cultures. This was borne out in tests for virus in "normal" cell cultures. During a one-month period, 10 rhesus and 10 cynomolgus monkey kidney cell lots prepared for use in ordinary Salk vaccine production were examined for presence of the vacuolating virus by passage to grivet kidney cultures. Each lot of monkey kidney was derived from a pool of kidneys from 2 or 3 monkeys. None of these "normal" cell lots presented any cytopathic change suggestive of the vacuolating virus. Of 10 lots of rhesus kidney, 7 yielded vacuolating virus. One lot was infected with a foamy-like virus (1) and 2 lots appeared free of simian contaminants. Only one of the 10 cynomolgus cell culture lots revealed vacuolating virus. However, it was not possible to carry out definitive tests of these materials since 8 of the lots were heavily contaminated with what appeared to be foamy virus. Thus, the vacuolating agent, if present, may have been overgrown and excluded by the foamy virus.

It is striking that the titer of vacuolating agent was as high as 10-6.0 in the fluids from the normal rhesus kidney cell cultures, even though these cultures showed no cytopathic change suggestive of the presence of a virus.

Cultures of grivet monkey kidney prepared in our laboratories have proved remarkably free of the vacuolating agent. During use of more than 1000 lots of individual monkey kidney cell cultures, presence of the vacuolating agent was suspect in only 4 lots and proved for only 2 lots.

Properties of vacuolating virus. Cytopathic effect. Fig. 1 shows the outstanding cyto

FIG. 1. Cytopathic changes in infected kidney cell cultures of Cercopithecus aethiops monkeys caused by strain 776 of vacuolating virus. a. Uninoculated control culture, unstained, day 6, 188 X mag. b. Infected culture, unstained, day 6, 188 X mag. e. Uninoculated control culture, day 6, 442 X mag., stained with H & E. d. Infected culture, day 6, 442 X mag., stained with H & E.

pathic changes caused by vacuolating virus in grivet kidney as observed in wet and in stained culture preparations. Grivet kidney cell cultures infected with about 1000 TCD50 of vacuolating virus show beginning cytopathic changes on day 3 or 4, when some of the cells may appear rounded or shrunken and the cytoplasm may be darkened. A few cells may show beginning vacuolation of the cytoplasm. With increased time, typical changes develop, consisting predominantly of ballooning of cells which are free or spread out on the glass together with intense vacuolation of the cytoplasm of such cells. These vacuoles are highly refractile in wet preparations and appear as "holes" with intensely stained boundaries in hematoxylin and eosin stained materials. In cultures stained with acridine orange and observed by ultraviolet fluorescent microscopy, the vacuoles likewise appear to be "holes" in the cytoplasm. We have not been able to demonstrate any stainable substance within the vacuoles nor has it been possible to demonstrate inclusion bodies

77-615 O-72-37

in such infected cells. Nuclei of infected cells may appear normal or disorganized internally. Limited fusion of the cytoplasm of adjoining infected cells may occur but, predominantly, each cell maintains its own integrity. Thus, the syncitium or giant cell formation, so characteristic of measles, foamy virus and certain other virus infections, is not prominent in vacuolating virus infection and this serves as a distinguishing feature. Terminal cytopathic changes occur in the cultures 5 to 10 days post-inoculation when the infected cells aggregate together and detach from the glass leaving behind only small islands of normal or degenerated cells. The cytopathic changes caused by vacuolating virus appear to be quite distinct from those described for Hull's 4 C.P.E. groups of simian viruses or for Malherbe's and Harwin's 7 simian agents.

Antigenic relationships. All strains of vacuolating virus studied to date have comprised a single immunologic group. Antisera against vacuolating virus strains 776 and 175 neutralized the homologous viruses and all of 3 additional strains as well (Table II). Additionally, and not shown in the table, all 3 of the Sabin vaccine isolates (S207, S211, S215) were neutralized by these sera.

Strain 776 of vacuolating virus was not neutralized by antiserum against C.P.E. group 4 strain S.V.6, S.V.26 or S.V.29 antisera furnished by Dr. Hull. Additionally, it was not neutralized by Myxovirus parainfluenza 1 or 3 antisera or by such SA, S.V.; or foamy virus antisera as were available to us. Vacuolating virus strains 776, 175, 953, 1095 and 584 were neutralized by pooled antisera against Hull's group 1B (S.V.20, 23, 25, 27) and the last three strains by group 3 (S.V.12. 28, 50). None was neutralized by group 2

423

(S.V.2. 10, 18, 19) or by group 4 (S.V.3. «. 20) antisera. In these tests, only a negative finding had significance since the antisera employed were prepared using virus grown in rhesus or cynomolgus monkey kidney cell cultures and these might also have contained vacuolating virus.

Host range. A preparation of vacuolating virus strain 776 which titered 10 in grivet kidney culture was tested in the various primary and line cell cultures shown in Table III. The virus was cytopathic for patas kidney and for rhesus monkey testicle, but these cells were about 30 to 300 times less sensitive than grivet kidney. Vervet monkey kidney has also been reported highly susceptible to the vacuolating virus(11). Grivet and vervet monkeys are separate races of the species Cercopithecus aethiops.

No definitive cytopathic change referrable to the vacuolating virus was observed in any of the remaining primary or line cell cultures, shown in the table, when observed for eight to twelve days. The virus persisted for at least 7 days in most of the cultures without increase in titer. Limited proliferation might have occurred. Titers as high as 108.5 have been obtained with vacuolating virus grown in grivet kidney.

The susceptibility of suckling mice and of embryonated hen's eggs to the vacuolating virus was tested. Suckling mice, inoculated intracerebrally, subcutaneously or intraperitoneally failed to develop illness within 21

TABLE III. Titration for Cytopathic Effect of Vacuolating Virus Strain 776 in Various Cell Cul

tures.

most viral agents (Table V). Fig. 2 presents data relating to inactivation at 37°C with 1:4000 formalin of the vacuolating agent. The slope of the inactivation curve indicates a slightly less rapid inactivation rate than that ordinarily found for poliovirus under the same conditions. Table VI presents a summary of known biological and physical properties of the vacuolating agent.

Occurrence of antibody against vacuolating virus in human and monkey sera. Table VII summarizes the results of tests for antibody against the vacuolating virus in sera of human subjects and in normal monkey sera. None of the normal or pre-vaccination sera

days following primary inoculation of vacuolating virus or following subpassage of brain or carcass carried out 7 days following primary inoculation. Embryonated hen's eggs inoculated after 8 days incubation via the allantoic or amniotic routes failed to develop manifest signs of disease within 6 days primary incubation. Additionally, none of the harvested fluids contained demonstrable hemagglutinins for chick erythrocytes.

Miscellaneous physical and biological properties of the vacuolating virus. The vacuolating agent was readily filtered through bacterial sterilizing filters (Table IV). The agent is relatively heat stable compared with

* We are indebted to E. Dwyer for these data.

425

from any of the human subjects contained antibody against the vacuolating virus nor was such antibody present in a lot of human gamma globulin. This suggests that natural infection in man with the vacuolating virus, if it occurs at all, is infrequent. However, all 9 persons given 2 doses of formalin-killed adenovirus vaccine prepared in rhesus or cynomolgus monkey kidney cell culture by one of us (M.R.H.) at the Walter Reed Army Inst. of Research (13), developed antibody against the vacuolating virus. Similarly, 7 of 13 persons given 2 doses of Salk poliomyelitis vaccine developed antibody against the vacuolating virus. These results are wholly consistent with the high frequency of vacuolating virus in monkey renal cell cultures, and indicate also the apparent high potency of vacuolating virus antigen in these vaccine preparations. None of 5 persons developed antibody against the vacuolating virus when fed Sabin live polio vaccine on 6 occasions. These data and those obtained more recently by Sabin (14) suggest lack of infectiousness of the vacuolating virus when administered by the oral route. This is in contrast to the apparent ability to infect man when the vi

: Sera tested for Dr. A. Sabin.

rus was administered by the respiratory route (12).

None of 14 normal sera from grivets representing 5 different shipments contained antibody against the vacuolating virus. By contrast, 12 of 18 sera from rhesus monkeys proved to have antibody to the virus. These findings are consistent with the very infrequent finding of vacuolating virus in grivet kidney cultures and the high prevalence in rhesus kidney.

Discussion. The discovery of the vacuolating virus represents the first demonstration of a hitherto hypothetical "non-detectable" simian virus, i.e., a virus present in monkey kidney cell cultures but not detectable by current procedures. All of the simian viruses which were recovered previously and which were of monkey kidney origin were detected in kidney cell cultures of the same monkey species, either in the primary culture itself or on sub-passage. In the present instance, the vacuolating virus was detected only following passage in the kidney cell cultures of a heterologous host. The demonstration of this first "non-detectable" simian virus raises the question of whether other as yet undetected