serotypes (1, 2, 3, 4, 5, 7, 14, and 21) following direct injection of the virus into the CNS of rhesus 
monkeys. Focal injection of the virus into the CNS (up to 0.5 ml in to each thalamus, and up to 10 7 
TCIDso/ml) elicited a spectrum of responses depending on the serotype of the virus. They demonstrated 
that, independant of serotype, viral antigen could be detected in the ependyma, choroid plexus and 
leptomeninges within 72 hours of injection into the thalmus [34]. This indicates that the wild-type 
virus is potentially capable of spreading within the substance of the brain. These experiments were 
done without the use of a stereotactic device (personal communication from Dr. Lucy Rorke, University 
of Pennsylvania) and that the localization of the injection site was confirmed retrospectively by the 
finding of an injection site granuloma at necropsy. Animals injected with various serotypes were 
observed for eighteen days and then necropsied. The histologic findings ranged from no significant 
pathologic findings to lymphocytic infiltration of the leptomeninges and choroid plexus, with 
ependymitis and focal neuronal destruction. Interestingly, this resulted in no overt symptoms of CNS 
disease. The authors note that "clinical observation disclosed no signs or symptoms among the monkeys 
with evidence of inflammatory lesions that distinguished them from those without such lesions" [34], 
Serotype 5 adenovirus produced lesions of mild to moderate intensity. The results were very consistent * 
within each serotype and appeared to be "an all-or-none phenomenon"[34]. CNS virus titers decline 
over a 21 day period, suggesting that viral infection is a self limited process. Furthermore, 
intranuclear inclusion bodies that are typically associated with adenoviral infections in permissive 
tissues are specifically absent in CNS. This finding is consistent with the relative lack of CNS tropism 
of adenovirus [34]. 
II. B. 5 Recombinant Adenovirus-Mediated Gene Transfer in Murine CNS 
Adenovirus-mediated gene transfer, unlike retrovirus gene transfer, has the potential advantage of 
transducing non-dividing cells. Brain tumor cells are heterogeneous with respect to the fraction in 
active cell cycle, yet this is unlikely to alter their transducability. Those cells undergoing gene 
transfer while in G 0 will be rendered sensitive to GCV provided it is given over a long enough period of 
time and that transgene expression persists over this period. We and other authors [36-39] have 
demonstrated the effectiveness of adenovirus vectors in the CNS. Direct injection of adenovirus into the 
brain parenchyma results in efficient gene transfer followed by gradual decline in transgene expression 
[39], but could be detected up to 30 days post injection [36]. 
Preliminary data suggest that significant replication and distant spread of the virus does not take place. 
The liver or spleen are readily transduced by intravenously administered adenovirus. However, when 
Davidson et. al. [39] injected recombinant adenovirus in the brains of rats, they were unable to identify 
virus in the liver or spleen (by X-gal staining). In addition, they were unable to isolate the virus from 
retro-orbital venous blood samples which drained the injected area. Local toxicity at the site of 
injection was observed. This was probably secondary to a reactive gliosis from the injection itself 
[34] (seen in sham injected control) and cytopathic effects in the neuronal cells exposed to a very high 
multiplicity of infection along the injection tract [37]. 
II. B. 6 Adenovirus-Mediated Gene Transfer In Non-Human Primate CNS 
As part of an investigation of hypoxanthine phosphoribosyl transferase (HPRT) gene therapy for the 
treatment of Lesch-Nyhan Syndrome, Davidson and colleagues have used adenovirus vectors to deliver 
the HPRT gene into the CNS of non-human primates [40]. This work was performed within the gene 
therapy program at the University of Michigan when Dr. Wilson was directing this effort. The 
information obtained on adenovirus vector CNS toxicity is relevant to the current study. The 
preliminary experiments in rhesus monkeys ( Macaca mulatta) are described in Appendix VLB, page 26. 
In these studies adenovirus H5.010RSVrhprt (Ad.RSVrHPRT) was injected (200pJ) into the CNS of 
rhesus monkeys at concentrations ranging from 10 9 particles/ml to 1.6x 10 11 particles/ml. There 
were no immediate postoperative complications and there were no signs of cerebritis, meningitis, or 
encephalitis. The animals groomed and fed normally and there were no obvious signs of impaired 
Recombinant DNA Research, Volume 20 
[357] 
