Reconstruction experiments were performed (Table 9, lower half) using SFV4 virus 
(injected i.n.) to assess whether die large excess of SFV-lacZ/Helper 2 could have 
masked the presence of low levels of replication-competent revertant virus. As few as ten 
infectious units of SFV4 were detectable in a population of 5x10^ activated SFV- 
lacZ/Helper 2 particles with this assay. Further, the LD^ for SFV4 injected i.n. into 40- 
day-old mice was shown to be only about 10 IU, demonstrating the high sensitivity of 
the mouse assay for replication-competent virus. 
Also presented in Table 9 is a comparison of the virulence of the prototype SFV strain, 
L10, and the progenitor clone for the SFV expression system, SFV4. Whereas the LD^ 
for SFV4 is approximately 10 4 IU, the LD 50 for L10 is clearly less than 10 4 . In fact, the 
LD 50 for SFV strain L10 by intraperitoneal injection of 40-day-old BALB/c mice is only 
14 PFU (Barrett et al., 1980). 
In summary, our inability to detect replication-competent SFV/Helper 2 virus either by 
in vitro infection of BHK cells, using a total of 2xl0 10 IU, or by injection of a total of 
2.4xl0 8 IU into the brains of neonatal mice, indicates that coincident SQL reversion and 
helper-vector recombination of the SFV-lacZ/Helper 2 virus must be exceedingly rare, 
probably well below the frequency of 10" 8 that one might predict from the individual 
apparent mutational rates. 
[458] 
Recombinant DNA Research, Volume 18 
