Unlike in studies by others, we used for gene transfer a plasmid vector system (pBMG-Neo) rather 
than a retroviral vector. The plasmid vector system pBMGNeo (44) is advantageous because it appears 
safer and, owing to its copy number of 20 -100 per cell, it allows higher levels of expression than single 
copy vectors. The vector system is described in detail below. Briefly, the cDNA is expressed under the 
metallothionein promotor, which is active in a wide variety of cells. The plasmid replicates as episome 
under the influence of the 69% fragment of the bovine papilloma virus type 1 (BPV1). The LI and L2 
late gene products of BPV1, coding for structural viral proteins, have been deleted, disabling the 
potential of the plasmid system to produce intact, infectious virus particles. An additional safety factor 
is the stringent species specificity of papilloma viruses. 
Transfection of murine LLC with pBMG-Neo-cDNA was achieved by the Lipofectin (BRL) 
technique. Stable and high levels of expression for all cytokines was observed using as inserted cDNAs 
murine 112, murine 114, and human 116 (Table 1, see Appendix C). These studies confirm the original 
description of pBMG-Neo as an excellent, eukaryotic expression vector, (Table 1, see Appendix C). 
The doubling time of transfected cells is not changed which demonstrates that the vector has no 
significant and consistent effect on the proliferation of LLC. 
TNF, contained in the vector pCDV-TNF (SV40 based) which integrates into the genome and 
expresses TNF under the SV40 promotor, showed variable levels of expression from one transfection 
to the next (Table 2, see Appendix C). In addition the levels of expression with this vector are low and 
sometimes undetectable. Table 2 (see Appendix C) also shows that in double transfection experiments, 
using pBMG-Neo-112 and pCDV-TNF, 112 expression is consistently high whereas TNF is quite 
variable. It is evident that the transfected and selected cells are able to secrete cytokines. Therefore a 
low level of TNF production is likely due to the low copy number of the vector and/or to the locus of 
insertion in the genome. For the purpose of gene therapy non integrating vectors with an intermediate 
copy number may be superior in the expression of the gene product, when compared to single copy 
Indeed, episomal plasmids usually produce 50 to 100-fold higher levels of the gene product than single 
copy vectors(45). In addition episomal vectors do not carry the risk of insertional mutagenesis. 
b. Rejection of LLC-I12 
Subcutaneous injection of C57B16 mice with 10 6 normal LLC or vector transfected LLC-Neo leads 
to tumor growth and death with a median survival of 33.3 days. Simultaneous, systemic application of 
recombinant 112 in various schedules and doses does not prolong survival. Subcutaneous injection of 112 
transfected LLC in contrast results in rejection of the 112 producing tumor cells and survival of the mice 
(Table 3, see Appendix C) up to a tumor cell dose of 10 7 /mouse. 
Only normal, immunocompetent mice were able to reject the 112 producing LLC cells. LLC-I12 was 
not rejected in syngeneic NK deficient mice (Beige) nor in T-cell deficient mice (Nude) nor in SCID 
mice (See manuscript in appendix). That the immunodeficient animals attempted to mount an immune 
response can be deduced from the fact that they survived longer upon injection of 112 producing LLC 
(70 days) when compared to the injection of the parent tumor (Table 4, see Appendix C). It is important 
to note that the vector or vector plus insert transfected cells do not decrease the survival time of injected 
mice. This indicates that the viral vector does not seem to increase the in vivo malignancy of the cells. 
This is in accord with a lack of influence on the doubling time in vitro and measurable growth rate in 
vivo. 
Recombinant DNA Research, Volume 18 
[247] 
