2JI PRE-CLINICAL STUPiES. 
3.1 CHARACTERIZATION OF THE HUMAN ylFN VECTOR PRODUCING CELL LINE, VCLHUy. 
The y-IFN proteins are highly species-specific and therefore the human y-IFN gene must be used for human 
clinical studies. The structure of the provector is illustrated in Fig.2 (Appendix). The clonal vector producing cell 
line, VCLHuy, has been carefully characterized with regard to the possible presence of adventitious agents 
(summarized in Table I). The cell-free supernatant results in an average titer of 3.7x10^ cfu/ml and several tests 
have failed to indicate the presence of RCRV. Co-cultivation of the VCLHuy vector producing cell line with Mus 
dunni cells (to amplify retrovirus, if present) also fails to detect RCRV. The cell line lacks adventitious agents as 
determined by broth and agar culture assays (for bacteria, fungi, and mycoplasma) as well as Hoechst staining (for 
mycoplasma). The cell line appears to have 1 provector with the appropriate genetic structure (Fig 3). Vector 
from VCLHuy results in expression of biologically active y-IFT^ in transduced cells (Table V) with the 
apjjrofHiate genetic structure (Fig.3). The VCLHuy master cell bank has been characterized in accwdance with the 
FTDA "Points to Consider in the Characterization of Cell Lines Used to Produce Biologicals (1987)". These and 
other tests are summarized in Table I. The high titer and lack of adventitious agents make VCLHuy ideal for the 
production of y-IFN vector for clinical trials. This vector producing cell line is described in a Master File (BB- 
MF51 10) submitted to the FDA and has already been approved by the RAC for use in human melanoma, in a trial 
under the direction of Dr. Hilliard Seigler (RAC #93064)43). 
3.2 Animal Model Studies. 
The feasibility of the overall approach has been tested in several murine tumor model systems using retroviral 
vectors that express murine y-IFN. Gene-modified murine tumor cell lines secreted y-IFN and exhibited 
substantial increases in the surface expression of Class I MHC proteins (28-30). The modified tumor cells were 
substantially less tumorigenic, tqjparently due to their ability to invoke increased anti-tumor immunity. Further, 
injection of y-IFN-expressing tumor cells resulted in potent CTL activity capable of specifically lysing the 
unmodified parent tumor cell (29). Unmodified tumor cells did not engender such a response. Pre-injection with 
irradiated y-IFN-expressing tumor cells, but not the unmodified tumor cells, protected animals from subsequent 
challenge with the unmodified tumw and has also “cured” metastatic disease in tumor bearing animals (29). This 
effect is presumably due to the observed induction of the tumor specific CTL. These data show that the naturally 
poor immunogenic properties of tumor cells can be improved by the constitutive expression of y-IFN delivered 
by retroviral vectors. The immune system, thus primed, can mount potent responses against the unmodified 
parent tumor. This conclusion suggests that a similar gene transfer bas^ immunotherapeutic approach to human 
cancer may substantially benefit cancer patients. 
3.2.2. MURINE y-IFN VECTOR EXPRESSION AND MHC. The retroviral vector containing the murine y- 
IFN gene was generated and introduced into several murine tumor cell lines; L33 (fibrosarcoma), B16F10 
(melanoma), CT26 (colo-rectal carcinoma), and Lewis lung carcinoma. All transduced lines produced 
measurable levels of biologically active y-IFN detectable in the growth media compared with the non- 
transduced parent tumor cell lines (Table IV). Transduction by murine y-IFN vector substantially increased the 
level of Class I MHC (Fig.4). Since these proteins present antigen to the immune system, the effect of 
vector transduction may be to improve immune responses, especially those that are Class I restricted such as 
CTL. 
3.2.3. TUMORIGENICTTY OF MURINE y-IFN-EXPRESSING CELLS. Improved immune responses against 
y-IFN-transduced tumw cells might be reflected in a decrease in the tumorigenic properties of the tumor. The 
effect of endogenous y-IFN expression on the tumorigenic properties of the vector modified tumor cells was 
therefore assessed in syngeneic animals. Vector-modified B16F10, CT26, and Lewis lung were all 
substantially less tumorigenic than the corresponding unmodified parent tumor cell lines. The data in Fig.6 
indicate that y-IFN transduced L33 fibrosarcoma cells are non-tumorigenic in immune competent animals. 
In addition, recombinant y-IFN protein added exogenously to the L33 cells did not affect tumorigenicity (Fig 
6), even though control experiments indicate that treatment resulted in high level MHC expression for as long 
as 3 days after treatment (data not shown). This indicates that the effect of constitutive expression of y-IFN 
obtained by gene transfer can be superior to the effect obtained by treatment with recombinant y-IFN protein. 
Control experiments with vectors which express only Neo*^ did not affect tumorigenicity (data not shown), 
indicating that the effect is due to y-IFN expression jjac ss. and not due to the vector transduction process in 
general. This suggests that effects are due to local secretion of y-IFN, rather than changes in antigenic 
phenotype due to exposure of cells to y-IFN. 
[94] 
Recombinant DNA Research, Volume 19 
