and 3) SCF and PIXY. In each case the enzyme activity in the MFG-GC infected group was more 
than 80% higher than the level of endogenous activity in noninfected cells grown under the 
same conditions. Control samples infected with a retroviral vector, containing the 
bacterial lacZ gene, did not differ significantly from noninfected cells, eliminating the 
possibility that increased activity from MFG-GC transduced cells was a result of nonspecific 
viral exposure or infection conditions. Infected cells were expanded in vitro for two weeks 
following infection, and a sustained elevation was observed in enzyme activity in the 
experimental group versus control cells (Figure 14). Southern blot hybridization of DNA 
from these cells suggested a transduction efficiency of 5 to 15%. 
In two subsequent experiments (Experiments #2 and #3), significantly increased GC 
expression was not observed following infection procedures. In these experiments, rat 
recombinant SCF (rrSCF) was used during infection and expansion of the cells, and although 
the biological activity of this factor was expected to be similar to that of human 
recombinant SCF (hrSCF) (Langly, et al , 1992) <E3) , it became clear in Experiment #4 that the 
cytokine activity was suboptimal using rrSCF from this particular source (FigUf# IS). In 
the latter experiment (#4) , cells were harvested at the point at which a decline in growth 
rate was observed. Analysis of glucocerebrosidase activity in these cells indicated that 
elevation of activity had occurred in each of several different groups following infection 
in comparison to similarly cultured noninfected cells (figure ^16) . These cells had been 
infected under a variety of conditions in an effort to uncover a reason for the lack of 
response in the two previous experiments. Two mixtures of cytokines were used; 1) IL-3, 
IL-6 and SCF, or 2) PIXY, IL-6 and SCF. The latter group was further subdivided into cells 
infected in the presence of a stromal cell layer, in the presence of human serum, on Petri 
dishes and on treated tissue culture dishes. The specific activity of these samples was 
lower across all the groups than in previous experiments (control cells -150 U/mg vs >200 
U/mg), perhaps as a result of the declining viability of these cells at the time, but the 
elevation for infected cells was about 70% above endogenous levels. 
In Experiment #5, for which hrSCF was used throughout infection and expansion phases, 
the specific activity was once again >200 U/mg, an increase of more than 100% above 
endogenous GC activity (Figure; 17). Southern blot hybridization suggests a transduction 
efficiency of >20% (Figurel$). In subsequent experiments, we have demonstrated that normal 
CD34 + cells are transduceci and express 2-4 times the activity of GC (Figure 19). We also 
have studied some of the variables that affect the transduction efficiency and resultant 
expression of GC in CD34 + cells. From current results, we have concluded that the 
combination of IL-3, IL-6 and SCF provides the most consistent results. Further, there is 
a linear relationship between total number of doses of viral supernatants and GC activity 
expressed in CD34 + cells (Figure 20). Using these conditions, CD34 + cells obtained from 
Gaucher bone marrow have shown an increase in GC activity of 20-40 fold essentially 
equalling the results in transduced CD34 + cells from controls (Figure 21). These cells kept 
in culture for 3 weeks maintain the elevated levels of glucocereSrosidase activity. 
We have recently demonstrated that Gaucher CD34 + cells can be rapidly evaluated for 
transduction efficiency using a conventional chromogenic immunocytochemical assay for 
glucocerebrosidase. Using this procedure, CD34 + cells from a Gaucher patient give no 
detectable color formation, whereas, many cells in the infected population gave a strong 
signal (see Figure 30). Counting ten high power fields and scoring for immunocytochemical 
positive celts, an estimated transduction efficiency of approximately 20% can be calculated. 
The advantage of this method is its speed. The analysis can be done within one day. We 
will utilize this method in characterizing cells to be reintroduced to the patients in the 
study (see Section V. I). 
The colony forming efficiency in methyl cellulose has been determined for freshly 
enriched CD34 + cells and for infected and control cells following infection procedures. 
Results from 3 experiments indicated that between 30-100% of the colony forming efficiency 
remains after infection. 
rect 
to | 
tot 
! )ft( 
t ieci 
j; Ok 
! steir 
cycl 
prep 
corr 
risk 
tran 
iicl 
[760] 
Recombinant DNA Research, Volume 17 
