6566 Genetics: KantofT et ul. 
Proc. Nall. Acad. Sci. USA 83 (1986) 
Kig. 4. Inhibition of proliferation of TJF-2 cells by 2’-deoxyad- 
enosine. c. TJF-2: a. SAX-transduced TJF-2: z. SAX-transduced. 
G-«18-selecied TJF-2. Points ■ and a represent 50%-inhibitory 
concentrations for K7 and HM. two nonleukemic ADA-positive 
T-cell lines derived by transformation with HTLV-I during one 
typical experiment, presented for reference. All points represent the 
geometric means of 3-5 separate experiments, each with triplicates. 
Individuals heterozygous for ADA deficiency, as well as 
rare individuals with partially defective ADA protein, may be 
immunologically normal and yet have only a fraction (5-50%) 
of the normal concentration of ADA enzyme activity (14, 17). 
Furthermore, transfusions with ADA-positive erythrocytes 
can occasionally improve the immune function of ADA- 
deficient SCID patients (31-33). It appears, therefore, that a 
small percentage of ADA-positive cells can partially correct 
the metabolic disturbances in ADA deficiency and restore 
immune function in some patients. To test whether a similar 
effect could be observed in vitro , we added increasing 
proportions of irradiated (10,000 rads: 1 rad = 0.01 Gy), 
SAX-transduced, G418-selected TJF-2 cells, as a nonpro- 
liferating source of detoxifying ADA activity, to ADA- 
deficient TJF-2 cells cultured in inhibiting concentrations of 
2'-dcoxyadenosine. Fig. 5 shows that in the presence of 300 
/xM 2'-deoxyadenosine. proliferation of the ADA-deficient T 
cells was restored to normal levels when 10-20% of the cells 
in the cultures were ADA-positive. This experiment not only 
demonstrates that functional ADA was obtained by gene 
transfer with SAX but also suggests that beneficial in vivo 
effects may be obtained when only a minority of the patient's 
lymphoid cells are transduced with the ADA gene. 
DISCUSSION 
Genetic diseases in which the primary pathology is localized 
to the lymphohematopoietic system are likely to be early 
candidates for gene therapy, since the tissue can be readily 
explanted, manipulated ex vivo, and reimplanted. Retrovi- 
ruses provide an efficient means by which this type of genetic 
manipulation can occur. Part of the retroviral life cycle 
involves the infection of susceptible cells, followed by the 
reverse transcription of its RN A into double-stranded circu- 
lar DNA and then the integration of this DNA into the host 
cell's genome. This system has been manipulated to facilitate 
the transduction of exogenous genes into hematopoietic cells 
several orders of magnitude more efficiently than other 
gene-transfer systems such as. for example, calcium phos- 
phate precipitation (1). Retroviral vectors have been utilized 
to successfully introduce the bacterial neo R (2. 4). human 
hypoxanthine phosphoribosyltransferase (3. 5. 7). human 
Fig. 5. Reversal of 2'-deoxyadenosine inhibition of proliferation 
of TJF-2 cells by addition of ADA-positive cells. The ADA-positive 
cells were TJF-2 transduced by SAX and selected in G418. Prior to 
addition of these cells to cultures of 50 x 10’ ADA-deficient TJF-2 
cells, they were lelhally irradiated (10.000 rads). 2’-Deoxyadenosine 
was present at 300 fiM. One hundred percent thymidine incorpo- 
ration was 35,000 cpm. Fifty thousand irradiated SAX-transduced 
TJF-2 cells incorporated 316 cpm. 
ADA (6). murine granulocyte/macrophage colony-stimulat- 
ing factor (8), and murine dihydrofolate reductase (13) genes 
into murine (2, 6, 8) or human (3, 7) hematopoietic cells in 
vitro or into mouse bone marrow in vivo (4, 5, 13). Retroviral 
vectors have also been used to introduce the neo R gene at 
high efficiency into long-lived stem cells of murine bone 
marrow in vivo (9-11). Recently, we have obtained prelim- 
inary data demonstrating that the SAX vector can success- 
fully transfer the ADA gene into primates by means of an 
autologous bone marrow transplantation protocol (12). 
Two prior reports (6, 34) have demonstrated construction 
of retroviral vectors containing the human ADA gene. Both 
of the vectors produced ADA transcripts using the Moloney 
LTR promoter. Introduction of these vectors into murine 
lymphoid (6) or 3T3 cells (34) resulted in levels of human 
ADA activity equal to or greater than the level of murine 
enzyme activity. In this report, we have constructed and used 
the SAX vector to deliver an S V40-promoted human ADA 
gene into human ADA-deficient T and B cells. Human ADA 
enzyme activity with a normal isozyme pattern was detected 
in the transduced T and B cells. Quantitation of the ADA 
activity in the total unselected population indicated an 
increase from 1.1% to 35% of normal for TJF-2 (T cells) and 
from 0.6% to 69% of normal for GM2756 (B cells). The 
G418-selected TJF-2 and GM2756 populations both demon- 
strated total ADA activity that reached or even exceeded the 
range of values for an identical number of normal human T or 
B cells. The gene transfer and conferred ADA enzyme 
activity in ADA-deficient cells led to functional correction of 
the hypersensitivity to 2’-deoxyadenosine toxicity character- 
istic of ADA deficiency. An example of the effects that 
genetically corrected cells have on ADA-deficient cells was 
shown when irradiated, SAX-transduced cells were used as 
a source of enzyme to detoxify 2’-deoxyadenosine in the 
medium of ADA-deficient cells. 
For gene therapy to be clinically useful, the transferred 
gene must be capable of expressing enzymatically active 
protein in physiologic quantities in the proper target cells. T 
cells or their precursors are the cells that are most severely 
affected in ADA deficiency. Peripheral lymphoid cells, es- 
pecially T lymphocytes, are not readily obtainable from 
patients with ADA deficiency due to the severe lymphopenia 
[248] 
Recombinant DNA Research, Volume 12 
