erythroid tells to produce normal amounts of normal 
beta globin 
Although IRIi s have faced questions of benefit ver- 
sus risk in new and potentially dangerous therapeutic 
regimens before, they have never had to evaluate 
treatment procedures designed to influence the geno- 
type of human cells. Correction of the germ line of the 
patient so that future offspring will be phenotypically 
normal may even be possible. Such new therapy may 
involve substantial risks, as we shall discuss, al- 
though gene therapy may turn out to be far safer and 
more effective than existing treatments for some dis- 
orders, including these lethal anemias. 
In summary, the conclusion drawn from consider- 
ation of these ethical codcs i lu is that before any gene 
is administered to a human being, animal studies 
must ensure that the new gene has a reasonable 
chance of producing the desired result and that there 
is a reasonable expectation that it will not produce a 
harmful result. 
Present State of the Art of Gene Therapy 
Our present concept of a gene is that it is a specific 
sequence of nucleotides in the l)NA of a cell. A gene is 
recognized by the product that it makes. Any gene 
coding for a molecule can, in theory, be identified in 
total cellular DNA. The gene can then be isolated, 
purified, modified, and attached to other DNA se- 
quences — for example, to other genes isolated from 
other sources. This final DNA fragment, perhaps 
containing several genes, can then be inserted into 
living mammalian cells in a number of ways: genes 
can be injected directly into .in individual cell with a 
micropipctte"; they can be introduced into a popu- 
lation ol cells in a culture dish by exposing the cells to 
a precipitate of DNA containing the desired genes 12 ; 
or they can be inserted into a viral particle (modified 
so that it will no longer be pathogenic), with the re- 
combinant virus then used to infect a population of 
cells. ,, - l< A number of genes have been inserted into 
living cells with these techniques. In summary, a gene 
can be isolated from normal cells, attached to other 
selected DNA sequences, and inserted into other cells 
in which, in some cases, the inserted gene will func- 
tion. 
With use of these techniques, experiments with 
genetic material inserted into intact animals have 
begun. It has been possible to introduce a gene that 
confers resistance to the toxic drug methotrexate into 
the bone-marrow cells of mice '- 2 The treated mice 
were more resistant to the drug as the result of the 
gene therapy. In another experiment, a viral gene was 
injected into mouse embryos, and a small number of 
newborn mice retained the viral DNA sequences in 
their cells ’ Clearly, research in the area of mamma- 
lian genetic engineering has progressed dramatically 
in recent years. Some workers may believe that the 
progress is so great that gene therapy in human beings 
should begin. Let us examine what we think should 
still be accomplished before experiments with human 
beings are justified 
Animal Studies Needed before Attempting Gene 
Therapy in Human Beincs 
Successful experimentation with genetic material 
inserted into laboratory animals needs to be accom- 
plished before that material is introduced into human 
subjects. At least three requirements should be met 
for each gene introduced: First of all, the new gene 
should be put into the proper target cells and should 
remain there. If, however, the gene also enters other 
cells, it should be shown that the presence of the gene 
or its product does not damage the non-target cells. 
Secondly, the new gene should be regulated appro- 
priately in the target cells. In other words, the gene's 
product should be made in amounts that are suffi- 
cient to correct the genetic defect but not so large as to 
have a detrimental effect. 
Thirdly, the presence of the new gene (and all other 
DNA sequences associated with it) should not harm 
the cell. Specifically, the inserted genetic material 
should not produce mutagenic events in the organ- 
ism. interfere with regulatory pathways within the 
cell, or otherwise adversely affect critical cell func- 
tions. 
We submit that these three criteria should be satis- 
fied to a considerable degree in animals before gene 
therapy is attempted in human beings. Although ad- 
vances may shorten the time or effort needed to meet 
these criteria, no advance can eliminate the necessity 
that all three requirements be met. However, we think 
that once all three are satisfied, attempts to cure 
human genetic diseases by treatment with gene ther- 
apy will be ethical The goal of biomedical research is, 
and has always been, to alleviate human suffering. 
Gene therapy is a proper and logical part of that 
effort. 
What Knowledge Is Necessary to Meet the 
Three Requirements? 
The New Gene Should Be Put Into the Target Cells and 
Remain In Them 
As yet, little is known about the cell's recognition 
signals. Ultimately, it may be possible to add nucleo- 
tide sequences to a gene that will allow that gene to 
enter or function only in a particular type of cell, but 
this is not feasible at present. Inserting a gene into the 
proper target cells may not be difficult in a few cases, 
however Lxperiments in mice have shown that bone- 
marrow cells can be removed, given a drug-resistance 
gene (see above), and replaced in the animal, with 
successful function of the gene . 1 - 2 Similarly, bone- 
marrow cells could be removed from a patient w-ith 
d-lhalasscmia or sickle-cell anemia and given the 
normal beta-globin gene; the genetically altered mar- 
row could then be introduced into the patient (non- 
erythroid marrow cells would also be exposed to the 
Recombinant DNA Research, Volume 12 
[ 315 ] 
