STANFORD U N I VE RS I TY A4 E D I C AL CENT E R

DEPARThfENT OF GENETICS

September 25, 1978

Dr. Maxine Singer
National Institutes of Health
Bethesda, Maryland 20014

Dear Maxine:

Here is a summary of my views of the non-E. coli HV1 problem,

--

and my proposed solution:

   As we have discussed, the proposed NIH Guidelines prohibit the
use of any and all host-vector systems (no matter how non-pathogenic
the organism)  that have not been specifically approved by the RAC.
In addition to requiring an unwieldy case-by-case review of each and
every instance where someone wants to use a non-E. coli K12 host,
this procedure makes the assumption that all non=E. coli systems are
dangerous and places the burden of proof for estaclishing safety on
the scientist who wants to use such a system. Omission of a non-
-- E. coli HV system from the "approved" list thus places the system in
the prohibited category.

   The modification I propose in the Guidelines is a minor addition
which addresses the issue of organisms that have not been specifically
dealt with by the committee. It does not, in my opinion, represent any
substantive change in what the RAC has done; it simply sets forth con-
ditions for carrying out experiments with organisms that the RAC has
not previously considered. There is a need for this addition, since
as noted above, simple experiments involving introduction of genes from
non-pathogens into other non-pathogenic non-E. coli hosts would other-
wise be placed in the "too dangerous to be done" category, despite the
fact that most scientists would agree that such experiments are among
the safest of all.

My suggested wording for this addition is: "Host-vector systems

involving organisms that have not undergone review and approval for HV1

                                                        shall
status, and which are not in etioloqic asent CIS s 2 or hisher,
be designated as HVO systems.
recipients for recombinant DNA molecules provided that the physical

containment conditions ern lo ed a e level higher than those
required for introductionnhost--VPc+nr s vstem.
Notwithstanding the above, the RAC shail be empowered to specifically
desianate any host-vector system, reyaruless of the etiologic agent class,
as unsuitable for use for any or all recombinant DNA experiments."

Such host-vector systems may be used as

1

Dr. Maxine Singer 2 September 25, 1978

This addition makes unnecessary a strain by strain or organism

by organism consideration of host-vector systems that involve non-
pathogenic organisms, while still allowing the RAC to preclude use
of any specific system that it considers to be unsuitable.
the basic principle that an HV system is innocent unless there is a
specific reason for anticipating a problem,rather than being guilty
unless innocence has been established by a specific vote of the RAC.
This presumption of innocence for certain prokaryotic HV systems is
based on rational judgment and scientific considerations; a similar
presumption has been applied in the proposed Guidelines to experiments
involving the cloning of eukaryotic viral DNA, and thus there is an
established precedent for it. In addition, the principle of requiring
a higher level of physical containment when biological containment is
reduced is also well-established in the Guidelines, and has been
followed for all sections except those where the committee has specifi-
cally recommended otherwise.

It follows

The wording I propose would assume that an HVO organism provides

no biological containment whatsoever, and would accordingly require
higher physical containment than an HV1 system. If an investigator
believes that an organism does in fact provide some biological contain-
ment, he then would have the burden of proving this to the committee,
in order to have the organism designated as HV1. In that case, the
physical containment level could be lowered.

non-E. coli non-pathogenic organisms to be considered as equivalents to
-- E. coli K12, experiments that employ B, subtilis, yeast, Neurospora,
Streptomyces species, and perhaps some other organisms as recipients
have been carried out with the approval of the NIH, NSF, or other grant-
ing agencies. In some instances, the timing and standards applied by
the different agencies in allowing experiments to be done have been
different. The wording of the new Guidelines removes ambiguity, and
because of this the Guidelines will require cessation of ongoing investi-
gations that involve the above organisms as recombinant DNA recipients.
The wording I suggest will deal with this problem.

Because of the ambiguity in the current Guidelines that permits

A "fall back" position would be to modify the end of my addition

to read, "such host-vector systems may be used as recipients for recombi-
nant DNA molecules that include components from organisms below etio-
logic agent class 2 under physical containment conditions, etc ....".
I believe the fall back position is far less satisfactory, and hope
that you will try for the first wording.
higher levels of physical containment for DNA from organisms and etio-
logic agents classes 2 and 3, and a further increase in containment level
when HVO hosts are used should be more than adequate to deal with bio-
safety concerns, regardless of the source of the DNA being cloned.

The Guidelines already require

Dr. Maxine Singer 3 September 25, 1978

I hope these comments will be useful to you. Please telephone

me if you want to discuss this further, and let me know what happens.

Best regards,

Sincerely yours,

SNC : kl