27926 
NOTICES 
E. coli K-12 did not appear to offer a signifi- 
cant hazard as a potential enteric pathogen 
even when It possessed well-defined determ- 
inants of pathogenicity It was emphasized by 
Orskov, Freter and Falkow that E. coli K-12 
strains carrying recombinant DNA molecules 
could still act as effective genetic donors in 
vivo and still posed a significant problem re- 
quiring control. E. Geldreich (UJ5. Environ- 
mental Protection Agency. Cincinnati, Ohio) 
discussed the possible outcomes of the re- 
lease of E. coli containing recombinant DNA 
molecules into the aquatic environment and 
concluded that total reliance cannot be 
placed on sewage treatment and the natural 
self-purification capacity of receiving waters 
to limit potential hazards. While these are 
realistic barriers to the dissemination of E. 
coli and associated fecal organisms via the 
water route, they are not infallible because 
of technological limitations, Improper op- 
erational practices and system overloading. 
Finally, M. Starr University of California, 
Davis) described the numerous genera of 
gram-negative bacteria found naturally oc- 
curring In the soil and on plants. He stated 
that most of these organisms do not appear 
to be a reasonable alternative to E. coli K-12 
as a host for recombinant DNA molecules. 
Indeed, Starr pointed out that since such 
genera as Erwinia, Rhizobium and Agrobac- 
terium are known to conjugate with E. coli, 
the potential dissemination of recombinant 
DNA molecule Includes a greater spectrum 
of microorganisms than Just enteric species. 
The fourth session of the workshop, 
chaired by R. Curtiss III (University of Ala- 
bama) , was concerned with the construction 
of safer bacterial hosts for DNA cloning. 
The goals In constructing safer host strains 
enumerated at the beginning of the session 
Included Introduction of mutations that 
would: (a) preclude colonization in normal 
ecological niches; (b) preclude cell wall bio- 
synthesis except In specially defined media; 
(c) cause degradation of genetic informa- 
tion In normal ecological niches; (d) cause 
vectors to be host-dependent; (e) minimize 
transmission of recombinant DNA to other 
strains In normal ecological niches; (f) in- 
crease usefulness for recombinant DNA 
molecule research; and (g) permit monitor- 
ing. 
Most of the progress in developing safer 
hosts has been achieved with E. coli K-12, al- 
though F. Young described a B. subtilis 
strain with a deletion for sp'orulatlon genes 
which readily undergoes autolysls. The 
strain also has defects In genes for purine 
and TTP biosynthesis and a mutation con- 
ferring a D-alanlne requirement can be In- 
troduced to cause cell wall biosynthesis to be 
defective. This strain may be defective In 
transformation, however, and therefore 
might be useful only with a phage vector 
which has yet to be developed and/or dis- 
covered. 
A. I. Bukhari (Cold Spring Harbor Labora- 
tory) described the use of the dapD8 muta- 
tion In E. coli K-12 to block cell wall biosyn- 
thesis and another non-reverting mutation 
which causes sensitivity to bile salts and de- 
tergents. The dapD8 allele Is the most stable 
dap point mutation known, although It does 
revert at frequencies of 10 8 to 10-». The mu- 
tation conferring bile salts sensitivity was ob- 
tained after Mu-1 infection of an Hfr strain 
and, although exhibiting the theoretically 
useful properties of ease of DNA Isolation 
and Inability to survive in the intestinal 
tract, might be due to Mu Insertion which 
would compromise its use for safe strain 
construction. 
Curtiss reported on the work performed by 
him and his coworkers In constructing and 
testing numerous strains with different mu- 
tations. Survival of strains in vivo was tested 
by feeding rats 10 10 cells In milk by stomach 
tube. A pur mutations did not reduce strain 
titers In feces whereas AthyA; A thyA drm; 
and AthyA dra mutations gave HP-fold, HP- 
fold and HP-fold reductions, respectively, In 
strain titers In feces. Strains with A thyA mu- 
tations also exhibited thymineless death In 
in vitro tests. Since strains with the dapD8 
allele can revert to Dap* strains were con- 
structed with both the dapDs and .AbioH- 
asd mutations. These strains have not been 
observed to revert to Dap* but can survive 
passage through the rat Intestine and in 
growth media lacking dlamlnopimellc acid 
but containing NaCl and 0.5% usable carbon 
sources. This survival was due to the produc- 
tion of the mucopolysaccharide, colanlc acid, 
which permits many of the cells to grow and 
survive as spheroplasts. A A gal-chl T muta- 
tion (also deletes \att, bio and uvrB genes) 
was introduced which blocks colanlc acid 
biosynthesis and leads to no detectable sur- 
vivors In media lacking diaminoplmelic acid 
or following passage through the rat In- 
testine. The dapD8 A bioH-asd Agal-chl' 
strains are more readily lysed, transform at 
higher frequencies and are conjugation-de- 
fective in matings with donors possessing 
conjugative plasmids in the P, W and O in- 
compatibility groups but Con* as recipients 
for F, I and T group plasmids when com- 
pared to the dap* gal + parent strain. Strains 
with endA mutations were also observed to 
exhibit increased transformation frequen- 
cies. Attempts to Introduce temperature- 
sensitive polA alleles into strains to block 
replication of ColEl cloning vectors at ele- 
vated temperatures and to cause DNA deg- 
radation at elevated temperatures in the 
presence of recA and A thyA alleles often do 
not have the same properties in the con- 
structed strains as In the strains in which 
the allele was originally Induced. Many mu- 
tations causing a Con- phenotype have been 
investigated, but many of these revert or do 
not exhibit a Con- phenotype in matings 
with donors possessing conjugative plasmids 
of the Incompatibility groups commonly 
found in enteric microorganisms. Some Con- 
mutants exhibit Increased sensitivity to bile 
salts; thus, the mutant described by Bukhari 
may also exhibit a Con- phenotype. All of the 
strains constructed by the Curtiss group are 
Sun* and most have mutations abolishing 
restriction alone or both restriction and mod- 
ification. Thus, sufficient information is now 
known to construct a usable safer E. coli K- 
12 host. Curtiss and collaborators are now 
Introducing the A thyA and dna mutations 
into their dapD8 Abio-asd Agal-chl T -uvrB 
hsr nalA r (for ase In monitoring) Su* X' 
4>80 r strain to accomplish this objective. 
The final session involved a general discus- 
sion of some of the major points raised pre- 
viously in the workshop. There was general 
agreement at this session that both plasmid- 
host and phage-host systems have been de- 
veloped that should meet the criteria of an 
EK2 system specified by the National In- 
stitutes of Health guidelines for research on 
recombinant DNA molecules. Additional test- 
ing is required to confirm the EK2 properties 
of these available systems, but It Is an- 
ticipated that these vector-host systems will 
meet these tests. 
Dr. Donald R. Helinski, University of Cali- 
fornia, San Diego. 
Dr. Stanley Falkow, University of Washing- 
ton. 
Dr. Roy Curtiss III, University of Alabama. 
Dr. Waclaw Szybalskl, University of Wiscon- 
sin. 
Appendix D. — Supplementary Information 
on Physical Containment 
I. Biological Safety Cabinets. 
Table I. 
II. Universal Biohazard Warning Symbol 
III. Laboratory Techniques for Biohazard 
Control : 
A. Pipetting. 
B. Syringes & Needles. 
C. Opening Culture Plates, Tubes, Bottles, 
and Ampoules. 
D. Centrifuging. 
E. High-Speed Centrifuges. 
F. Blenders, Ultrasonic Disintegrators, Col- 
loid Mills, Ball Mills, Jet Mills, Grinders, Mor- 
tar and Pestle. 
G. Miscellaneous Precautions & Recommen. 
datlons. 
IV. Personal Hygiene Habits, & Practices 
V. Care & Use of Laboratory Animals: 
A. Care & Handling. 
B. Cages . Housing Infected Animals. 
C. General Guidelines that Apply to Ani- 
mal Room Maintenance. 
D. Necropsy Rules for Infected Animals 
VI. Decontamination & Disposal : 
A. Introduction. 
B. Decontamination Methods. 
C. Laboratory Spills. 
D. Disposal. 
E. Characteristics of Chemical Decontaml- 
nants in Common Use In Laboratory Opera- 
tions. 
F. Properties of Some Common Decontami- 
nants. 
G. Vapors and Gases. 
H. Residual Action of Decontaminants. 
I. Selecting Chemical Decontaminants for 
Research on Recombinant DNA Molecules 
Table H. 
VII. Housekeeping: 
A. Introduction. 
B. Floor Care. 
C. Dry Sweeping. 
D. Vacuum Cleaning. 
E. Selection of a Cleaning Solution. 
F. Wet Mopping — Two-Bucket Method 
G. Alternative Floor Cleaning Method for 
Animal Care Areas and Areas with Monolithic 
Floors. 
VIII. Clean-Up of Biohazardous Spills' 
A. Biohazardous Spill in a Biological Safety 
Cabinet 
B. Biohazard Spill Outside a Biological 
Safety Cabinet. 
C. Radioactive Biohazard Spill Outside a 
Biological Safety Cabinet. 
IX. A Secondary Reservoir and Filtration 
Apparatus for Vacuum Systems. 
X. Packaging and Shipping: 
A. Introduction. 
B. Packaging of Recombinant DNA Ma- 
terials. 
C. Labeling of Packages Containing Recom- 
binant DNA Materials. 
Table ni. 
Table IV. 
XI. Training Aids, Materials and Co arses' 
A. Slide-Tape Cassettes. 
B. Films. 
C. Courses. 
XII. Outline of a Safety and Operation 
Manual for a P4 Facility. 
References. 
1. BIOLOGICAL SAFETY CABINETS 
Biological Safety Cabinets suitable for con- 
fining operations involving recombinant DNA 
molecules are described below: 
1. Class I. A ventilated cabinet for per- 
sonnel protection only, with an unrecircti- 
lated Inward flow of air away from the op- 
erator. The exhaust air from this cabinet 
may be filtered through a high-efficiency oi 
high-efficiency particulate air (HEPA) filter 
before being discharged to the outside at- 
mosphere. This cabinet is suitable for re- 
search work with the Center for Disease Con- 
trol (CDC) classes of etlologic agents 1, 2 and 
3 where no product protection Is required 
This cabinet may be used in three opera- 
tional modes: (i) with an eight-inch high, 
full-width open front; (ii) with an installed 
front closure panel (having four, eight-inch 
diameter openings) without gloves; and (iii) 
with an Installed front closure panel 
equipped with arm length rubber gloves. See 
Table I for ventilation requirements, agent 
use limitations, and minimum performa’nce 
requirements. 
2. Class II. A ventilated cabinet for per- 
sonnel and product protection having an 
FEDERAL REGISTER, VOL. 41, NO, 131 — WEDNESDAY, JULY 7, 1976 
[ 26 ] 
