FILE PROGRESS REPORT

Grant Number : NIH 5 ROT A1-05160

Principal Investigator : Joshua Lederberg

Grantee Institution : Stanford University School of Medicine
Project Title : Genetics of Bacteria

Grant Period : 09/01/68 through 08/31/73

1. Statement of Progress

Background: The general features of transformation are well known,
and discussed in much detail in numerous reviews and texts (e.g.,
Stent ( 45 ), Watson ( 46 ), Hayes ( 32 ). A review by Hotchkiss and
Gabor ( 33 ) cites 241 references through 1969, almost all of them
relevant to the present discussion. We will then not undertake a
monographic survey of complete bibliography here.

The specificity of transformation, e.g. the rejection of E. coli DNA

by B. subtilis, pervades the literature. However, few published reports
are directed at pushing the empirical upper limits, perhaps <10-9 , on
the relative efficiency of gene transfer between these species, or simi-
lar situations. Undoubtedly, we have followed a common procedure in not
bothering to publish such negative results in detail.

 

Gene transfer between B. subtilis and other B. spp. is greatly hin-
dered, but nevertheless does occur (cf. Wilson and Young, 1972 ( 48 ),
whose results concur with our own, R. Harris, unpubl.) The behavior of
"hybrids" is consistent with the model of a requirement for regular
duplex formation in local segments mentioned above (in contrast, for
example, with a restriction and host-modification system as is reported
for Hemophilus (Smith and Wilcox, 1970 ( 44 )). Hotchkiss ( 33 ) has
discussed the relative exclusion of some markers arising by mutation
possibly involving heterogeneity at a single base pair. These findings
help to delineate, rather than to solve, the problem of achieving mole-
cular translocation involving arbitrary, alien Sequences.

An overview of mutation in bacteria compared to eukaryotes also sug-
gests a basic difference in the role of chromosome inversion and trans-
location. These processes dominate variety and species formation in
eukaryotes; they occur only exceptionally in bacteria, and then perhaps
mainly in relation to the integration and de-integration of episomes.

For this reason, in previous grant applications, I had postulated __
that translocation-mediating enzymes were a later evolution, connected
with mechanisms of gene-regulation more complex than the sequential
expression of markers in a linear ( or circular, or simple multi-seg-
mented) pattern - the concept associated with the operon model of gene
regulation in prokaryotes. (One could also argue that eukaryote chromo-
somes have evolved multiple recognition sites, i.e., that this is

one role of "redundant-sequence DNA" across which translocations

could be sealed without invoking new enzymes. The report of trans-
2

locations between mouse and human chromosomes in somatic cell hy-
brids ( Ruddle, 1972 ( 41 )), tends to argue against such a role.
For this reason, some of our previous efforts were directed at the
examination of DNA-repair mechanisms of a eukaryote in vivo, that
is in eggs of Xenopus injected with bacterial DNA. These studies,
reported further in "4. Progress Report," have, however, been over-
taken by Sgaramella's findings with T4-ligase.

 

This recent paper, which is appended, is the main foundation

of our intended further work. Previous work had indicated terminal
joining of two synthetic polynucleotides (Sgaramella et al, 1970

( 43 )), each of which had based-paired ends ( a deoxynucleoside

5' phosphate paired with a complementary 3' hydroxy1), a condition
of the ends that may be termed "flush". The present paper takes
advantage of the known condition of phage P22 DNA as a flush clipped
ensemble of segments which are terminally redundant, but produced as
if by a random cut in a circular molecule. Therefore, almost every
end is different, and the ease of terminal joining then argues against
sequence-specificities for the ligase.

Furthermore, P22 did not join terminally to linearized SV-40 DNA,

the latter having thus been shown to have not a clipped but a cohe-

sive end, i.e., an overlapping simplex that would seeks its complement
on another strand. This SV-40 DNA (obtained by the action of a sequence-
specific nuclease restriction enzyme on circular SV-40) could, however,
be homo-oligomerized either with T4 or E. coli ligase, both enzymes
being able to seal a contrived, doubly nicked duplex. The paper itself
gives a more complete, md possibly clearer, account.

The problem of molecular translocation can then be reduced to that of
securing flush ends on biologically interesting DNA. Alternatively,
various restriction enzymes might have just the appropriate specificity
to generate useful cohesive ends, or terminal polymerases may be used
to add synthetic homo-polymer cohesive ends to existing DNA molecules,
an approach also contemplated in the previously submitted applications
in this series. Berg's group will be emphasizing the second and third
of these approaches, and have, of course, already achieved an outstan-
ding result with SV-40 and lambda; we will be concentrating on ter-
minal joining.

In order to pursue the biological activity of the oligomers, we have
developed a transfection system (see Progress Report) for P22 DNA.
(In the light of unforeseeable hazards with derivatives of SV-40,

we prefer not to pursue work with this as an animal virus in tissue
culture). Although most promising, the transfection system needs
further improvements (which we foresee should be possible )before we
can efficiently test the oligomers for biological activity.
Summary: We tried various approaches to the chemical cross-
linging molecules, the addition of synthetic cohesive ends,
and alying the basis for identifying a terminal-looking
enzyme in frog eggs. These false starts were superseded by
the identification of terminal-joining activity in T4-ligase
and its use in the formation of oligomers of P22 DNA ( 24 ).

We have also searched for deletion mutations in B. subtilis ‘and
were surprised to find that they are very rare, if they occur

at all in our strain. Heterospecific transformations (B. subtilis
x B. globigii) were studied and the model that sequence hemology

is required for efficient transformation was supported.

We have made a detailed analysis of the effects of chlorine on

DNA, but eventually concluded that its effects on DNA breakage

were incidental to effects on proteins, and that the latter was

the principal site of cell damage. However, chlorine is a (feeble)
mutagen, and has an enigmatic effect on the burst size as well as
viability of treated phage (lambda). It therefore deserves further
study as a possible environmental pollutant of biological consequence.

A variety of other incidental findings are indicated in the list
of published titles.

The most salient findings have already been outlined under "“back-
ground" and in the Sgaramella paper (1972, ( 24 )), which should be
regarded as part of this application.

A range of other studies avowedly of lesser importance is detailed
in the list of publications.

Work not yet published includes:

i. (R. Harris). A study of the specificity of a nuclease produced

by B. globigii and not by B. subtilis. This nuclease attacks B. sub-
tilis DNA more rapidly than that from B. globigii and may therefore
resemble "restriction enzymes" in specificity. The DNA from dif-

ferent hybrids is being examined to look into the role and nature of

a host-modification system. However, there is no evidence of such an
enzyme with differential specificity in the competent strain of B. sub-
tilis as might be hypothesized to account for the specificity of DNA

in transformation.

ii. (I. Majerfeld). The cross-linking of DNA with various agents, of
which glyoxal and glutaraldehyde appear to be the most promising. (The
rationale was that chemically linked DNA might be copied with DNA poly-
merase with a mere skip across the link in the template. However, ma-
terial isolated so far has not been well enough defined for a cogent
test of the concept). The work was interrupted by Mrs. Majer-
feld's emigration to England with her husband prior to the
completion of her Ph.D. research. She may, however, resume it
at Sussex.

iii. (B. Brandt and J. Wachtel, in press). On the hypothesis

that eukaryotes possessed an enzyme for molecular translocation,

we tried to develop a system for studying the effect of frog egg
enzymes, in vivo, on injected bacterial DNA. Following Gurdon,

( 31), DNA synthesis was demonstrable on the injected templates.
However, it proved to be too difficult to recover workable amounts

of "repaired" DNA from injected eggs, and attention was then directed
to the enzymes in the extracts. These have been shown to contain a
DNA-polymerase with variations in template specificity similar to those
reported for the intact eggs by Guurdon. That is, Xenopus laevis DNA,
undenatured (but treated with DNAse I) was a preferred primer, com-
pared to E. coli DNA or to poly d(A,T). E. coli polymerase I prefers
denatured DNA from various sources. Larvae and immature ovaries yiel-
ded enzymatic activities with still different patterns of preference.
However, on 50-fold purification, the egg extract enzyme was less dis-
criminating, perhaps owing to the removal of a DNAse-inhibitor. A
ligase has also been demonstrated in these extracts, but has not yet
been characterized, i.e. for terminal joining activity.

 

iv. (V. Sgaramella and H. Bursztyn, manuscript in preparation). Trans-
fection in P22. The more or less fortuitous amenabi lity of P22 DNA to
terminal joining by T4 ligase naturally ted to an inquiry on the biolo-
gical activity of the oligomers. Published work in transfection with
P22 DNA has been remarkably discouraging, efficiencies about 10-9
having been reported with spheroplasted hosts ( 28 ). In order to
attempt correlative transduction, we would prefer intact bacteria.
Attempts to condition Salmonella by cold shock and CaClo which has
given spectacular results with E. coli ( 29 ) were unrewarding. However,
supernates of shocked E. coli cells were found to condition an R strain
of S. typhimurium LT2 to a low rate of transfective competence. (The
supernate factor is, of course, under close study). Subsequently it was
found that limited treatment of P22 DNA with exonuclease would further
augment the efficiency of transfection toa level now in the range of
10-7 to 10-8. We have not evidently exhausted the possiblity of further
improvement which would furnish another valuable tool. These rates are
still too low to expect the transduction of bacterial markers to be
observed - nor have we done so as yet.

Staffing: (1968 - 1972)

Joshua Lederberg, Principal Investigator, Professor of Genetics
Professors A.T. Ganesan and Walter Bodmer (until the latter left
Stanford for Oxford in December 1970 ) also played important consul-

tative roles in this work; however, their research is supported and
reported independently.
Sergio Barlati, Research Associate 1966-1969

Orio Ciferri, Research Associate. (Professor of Genetics,
University of Pavia, Italy. Dr. Ciferri played an active
role in our research and was in the Department of Genetics
at Stanford from July to December 1966, September to December
1971 and October to November 1972).

Vittorio Sgaramella, Research Associate 1971 - present
Alan Duffield, Sr. Research Associate 1970 - present
Silvano Riva , Research Associate 1969 - 1970
Katherine Shih , Postdoctoral trainee
NIH training grant 1969 - present
Bruce Brandt, Postdoctoral trainee
NIH training grant 1970 - 1972

2. Publications

Barlati, S. and Ciferri, 0. 1970. Incorporation of 5-Methyl and 5-
Hydroxy-Tryptophan into the Protein of Bacillus subtilis. J. Bact.
101, 166-172.

Barlati, S., 1970. Incorporation of Uridine into Bacillus subtilis
and SPP] Bacteriophage Deoxyribonucleic Acid. J. Bact. 101, 330-332.

Majerfeld, I., Barlati, S. and Ciferri,O., 1970. Tryptophaniless
Death in Bacillus subtilis. J. Bact. 101, 350-354.

Barlati, S. and Majerfeld, I., 1970. Partial Characterization of the
Factor Responsible for Tryptophanless Death in Bacillus subtilis.
J. Bact. 101, 355-360.

Barlati, S., 1970. Polyribosomes from Bacillus subtilis during Amino
Acid Starvation in the Presence and in the Absence of Actinomycin.
J. Bact. 101, 925-930.

Barlati, S., 1972. DNA Replication during Development of Competence
in Bacillus subtilis. Molec. Gen. Genet. 118, 327-333.

Ciferri, 0., Barlati, S., and Lederberg, J., 1970. Uptake of Synthetic
Polynucletides by Competent Cells of Bacillus subtilis. J. Bact. 104,
684-688.

Heineman, S.F. and Spiegelman, W.G., 1970. Control of Transcription
of the Repressor Gene in Bacteriophage Lambda. P.N.A.S. 67, 1122-1129.
Heinemann, S.F. and Spiegelman, W.G., 1970. Role of the Gene N
Product in Phage Lambda. Cold Spring Harbor Symp. XXXV, 315-318.

Spiegelman, W.G., 1971. Two States of Expression of Genes cl,
rex, and N in lambda. Virology 43, 16-33.

Patton, W., Bacon, V., Duffield, A.M., Halpern, B., Hoyano, Y.,
Pereira, W. and Lederberg, J., 1972. Chlorination studies. I.

The Reaction of Aqueous Hypochlorous Acid with Cytosine. Biochem.
Biophys. Res. Commun. 48, 880-884.

Sgaramella, V., 1972. Enzymatic Oligomerization of Bacteriophage
P22 DNA and of Linear Simian Virus 40 DNA. P.N.A.S. 69, 3389-3393.

Lederberg, J., 1972. Biological Innovation and Genetic Intervention
in Challenging Biological Problems, Ed. Behnke, J.A., Amer. Inst.
of Biol. Sci. 25th Anniversary Volume, Oxford University Press, N.Y.
pp. 7-27.

Mazza, G., Eisenstark, H.M., Serra, M.C. and Polsinelli, M., 1972.
Effect of Caffeine on the Recombination process of Bacillus subitilis.
Molec. Gen. Genet. 115, 73-79.

Harris-Warrick, R.M., Lederberg, J., 1977. Interspecies Transformation
in Bacillus: Mechanism of Heterologous Intergenote Transformation.
J. Bact. 133, 1246-1253.

Harris-Warrick, R.M., Lederberg, J., 1977. Interspecies Transformation
in Bacillus: Sequence Heterology as the Major Barrier. J. Bact. 133:
1237-1245.