Department of Bacteriology, ;
London School of Hygiene and Tropical Medicine,
Keppel Street (Gower Street),

LONDON, WeCle

Sherbet” Muy, wi pratka Ly Set 30th June, 1953.

prt a
Prt

Lift int Ce pegen 2el-uge | Ar,

Cnt ar < get Légeky. (He Dene proep [3 V5 tard by ac Aa Zy,)

6p ee ey

f ;

‘ Ce pe em EES Von ft i

De ‘ | f . 2 Coan. & ton. ae } : x / : 3 f Lf oye \.
ar Te AL

Thanks for your air letter of the 30th May, and for
suggestions in it about paper, The paper went in in the middle
of May, and was accepted after some weeks {@® cogitation. I just
managed to get the typescript back yesterday, the editors have me&

iyetdonly a few minor alterations of wording. I have been over it

again and have dealt with all the noints you mention (except ones
where the matter criticised has beencut out anyway). I enclose
a few pages which I had retyped. I gather the stuff goes to the
printers in a few days, but if you see anything that should be
corrected in proof I hope you will voint it out,

I saw Felix today, to let him look over, at his
request, the bi ts in which he is quoted, All went well as to
this. I have restored the "Dr." ete., he agreed that the new
wording re B Vi was more suitable; he rather wanted me to put
in a bit about how stable Glasgow O was as indicated by use in
serodiagnosis, etc, but agreed it was non-essential. Actually he
was quite affable, though I always approach these interviews with
some trepidation. He was very disappointed to hear you are not
going to-be at Rome,

Kauffmann has toldme about his and your strains which
acquired I at the same time as an H antigen, this is a surprising
thing. It would be interesting to try and get a I - form of
SL13, serial enrichment in broth with anti-I as worked out by
Clive S oft to work, one would then perhaps be able to get triple
transfers (if the I is really a transduction, not selection of
I+ mutant), I saw Clive today, seems well, not doing any trans-
duction stuff as he_is involved in locum duties,

(K. fomed oe Kuo w $113)

I have done a few more micro-manipulation experiments,
It seems that the majority, or half anyway of motile cells picked
early (about 3 hours at 37° after mixing with phage) give rise,
when incubated at room temperature to the equivalent of more than
1 trail, The maximum so far is the equivalent of S
wt lar A drow Bal (tod | homtt it): Prue lene, Te Sat,
rape 4 wit ww Corper ay 1C-/2, Y Am Omiged lo Lec Fraf-

4 , Pane & - wiley Thue Mr cle. Bt niut- Lo s

y “ trails from one cell, It is now clear that there is a discrepancy
Wa with results on semi~solid medium at 37° on which I am no longer

: q satisfied that I can get even twin trails with the same combination

(TM2 - X SW541). Have done a few oxperiments using semi-golid

agar (no gelatin) incubated at 23°, but again no definite multiple

trails. I suspect reason may be that cells concerned are insuff-

iciently motile to move through soft agar, from a sort of dose

} eftect. Possibly also the disintegration of clump of replicated
chromosome fragment only cecurs at low temperature. All this is

on the assumption that there is either normal or no replication

4

NX

e
-T

of the imported gene; it's hard to believe there {fs an inter-
mediate cond“ition, Also so far as the results go there igs no
evidence yet of segregation of motility preceding clcreiirat oe
I have separated products of first division after isolation of
various celis, two or three of which gave all (or nearly all)
i motiles from both progeny, never motiles from one and non-motile
from the other; some of these single-cell isolations were made
x using non-log-phase cultures, so that the population at the time
the cells were captured was only about X2 that present when
phage added, This ks difficult to account for on a multi-strand
theory. Maybe there is a master strand which can pair with chromo~
5 some and a lot of others which can't, A propos this Z have you

7 Aa Qrelvcs

Of Comme J Lint Ata

seen any significant instances of swarm arising from a trail?
“\ I have seen it several times, but inretrospect I suspect that it

~w 8 may have been mutation, as the most clear-cut case was in SW573,
Sf where the trail was probably made of varalysed but H mutants.

&
+2 Te Another odd thing is that either gene is more than

. , S=plicate at times, or else phenotypic lag can be up to 5 generat-
AX a i ions, for in one of the cases where I had 5 trail-equivalents I

picked 18 motile cells from @® drop and got only non-motile progeny
from them all. Mortality still a nuisance, even in lysogenic

' § recipient.

Ky Vy

3 Y y I don't understand what you say (May 30th) about tri-
oN 3 phasics, but the ~n.o of your and Edwards paper has just

5 wi arrived, perbaps I will grasp it after reading that,

N Ou naciq yey Ma ato fle Euypriig, At Jeeny, the Mg
—_ Act iveh nm erat He Braue A Ob epere.. Ven ePumtin, Fie
we fra ye Jr rue, Zr / POE prrtli, Pre, tal Cefn I tuted,
Punt ceva tiny uryr—tel Hee pret On bgp Toa 7 Lege Gig
(have Gotint te Mere S. Ore 260% eet miele fort Ci
Yen , J Arcee)

Hare sat ator af Tom 4f- Ora lo Ccve § Le Caran Q<€.
eat 49 I hae wo Gal, nreckiinan, Were wy,
so that motile strains spread through it as they grow; but
because of its high gelatin content it solidifies at room
temperature, ‘To economise materials, this medium was used in
5 om. Petri dishes, about 10 ml. per dish. Inoculated plates
were inoubated uninverted for 10 to 18 brd, at 37°, allowed
to solidify at room temperature and examined for swarms or
other evidence of bacterial movement by inspection with a
Plate microscope and oblique transmitted light. In plates
incubated for 10 hrg, swarms were usually only about 5 em. in
diameter and, if not numerous, were still discrete, But after
18 hr. incubation even a single swarm wo..ld generally have

spread right across the plate,

BEHAVIOUR OF UNTREATED NON-MOTILE STRAINS GKCWN ON
i SEMI-SOLID MEDIUM

The growth of most of the non-motile strains on semi-
solid agar was strictly sonfined to the site of inoéulation ,
even when incubation at 37° was continued for 48 hr. or .
longer (Pl. 1, fig. 1). ‘Some strains, however, gave rise to
spontaneous swarms which on sub-culture were found to consist
of stable, motile, Heagglutinable mutants. In many such
strains this mutation was a rare event, so that only a few
plates showed swarms. In others the mutation was more frequent,
so that most or all plates showed spontaneous swarms, but they
usually appeared rather late, that is later than the 1£th hour
of incubation; this presumably indicates that only after
several hours of incubation waa the population of cells large
enough, in relation to the mutation rate, for there to be more

than a small probabLllity of a mutation oocurring. ~

 

CA few 0 strains were discarded because of their inconveniently
high rate of mutation.
Se himuri O strain S545 grew on the surfsce of
the sémiesolid medium at the site of inoculation and also

produced large numbers of micro-colonies, singly or in small
é
~

Groups, one mm. or less below the surface of the agar (Pl. 1,
Afig. 2). Sub-culture from these colonies yielded only stable
O forms similar to the original strain, and indeed H mutants
have never been obtained from strain SW545, This phenomenon
is being further investigated, but it seems probable that it
results from the spontaneous occurrence in this strain of a
very mall fraction of cells which are motile, and therefore
migrate through the agar, but which produce only non-motile
progeny. Micro-colonies separate from and deep below the

site of inoéulation have been seen in several other © strains,
but in smaller numbers than in SW545.

The centre, and therefore presumed point of origin, of
@ swarm Was frequently occupied by an area wf a few mm. in
diameter in which there was a dense crowd of micro-colonies,
of diameters inversely related to their distance from the
centre of the area, This appearance, which is illustrated
in Pl. 1, fig. 3, will be termed a "flare", Numerous single-
colony isolations from flares crowded with micro-colontes
always gave stable motile cultures similar to tlose obtained
from the periphery of the swarm, which indicates that the
mioro-colonies consist of cells which give rise to motile
progeny. Flares have also been obtained by inoculating a few
ceélla of certain partly rough motile straina onto semi-solid
agar. ‘The mechanism of the formation of mioro-colontles near
the point of origin of swarms is not clear; the trapping
of motile cells amongst agar fibrils perhaps plays a part, but
partial roughness or other physiological or genetic factors
which hinder the progression of the initial motile cells may
be involved,

The appearance of non-motile strains grown on semi-solid
agar has remained constant from experiment to experiment.
When a strain was treated with phage 22 propagated on that
same strain, the behaviour of the phage-treated célls, as to
production of swarms or isolated deep colonies, was exactly
the same as that of the untreated culture. This would bs
predicted if phage in itself has no effeat on hereditary
properties, such as motility, but can carry across genetic
material from its last host; when the genetic material of
the last host and the recipient is identical, its transfer
from the one to the other will have no detectable effect,

Behaviour of non-motile pireine trea ted wi th
transducin sates and incubated on semi-solid ar

In tests for transduction of motility a broth culture
of a non-motile strain was incubated with an equal volume
of a lysate for 30 min, to allow phage absorption. Three
loopfuls of the mixture, or the whole of the deposit obtained
vy centrifuging 1 ml. of the mixture, were then placed on
semi-solid agar and incubated at 37° for 10 to 18 hr. ‘The
appGarances produced when transuc tion of motility occurred
are illustrated in Pl. 1, fig. 4, which may be compared with
Ple 1, fig. 1, which shows the same recipient strain plated
without phage treatment, With combinations such aa that
illustrated, the inoculated area was occupied by dense growth,
and surrounded by a wide border of confluent swarming which
extended throughout the depth of the agar; when a smaller
number of phage-treated cells were inoculated, thia contin-
uous border was replaced by discrete swarms (Pl. 1, figs 5).
The centres of the swarms were sometimes marked by flares,
as described in the previous section. Unlike spontaneous
swarns, these induced swarms were for the most part well
developed by the 12th hour of incubation. Sub-culture from
the edge of a swarm always gave a pure growth of motile
organisms, and motility was retained on further sub-culture,
The growth obtained by sub-culture from the swarms differed
from the parent non-motile strain by its motility and H
agglutinability, and,
amounts detectable by agglutination teats) and carry phage Bz

only. This evidence is not unequivocal for the isolation of
both V-positive and Venegative strains from the same patient
(Kauffmann, 1934) suggests that antigen V may appear” or
disappear by mutation; and strains may be made lysogenic for
phage A2 in the laboratory, and presumably th«refore this may
also occur in nature, It is therefore uncertain whether this
group of O strains are all derived from a common 9 ancestor,
or whether they have arisen on two or more occasions by
mutations perhaps affecting tie same gene.

The high frequency with which lysates of 0 strains render
motile other © strains might suggest that the production of H
forma under tiese conditions is not dependent on the transfer
of genetio material from the lysed strain, as has been argued
above, but is a conseqience of exposure of an © strain to phage
22 grown on any heterologous hoat. However phage grown on a
spontaneous H mutant of 0 strain SHBAS evoked numerous swarms
from its pres parent 0 strain, Similarly a lysate of an
induced H form obtained from 0 strain SL15 evoked swarns from
its parent 0 strain. In each case a lysate of the 0 strain
itself had no effect. Tents with other © strains and their
spontaneous or induced H derivatives have given similar results.
Lach of these H derivatives, so far as we know, differed from
ita 9 parent only by the presence of flagella, and of the gene
determining this; but phage grown on the E derivative could
confer motility on the 0 parent, Thus motility may result
from exposure of an © strain to phage grown on a strain which
ig not an unrelated atrain, provided that 1t possesses the

gene which is lacking in the reolpient © strain.
Flagellar Antigenic Phase Latent in Cells of 0 Strains
Motlie forms of S. typhimurium, obtained from 0 strains

by transduotion, underwent diphasic flagellar variation
exactly as do normal H cultures of SS. typhimurium. ‘The first
culture from any one swarm gave growth which was agglutinated
atrongly by antiserum for one of the two H antigens, weakly
or not at all by antiserum for the other; that 1s, the swarm
was (predominantly) in one phase, The phase of indueed
swarms was determined by the recipient O culture, and not by
the inducing lysate. This was demonstrated by treating single
colony broth cultures of © straina with lysates of each of a
number of different strains. In each such experiment all or
nearly all of the swarms resulting from transduction (or
arising £8 by spontaneous mutation), proved to be in one and
the same flagellar phase, even when the lysate used was made
from an H strain known to have been almost entirely in the
opposite phase at the time of lysis, A single colony broth
culture, as used in these experiments, would heve been derived
from one cell by a sm&ll number of generations, and would
therefore be nearly homogeneous, even for e character which,
like flagellar antigenic phase, mute tes ata very high rate
(Stocker, 1949). Cells of an O atrain may tien be considered
to be in one flagellar antigenic phase or the other, though

this character is unexpressed in the absence of flagella.

ixperiments on Non-Mo tile Flegellated ("Paralysed") Strains
All motile Salmonelia strains have flagella, but several
non=motile fla gellated strains have been reported (
Colquhoun & Kirkpartick, 1932; Kauffmann, 1939; Edwards,
Moran & Bruner, 1946; Hirsch, 1947; Byew Priewer & Leifson,

1952). Hirsch's strain of S. paratyphi B and Friewer &
Leifson's strain of S, typhimurium