AGGLUTINATION OF MOTILE SALMONELLAS BY ACRIDINES

 

, ALECK BERNSTEIN ann JOSHUA LEDERBERG
Department of Genetics, University of Wisconsin, Madison, Wisconsin

Reprinted from Jounnat or Bacrmrio.ocy
Vol. 69, No. 2, pp. 142-146, February, 1985
Printed in U.S.A.

3é
Reprinted from JounNnaL or BacTERIoLocy
Vol. 69, No. 2, pp. 142-146, February, 1955
Printed in U.S.A.

AGGLUTINATION OF MOTILE SALMONELLAS BY ACRIDINES!
ALECK BERNSTEIN? anp JOSHUA LEDERBERG

Department of Genetics, University of Wisconsin, Madison, Wisconsin

Received for publication July 6, 1954

Genetic studies of the determination of Sal-
monella flagella have suggested that the alter-
native antigenic phases correspond to two dis-
tinct homology groups, ie., to factors at two
distinct genetic loci, one for phase 1 (‘specific
phase’), and one for phase 2 (“group phase’)
(Lederberg and Edwards, 1953). It has, however,
been generally understood that the two phases
were equivalent to one another in all respects
save their detailed antigenic structure and thus
would give no hint of such a dual determination
(Stocker, 1949). Sertic and Boulgakov (1936a, b)
had, however, reported slide agglutination of
smooth phase 2 cultures by acriflavine, while
phase 1 did not react. We have sought, therefore,
to confirm this generally overlooked observation
and to look for any further clues that might bear
on the reversible differentiation of the two loci
which control flagellar phase variation (Leder-
berg, 1954).

MATERIALS AND METHODS

Various serotypes were selected from stock
cultures listed previously (Lederberg and Ed-
wards, 1953). Each was carefully examined
for roughness which might interfere with the
flagellar reaction. The alternate phases were
obtained either by colony selection (Andrewes,
1922) or by selective swarming in motility agar
containing absorbed anti-H sera (Edwards and
Ewing, N.D.).

Previous investigators have used the slide
agglutination technique. Since this method does

1 Paper no. 558 of the Department of Genetics.
This work has been supported in part by grants
(C-2157) from the National Cancer Institute,
National Institutes of Health, Public Health
Service, and from the Research Committee,
Graduate School, University of Wisconsin, with
funds provided by the Wisconsin Alumni Re-
search Foundation.

2 Assistant bacteriologist, on leave of absence
from the Public Health Laboratory Service,
England.

not lend itself to quantitative titrations and is
liable to give false reactions, tube agglutination
procedures were preferred. Bacterial suspensions
were prepared from overnight cultures in Difco
penassay broth inoculated with single colonies
of the desired phase and incubated at 87 C.
These cultures were diluted to approximately
10° cells per ml and used alive, unless otherwise
stated.

The agglutinations were conducted in 10 by 75
mm culture tubes held at 37 C and read after 2
to 4 hours. Except for a slight increase in titer,
the readings were unchanged after the longer
period.

For precipitation tests on isolated flagella, the
double diffusion agar gel technique (Ouchter-
lony, 1949) was chosen after preliminary trials.
Flagellar suspensions were isolated from Roccal
treated bacteria by method 2 of Uchida, Sunakawa,
and Fukumi (1952) from the 6 and /,2.. phases
of Salmonella paratyphi B.

The acridines tested and some analogues are
listed in table 3. They were generally made up as
0.2 per cent solutions in distilled water with the
addition of dilute acetic acid if required.

RESULTS AND DISCUSSION

The experiments fall into three main groups:
those conducted on broth suspensions, those
performed with washed aqueous or saline prepara-
tions, and those utilizing isolated flagella.

Agglutination of living broth suspensions. The
strains were tested in parallel with specific
(absorbed) anti-H sera, acriflavine (0.2 per cent
in 1 per cent saline), and with Millon’s reagent
(White, 1929). The results are presented in table
1. Phase reversal was reiterated with several
strains, and only phase 2 was agglutinated so
long as the strain remained smooth.

It was also found that, unless fully motilized,
cells were usually inagglutinable by acriflavine
although still capable of reacting with anti-H
sera.

Besides confirming Sertic and Boulgakov’s

142
 

 

 

 

1955} AGGLUTINATION OF MOTILE SALMONELLAS BY ACRIDINES 143
TABLE 1
Agglutination reaction of strains in broth
Phase 1 Phase 2
Strain H Antigens Serum Acti Serum Acti
| Arie yin, ¢ Mill
Phi | paz} He |") pad] pha | dine “
S. typhimurium 4:1,2 + ~ - 8 ~ + + iS)
S. stanley a:1,8 + - - 8 _ + + 58
S. zega dizs + _ _ 8 - + + 8
S. london 1,v:1,2 + - - 5 _ + + 8
S. dar-es-salaam l,wre,n,Z18 + - - 8 - + + 8
S. wien bil,w + - - 5 _ + - 8
S. abony bie,n,x + _ + R ~ + + R
S. paratyphi B b:1,2 + _ _ 8 - + + 8
S. typhi O 901 — {d) - - 8
SWwo6é6 — {[d] _ - 8
SwW926 1,2:¢,n,x + 7 | + iN) - + + 8

 

 

 

 

 

 

 

 

 

Anti-H sera were used at a final dilution of 1 in 1,000. The titers of the sera ranged from 1 in 3,000
to 1 in 300,000. The acridine dye was 0.2 per cent acriflavine.

S and R indicate smooth and rough, respectively, as determined by the Millon reaction (White,
1929). Precipitation alone was taken to indicate roughness.

+ and — indicate presence or absence of agglutination.

findings, these results raise one other point. The
antigenic complex J,w which appears as phase 1
in some, and as phase 2 in other, serotypic for-
mulae reacts with acriflavine in all cases as if it
were a phase 1 complex. The complex / ,2 . . usu-
ally represents phase 2 but is present as an anom-
alous phase 1 in some bacteria (Lederberg and
Edwards, 1953). These organisms are agglu-
tinated by acriflavine in both phases. With these
exceptions phase 1 homologues are stable and
phase 2 homologues are agglutinated by the dye.

Conditions of agglutination in broth. Cells of
both phases were washed and suspended in
broths which had been adjusted to pH values
from 2 to 10 inclusive by the addition of concen-
trated HCi or NaOH. Both phases agglutinated
spontaneously at pH 5; above this value the
reaction was unaffected. At pH 4 and below no
agglutination of phase 2 cells occurred. If cells
exposed to these low pH values were washed
and resuspended in normal broth at pH 7, they
were still inagglutinable. This suggests that some
irreversible change has taken place in the cells
during their exposure to these low pH’s. It is
likely that the flagella are the site of this action
since they are known to be acid sensitive (Weibull
and Tiselius, 1945).

Organisms heated to 100 C for 15 minutes
were no longer agglutinable by either anti-H

sera or acriflavine, a correlation which would also
implicate the flagella. Furthermore both macro-
scopically and microscopically the agglutinates
resembled those normally formed by anti-H
sera,

Since it is inconvenient to work with large
volumes of living pathogens, some antiseptics
were tested for interference with the dye agglu-
tination reaction. Mercurials and formalin
reacted directly with acriflavine to produce a
chrome yellow precipitate. If the cells were
then washed and resuspended in fresh broth,
there was no interference with the reaction.
Phenol tended to inhibit the agglutination unless
removed by washing. Roccal (benzalkonium
chloride, Winthrop-Stearns, Inc., N. Y.) had
no effect on the reaction even at bactericidal or
higher concentrations, and was used whenever
living cells were contraindicated (Anderson
et al., 1952).

The effect of broth on the agglutination. The
previous experiments were conducted in broth.
To determine whether electrolytes were needed,
twice washed cells were suspended in distilled
water and tested with 0.2 per cent aqueous
acriflavine, Contrary to the results in broth all H
suspensions were agglutinated equally by the
dye, but O cells were unaffected. Apparently
this reaction, unlike serum agglutination, does
144

not require the addition of electrolytes for visible
aggregation unless the dye itself (at m/4,000)
acts in this respect.

This loss of specificity for the two phases in
distilled water might be ascribed to (1) a reagent
in the broth, (2) a bacterial metabolic product,
or (3) the washing procedure.

Phase 1 cells were washed twice with distilled
water, and four aliquots were then resuspended
in distilled water, fresh broth, the supernatant
broth from the original culture, and the super-
natant broth from a phase 2 culture, respectively.
Only the aqueous suspension was found to be
agglutinable by acriflavine. This supports the
hypothesis that normal broth contains an agent
capable of inhibiting the agglutination of phase 1
cells by acriflavine. It remains to determine
what the agent is and whether the inhibition is
absolute or not.

Inhibiting agents. An empiric examination of
certain constituents of penassay broth was made
to see whether they would inhibit the reaction.
Aqueous phase 1 cells were used as indicators.
The results are presented in table 2: the purines
and pyrimidines in broth could be part at least
of the inhibitor. Broth, however, is a very complex
medium, and the participation of other com-
ponents is not excluded.

It was further found that one could overcome
the inhibition of any of these agents by adding
further acridine. In a titration involving adenine

TABLE 2
Activity of inhibiting agents

 

 

 

 

 

 

 

Agent Dilution | Effect
Saline .....0 000.6 40/ +
Yeast extract (Difco)............ 1,600 | +
Yeast nucleic acid............... 3,200 | +
Peptone (Difco)................. 200 +
Tryptone (Difco)................ s00 | —
Guanosine....................005 6,400 | +
Xanthine...................000.. 12,800 +
Thymine........................ 12,800 | +
Adenine.................00000005 12,800 | +
Glycine... 0.0.0... eee eee 80; —

 

The dilution is expressed as the reciprocal of
the lowest concentration inhibiting the reaction,
or of the highest concentration not causing pre-
cipitation of acriflavine in the case of inactive
compounds.

+ and — indicate presence or absence of in-
hibition of acriflavine agglutination.

ALECK BERNSTEIN AND JOSHUA LEDERBERG

[von. 69

hydrochloride and aminacrine (see below) with
aqueous phase 2 cells as indicator, it was found
that for every mole of adenine two moles of
acridine had to be added to overcome the inhibi-
tion. That there is a direct reaction between
acridine and inhibitor is suggested, but competi-
tion for bacterial sites cannot, on the basis of this
evidence alone, be excluded.

Titration of acriflavine activity. Doubling dilu-
tions of acriflavine were titrated against suspen-
sions of phase 1, phase 2, and O cells. Preliminary
titrations had shown that, unlike serum reactions,
the concentration of cells did not, within broad
limits, affect the end point.

Tn broth, O cells were not agglutinated by con-
centrations up to 1 per cent acriflavine, this being
near the limit of solubility of the acridine. Phase
1 cells were agglutinated by 1 in 100 acriflavine,
and phase 2 cells were agglutinated at concentra-
tions as low as 1 in 3,000.

In water, the O cells still remained inagglu-
tinable, but the end points for the other two cell
types were changed. Phase | cells were now ag-
glutinated by 1 in 6,400 and phase 2 cells by 1
in 12,800 final concentrations of acriflavine,
respectively.

Activity of other acridines. Several other acri-
dines and some analogues were examined for
their effects in the presence and absence of broth.
The results are shown in table 3 from which it
can be seen that the majority of the acridines
behave like acriflavine.

Certain of the reactions are of interest when
one considers the theory of acridine bacteriostasis
postulated by Albert (1951). He has suggested
that a certain “envelope” size is required for the
acridine to be effective as an antibacterial agent.
Tetrahydroaminacrine would be inactive because
the saturation of one of the aromatic rings results
in a nonplanar molecule. Similarly, aminoquino-
line is too small, but the addition of a styryl
side chain increases the envelope to the size re-
quired for an active antibacterial agent. The
inactivity of the hydroxy compound, on the other
hand, was thought to be due to its zwitter-ion
structure.

In general the broth reactions agree with this
theory; the action of tetrahydroaminacrine on
aqueous phase 2 cells is, however, discordant.
Whether there is a mechanism of acridine ac-
tivity on flagella different from that involved in
bacteriostasis, or whether the acridine compounds
 

 

 

 

1955) AGGLUTINATION OF MOTILE SALMONELLAS BY ACRIDINES 145
TABLE 3
Acridine dyes and analogues investigated with their activities
Suspensions
Compound Formula (qumbering. according to Broth Water
QO {Phase 1 |Phase 2 OQ |Phase 1 |Phase 2
Acriflavine* mixture of proflavine and | — - + - + +
euflavine
Proflavine* 2,8-diaminoacridine - _ + _ + +
Acridine orange (a)* 2,8-diamino-bzs-dimethyl- _ + + - + +
acridine
Acridine orange (b)* 2,8-diamino-bis-dimethyl- - - - _ + +
acridine
Euflavine 2,8-diamino-10-methylacri- - _ + ~ + +
dinium hydrochloride
Aminacrine 5-aminoacridine - - + _ + +
Tetrahydroaminacrine 1,2,3,4-tetrahydro-5-amino- - - - _ + +
acridine
Hydroxyaminacrine 2-hydroxy-5-aminoacridine t t tT - + +
Aminoquinoline 4-aminoquinoline _ 7 - - - -
Styrylaminoquinoline 4-amino-2-styrylquinoline t t t _ + +

 

 

 

 

 

 

 

 

* These compounds were obtained from commercial sources and the remainder through the kindness

of Dr. A. Albert and Dr. 8. D. Rubbo.

+ and — indicate the presence or absence of agglutination.
¢ Indicates that the compound was precipitated in the presence of broth.

would display different antibacterial activities
if examined in buffered saline rather than in
broth has not been decided.

% The reason for the difference in activity of the
two preparations of acridine orange is not certain.
These were both commercial preparations and
may well have contained impurities or have dif-
fered considerably in dye content.

Bactericidal tests with acriflavine and euflavine
were conducted on phase 1 and 2 Salmonella
typhimurium cultures. No significant difference
in the rate of killing was found.

Absorption of acriflavine by bacteria. Concen-
trated suspensions of O, phase 1, and phase 2
cells were added to acriflavine solutions of dif-
ferent concentrations and pH. Roccal treated
suspensions had been estimated photometrically
so as to contain equivalent amounts of bacterial
substance. After a 5 minute exposure at 37 C
the cells were removed by centrifugation, washed,
and the acriflavine content of both cells and
supernates estimated with the Coleman spectro-
photometer at 445 my.

Spectrophotometry with acridine compounds is
liable to give misleading results owing to their
high degree of fluorescence (Albert, 1951).

However, no marked differences could be found
in the amount of dye retained by the different
types of cells under identical conditions.

Reactivity of isolated flagella. Tube agglutination
tests were performed with aqueous suspensions
of phase 1 and phase 2 flagella. Acriflavine at a
concentration of 1 in 6,400 agglutinated phase 2
flagella, but a concentration of 1 in 1,600 was
needed for the agglutination of phase 1 flagella.
It was found that more concentrated flagellar
suspensions were needed for agglutination by
acriflavine than by specific antisera. This is in
accord with the earlier findings on poorly motile
cultures prior to the selection of full motility by
passage through semisolid agar.

Plate precipitation tests with acriflavine or
aminacrine were uniformly negative although
similar tests with specific sera gave strong pre-
cipitin lines. Similarly when flagella were disag-
gregated with dilute HC] (Weibull and Tiselius,
1945), they no longer reacted with either serum
or acriflavine in tubes. The treated flagella still
responded to serum in plate precipitin tests. The
merging of precipitin lines showed no difference
between the native and treated flagella prepara-
tions.
146

These results confirm the impression gained
earlier that the site of acridine action is the
flagellum. It is probable that acid treatment
gives rise to particles too small to form sedimen-
table aggregates with acridine but not with sera.
This would also explain the lack of acridine
precipitin lines since intact flagella would be too
large to diffuse through the agar whereas the
degradation products would be too small to be
sedimented (Gard et al., in press).

These studies have not settled the physico-
chemical basis of flagellar differentiation. Thus
while Sertic and Boulgakov’s original observa-
tions are confirmed, closer study has shown that
the differences between the phases must be con-
sidered on a quantitative, not qualitative, basis.
However, the genetic differentiation of the phases
as the products of distinct loci is paralleled by
this difference in agglutinability, whatever its
physical foundation.

Acriflavine has been widely used as a criterion
of smoothness (Braun and Bonestell, 1947). The
present findings reinforced the caution with
which this criterion must be interpreted, espe-
cially when dealing with Vi+ (Hirsch, 1937) or
with motile organisms.

SUMMARY

Sertic and Boulgakov reported that motile
Salmonella in flagellar phase 2 were agglutinated
by acriflavine. Their observations have been
confirmed for the specific conditions they de-
scribe.

Motile smooth Salmonella sp. are agglutinable
by acridine dyes. Phase 2 reacts at distinctly
lower dye concentrations than does phase 1. The
agglutination resembles the reaction with anti-H
serum in appearance and sensitivity to heat.

Examinations of the reaction in penassay broth
and in aqueous suspensions show that yeast
extract contains agents that inhibit agglutination,
probably purines or pyrimidines. Electrolytes,
in excess of the dye itself, are not required.

Experiments on isolated flagella indicate that
these are the site of the reaction.

REFERENCES

AuperT, A. 1951 The acridines; their prepara-
tion, physical, chemical, and biological proper-
ties, and uses. EH. Arnold, London, England.

ALECK BERNSTEIN AND JOSHUA LEDERBERG

[voL. 69

ANDERSON, R. E., Stren, L., Moss, M. L, aNnp
Gross, N. H. 1952 Potential infectious
hazards of common bacteriological techniques.
J. Bacteriol., 64, 473-481.

ANDREWEs, F. W. 1922 Studies on group ag-
glutination. J. Pathol. Bacteriol., 25, 515-
521.

Braun, W., aNnD Bonestety, A. E. 1947 Inde-
pendent variation of characteristics in Bru-
cella abortus variants and their detection.
Am. J. Vet. Research, 8, 386-390.

Epwarps, P. R., anp Ewinc, W. H. N.D. A
manual for enteric bactertology. Federal
Security Agency, Public Health Service,
Communicable Disease Center, Atlanta, Ga.

Hrascu, W. 1937 The agglutinability by Trypa-
flavine of B. typhosus and its relation to the Vi
antigen. J. Pathol. Bacteriol., 44, 347-355.

Leprersere, J. 1954 Phase variation in Salmon-
ella. Rec. Genetics Soc. Amer., 28, 51.

LEDERBERG, J., AND Epwarps, P.R. 1953 Sero-
typic recombination in Salmonella. J. Im-
munol., 71, 507-515.

OucHTERLONY, O. 1949 Antigen and antibody
reactions in gels. Acta Pathol. Microbiol.,
26, 507-515.

Sertic, V., AND Boutcakov, N-A. 1936a@ Sur la
sensibilité des souches d’£berthella typhi au
bactériphage, en relation avec les caractares
antigénique. Compt. rend. soc. biol., 122,
35-37.

Sertic, V., aNnD BouLtcaKkov, N-A. 1936) L’ag-
glutination par la Trypaflavine des Salmonella
de structure antigénique flagellaire non-

spécifique. Compt. rend. soc. biol., 123,
951-952.
Srocxer, B. A. D. 1949 Measurements of rate

of mutation of flagellar antigenic phase in
Salmonella typhimurium. J. Hyg., 47, 398-
413.

Ucuipa, H., Sunakawa, 8., anp Fuxumi, H. 1952
Studies on the bacterial flagella. I. Methods
of purification. Japan. J. Med. Seci., 5,
351-355.

WEIBULL, C., AND TisEuius, A.
hydrolysis of bacterial flagellae.
Mineral. Geol., B20, No. 3.

P. B. 1929 The Salmonella group.

System of bacteriology. Med. Research

Council (Brit.), 4, 86-158.

1945 Acid
Arkiv Kemi,

Wuite,