Paper Chromatographic Identification of Polypeptidic Gram Positive Inhibiting Antibiotics1

NEVA SNELL, KoSUKE IJICHI

AND

J. C. LEWIS

W1'estern Utilization Research Branch, Agricultural Research Service, United States Department of Agriculture, Albany, California

Received for publication August 1, 1955

One-directional paper chromatography followed by bioautography has proved useful for differentiating antibiotics produced by Bacillus species in connection with a screening program. For purposes of comparison with unidentified antibiotics, Rf values have been determined with a number of solvents for most of the available Bacillus antibiotics active against gram positive bacteria and several of the polypeptidic antibiotics produced by other microorganisms. Goodall and Levi (1946) first used this technique in separating the penicillins. Since that time, it has been used with increasing scope. Several of the polypeptidic, gram positive inhibiting antibiotics have previously been separated chromatographically (bacitracin, tyrothricin, and viomycin by Fischbach and Levine, 1953; licheniformins A, B, and C by Callow and Work, 1952; catenulin and neomycin by Davisson et al., 1952; the various actinomycins by Vining and Waksman, 1954; and actinomycin and streptothricin sulfate by Ammann and Gottlieb, 1955). EXPERIMENTAL METHODS Ascending chromatograms on Whatman No. 1 paper were developed in glass jars containing 1 cm of solvent mixture. Solvent compositions are based on volume measurements of each of the components. Effects of small changes in solvent concentration were sufficiently large to warrant preparation of fresh mixtures for each use. A few ml of ammonia in a small beaker provided ammonia in the atmosphere when needed. Samples were spotted 2 cm above the base of the paper, with a hypodermic syringe and 25-gauge needle, and dried thoroughly before placing in the solvent. After the solvent had migrated to the top of the paper (usually 12 cm) the chromatograms were dried thoroughly. Exposure to steam intermittently during the drying ensured adequate removal of acetic acid. The migrated antibiotics were detected bioautographically in a manner similar to that of Goodall and Levi (1946). Papers were placed for one-half hour on Difco nutrient agar plates seeded with the test organ1 Presented in part before the 53rd General Meeting, Society of American Bacteriologists, August 1953.
13

ism, usually Bacillus megaterium NRRL B-938. The plates were incubated at 35 C for 17 hours. Purified antibiotics were dissolved in water, except for gramicidin and tyrocidine* HCl, which were dissolved in 95 per cent ethyl alcohol. Aterrimin (Alderton and Snell, unpublished work), available only in partially purified form, was dissolved either in dry butyl alcohol or in absolute ethyl alcohol. Crude Bacillus cultures were tested directly or as crude extracts. One mg of the purified antibiotics per ml was a practical concentration for detection with Bacillus megaterium, with the followving exceptions (mg per ml): actinomycin, 0.1; cinnamycin, 0.5; gramicidin, viomycin, and vivicil, 2; licheniformin A5 and tyrocidine. HCl, 4. Sources of antibiotics are as follows: actinomycin, cinnamycin, grisein, laterosporins A and B, polypeptin, streptolin, viomycin, and vivicil, Northern Utilization Research Branch; catenulin, Chas. Pfizer Co.; licheniformin AS, Dr. L. H. Kent, Ministry of Supply, England; nisin, Bengers, Ltd.; subtilin and aterrimin, this laboratory; others, commercial products.

RESULTS Various concentrations of t-butyl alcohol, n-butyl alcohol, methanol, or acetone with 2 to 10 per cent acetic acid have given the best differentiations of the purified antibiotics included in this study, t-butyl alcohol being especially useful. Ammonia in the at-

S2
0
55:6:
39

60:6
34

65:6 29

70:6:76
24

20

t-butyl alcohol: ocetic ocid: H20

FIG. 1. Effect of solvent composition on Rf values for subtilin

14
I.C
rv

NEVA SNELL, KOSUKE IJICHI AND J. C. LEWIS

-

[VOL. 4
1.0
f. .

.42cinomycin
AAerrimin
1

Gromicidin

kjfenicilfin G
5
.9c
I

0^e
o (a
_ o,
0

I.0

SActinomycin _ Aterrimin (also at

0.r Penicillin G
0) 0)
J

-,of'
o 0\

1Gromicidin
IeniciIin

(also at
G

.85

Act/inom,cin
(also at .1)

a
N4
I

.7 Actinomy,cn (also at 0)
-

0
00
-

I;
N

.{Loferosporin A (ryrocidine -HCI

Q.

,
0

-|
0

Aterrimin (streaked 0-.35)

/

ryrocidine HCI
-

.8a

Loterosporin B

50 (no0

tcAerrimin (also at I)
Loterosporin

rActinomycin I(also at.7)

0I -(GromicidMn
I.0

U
a

Actinomycin m t.85) .L (also

Gromicidin (also at I) Loterosporin A

\

9
o

C0-Gromici'dln
1.0 -Loterosporin A

)
0

.90 .85

Laterosporin

Loterosporin 6
A

I)E
o
0
0

ON
.45
0

I
Tyrocidine * HCI
Aterrimin
(streaked 0-.35
0L Gramicidin
10

.7C - Polypept/n
)
.6'
IC) c\j
-

1.0r
.90 -Tyrocidrne HCI
0 o

E I

o Z

L oersoi
1.0 Polypeptin

-OterospoMin

Socltrocln

e
N

.75 -Loterosporin A -Loterosporin

(streaked)
Bocitrocin

0 0 0

IIo.60 CQ

8oci/racin
Vv/vc/I

.0

ezo.2
0

o .35

Volypeptin

Ci.
0
c

-C
I z
.

.55 -Bacitrocin

N
0

.25 - Vivicil

a)
0

0I

oI
0I

OLoterosporin

O0

.-C

0 0
I-)

J
.45
-

o~~~~0

V/ivci/

a)

-a
U4)

C..

o0

o eD I jI o~

oo
-

.6k Sub/il/n 4 O * innomycin

0
= CY 0

Oi

O-

Vivicid

.3C

CInnomycMn

'

~~Of
1.0
0

/.
Cinnomycin

.31- Subti/in
'

Subtilin .0 _ =

(streaked))
I

90F Subti//n
I

.0 L

Nisin

at .25)
-O

.98
0 -

1.0

,o
-

Nisil, Subti/ij

Subti/in

I
.08
-

cidi
cJ
o

t- .85 ,-Licheniformin i-Grisein

.65
- Viomycin

.75

.35 Streptothriciin h/CI N25isin (also I

.20 Licheniform,pin A5 }
OL

Lr

7-

HC;I

31

f
c

*-

'4

Grisein
Nisin
1

-Cotenulin
-Strepto/in

.o~~~~~~~~~ta. 3 F. 0~~~~C
.65L Streptothricmn .C/

WC .X03 L/..bOenuln Licheniformin A CI . I0M.

15 1

o
a

.60r Viomycin

.251- Nisin
0
L

1/omycin

Streptothricln *H

streptolin

i~

CaI CotFnv/in

;g

Licheniformin A5

FIG. 2. Key to chromatographic identification of gram positive inhibiting polypeptidic antibiotics and the nonpolypeptide aterrimin. The order for inspection of results with various solvent mixtures is from left to right as is indicated by the arrows.

1956]
. .

PAPER CHROMATOGRAPHIC IDENTIFICATION OF ANTIBIOTICS

-

10~

.>

6 C; C;

0t

00 d

> 666co

o.3-

%-~~~~~~~t0
1.4
I

0C; '-C

10

I
o

1-

l
~ ~ ~ ~

0 0 0 Ot-o o 6o o 0 1010~~~~~~~~~~~~~~~~~~~~~~~~~O1
0. 0
~

00000

;>

O I

000000- d

'-4

@~~~~~~~Cto
D
1.

0O

_; * * 0* 0 0*_{
15
am

10-

xo 10

xo x CO N --0 C

kI; O 0)CD
.. .

toO a L
0)t-

o 010
O-4

10

CO

CD

C0

10

0)

~~~~~~~~~~C t'- co

.C)C 00

0)

m CO

C) .Om

LO 0 10

(M =

10

000 000o
CO

10

'" COC5)

1C

10O

co

10

CD1 CO0 )-4 0 &~C

0

g~~~~~~~~~~~~~~~~~0
e o
.

0 -

U:

00

~~

~~~~~~~~~~~~~~~~~~c

LO

LO

LO

010*

0
4

0cO et t-

--0

T"4*-

CO-0

'-4

I0

'-4'-4-

r-

'--q

r-

-_

'_1

C
_ C

0 _ 00

0
1-

Ct100
~

10
* O 1
O
s b * O-

10
~

co ~

A*

O ~~

~~~~~ ~ OC ~~~~~~MX OCXo Co

* _LO 0* 1

O Co 00000
I

0 00.1

0o CO *

. 00

1I

fs

s A~~~~

1010~~~~~~~~~~~~~~~~0 O 0)* t OC OCO O OC *e

i
OCO
|
0)

,_

|_ -

0)

_ _ -OO

0)0

eIas
0 1
bt~O

._
.9
C

0 C)e1 ) C)C) 00 t_ 3

O0o 10
O O

10 Lo-CbC
s__0

Oc
00

.O

u;
.

~ ~

C; '4-4 CQ

U: '-.

UU e-D C'*

o e.4

14~ ~ ~ ~ ~ ~ ~ ~ 0

4)~ ~ ~ ~ ~ ~ ~0k ~k

kt

'g C)
-

Qt

4~

_O|

~~

~~

CS1 O O

~~~ ~

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4

CO

Cl

O- *

C _

00

_;acO _
00 S*o

_COO;|
0

a,

f.

. Z000 rn ~ 000

0C

34
-4

CO

"I

I
D

PA

I 1

E
10
c: t-m
C)

*- 4
4)

-4

103
a)

4-

DL
.-

0 3
---*
0 2

Go 3 a)4

v)
.....

....
0

-a-

c*...R-

.q
-4- D -4
-4

.04 1,4 Soa)00 aaSs s to 09 oo

a SO

16

NEVA SNELL, KOSUKE IJICHI AND J. C. LEWIS

[VOL. 4

mosphere in the solvent jar has made possiible a few differentiations which were not obtained with the acidic solvents. Careful control of solvent conLcentration is important. Figure 1 illustrates an effect of relatively small changes in t-butyl alcohol concentratic)n. Similar effects may be noted in many instances, awithin the critical range of other solvents. The presence of acetic acid tends to imprc)ve resolution of some antibiotics, but the acid must b e removed before bioautographic analysis. Interfering quantities of acetic acid for three test bacteria (as nnuch as 2 microequivalents per square centimeter) wer e retained by paper dried in a hot-air cabinet, or at roo:m temperature when the relative humidity was low. ThLe titration curve was typical of acetic acid, which indiicates that the inhibition of the bacteria was due to res3idual acid rather than to the production of toxic substances. However, paper saturated with 10 per cent facetic acid has been rendered nontoxic and acid-free within 3 hours by evaporation in humid laboratory air or by exposure to steam during drying. Streaking of the antibiotic over part or even the entire length of the paper occurs under certasin circumstances. This may be avoided sometimes b3y use of a lower concentration of the antibiotic and sometimes by choice of a more appropriate solvent mixlture. A key to some of the gram positive inhibiiting polypeptidic antibiotics is shown in figure 2. Variious others were not readily available. Solvent 1 provedI the most satisfactory for giving a good initial spread of the entire group of antibiotics. After determination of rate of movement in Solvent 1, other appropriate cc)mparisons may be made as is indicated by the arrows. Table 1 gives the Rf values for these antib: iotics with a larger number of solvents. Two distinct inhibition zones have beeri observed for a few of the antibiotics with certain soilvent mixtures, as is evident from the data in table 1. Lack of
Orgonism Inhibited
-

homogeneity, the most obvious explanation, may be responsible in at least some of these cases. Inasmuch as our interest was not in differentiating within families of antibiotics (for example, between the bacitracins), the use of such solvent mixtures was generally avoided for the antibiotics in question in preparing the key (figure 2). By the use of several test bacteria and solvents, various stocks of Bacillus have been shown to produce 2 to 4 antibiotics simultaneously. The production of only one antibiotic appears to be the exception rather than the rule. Figure 3 illustrates the presence of at least 3 antibiotics in each of three typical Bacillus whole cultures. Fractionation has further established at least two of these activities for B. subtilis NRRL B-1471 (aterrimin-producer) and 4 distinct activities for B. subtilis NRRL B-1474.
DISCUSSION Multiplicity of antibiotic production obviously complicates attempts to identify antibiotics in unfractionated material. However, even in whole cultures similarities or differences between known and unidentified antibiotics may be noted by means of chromatography. Two-dimensional chromatography was useful in clarifying results with crude preparations. Exact Rf values should not be credited with significance in evaluating impure preparations, as movement rates of antibiotics may be influenced by the presence of other constituents of the culture. The viscosity of whole cultures may cause failure of the sample to sink completely into the filter paper, and thus retard movement because of delayed contact with the solvent. In our experience, the combination of antibiotically inactive whole culture broth with pure antibiotic preparations has had a relatively minor effect on movement rates (0.2 Rf or less), but the possibility of a greater effect should be kept in mind. Even preliminary fractionation of cultures may increase precision of chromatographic results. In view of the many factors which may influence exact Rf values, we have attempted to find solvent mixtures which distinguish by wide differences. This goal was not realized to the desired degree in some instances, most notably streptothricin HCl vs. either grisein or viomycin, catenulin vs. streptolin, and aterrimin vs. gramicidin or tyrocidine HCl. Inasmuch as most Bacillus antibiotics which have been characterized to date are polypeptides (Fuller, 1955), unidentified antibiotics from Bacillus probably should be compared also with polypeptidic antibiotics from other sources. For this reason, we have included other readily available gram positive inhibiting polypeptidic antibiotics in this study. Aterrimin, although not a polypeptide (Alderton and Snell, unpublished data), has been included because it is produced by a Bacillus.

Organism Inhibited
-B

Organism Inhibited
-B megoteriumwhc

B.
v.

megoterium ond Mpyogenes

/bus

megoterium

only

-M. flavus only

r5
Mpyogenes
v

a/bus anly

-a.megoternLum
B. megoterium
.e _B and Mf/ovus B-1474

-c-B. megoterium
~
only

megoterivm

B-1471

6633

FIG. 3. Multiple antibiotic production by thLree Bacillus subtilis cultures: NRRL B-1471 (aterrimin-produtcer), NRRL B-1474, and ATCC 6633 (subtilin-producer). (Solb vent: t-butyl alcohol, acetic acid, H20-74:3:25.)

19561

CLOUD CHAMBER TECHNIC FOR STUDYING AEROSOLS

SUMMARY

17

A key based on paper chromatography and bioautography is proposed for differentiating most of the gram positive inhibiting antibiotics produced by Bacillus, as well as polypeptidic antibiotics produced by other microorganisms.
REFERENCES AMMANN, A., AND GOTTLIEB, D. 1955 Paper chromatography of antifungal antibiotics. Appl. Microbiol., 3, 181-186. CALLOW, R. K., AND WORK, T. S. 1952 Antibiotic peptides from Bacillus licheniformis. Lichenformins A, B, and C. Biochem. J., 51, 558-567.

DAVISSON, J. W., SOLOMONS, I. A., AND LEES, T. M. 1952 Catenulin, a new antibiotic. Antibiotics and Chemotherapy, 2, 460-462. FISCHBACH, H., AND LEVINE, J. 1953 The identification of the antibiotics. Antibiotics and Chemotherapy, 3, 1159-1169. FULLER, A. T. 1955 A new antibiotic of bacterial origin. Nature, 175, 722. GOODALL, R. R., AND LEVI, A. A. 1946 A microchromatographic method for the detection and approximate determination of the different penicillins in a mixture. Nature, 158, 675-676. VINING, L. C., AND WAKSMAN, S. A. 1954 Paper chromatographic identification of the actinomycins. Science, 120, 389-390.

Studies of Aerosols with

I.

Simple Cloud-Chamber Technic'

The Evaluation of a Technic for the Rapid and Convenient Determination of the Survival of Air-Borne Microorganisms
AND

CHARLES W. GRIFFIN, H. LUCILLE KANTZES, PAMELA M. LUDFORD

Received for publication July 29, 1955

MICHAEL J. PELCZAR, JR.

Department of Bacteriology, University of Maryland, College Park, Maryland

Since the realization that the atmosphere could serve as a vector for the transmission of disease-producing agents, many investigations have been performed to determine the microbial flora of air as well as the environmental factors which affect survival. Bacteriological evaluation of the latter presents a special problem in instrumentation and technic. The purpose of this investigation was to develop a simple, small-scale, cloud chamber technic for the study of bacterial aerosols as influenced by physical environmental conditions such as relative humidity, temperature, and solar radiation. To accomplish this, a small portable chamber was desired, one that was simple in design, convenient for adjustment of or exposure to variations in physical conditions and capable of providing reproducible results. The practicability of such equipment was suggested by the work of Van den Ende et al. (1948), who used small, round-bottom, quartz flasks to study the bactericidal effect of ultraviolet light on bacterial aerosols. For the analysis of the concentrations of air-borne bacteria, a number of air-sampling technics (based on two methods of collection) have generally been employed by various investigators. The first includes those devices which collect organisms in a liquid and break up any clumps of bacterial cells, such as the bubbler-pump method of Wheeler et al. (1941) and
' This investigation was perfonned under contract for the Chemical Corps, Camp Detrick, Frederick, Maryland.

the atomizer-bubbler apparatus of Moulton et al. (1943). The second type of sampler impinges the bacteria directly onto a solid medium without attempting to break up clumps and is represented by the Wells' air centrifuge (1933), the funnel device of Hollaender and DallaValle (1939), the slit sampler developed by Bourdillon et al. (1941), the sieve device of DuBuy and Crisp (1944) and, more recently, the application of the molecular filter membrane by Goetz (1953). Another recent technic used for analyzing aerosols is the electronic counter developed by Gucker and O'Konski (1949). In the present investigation, a syringe-dilution method (a liquid-collection technic) and the slitsampler method (a solid-impingement technic) were evaluated for their suitability in the analysis of static bacterial aerosols produced in the small cloud chamber to be described. These technics have been adapted from those investigated and recommended by personnel of the Chemical Corps, Camp Detrick.
EXPERIMENTAL METHODS The cloud-chamber technic employed in our studies consisted essentially of the production of a bacterial aerosol in an inverted, one-liter, round-bottom flask (aerosol chamber) and the transfer of a portion of this aerosol to another inverted one-liter flask (transfer chamber) from which analyses of bacterial survival were made. This type of chamber was selected because