0022-202X/78/7106-0363$02.

00/0

Vol. 71, No. 6

Printed in U.S.A.

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY, 71:363-365,1978

Copyright 1978 by The Williams & Wilkins Co.

Differential Susceptibility of Propionibacterium acnes, P. granulosum and

P. avidum to Free Fatty Acids
HONG LIOE KO, M . S C , PIOTR B . HECZKO, M.D., AND GERHARD PULVERER, M.D.
Institute of Hygiene, University of Cologne, Cologne, West Germany, and Institute of Microbiology, Medical Academy, Krakow, Poland

The susceptibility of 98 Propionibacterium acnes, P.

granulosum and P. avidum strains isolated from acne
cases and appropriate reference strains to decanoic, dodecanoic, octadeca-9,12 dienoic, and octadeca-9,12,15 trienoic acids was investigated using minimal inhibitory
concentration estimation by an agar dilution technique.
The tested acids showed their relative antibacterial activity in the following order: C18:3 A 9,12,15 > C18:2 A
9,12 > C12:0 > C10:0
Differences between minimal inhibitory concentration
values for P. acnes, and P. avidum versus P. granulosum
strains were demonstrated in the cases of octadeca-9,12
dienoic, and octadeca-9,12,15 trienoic acids.
The anaerobic diphtheroids are typical resident bacteria of
the human sebaceous follicles [1,2], and are almost certainly
involved in the pathogenesis of acne vulgaris [3]. After multiplication in the lumen tbey move out from the follicles over the
skin surface with the sebum forming dense populations in
sebum-rich areas of the skin [4].
Strains of the anaerobic diphtheroids isolated from the human skin are at present uniformly classified into the genus
Ptopionibacterium [5]. However they have shown biochemical,
serological, and genetic differences that enabled the definition
of 3 species, listed in the latest edition of Bergey's Manual
[5-8]. Representatives of all 3 species can be found on the skin
surfaces of normal individuals and also of acne patients in
different proportions related to the sampled skin region, age,
and to the presence of acne lesions [2,3].
Human skin surface lipids have been found to contain a
complex mixture of different fatty acids of various chain structures and position of unsaturation [9]. They originate from
sebum triglycerides hydrolyzed in the pilosebaceous follicles by
the action of the Upases of Propionibacterium [10] and may
play a significant role in the etiology of the inflammatory lesions
of acne vulgaris [11].
On the other hand human skin surface lipids possibly play a
role as antimicrobial substances controlling colonization of the
skin by pathogens like Staphylococcus aureus. Streptococcus
pyogenes, and Candida albicans [12]. The antimicrobial activity of the lipids is most probably related to the free fatty acids
fraction of the sebum since the in vitro bacteriostatic activity
of these acids against a variety of bacteria is well known [13,
14]. On the contrary fatty alcohols and glyceryl esters show
very weak inhibitory properties [15].
In our previous studies based on the testing of minimal
inhibitory concentrations of various saturated and unsaturated
fatty acids for several hundreds of the skin hacteria we found
Manuscript received September 8, 1977; accepted for publication
May 24, 1978.
Tbese investigations were performed during the tenure as a senior
scientist, Alexander von Humboldt Foundation, and partially supported
by a grant 05-339-C from tbe Center for Disease Control, U.S. Public
Healtb Service, Atlanta, GA, U.S.A.
Reprint requests to: Dr. P. B. Heczko, Institute of Microbiology,
Medical Academy, 18 Czysta Street, 31-121 Krakow, Poland.
Abbreviations:
MIC: minimal inhibitory concentration

that 4 of them: decanoic, dodecanoic, octadeca-9,12 dienoic, and

octadeca-9,12,15 trienoic acids were the most active against

strains of Staphylococcus aureus, Staphylococcus epidermidis,

and Propionibacterium acnes [16]. In this study, our interest
has focused on the differences in susceptibility of the representatives of three Propionibacterium species to tbe selected
free fatty acids.
MATERIALS AND METHODS
Bacterial Strains
Ninety-eight Propionibaeterium strains were tested. Tbey were 69
P. aenes, 20 P. granulosum,, and 9 P. avidum, strains. This set consisted
of botb well-cbaracterized reference strains and fresh isolates. Twentyone P. acnes, 15 P. granulosum, and 4 P. avidurri strains bave been
described previously [17] while the remaining 48 P. aenes, 5 P. granulosum, and 5 P. avidum strains were freshly isolated from acne pustules
of young individuals [18] and identified according to Bergey's Manual
[5] with additional serotyping [19].
Culture Mediums
A-broth and A-agar were used throughout. The composition of
A-media was as specified before [19,20].
Free Fatty Acids
Decanoic, dodecanoic, octadeca-9,12 dienoic, and octadeca-9,12,15
trienoic acids were purchased from Nu-Cbec-Prep., Inc., Elysian, MN,
U.S.A.
Procedure
Twofold dilutions of the fatty acids were made in n-bexane (Baker)
immediately before use. The solutions were added in appropriate
amounts to molten agar cooled to 48 C with constant agitation and
then poured into Petri dishes. After solidification the plates were dried
open for 3 hr at 37C and then inoculated by a spot technique placing
10 drops of bacterial inocula (10 ^I/drop) in 1 plate. Bacterial inocula
were obtained by 1:10 (for P. acnes and P. granulosum) or 1:100 (for P.
avidum) dilutions of 48-br-old A-broth cultures with sterile A-brotb.
Tbey contained about 3 x 10" colony forming units (c.f.u.)/ml of P.
aenes, 4 x 10* c.f.u./ml of P. granulosum, and 6 X 10* c.f.u./ml of P.
avidum by viable count. Inoculated plates were incubated for 48 hr at
37 C in an atmospbere of 95% N2 and 5% CO2 in tbe Heraeus type
VT/N2 incubator and the results were read immediately tbereafter
using ligbt transmitted tbrougb tbe base of the plates. Tbe lowest
concentration of fatty acid tbat resulted in no visible growth was taken
as tbe minimal inbibitory concentration (MIC). Growtb controls of tbe
strains tested on tbe A-agar alone or tbis containing n-bexane were
included witb each assay.

RESULTS
In order to compare the activities of the fatty acids that vary
in molecular weight, the results have been expressed as molar
concentrations (Table).
Analysis of MIC values for aU tested bacteria showed both
unsaturated acids to be more inhibitory than the saturated
ones. Octadeca-9,12,15 trienoic acid appeared to be the most
inhibitory, while decanoic acid had the lowest bacteriostatic
activity.
P. granulosum strains were more susceptible to octadeca9,12 dienoic acid (median MIC, 0.7 fiM/nil) than P. acnes
(median MIC, 4.1 /iM/ml) or P. avidum (median MIC, 5.5

363

364

Vol. 71, No. 6

KO, HECZKO, AND PULVERER

Fatty acid

TABLE. Susceptibility of Propionibacterium species to free fatty acids

No. of strains inhibited at MIC (/iM/ml) of:
Bacterial species
1.56
0.78
0.39
12.5
6.25
3.12
1
12
56
aenes
9
10
1
granulosum
9
avidum
34
4
31
acnes
11
9
granulosum
2
6
1
avidum
22
19
13
6
1
5
acnes
6
4
5
granulosum
1
4
4
avidum
_
_
24
24
5
9
acnes
2
8
granulosum
7
2
avidum
-

Decanoic
P.
(C10:0)
P.
M.W. = 172.27
P.
Dodecanoic
P.
(C12:O)
P.
M.W. = 200.35
P.
Octadecadienoic
P.
(C18:2 A 9,12)
P.
M.W. = 280.48
P.
Octadecatrienoic
P.
(C18:3 A 9,12,15)
P.
M.W. = 278.48
P.
Minimal inhibitory concentration

Median MIC
0.19
1
4
7
10
-

2
1

7.2
8.8
6.2
4.5
9.6
3.5
4.1
0.7
5.5
1.7
0.3
2.7

was in this study a little lower than in our previous investigations [16]. This may be caused by some differences in media
composition since even minute alterations in concentrations of
Mg** and Ca** in a medium influence the bacteriostatic properties of the fatty acids [24]. Our data are in agreement with
earlier findings of Puhvel and Reisner [25] although their
studies have been based on only 2 strains of P. acnes isolated
from non-acne sources. The authors have found that octadeca9,12 dienoic acid exerted no inhibitory but slightly stimulatory
effect on the strains used at highest concentration applied of
Decanoic and dodecanoic acids were nearly equally active 500 |Ug/ml i.e. about 2 juM/ml. Contrary to other studies based
against all 3 Propionibacterium species tested, however, P. on a testing of single representatives of various bacteria [15] we
avidum strains appeared to be a little more susceptible and P. have found octadeca-9,12,15 trienoic acid to be more active
against propionibacteria than octadeca-9,12 dienoic acid. Algranulosum strains more resistant to both acids.
though the octadeca-9,12,15 trienoic acid occurs only in trace
amounts in human skin surface lipids [23] it is probable that it
DISCUSSION
also plays a role in maintaining ecological relationships between
The factors that might contribute to the pathogenesis of acne various constikients of the skin microflora.
are the rate of sebum production, the number of propionibacT h e composition of t h e human skin lipids with particular
teria, and the free fatty acid composition of the sebum lipids
[21]. The last 2 factors seem to be closely related to each other. regard to free fatty acids, seems to be one of t h e physiologic
The elevated numbers of propionibacteria on the skin of acne factors of the host organism determining the kinds and numbers
patients [3] could be caused by changes of octadeca-9,12 dienoic of organisms t h a t form stable communities in t h e skin surfaces
acid levels in total skin lipids. It has been recently demonstrated of a given individual.
that the amount of this acid in the skin surface lipids is
We would like to thank Dr. R. R. Marples for his advice and for his
significantly diminished in acne patients in comparison to nor- critical
review of the manuscript.
mal subjects [22]. Our in vitro findings indicate that the majorREFERENCES
ity of the human skin propionibacteria are susceptible to 6.25
/iM/ml or less of this compound. Therefore it seems probable
1. Marples M: The Ecology of Human Skin. Springfield, Thomas,
1965
that octadeca-9,12 dienoic acid present in large quantities on
2. Marples RR, McGinley KJ: Corynebacterium acnes and other
the human skin surfaces [23] could play a role in maintaining
anaerobic diphtheroids from buman skin. J Med Microbiol 7:
a relatively stable population of the propionibacteria on the
349-357, 1974
skin of healthy individuals and that a decrease of its level in
3. Leyden JJ, McGinley KJ, Mills OH, Kligman AM: Propionibacterium levels in patients witb and without acne vulgaris. J Invest
the sebum of acne patients might permit propionibacteria to
Dermatol 65:382-384, 1975
multiply over normal numbers. The same mechanism may be
4.
Evans
CA: Persistent individual differences in the bacterial flora of
generally true for all young adult individuals since it has been
the skin of the forehead: Numbers of propionibacteria. J Invest
shown that levels of the propionibacteria on their skin surfaces
Dermatol 65:42-46, 1975
are as high as those of acne patients [3].
5. Moore WEC, Holdeman LV: Propionibacterium. Bergey's Manual
of Determinative Bacteriology, 8th edition. Edited by RE BuThe latest classification of the anaerobic coryneforms of the
chanan, NE Gibbons. Baltimore, Williams & Wilkins, 1974, pp
skin enabled the study of possible ecologic differences between
633-641
resident sites and biochemical effects of different Propionibac6. Voss JG: Differentiation of two groups of Corynebaeterium aenes.
J Bacteriol 101:392-397, 1970
terium species [2,5]. It appears that P. acnes dominates on the
7. Johnson JL, Cummiris CS: Cell wall composition and deoxyriboskin of both normals and acne patients while P. granulosum
nucleic acid similarit;ies among the anaerobic coryneforms, clasoccurs on the skin of only 2% of normal subjects but 42% of
sical propionibacteria, and strains of Arachnia propionica. J
those with acne representing 16% of total anaerobic flora of
Bacteriol 109:1047-1066, 1972
8. Cummins CS: Identification of Propionibaeterium acnes and reacne subjects and less than 1% in normals [3]. These differences
lated organisms by precipitin tests with trichloroacetic acid exin density of both species and their changes in acne patients
tracts. J Clin Microbiol 2:104-110, 1975
could be explained on the base of unusual susceptibility of P.
9. Downing DT, Strauss JS: Synthesis and composition of surface
granulosum strains to octadeca-9,12 dienoic acid found in this
hpids of human skin. J Invest; Dermatol 62:228-244, 1974
10. Marples RR, Downing DT, Kligman AM: Control of free fatty in
study.
human surface lipids by Corynebacterium aenes. J Invest DerDifferential susceptibility to octadeca-9,12 dienoic and octamatol 56:127-131, 1971
deca-9,12,15 trienoic acids may also serve as an additional 11. Strauss JS, Kligman AM: The pathologic dynamics of acne vulgaris.
characteristic for the classification of these organisms.
Arch Dermatol 82:779, 1960
The sensitivity of the tested propionibacteria to 4 fatty acids 12. Aly R, Maibach HI, Rahman R, Shinefield HR, Mandel AD:
/iM/ml). No correlation between the resistance to octadeca-9,12
dienoic acid and the origin of the tested strains was found.
Octadeca-9,12,15 trienoic acid was the most active compound
against the propionibacteria tested. P. acnes and P. avidum
strains appeared to be more resistant to octadeca-9,12,15 trienoic acid than P. granulosum strains. Differences in medians
of MIC values for these 3 species were even more clear-cut than
in the case of the octadeca-9,12 dienoic acid (P. acnes = 1.7
P. avidum = 2.7 juM/ml, and P. granulosum = 0.3

Dec. 1978

13.
14.
15.
16.

17.
18.

SUSCEPTIBILITY OF PROPIONIBACTERIA TO FATTY ACIDS

Correlation of buman in vivo and in vitro cutaneous antimicrobial

factors. J Infect Dis 131:579-583, 1975
Glasman HN: Surface active agents and tbeir application in bacteriology. Bacteriol Rev 12:105-148, 1948
Nieman C: Influence of trace amounts of fatty acids on tbe growtb
of microorganisms. Bacteriol Rev 18:147-163, 1954
Kabara JJ, Swieczkowski DM, Conley AJ, Truant JP: Fatty acids
and derivatives as antimicrobial agents. Antimicrob Ag Cbemother 2:23-28, 1972
Heczko PB, Kasprowicz A: Epidemiological and ecological studies
on mechanisms of stapbylococcal carriage. Stapbylococci and
Stapbylococcal Diseases. Edited by J Jeljaszewicz. Stuttgart &
New York, Fiscber, 1976, pp 935-940
Hoeffler U, Ko HL, Pulverer G: Antimicrobial susceptibility of
Propionibaeterium acnes and related microbial species. Antimicrob Ag Cbemotber 10:387-394, 1976
Ko HL, Kasprowicz A, Heczko PB, Pulverer G: Properties of
Propionibacterium strains isolated from acne lesions. J Clin
Microbiol, in press

365

19. Pulverer G, Ko HL: Fermentative and serological studies on Propionibacterium acnes. Appl Microbiol 25:222-229,1973
20. Jong EC, Ko HL, Pulverer G: Studies on bacteriopbages of Propionibacterium aenes. Med Microbiol Immunol 161:263-271, 1975
21. Kligman AM: An overview of acne. J Invest Dermatol 62:268-288,
1974
22. Morello AM, Downing DT, Strauss JS: Octadecadienoic acids in
the skin surface lipids of acne patients and normal subjects. J
Invest Dermatol 66:319-323, 1976
23. Ansari MNA, Nicolaides N, Fu HC: Fatty acid composition of the
living layer and stratum corneum lipids of buman sole skin
epidermis. Lipids 5:838-845, 1970
24. Galbraitb H, Miller TB, Paton AM, Thompson JK: Antibacterial
activity of low cbain fatty acids and tbe reversal witb calcium,
magnesium, ergocalciferol and cbolesterol. J Appl Bacteriol 34:
803-813, 1971
25. Pubvel SM, Reisner RM: Effect; of fatty acids on tbe growtb of
Corynebaeterium acnes in vitro. J Invest Dermatol 54:48-52,
1970