Skip to main content
Open Access Publications from the University of California

Anti-inflammatory effects of Erythromycin and Tetracycline on Propionibacterium. acnes induced production of chemotactic factors and reactive oxygen species by human neutrophils

  • Author(s): Jain, A
  • Sangal, L
  • Basal, E
  • Kaushal G, P
  • Agarwal, SK
  • et al.
Main Content

Anti-inflammatory effects of Erythromycin and Tetracycline on Propionibacterium acnes induced production of chemotactic factors and reactive oxygen species by human neutrophils
Jain A1, Sangal L1, Basal E1, Kaushal GP2 and Agarwal SK1
Dermatology Online Journal 8(2): 2

1. Department of Microbiology 2. Department of Medicine, King George's Medical College, Lucknow-226003, INDIA


Propionibacterium acnes (P. acnes), an anaerobic pathogen, plays an important role in the pathogenesis of acne and seems to initiate the inflammatory process by producing neutrophil chemotactic factors (NCF). Once neutrophils attracted by bacterial chemoattractants reach the inflamed site, they release inflammatory mediators such as lysosomal enzymes and reactive oxygen species (ROS). Previously, it has been shown that antibiotics may affect acne by means other than their anti-bacterial effects. Thus, we investigated the effect of subminimal inhibitory concentration (sub-MIC) of tetracycline and erythromycin on production of NCF and ROS. NCF was tested in vivo in a mouse model and ROS was estimated on human PMNL in vitro, by nitroblue tetrazolium dye reduction test (NBT) and cytochrome-C reduction test. Tetracycline (CS-T) and Erythromycin (CS-E) treated cultures showed a significant reduction of 35.8% and 58.3% in NCF production respectively, as compared to P. acnes stimulated cultures. Tetracycline and erythromycin at their sub-MIC also significantly inhibited release of ROS from human PMNL. Thus, tetracycline and erythromycin, besides having antibacterial activity, also have an anti-inflammatory action. These antibiotics reduce the capacity of P. acnes to produce NCF, as well decrease its ability to induce ROS from PMNL.


Systemic administration and topical application of certain antibiotics have been shown to be effective in the treatment of acne vulgaris. Acne formation can be conceptualized as a two-stage process, i.e. comedo formation and inflammation [1]. Propionibacterium acnes appear to play an important role in the inflammatory process. P. acnes mediated generation of free fatty acids may play a lesser role in the pathogenesis of acne inflammation [2] than previously believed [3,4], but it has been reported that P. acnes produces low-molecular-weight chemotactic factors [5] that can diffuse through intact follicular epithelium and attract polymorphonuclear leucocytes (PMNLs). Once PMNLs reach the sebaceous follicles, they are considered to release reactive oxygen species (ROS) besides other inflammatory mediators, like lysosomal enzymes. These inflammatory mediators in the presence of anti P. acnes antibodies and complement cause damage to follicular epithelium and extrusion of follicular contents into the dermis.[6,7] Successful treatment with antibiotics has been attributed to the reduction in number of P, acnes and inhibitory effects on the production of P. acnes-associated inflammatory mediators. Martin et al. have demonstrated that tetracycline in lower concentrations comparable to common drug therapy, markedly depressed migration of human leukocytes in vitro.[8] This finding has fostered the concept that antibiotics may affect acne by means other than direct antibacterial effects.

In this study we investigated the effect of subminimal inhibitory concentrations (sub-MIC) of tetracycline and erythromycin on the production of neutrophil chemotactic factors and generation of ROS both by Nitro Blue Tetrazolium (NBT) dye reduction and cytochrome-C reduction tests. We observed that both antibiotics posses anti-inflammatory activity as they suppressed production of NCF and ROS at their sub-MIC.



Tetracycline hydrochloride and erythromycin were purchased from Himedia (Mumbai, India). Ficoll-Hypaque, fetal calf serum, zymosan, cytochrome C-reductase and super oxide dismutase were obtained from Sigma (St. Louis, USA).

Standard strain

The standard strain of P. acnes MTCC 1951, obtained from Microbial Type Culture Collection and Gene Bank, Institute of Microbial Technology, Chandigarh, India was used throughout the study. It was cultured in 10% serum nutrient broth under anaerobic atmosphere of N2/H2/CO2::80/10/10/ (v/v), maintained by Anaxomat (Yunimydex, Germany), an automated system.

Culture supernatant (CS) preparation:

MIC of tetracycline and erythromycin for P. acnes was evaluated by broth dilution method. Sub-MIC (1/10th MIC) of antibiotics was added to appropriate cultures at the time of P. acnes inoculation and was incubated at 37()()C anaerobically for 72 h. Culture supernatant with P.acnes alone was labeled as CS-P, while culture supernatant of P. acnes with tetracycline and erythromycin were labeled as CS-T and CS-E respectively. Cultures without antibiotics were set up as negative control. One tube, each containing 1/10th MIC of tetracycline and erythromycin without P. acnes culture was incubated and was used as drug control to check the effect of residual drug on ROS production. The log phase culture was centrifuged at 5,000 x g for 15 min at 4()()C. The supernatants were harvested, filtered through a 0.22-mm-pore-size filter and stored at -200C till used. To rule out bactericidal or bacteriostatic effect of antibiotics, the growth of P, acnes culture was monitored in the presence of sub-MIC of the drugs. Both antibiotics at their 1/10th - MIC did not have any significant effect on bacterial growth as evident from sample growth presented in (Fig 1).

Neutrophil chemotactic factor (NCF) assay:

NCF was estimated by modified method of Reebuck.[9] Briefly, skin above the dorsal tail vein of mouse was scrapped with the help of a surgical blade, and a drop of culture media with and without stimulant was added with the help of syringe at an interval of 5 minutes for 1 hr. Peripheral blood smears were prepared, before and after the stimulation and stained with Leishman stain. Percentage of attracted neutrophils was performed by differential count.

Superoxide production assay:

Nitroblue tetrazolium dye (NBT) reduction test -

Test was done by method of Park et al [10]. Briefly, to 0.5 ml of whole blood, 0.05 ml of stimulants/ controls were added and incubated at 200 C for 15 minutes. A 100ml of NBT solution (1mg/ml of PBS) was added to the above 0.1 ml primed blood and incubated at 370C for 30 minutes and then again at 200C for 20 minutes. Finally smears were prepared and stained by Leishman stain for differential counting of formazan deposits in PMNLs.

Cytochrome-C-reduction test

Heparinized blood was collected from volunteers by venipuncture. PMNLs were purified from whole blood by Ficoll-Histopaque (Sigma) gradient and were suspended at concentration of 107 PMNL/ml.[11] The viability count of more than 95% was taken as satisfactory. Opsonized zymosan was prepared; freshly each time by incubating 11mg of zymosan with 1ml of autologous serum at 37()() C for 30 min. Superoxide (O2') generation was assayed as superoxide dismutase (SOD) inhibitable reduction of cytochrome-c by method of Weening et al.[12] Briefly, two sets of reaction mixture were prepared by adding 50ml of cytochrome-C (12.5 mg/ml) followed by addition of 50ml of PMNL (1x 107 cells/ml). In one set SOD (1 mg/ml) was added to see the inhibition of superoxide production by oxygen burst in PMNL. The reaction was started by addition of 50ml of stimulant (CS-P, CS-T and CS-E); or zymosan or culture media as negative controls. Reaction mixture were mixed rapidly and incubated at 37()()C for 25 min, followed by harvesting of supernatant by centrifugation at 900 x g for 20 min at 4()()C. Reduced cytochrome-C in cell free supernatant was measured at 550 nm on a double beam spectrophotometer. The difference between the amount reduced in the presence and absence of SOD represents the amount of superoxide generated during the incubation. Superoxide generation was calculated on the basis that a change in absorbance of 1.0 at 550 nm corresponds to the presence of 47.4 nmol of superoxide.

Statistical analysis

Mean, standard deviation and source of error were calculated for all the stimulant groups. The mean values of different groups were compared by analysis of variance (ANOVA). On the evidence of significant F-ratio, multiple comparisons were done by Newman Keuls' test[13].


Figure 1
Figure 1. Growth curve of P. acnes, 1951 in the presence and absence of sub-MIC of antibiotics. Growth was monitored nephelometrically at 12, 24, 48 and 72 h. Symbols: *, no antibiotics; p, 0.3 mg/ml of erythromycin; (R), 0.1 mg/ml of tetracycline. Results are representative of three independent experiments.

A: Effect of tetracycline and erythromycin on production of NCF by P. acnes

Figure 2
Figure 2. Effect of erythromycin and tetracycline on neutrophil chemotactic activity. P. acnes was cultured in the absence or presence of sub-MIC of erythromycin and tetracycline for 72 h and culture supernatant (CS) was collected. NCF in these CS was estimated as mentioned in material and method. Results are expressed as percent increase in NCF activity over controls (culture media). The mean difference between CS, CS-T and CS-E was compared using Newman Keuls¹ Test and it was seen that both antibiotics caused a significant decrease in NCF activity in P. acnes stimulated cultures. Data are representative of three independent experiments

The chemotactic activity of culture supernatants (CS) tested, was expressed as percent increase or decrease over the basal PMNL count at the site of stimulation. Table-1 summarizes the result of NCF assay. The chemotactic activity caused by the local tissue injury was minimal (14.9%) and insignificant. CS of P. acnes (CS-P) showed the maximum capacity to attract the PMNL towards the site of stimulation (118.5%). Culture media was taken as negative control. Tetracycline and erythromycin caused a significant decrease of 35.8% and 58.3% {Fig 2} respectively as compared to that expressed by CS-P. When these observations were analyzed by applying two way ANOVA test, the CS-T and CS-E showed a significant decrease (p<0.05). The inhibition in NCF activity was more with CS-E [22.5% (P<0.05)] than that seen with CS-T, implying that although both the antibiotics were effective in reducing the amount of NCF secreted by P, acnes significantly, erythromycin was more effective than tetracycline.

Effect of tetracycline and erythromycin on production of superoxide by PMNL

NBT test

Figure 3A

Figure 3B
Figure 3. Effect of erythromycin and tetracycline on superoxide production by PMNL by (A) NBT dye reduction test and (B) cytochrome-C reduction test. P, acnes was cultured in the absence or presence of sub-MIC of erythromycin and tetracycline for 72 h and culture supernatant (CS) was collected. ROS inducing capacity of these CS was estimated as mentioned in material and method. Results are expressed either as percent of PMNL showing formazan deposits in case of NBT dye reduction test (A), or in nanomoles/107 cells/min in case of cytochrome-C reduction test (B). The mean difference between CS, CS-T and CS-E was compared using Newman Keuls¹ Test and it was seen that both antibiotics caused a significant decrease in ROS inducing activity of P, acnes CS. Results are representative of three independent experiments.

The superoxide radical anion production in NBT dye reduction test was approximated by counting all the PMNL containing formazan deposit. When CS-P was used as a stimulant 58.67 +5.89% (Figure 3A) of PMNL were containing the deposit of formazan, while stimulation by culture media (negative control) showed that only 20.33% {fig 3A} of PMNL have formazan deposit. Normal saline (NS) was also used a negative control and it showed less than 5% formazan deposit. When CS-T and CS-E were tested, the rise obtained in PMNL containing formazan deposit was 42.62+3.4% and 40.50+3.8% (Figure 3A) respectively, which was significantly less than that obtained by CS-P. On comparing the result by one-way ANOVA test, a significant suppression of 23.86% by CS-T and 17.33% (Figure 3A) by CS-E was noticed as compared to CS-P. The difference between the inhibitory effect of CS-T and CS-E was not significant. The antibiotic controls (tetracycline and erythromycin) in their sub-MIC concentration after over night incubation at 37()()C in CM were unable to significantly inhibit the burst caused by CS-P and zymosan.

Cytochrome-C reduction test

The generation of superoxide anion was also quantitated by Cytochrome-C reduction test. Culture media (negative control) produced the negligible amount of superoxide anion (10.1+1.4 nmols/107cells /min) with relatively small standard deviation, CS-P produced superoxide anion (55.2+4.24 nmols/107cells /min) (Figure 3B) that was significantly higher than CM. The amount of ROS generated by CS-E and CS-T was 28.33+3.23 nmols/107cells/min and 26.56+2.78 nmols/107cells /min (Figure 3B) respectively, which was significantly lower than that produced by CS-P. To evaluate the actual inhibition produced by antibiotics, multiple comparisons were drawn between the values obtained by various stimulants. Significant decrease of 37.78% by CS-T and 42.2% (Figure 3B) by CS-E was observed. There was no significant difference between the inhibitory effects of two antibiotics on superoxide.


We observed that erythromycin and tetracycline inhibit production of NCF and ROS induced by P. acnes, by PMNL at their sub MIC. Propionibacterium acnes is considered to be the most important microorganism in the development of acne lesions.[14] The clinical observation that inflammatory acne can often be improved by administration of antibiotics, implicates sensitive microorganisms as having a role in the pathogenesis of the inflammatory phase of the disease. Several hypotheses have been postulated to explain the mechanism by which P. acnes produces inflammation. These may include production of or induction of exocellular enzymes,[14] complement activating compounds, [15] chemotactic substances, [16,17] lysosomal hydrolases, [18] and ROS. [19] The soluble factors produced by P. acnes act as a chemoattractant for PMNL and these substances can be retrieved from acne lesions suggesting their role in inflammation of acne.[5]

The present study demonstrates that the CS of P. acnes induces the maximum attraction (118%) of PMNL towards the site of stimulation, thus suggesting the release of certain NCF by P. acnes, during its growth in culture media. The sub-MIC of tetracycline and erythromycin caused a marked decrease in NCF produced by P. acnes. A significant reduction of 27 % by CS-T and 47% by CS-E was noticed from the peak values of CS-P. Sub-MIC of erythromycin and tetracycline may have an anti-inflammatory action by inhibiting NCF production. Webster et al. also reported a significant suppression of NCF production at sub-MIC of tetracycline, minocycline and erythromycin, while ampicillin failed to do so.[20] A similar observation was also reported by Akamatsu et al. who showed that sub-MIC of erythromycin inhibits NCF production in susceptible strains of P. acnes, P. granulosam, and coagulase negative staphylococci, which are mainly isolated from acne lesions present on the face.[21] In a different study it was also shown that the mean chemotaxis in subjects receiving topical tetracycline was significantly low compared to that of untreated subjects.[22] Thus our data, together with these observations suggest the anti-inflammatory nature of these antibiotics at their sub-MIC. In our study, the efficiency of erythromycin was found to be better than tetracycline, when NCF was taken into account. This may correlate to an inhibitory effect on protein synthesis by erythromycin, because NCF are presumably derived from post-synthetic protein processing.[5,23]

Another potent stimuli for inflammation is ROS and it is seen that the ability of PMNL to produce ROS varies according to the severity of acne. [19] Patients with papulopustular acne showed a significantly increased level of ROS production from PMNL as compared to patients with comedonal acne and healthy controls. It is considered that NCF released from P. acnes may accumulate and induce the migration of PMNL to the site and subsequently lead to release of ROS. These ROS are supposed to play some role in pathogenesis of acne. We found that P. acnes caused a significant increase in ROS production from PMNL as compared to negative control (CM), both in the nitroblue tetrazolium dye reduction and cytochrome-C reduction tests.

We also found that both antibiotics caused a significant suppression in production of ROS from PMNL. The inhibitory effect of tetracycline and erythromycin was more or less the same by the two methods used. Our data is supported by a similar study of Miyachi et al. who showed that eythromycin, tetracycline, and minocycline reduce ROS generation from zymosan stimulated PMNL.[24] These data suggest that the efficacy of antibiotics in acne cases was not due to their microbicidal effect and may be attributed to their inhibitory effect on ROS generation by PMNL. This hypothesis is further supported by a report that metronidazole, having no antibacterial activity, produced improvement in eruptive lesions of patients with inflammatory acne. [25,26,27,]

From these observations we conclude that these antibiotics have favorable effects in resolving inflammatory acne lesions not only by reducing the P. acnes density but also by directly inhibiting NCF and ROS production. This may also suggest a new role for antibiotics as adjunctive therapy in atopic dermatitis and other diseases whose severity is increased by bacterial colonization of the affected areas. A reduction in the inflammatory capacity of the colonizing organisms could enhance clinical improvement.


1. Webster GF, Leyden JJ, Mechanisms of Propionibacterium acnes mediated inflammation in acne vulgaris, Semin Dermatol. 1:299-304, 1982.

2. Puhvel S M, Sakamoto M, A re-evaluation of fatty acids as inflammatory agents in acne, J Invest Dermatol. 68:93-99, 1977.

3. Strauss JS, Pochi PE, Intracutaneous injection of sebum and comedones, Arch Dermatol. 92: 443-456, 1965.

4. Freinkel RK, Strauss JS, Yip SY, Pochi PE, Effect of tetracycline on the composition of sebum in acne vulgaris, N Engl J Med. 273:850-854, 1965.

5. Puhvel SM, Sakamoto M, The chemoattractant properties of comedonal contents, J Invest Dermatol. 71:324-329, 1978.

6. Webster G F, Leyden J J, Tsai CC, Bachni P, McArthur WP, Polymorphonuclear leukocyte Lysosomal release in response to Propionibacterium acnes in vitro and its enhancement by sera from patients with inflammatory acne, J Invest Dermatol. 74: 399-401, 1980.

7. Webster GF, Kligman AM, A method for the assay of inflammatory mediators in follicular casts, J Invest Dermatol. 73: 266-268, 1979.

8. Martin RR, Warr JA, Couch RB,Yeager H, Knight V, Effects of tetracycline on leukotaxis, J Infect Dis. 129: 110-115, 1974.

9. Reebuck J W, Crowley J H, A method of studying leukocyte function in vivo, Ann. N.Y. Acad. Science. 59: 757,1955.

10. Park B H, Fikrig S M, Smithwick E M, Infection and nitroblue tetrazolium reduction by neutrophils, Lancet. 532,1968.

11. Boyum A, Isolation of mononuclear cells and granulocytes from human blood, Scand. J. Clin. Lab. Invest. 21(Suppl-97): 77-89,1968.

12. Weening R S, Wever R, Ross D, Quantitative aspects of the production of superoxide radical by phagocytosing human granulocytes, J. Lab. Clin. Med. 85: 245,1975.

13. Zar J H, Biostatistical Analysis, Prentice. 1974.

14. Holland K T, Ingham E, Cunileffe W J, A review-The microbiology of acne, Journal of Applied Bacteriology. 51: 195,1981.

15. Massey A, Mowbray J F, Noble W C, Complement activation of Corynebacterium acnes, Br. J.Dermatol. 98: 583-584,1978.

16. Russel R J, McInroy R J, Wilkinson P C, White R G, A lipid chaemotactic factor from an aerobic coryneform bacteria including Corynebacterium parvum, with activity for macrophages and monocytes, Immunology. 30: 935,1976.

17. Gould D J, Cunileffe W J, Holland K T, Chemotaxis and acne, J Invest. Dermatol. 68: 251,1977.

18. Lee W L, Suntharalingam K, Fiking S M, Shalita A R, Neutrophil chemotaxis by P. acnes, Clin. Res. 25, 283A, 1977.

19. Akamatsu H, Horio T, The possible Role of reactive oxygen species generated by neutrophils in mediating Acne Inflammation, Dermatology. 196: 82-85,1998.

20. Webster G F, Leyden J J, McGinley K J, and McArthur W P, Suppression of polymorphonuclear leukocyte chemotactic factor production in Propionibacterium acnes by subminimal inhibitory concentrations of Tetracycline, Ampicillin, Minocycline, and Erythromycin, Antimicrobial Agents & Chemotherapy. 21: 770-772, 1982.

21. Akamatsu H, Kurokawa I, Nishijima S, and Asada Y, Inhibition of neutrophil chemotactic factor production in comedonal bacteria by subminimal inhibitory concentrations of Erythromycin, Dermatology. 185: 41-43,1992.

22. Elewsky B E, Sam M Jr, Gammon W R, In Vivo suppression of neutrophil chemotaxis by systemically and topically administered tetracycline, J Am Acad Dermatol. 8:807-812,1983.

23. Webster G F, Leyden J J, Production of serum independent polymorphonuclear leukocyte chemotactic factors by Propionibacterium acnes, Inflammation. 4: 261-269,1980.

24. Miyachi Y, Yoshioka A, Imammura S, Niwa Y, Effect of antibiotics on the generation of reactive oxygen species, J. Invest. Dermatol. 86: 449-453,1986.

25. Akamatsu H, Nishijima S, Takahashi M, Ushijima H, Asada Y, Oguchi M, Niwa Y, The inhibition of free radical generation by human neutrophils through the synergistic effect of metronidazole with palmitoleic acid; possible mechanism of action of metronidazole in rosacea and acne, Dermatol Res. 282:449-454,1990.

26. Oguchi M, Akamatsu H, Asada M, et al., Treatment of acne with oral metronidazole, Skin Res. 28: 995-1000,1987.

27. Pye R J, Burton J L, Treatment of rosacaa by metronidazole, Lancet. 1211-1215,1976.

© 2002 Dermatology Online Journal