Modulation of Comedonal Levels oflnterleukin-l in Acne Patients Treated with Tetracyclines

E. Anne Eady, * Eileen Ingham,* Christina E. Waiters,* Jonathan H. Cove,* and William J. Cunliffe'I' "Department of Microbiology, University of Leeds; and 1"Leeds Foundation for Dermatological Research (WJC), Leeds General Infirmary, Leeds, U .K.

------------------------------------------------------------------------------------------------------, To understand the basis for the anti-inflammatory activity of tetracyclines in acne, we compared the cytokine profiles [in­terleukin 1 (IL-l) alpha and beta, tumor necrosis factor (TNF) alpha, and IL-6] and bacterial flora of 66 open come­dones removed from eleven patients before and after at least 8 weeks treatment with either tetracycline or minocyc1ine.

Pre-treatment, the only cytokine regularly recovered from comedones was bioactive IL-1alpha-like material. The mean concentration ofIL-lalpha - like bioactivity / mg come­donal material rose from 272.0 ± 88.6 pg to 844.3 ± 196.7 pg following treatment (p < 0.05, Wilcoxon matched pairs). All six minocycline-treated patients showed an in­crease in bioactive IL-lalpha -like material compared with three of five tetracycline-treated patients. The incidence (p < 0.001, X2) and concentration (p < 0.05, Wilcoxon) of immunochemical IL-beta were also raised post-treatment, although significantly more patients assigned to minocyc1ine therapy had detectable levels of this cytokine before therapy was initiated. However, the mean concentration of IL-1beta/mg comedonal material post-treatment was similar in

both groups (72.5 ± 23.3 pg for tetracycline-treated com, pared with 78.6 ± 41.9 pg for minocydine-treated patients) The other cytokines were either absent (IL-6) or present i~ < 10% of comedones (TNFalpha) before and after therapy

Following treatment, only three of 11 patients showed ~ decrease of;;::: 110glO in propionibacterialnumbers/mg come. donal material, whereas six patients showed an increase 01 > 0.51og1o m numbers of staphylococci. In eight patients, th~ increase or decrease in staphylococcal numbers correlate~ with the change in concentration of IL-lalpha -like bioac, tivity.

This is the first study to show an effect of antibiotic theraP1-on cytokine levels in vil/o. Increased levels of IL-l in come, dones destined to become inflamed may enhance resolutio~ and promote repair of the damaged follicular epithelium. Hence, these results provide further evidence of the augmen~ tation of immune responses by tetracyclines and support th~ hypothesis that epidermal IL-l plays a physiologic role i~ wound healing.] Invest DerrnatoI101:86-91, 1993

-------------------------------------------------------------------------------------------------------,

Acne vulgaris is an inflammatory disease of the pilose­baceous follicles of the face, back, and chest, charac­terized by a variety of lesions showing various degrees of visible erythema [1 J. Little is known of the inflam­matory mediators other than complement [2,3]

present in acne lesions. Two lines of evidence, namely, the histo­logic demonstration of non-random helper T-Iymphocyte infiltra­tion [4J and the secretion of pro-inflammatory cytokines by kerati­nocytes in vitro in response to a number of different environmental stimuli [5], suggested that the estimation of cytokines in acne lesions might yield information pertinent to the pathogenesis and treat­ment of the disease. We have recen tly shown that the majority of open comedones contain biologically active IL-1alpha-like mate­rial in concentrations sufficient to promote visible erythema if re­leased into the dermis (6J .

Successful anti-acne therapy with antibiotics is thought to be due, at least in part, to their inhibitory effect on the growth and metabo­lism of Propionibacterium awes [7 -9). Micro-organisms, especially P. awes, have been strongly implicated in the pathogenesis of acne

Manuscript received May 14, 1992; accepted for publication March 9, 1993.

A preliminary report of this work was presented at the Annual Meeting of the European Society for Dermatological Research, Copenhagen, June 1 - 3, 1991.

Reprint requests to: Dr. E. Anne Eady, Department of Microbiology, University of Leeds, LEEDS, LS2 9JT. U.K.

[10,11] and may release microbial mediators of inflammation intI:< the dermis or trigger the release of cytokines from ductal keratino, cytes. However, despite clinical improvement, the numbers of pro, pionibacteria do not always fall during antibiotic therapy (7) . Anti, inflammatory activity has been proposed as an alternative a~ complementary mode of action of antibiotics in acne [7] and h~ been demonstrated ill IIillo, for both tetracyclines and erythromycil\ [1 2). There is currently much interest in the possible effects at antibiotics on cytokine production and upregulation of IL-l an~ IL-6 production by monocytes has been observed in vitro [13-15 ], The aims of this study were to determine whether oral acne theraPl' with tetracyclines altered the cytokine content of open comedone~ and to identify any concomitant changes in the comedonal micro, flora.

MATERIALS AND METHODS

Patients Open comedones (blackheads) were donated by 11 p.h tients with mild to moderate acne vulgaris (grade 0.5 to 3.0 on tlw scale of Burke and Cunliffe [16]) before and after therapy for ~ minimum of 8 weeks with minocycline (four male and two femal~ subjects, 15-29 years old, 100 mg/d), tetracycline (two male an~ one female, 13 -26 years old, 500 mg twice daily) or Mysteclin (one male, 18 years old and one female, 15 years old, tetracycline 500 mg twice daily plus nystatin 500,000 units twice daily). Treatment was selected on the basis of clinical criteria only and two patients who had received tetracycline previously were given Mysteclin for rea· sons of compliance. No patient had received any anti-acne therapy

0022-202Xj93jS06.00 Copyright © 1993 by The Society for Investigative Dermatology, Inc.

86

VOL. 101, NO.1 JULY 1993

for 6 weeks prior to, nor any natural ultraviolet irradiation during, the study.

Recovery and Extraction of Comedones Six open comedones were collected aseptically from the same sample site before and after therapy using a comedone extractor (Thackray) after swabbing the skin surface with isopropanol. Four patients were sampled from the face and seven the back. No sample was obviously contaminated with blood. The wet weight of each comedone was determined prior to gentle homogenization for 1 min in a micro-tissue homo­genizer (Thackray) in 250 III of Dulbecco's modified Eagle's me­dium supplemented with 2 mM L-glutamine, 0.375% (w Iv) NaHC03, 20 mM Hepes, and 10% (v/v) fetal calf serum. The homogenate was centrifuged (10,000 X g; microfuge; MSE Micro­centaur) for 10 min. The supernatant (approximately 230 Ill) was removed and diluted 1: 10 with base medium containing penicillin (100,000 V/I) and streptomycin (100 mg/I), aliquotted and stored at -20°C for the determination ofbioactive and immunochemical IL-l alpha and beta. The pellet was resuspended in 200 III of wash fluid (0.075 M sodium phosphate buffer containing 0.1 % v /v Tri­ton-XI00, pH 7.9) for microbiologic determinations.

Cytokine Bioassays Bioactive IL- lalpha and beta were esti­mated as previously described using a modification of the C3H murine thymocyte proliferation assay in the presence of saturating levels of IL-2 [6]. Specificity was accomplished by pre-incubating the supernatants in the presence and absence of anti - IL-lalpha and anti-IL-1beta antibodies (British Biotechnology) separately and together prior to assay. Allowing for the dilution of the comedonal material in the assay, the sensitivity of the IL-l bioassay was 10 pg/comedone. Bioactive TNFalpha was estimated using a conven­tional L929 cytotoxicity assay as previously described [6] and the lower limit of detection was 50 pg TNFalpha/ comedone. Bioactive IL-6 was quantified by the method of van Oers et al [17] using IL-6-dependent B9.9 hybridoma cells and substituting MTT con­version instead of 3H-thymidine incorporation as a measure of cell viability. All bioassays were validated "in house" for use with come­donal material as described in our previous report [6].

Cytokine Immunochemical Assays Immunochemical IL­lalpha, TNFalpha, and IL-6 were determined using Quantikine enzyme-linked immunosorbent assay kits (British Biotechnology). The sensitivity of the kits varied between 17 and 31 ng/I (170-310 pg/comedone) for IL-lalpha, and between 31-48 ng/l (90 pg/ comedone) for TNFalpha and IL-6. IL-l beta was determined using Cistron enzyme-linked immunosorbent assay kits supplied by T­cell Sciences. The sensitivity was 4 ng IL-l beta/I (20 pg/ come­done). All assays were carried out according to the manufacturers' instructions.

Enumeration of Microorganisms Staphylococci and propion­ibacteria were enumerated in the pellets from the centrifugation of comedonal homogenates after resuspension in wash fluid . Viable counts [expressed as colony-forming units (cfu)] were carried out as previously described [6]. The lower limit of detection was 4 cfu/ comedone.

Statistical Analyses These were carried out according to the recommendations of Sokal and Rohlf[18].

RESVLTS

Effect of Therapy on Acne Grade Acne is a slowly responding disease and the patients were deliberately sampled early (i.e., approx­imately one third of the way through a standard 6-month course of treatment) to determine the likely initial events leading to clinical improvement. At the time of sampling, two of the 11 patients had a > 50% improvement in their acne grade (both on minocycline), four had an improvement of < 50% (two on minocycline and two on tetracycline), one tetracycline-treated patient experienced a marked deterioration in acne severity that necessitated a change of therapy, and the remainder showed no detectable change. At the end of the standard 6-month course, four of six minocycline-treated

MODULATION OF INTERLEUKlN-I BY TETRACYCLINES 87

Table I. Changes in Comedonal Weight Following Treatment

Mean Weight of Comedones (mg) ± 95% Confidence Limits'

Treatment Group

Minocycline (n = 36)' Tetracycline (n = 30) All patients (n = 66)

Pre-Treatment

0.86 ± 0.43 0.90 ± 0.49 0.88 ± 0.30

• Calculated from student t X SEM. • Student t tcst. NS, non significant (p > 0.05) . , Number of comedones.

Post-Treatment

0.93 ± 0.77 0.63 ± 0.29 0.79 ± 0.42

p Valueb

NS NS NS

patients were showing a > 50% improvement in acne grade, one patient showed a < 50% improvement, and one patient failed to return for their final assessment. Two of five tetracycline-treated patients were showing a < 50% improvement at 6 months, one had withdrawn from the study (see above), and two failed to return for follow-up appointments.

Effect of Therapy on Comedone Weight There was no signif­icant difference (Student t test) between the weight of the come­dones before and after treatment with either minocycline or tetracy­cline although there was reduction of 30% in the mean comedonal weight following tetracycline treatment (Table I).

Effects of Therapy on Bacterial Numbers Changes in pro­pionibacterial numbers were minimal with only three of 11 patients showing a ~ 1 loglo reduction in mean count expressed as cfu/mg comedonal material after at least 8 weeks oral therapy with either minocycline (Fig 1) or tetracycline (Fig 2). Changes in staphylococ­cal numbers varied from individual to individual but surprisingly six of 11 patients showed an increase of > 0.5 loglo (Figs 1 and 2). Overall, there was no change in the mean counts of staphylococci or propionibacteria following treatment with either antibiotic (Table II). Pre-treatment, 37% of comedones were not colonized (i.e., < 10- 2 cfu/mg) by staphylococci and 34% were not colonized by propionibacteria. Following treatment, there was no alteration in the proportion of uncolonized follicles: 39% were not colonized by staphylococci, and 36% were not colonized by propionibacteria.

Effect of Therapy on Cycokine Levels Bioactive IL-lalpha ­like material was detected in 76.9% of open comedones before treatment and in 75.8% after treatment. However, the mean con­centration of IL-lalpha -like bioactivity per mg comedonal weight was significantly increased post-treatment in minocycline-treated patients (p < 0.01, Fig 1) and when data from both treatment groups were combined before analysis [p < 0.05, Wilcoxon matched pairs (Table III)]. In eight patients (three of six on minocy­cline and five of five on tetracycline), the increase or decrease in IL- lalpha-like bioactivity corresponded with an increase or de­crease in numbers of staphylococci (Figs 1 and 2) . Bioactive IL­l beta was not detected before or after therapy with either antibiotic.

Following treatment, the incidence ofimmunochen'lically detect­able IL-lalpha rose from 12.1% to 28.8% of comedones and the mean concentration/mg comedonal material was significantly raised [p < 0.05, Wilcoxon (Table IV)] . Both the incidence (p < 0.001, X2) and mean concentration/mg comedonal material (p < 0.05, Wilcoxon) of immunochemical IL-lbeta rose significantly post-treatment when data from both treatment groups were com­bined (Table V). The greatest increases were recorded in tetracy­cline-treated patients. This difference between treatment groups was due to the significantly higher baseline incidence of immuno­chemical IL-lbeta-containing comedones in the group assigned to minocycline therapy (p < 0.001, X2). In this group, all six come­dones from two patients contained immunochemical IL-l beta be­fore therapy was initiated. However, the incidence and mean con­centration of immunochemical IL-l beta post-treatment was similar in both treatment groups (Table V).

TNFalpha-like bioactivity was detected in 4.5% of comedones

88 EADY ET AL

A

"0 .;: ~ o E o c: o "0

" E o u

'" E '­'" c >-

1 v o

.2

.a

" ~ I

" I

8

7

6

5

4

2

d 0 ~~"~~~-L ____ J-~ __ L-~L-L-"~

B

c

~ " o E o c: o

"0

" E o u

'" E '­'u u o CJ o >. .<: a. o

-,;;

a

B

7

6

5

4

.3

2

B r

~ 6 o "0

" E 5 o u

'" E '-.g " :3 U o .-e ~ 2 '0.

2 :3 4

2

L~rl o 2

Patient number

5

5

5

6

6

'-I-

6

Figure L Effect of minocycline therapy on IL- l alpha - like bioactivity (A) and on vIable counts of staphylococci (B) and propionibacteria (C) per mg comedonal materia l. Each pair of bars shows the mean ± 95% confidence limits (Student t X SEM) for six open comedones collected from each pa­tient before (II) and after (0) treatment.

A

o .;;:

~ o E "0 c o

"0

" E o v <J> E '-'" c

?: :~ u o o

:D

" ~ I

" I

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

B

7

6

5

.3

2

~ 0 L-.---~L-L-~~J-__ ~-L __ ~~-J"~ a

8

B .g 7

c

" " E "0 c o "0

" E o U

."

E '­'u u o u o >. .<: Q

" v; a

'"

6

5

4

J

2

.2 0

~ " "0 E "0 c o

"0

" E o u

'" E '­.2 'i:i U o ., 'c o '0. ~ a.

'" o

o

8

7

6

5

4

J

2

o L---L--L­o

2 5

2 4 5

Patient number

Figure~. Effect of tetracycline therapy on IL-l alpha - likc bioactivity (A) and on vIable counts of staphylococcI (B) and propionibacteria (C) per mg comedonal material. Each pair of bars shows the mean ± 95% confidence limits (Student t X SEM) for six open comedones collected from each pa­tient before (II) and after (0) treatment.

VOL. 101, NO. 1 JULY 1993 MODULATION OF INTERLEUKIN-I BY TETRACYCLINES 89

Table II. Microbial Numbers Before and After Antibiotic Therapy

logto cfu/ m g Comedonal Material

Number of % Mean' ± 95% Micro-Organism Therapy Comedones Positive Range" Confidence Limi ts'

Staphylococci Pre-treatment Minocycline Post-treatment

Propionibacteria Pre-treatment Minocycline Post-treatment

Staphylococci Pre-treatment T etracycline Post-treatment

Propionibacteria Pre-treatment Tetracycline Post-treatment

• Range of positive values. , Mean for all comedones. , Calculated from Student t X SEM.

pre-treatment and in 6.1 % of comedoI?es following treatment. Im­munochemical TNFalpha was present 111 9.1 % of comedones before treatment but was not detected following treatment. Immunochem­ical IL-6 and IL-6 bioactivity were undetectable both before and after treatment.

DISCUSSION

The tetracycline antibiotics are used extensively in the therapy of acne vulgaris. Although the.y i~hib~t ~he gr?wth of Propi?/Iibacl~r­ium acnes both in vitro and III VIVO, It IS un!tkely that antibacterial activity alone is ~esponsible for their therapeutic. efficacy bec~u~e better antimicrobial agents such as benzoyl perOXide are less c!tnt­cally effective [1 9,20]. Both tetracycline and erythromycin have been shown to reduce l20tassium iodide - induced cutaneous inflam­mation in human skin l1 2]. In vitro, the tetracyclines exert a number of effects on host defense mechanisms such as the down regulation of cell-mediated immune responses and the inhibition of neutrophil functions [21-25]. More recently, we have shown that minocycline and to a lesser extent, tetracycline enhanced IL-l beta secretion by LPS-stimulated mononuclear cells from five different donors [14]. Therefore, it appears that there are a number of possible mecha­nisms whereby the tetracyclines could modulate immune responses and hence inflammation in vivo. However, evidence to show that the tetracyclines interfere with any of these physiologic processes during therapeutic use has been lacking.

We have previously demonstrated the presence of bioactive IL­lalpha-like material in the majority (76%) of open comedones from untreated acne patients [6] . The identity of the comedonal mediator with monocyte-derived IL-lalpha has not been con­firmed; we have followed the recommendation of Camp et at [26] and referred to the activity as IL-lalpha - like. We have now shown that treatment with tetracycline antibiotics upregulates the produc-

35 80 1.00 - 6.36 3.82 ± 0.78 36 83.3 1.1 2 - 6.6 1 3.37 ± 0.75 35 80 1.30 - 6.77 4.20 ± 0.88 36 83.3 0.70 - 7.21 4.11 ± 0.84 30 76.7 0.32 - 5.84 2.40 ± 0.77 30 80 0.90 - 6.33 2.56 ± 0.72 30 83.3 0.90 - 6.79 2.67 ± 0.82 30 70 1.06 - 6.53 2.68 ± 0.92

tion ofbioactive IL-1 alpha - like material and immunochemical IL­Ibeta. As in our earlier study , we found no correlation between bioassay and immunoassay data for IL-lalpha, suggesting possible differences in the epitopes expressed by the skin-associated and the monocyte-derived cytokine. However, the activity in the IL-l bioassay was completely neutralized by anti - IL-lalpha antibody. The observed increase in com.edonal IL-l is difficult to reconcile with the anti-inflammatory effects of the tetracyclines because IL-l is usually considered to be a pro-inflammatory cytokine. Camp et at have already demonstrated that the injection of ~ 100 pg of IL­lalpha - like material from human skin prom.otes dose-related visi­ble erythema lasting up to 48 h [26]. Using this criterion, 58.5% of comedones removed pre-treatment and 65.2% removed post-treat­ment contained sufficient IL-l alpha- Iike material (> 100 pg/mg) to promote visible inflammation if released into the dermis follow­ing spongiosis or rupture of the follicle wall. Furthermore, the mean concentration of IL-lalpha - like material was four times higher following therapy. In follicles in which the initial stimulus (still unidentified but possibly bacterial) has set the inflammatory cascade in motion, the normal chain of events leading to eventual resolution of the lesions will occur irrespective of whether treatment inter­venes. The enhancement ofIL-l production within such follicles by tetracycline therapy may accelerate resolution by decreasing both the extent and duration of the inevitable inflammatory stage of the disease. In normal follicles, in which there is no breach in the folli­cle wall, increased levels of IL-1 resulting from antibiotic therapy will be sequestered within the duct and thus unable to promote inflammatory changes. IL-1 is also recognized to play a role in wound healing following thermal or traumatic damage to the epi­dermal barrier [27,28] and may thus facilitate repair of the damaged follicular epithelium. On the other hand, modulation ofIL-l levels may not be related to the anti-i nflammatory action of the tetracy-

Table III. Levels of Bioactive IL-lalpha - like Material before and After Treatment

pg IL-lalpha/mg Comedonal Material

Number of % Therapy Comedones Positive Range'

Pre-treatment Minocycline (n = 6)J 36 91.7 31-1170 Post-treatmen t 36 77.8 41- 11050 Pre-treatment Tetracycline (n = 5) 29 58.6 23 - 1750 Post-treatment 30 73.3 51 - 11 27 Pre-treatment All patients (n = 11) 65 76.9 23 - 1750 Post-treatment 66 75.8 41 - 11050

• Range of positive values. • Mean for all comedones. , Calculated from Student t X SEM. 'Number of patients in each group. IL-lalpha - like bioactivity was measured in six open comedones extracted from each patient 'J Post-treatment means were significantly higher than pre-treatment means for minocycline-treated patients (p < 0.01 ') and when (

analysis (p < 0.051', Wilcoxon matched pairs).

Meanb ± 95% Confidence Limits'

308.2 ± 101.8 1267.2 ± 695.2'

227.0 ± 158.6 336.9 ± 137.4 272.0 ± 88.6 844.3 ± 196.71

90 EADY ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOG'I;

Table IV. Levels of lmmunochemical IL-1alpha Before and After Treatment

Number of % Therapy Comedones Positive

Pre-treatment Minocycline (n = 6)J 36 16.7 Post-treatment 36 36.1 Pre-treatment Tetracycline (n = 5) 30 6 .7 Post-treatment 30 16.7 Pre-treatment All patients (n = 11) 66 12.1 Post-treatment 66 28.8

"' pg IL-l al pha/ mg Comedonal Materia'!

"' Meanb ± 95% Range' Confidence Limits\

"' 119 - 427 47.3 ± 38.8 78 - 2040 239.4 ± 174.4

279 - 377 21.9 ± 31.4 148 - 1020 72.07 ± 75 .2 119 - 427 35.7 ± 25.1 78 - 2040 163.3 ± 101.1 '

---• Range of positive values. b Mean for all comedones. Comedones in which the level of Ilo 1 alpha was below the detection limit were considered not to contain the cytokine. , Calculated from Srudent t X SEM. i Number of patients. 'The concentration of IL-laipha was significantly higher post-treatment when data from both groups were combined before analysis (p < 0.05, Wilcoxon matched pairs) .

clines, which cou ld instead be mediated by one of several other well documented effects on cellular immunity (vide supra) or could be a secondary consequence of changes in microbial numbers (vide i /lfra).

In a previous study, we found an association between the lower limit of microbial density and comedonal levels of IL-lalpha-like bioactivity although there was no evident correlation between numbers of individual microbial species and concentrations of this cytokine [6] . In addition, three comedones that contained no micro­organisms contained no detectable IL-1alpha-like bioactivity. We believed that any real relationship that may exist between microbial numbers and cytokine content of open comedones might be re­vealed by concomitantly measuring changes in both variables as a result of antibiotic therapy. We confidently expected numbers of propionibacteria to fall based on our own previous data [9] and those of others [8]. Leyden et al [8] measured decreases of> 2 loglo in propionibacteria I numbers in follicular casts following 6 weeks of oral therapy with either 1000 mg of tetracycline or 200 mg of mino­cyeline. In the present study, we found no overall decrease in numbers of propionibacteria/mg comedonal material following ~ 8 weeks of oral therapy with an equivalent dose of tetracycline or 100 mg minocyeline. The reasons for this difference are unelear. It is possible that bacteria in acne lesions respond more slowly to antibiotic therapy than surface organisms or those in normal folli­cles due to interference with antibiotic penetration by hypercornifi­cation and blockage of the follicular opening. Paradoxically, staphy­lococcal numbers increased during therapy in six of 11 patients and in eight patients the increase or decrease in staphylococcal numbers correlated with a corresponding increase or decrease in the concen­tration of IL-1alpha-like bioactivity. Therefore, comedonallevels ofIL-1alpha-like bioactivity may, at least in part, be determined by

staphylococcal and not propionibacteria I population densities. This possibility is supported by the demonstration that as few as 10 cells per monocyte of "Staphylococcus albus" (now Staphylococcus epiderm i­dis) stimulated production of high levels ofIL-l in culture [29]. We must be careful not to overinterpret our preliminary observations, which are based on the measurement of bacterial densities and cyto_ kine levels at one time point only. It would be pertinent now to follow the changes in IL-1 concentration and microbial numbers at several time points during a standard 6-month course of therapy. In this way we should obtain a much clearer picture of how the changes resulting from treatment with the tetracyclines relate to each other, to the therapeutic outcome, and to the anti-inflamma­tory effects of the antibiotics .

The cellular origin of the IL-1 detected in acne comedones was not investigated in the present study. The major species of IL-l found in normal human epidermis appears to be IL-1alpha [30,31]. Cultured human keratinocytes secrete bioactive IL-lalpha but not beta, whereas monocytes secrete both cytokines in a bioactive form [32]. Keratinocytes secrete inactive IL-l beta because they lack the necessary proteinase to convert it into the active form [32]. There­fore, the increase in bioactive IL-l alpha and immunodetectable but biologically inactive IL-l beta fo llowing treatment with tetracy­clines is consistent with their production by ductal keratinocytes.

As far as we are aware, this is the first report to demonstrate an effect of antibiotic therapy on cytokine levels i,1 vivo. Whether or not modulation of IL-l production contributes to the therapeutic efficacy of the tetracyclines in acne cannot be predicted from the results of this study. The demonstration of a wide range of effects of so-called "antibacterial" antibiotics in physiologic concentrations on a variety of functions in eukaryotic cells should at least lead to a

Table V. Levels of Immunochemical IL-l beta Before and After Treatment

pg IL-1beta/mg Comedonal Material

Number of % Therapy Comedones Positive Range'

Pre-treatment Minocycline (n = 6)J 36 52.8' 14 - 104 post-treatment 36 83 .3 4.3 - 540 Pre-treatment Tetracycline (11 = 5) 30 0' Post-treatment 30 93.31 18 - 270 Pre-treatment All patients (n = 11) 66 28.8 14 - 104 post-treatment 66 87.91 4.3 - 540

• Range of positive values. b Mean for all comedones. Comedones in which the level of IL-I beta was below the detection limit were considered not to contain the cytokine. , Calculated from Student t X SEM. i Number of patients. , The number of comedones containing IL-1 beta pre-treatment was higher in the group assigned to minocycline therapy (p < 0.00 I, X2).

Meanb ± 95% Confidence Limits'

22.7 ± 9.4 78.6 ± 41.9

72.5 ± 23.3 12.4 ± 5.8 75.9 ± 24.7x

IThe number of comedones containing IL-lbeta was significantly higher post-treatment in tetracy.cline-treated patients and when data from both groups was combined before analysis (p < 0.001, X2 ) .

• The concentration ofIL-lbeta was significantly higher post-treatment when data from both groups were combined before analysis (p < 0.01, Wilcoxon matched pairs).

VOL. WI, NO. I JULY 1993

reaypraisal of their mode of action in the therapy of acne and other inflammatory dermatoses.

This lVork was supported by Cyanamid (Lederle Laboratories, Hampshire, U.K.).

REFERENCES

1. Cunliffe WJ: Clinical Features of Acne. In: Acne. Martin Dunitz Ltd., London, 1989, pp 11-75

2. Scott DG, Cunliffe WJ, Gowland G: Activation of complement-a mechanism for the inflammation in acne. Br J Dermatol 101:315-320,1979

3. Dahl MGC, McGibbon DH: Complement C3 and immunoglobulin in inflammatory acne vulgaris. Br J Dermatol 101:633-640, 1979

4. Norris JFB, Cunliffe W]: A histological and immunocytochemical study of early acne lesions. Br] Dermatol 118:651 - 659, 1988

5. Barker ]NWN, Mitra RS, Griffiths CEM, Dixit VM, Nickoloff B]: Keratinocytes as initiators of inflammation. Lancet 337:211 - 214, 1991

6. Ingham E, Eady EA, Goodwin CE, Cove ]H, Cunliffe WJ: Pro-in­flammatory levels of interleukin-1 alpha - like bioactiviry are present in the majoriry of open comedones in acne vulgaris.] Invest Derma­tol 98:895 -901, 1992

7. Eady EA, Holland KT, Cunliffe WJ: The use of antibiotics in acne therapy: oral or topical administration? ] Antimicrob Chemother 10:89-115,1982

8. Leyden JJ, McGinley K], Kligman AM: Tetracycline and minocycline treatment. Effects on skin surface lipid levels and Propiotlibacteritlm awes. Arch DermatoI118:19-22, 1982

9. Eady EA, Cove ]H, Holland KT, Cunliffe W]: Superior antibacterial action and reduced incidence of bacterial resistance in minocycline compared to tetracycline-treated acne patients. Br ] Dermatol 122:233-244,1990

10. Holland KT, Ingham E, Cunliffe WJ: A review. The microbiology of acne.] Appl Bacteriol 51:195-215, 1981

11. Holland KT: Microbiology of Acne. In: Acne. Martin Dunitz Ltd., London, 1989, pp 178-210

12. Plewig G, Schopf E: Anti-inflammatory effects of antimicrobial agents: an in vivo study. J Invest Dermatol 65:532- 536, 1975

13. Roche Y, Fay M, Gougerot-Pocidalo M-A: Interleukin-l production by antibiotic-treated human monocytes. ] Antimicrob Chemother 21:597-607,1988

14. Ingham E: Modulation of the proliferative response of murine thymo­cytes stimulated by IL-1, and enhancement of IL-l beta secretion from mononuclear phagocytes by tetracyclines.] Antimicrob Che­mother 26:61-70, 1990

15. Bailly S, Pocidalo J-J, Fay M, Gougerot-Pocidalo M-A: Differential

MODULATION OF INTERLEUKIN-I BY TETRACYCLINES 91

modulation of cytokine production by macrolides: interleukin-6 production is increased by spiramycin and erythromycin. Antimi­crob Ag Chemother 35:2016-2019,1991

16. Burke BM, Cunliffe W]: The assessment of acne vulgaris - the Leeds technique. Br] Dermatol 111 :83 - 92, 1984

17. van Oers MH], van der Heyden AAPAM, Aarden LA: Interleukin 6 (IL-6) in serum and urine of renal transplant recipients. Clin Exp Immunol 71:314-319,1988

18. Sokal RR, Rohlf F]: Correlation. In: Biometry, 2nd cd., Chapter 15. WH Freeman & Company, San Francisco, CA, 1981, pp 561- 616

19. Kligman AM, Mills OH, McGinley K], Leyden]]: Acne therapy with tretinoin in combination with antibiotics. Acta Dermato Venereol (Stockl) 55(suppl 74):111-115, 1975

20. Leyden]], McGinley K], Mills OH, Kligman AM: Topical antibiotics and topical antimicrobial agents in acne therapy. Acta Dermato Venereol (Stockl) 60(suppl 89):75-82, 1980

21. Munster AM, Loadholdt CB, Leary AG, Barnes MA: The effects of antibiotics on cell-mediated immuniry. Surgery 81 :692 - 695, 1977

22. Forsgren A, Banck G: Influence of antibiotics on lymphocyte function ill "itro. Infection 6(suppl 1):91 -97, 1978

23. Hauser WE, Remington ]S: Effects of antibiotics on the immune re­sponse. Am J Med 72:711 - 716, 1982

24. Thong YH, Ferrante A: Inhibition of mitogen-induced human lym­phocyte proliferative responses by tetracycline analogues. Clin Exp Immunol 35:443-446, 1979

25. Esterly NB, Furey NL, Flanagan LE: The effects of antimicrobial agents on leucocyte chemotaxis.] Invest Dermatol 70:51- 55, 1978

26. Camp R, Fincham S, Ross J , Bird C, Gearing A: Potent inflammatory properties in human skin of interleukin-1 alpha-like material iso­lated from normal skin. J Invest Dermatol 94:735- 741, 1990

27. Gahring LC, Buckley A, Daynes RA: Prescnce of epidcrmal-derived thymocyte activating factor/intcrleukin-1 in normal human stra­tum corncUln.] Clin Invcst 76:1585-1591,1985

28. Sauder DN, Orr FW, Matic S, Stetsko D, Parker KP, Chizzonite R, Kilian PL: Human interleukin-l alpha is chemotactic for normal human keratinocytes. Immunol Lett 22:123-128,1989

29. Chu E, Roscnwasser L], Dinarello CA, Lareau M, Geha RS: Role of interlcukin 1 in antigen-specific T-cell proliferation. ] lnununol 132:1311 - 1316,1984

30. Gruaz D, Didierjean L, Grassi], Frobcrt Y, Dayer J-M, Saurat J-H: Interleukin-1 alpha and beta in psoriatic skin: enzymoinununoassay, immunoblot studies and effect of synthctic rctinoids. Dermatologica 179:202-206, 1989

31. Cooper KD, Hammerberg C, Baadgaard 0, Elder JT, Chan LS, Sauder DN, Voorhees JJ, Fisher G: IL-1 activiry is rcduced in psoriatic skin. J ImmunoI144:4593-4603, 1990

32. Mizutani H, Black R, Kupper TS: Human keratinocytes produce but do not process pro-interleukin-l (IL- l) beta. J Clin Invest 87: 1066-1071, 1991

CALL FOR PAPERS

Tokyo Symposium: New Frontiers in Hair Research will be held October 25-26, 1993 in Tokyo, Japan. The meeting will be immediately before the Second Tricontinental Meeting of the Japanese Society for Investigative Dermatology (JSID), the Society for Investigative Der­matology (SID), and the European Society for Dermatological Research (ESDR), Kyoto,Japan.

President: Hideoki Ogawa, M.D., Department ofDermatology,Juntendo University School of Medicine.

For further information please contact: Symposium Secretariat Ryusuke Imai, M.D. Fax, 81-3-3813-2205.