| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Department of Molecular Preventive Medicine and Solution Oriented Research for Science and Technology,* The University of Tokyo School of Medicine, Tokyo; the Department of Respirology, Chiba University School of Medicine, Chiba; the Department of Human Pathology, School of Medicine, Tokyo Medical and Dental University, Tokyo; and the Second Department of Internal Medicine, Nara Medical University of Medicine, Nara, Japan
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Because of their clinical and immunopathological similarities, it has been suggested that the most common mycobacterial infection, tuberculosis, might be related to sarcoidosis. However, despite the use of bacterial culture systems, and histochemical and polymerase chain reaction (PCR)-based methods, an association between Mycobacterium tuberculosis and sarcoidosis remains controversial.4-6 Propionibacterium acnes, which is ananaerobic nonspore-forming gram-positive rod bacterium that exists indigenously on the skin or mucosal surfaces,7 has been reported as one of the suggested causative antigens of sarcoidosis.8 Some studies usingquantitative PCR have revealed markedly higher levels of P. acnes genomes in the mediastinal or superficial lymph nodes (LNs) of sarcoid patients than in those of controls, suggesting that there is an intrinsic infection because of P. acnes in patients with sarcoidosis.9-11
The triggering process in pulmonary granuloma formation is thought to consist of airborne or blood-borne antigens anchoring in the lung, and antigen-presenting cells (APCs), such as macrophages or dendritic cells,12 accumulating and surrounding them for phagocytosis and subsequent antigen presentation.13 Based on this view, a number of animal models of pulmonary granuloma have been proposed, notably murine schistosomiasis, with antigen embolization to hold antigens in the lung.14,15 However, long-term antigen deposition on the pulmonary interstitium is not suitable for clinical studies, and it is unlikely that disseminated blood-borne antigens are responsible for all cases of pulmonary granuloma.
We observed that an immune response against indigenous P. acnes already exists in the normal lung, which is believed to be germ-free. We showed that expansion of the numbers of recirculating P. acnes-primed cells, produced by extrapulmonary sensitization, can specifically cause granulomatous changes of the lung sharing several similarities with pulmonary sarcoidosis, even in the absence of antigen anchoring. Furthermore, antibiotic treatment to eliminate pre-existing P. acnes substantially reduced pulmonary dysfunction. Based on these observations, we propose a novel view of the pathogenesis of unresolved pulmonary granulomatosis such as sarcoidosis, as well as a potentially effective treatment for this condition by means of common antibiotics.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Female 5- to 7-week-old C57BL/6J mice were obtained from CLEA Japan (Shizuoka, Japan) or Japan SLC (Tokyo, Japan), and kept under specific pathogen-free conditions in the animal facility of the Department of Molecular Preventive Medicine, University of Tokyo School of Medicine, Tokyo, Japan. All animal experiments complied with the guidelines of the University of Tokyo.
Immunostaining
The following anti-mouse monoclonal antibodies were used: CD4 (clone; RM4-5), biotinylated interferon (IFN)-
(XMG1.2), and biotinylated interleukin-4 (BVD6-24G2), all from BD PharMingen (San Diego, CA); biotinylated F4/80 (CI:A3-1) and CD11c (N418), both from Serotec (Oxford, UK); DEC-205 (NLDC-145; BMA Biomedicals, Augst, Switzerland); and mouse monoclonal antibody to P. acnes recognizing lipoteichoic acid of the plasmalemma.16
The secondary antibodies were as follows: an alkaline phosphatase-labeled anti-rat (Jackson ImmunoResearch Laboratories, West Grove, PA) or hamster (Cedarlane, Ontario, Canada) IgG (IgG), or avidin (Nichirei, Tokyo, Japan), and a horseradish peroxidase-labeled anti-rat (BioSource, Camarillo, CA) or mouse (DAKO, Carpinteria, CA) Ig.
Single and double immunostaining were performed by the indirect immunoalkaline phosphatase and immunoperoxidase methods.17
For double-immunofluorescence staining, acetone-fixed 6-µm fresh-frozen tissue sections were incubated with anti-CD4 and then with Alexa Fluor 488 anti-rat Ig (Molecular Probes, Eugene, OR). They were next incubated with biotinylated IFN- or biotinylated interleukin-4 followed by Alexa-594-conjugated avidin (Molecular Probes), and were observed using fluorescence microscopy.18
Reverse Transcriptase (RT)-PCR
Samples of 1 µg of total RNA were isolated from the lungs and secondary LN specimens of the lung, liver, skin, and pancreas of specific pathogen-free mice using Trizol (Invitrogen, Groningen, The Netherlands) according to the manufacturer’s instructions. The RNA samples were then reverse-transcribed into complementary DNA (cDNA)16,19 and amplified. PCR products of the 16s ribosomal RNA of P. acnes were electrophoresed on 2.5% agarose gels. Bands visualized by ethidium bromide staining were of the expected size for each mRNA product. The oligonucleotide primers for P. acnes, M. tuberculosis, and Propionibacterium granulosum were designed as described previously:10 P. acnes, forward 5'-GCGTGAGTGACGGTAATGGGTA-3' and reverse 5'-TTCCGACGCGATCAACCA-3'; M. tuberculosis, forward 5'-TCCTATGACAATGCACTAGCCG-3' and reverse 5'-GCCAACTCGACATCCTCGAT-3'; and P. granulosum, forward 5'-ACATGGATCCGGGAGCTTC-3' and reverse 5'-ACCCAACATCTCACGACACG-3'. Contamination by P. acnes during the course of the experiment was checked with buffer controls (data not shown). The primers for GAPDH as an internal control were described previously.19 The PCR conditions were denaturation at 95°C for 5 minutes, followed by 40 cycles of 95°C for 30 seconds, 58°C for 60 seconds, and 72°C for 90 seconds, with a final step of 72°C for 10 minutes.
Antigen-Specific Proliferation Assay
In vitro cell-proliferation assays were performed as described previously.20 Briefly, peribronchial, inguinal, hepatic, and pancreatic LN cells (105 cells/190 µl/well) from normal mice were stimulated with antigens (P. acnes and OVA; 10 µg/10 µl of culture medium) at 37°C for 72 hours. After incubation, cell numbers were measured using the Premix WST-1 cell-proliferation assay system (Takara Bio, Shiga, Japan), according to the manufacturer’s instructions.
Adoptive Transfer of P. acnes-Primed Helper T Cells
P. acnes-sensitized CD4+ T cells were obtained from the inguinal LN of the mice immunized three times. Immunization was performed by subcutaneous injection of 400 µg of heat-killed P. acnes (ATCC 11828; American Type Culture Collection, Manassas, VA) with complete Freund’s adjuvant (CFA) (Difco, Detroit, MI) or CFA alone into the footpad at 2-week intervals. CD4+ cells were isolated using the MACS system (Miltenyi Biotech, Bergisch Gladbach, Germany), according to the manufacturer’s instructions. The purity of the CD4+ cell populations was 94% or more, as confirmed by immunofluorescence flow cytometry. Isolated CD4+ cells [2 x 106 cells/200 µl of phosphate-buffered saline (PBS)] were injected into the tail vein of normal mice, and 2 weeks later the lungs were examined histologically.
Induction of Chronic Pulmonary Granulomatosis
Lung granulomatosis was induced by multiple immunizations, which were performed by subcutaneous injection of 400 µg of heat-killed P. acnes with CFA into the footpad at 2-week intervals.
Flow-Cytometric Analysis of Cells
Bronchoalveolar lavage (BAL) cells were collected by five injections of 0.8 ml of sterile PBS containing 2% fetal calf serum (Sigma, St. Louis, MO) and 2 mmol/L ethylenediaminetetraacetic acid. The total number of BAL cells was counted with a hemocytometer. BAL cells were analyzed using an EPICS Elite instrument (Beckman Coulter, Miami, FL) and preincubated with rat anti-mouse CD16/CD32 (clone 2.4G2) monoclonal antibody to block FcR-mediated binding, followed by incubation with fluorescein isothiocyanate-conjugated anti-CD4 (H129.19) and phycoerythrin-conjugated anti-CD8
(53-6.7) monoclonal antibodies (both from BD PharMingen) for 25 minutes at 4°C.
Serological Analysis
Serum calcium levels were determined using a Fuji DRI-CHEM 5500V (Fuji Medical System, Tokyo, Japan), and angiotensin-converting enzyme activities were measured by angiotensin-converting enzyme color (Fuji REBIO, Tokyo, Japan) according to the manufacturer’s instructions.
Antigen-Preloading Experiment
Doses of 10 µg, 1 µg, and 0.1 µg of P. acnes and PBS were intratracheally administered to mice 1 week before the first immunization of the lung granulomatosis model; 1 µg of P. acnes contained 2.5 x 105 organisms. In the control experiment, Lactobacillus gasseri (ATCC 33323) was used as the antigen in place of P. acnes.
Antibiotic Treatment
Minocycline hydrochloride (MINO) (Wyeth Lederle, Tokyo, Japan), clindamycin (CLDM) (Pharmacia, Tokyo, Japan), and gentamicin sulfate (GM) (Schering-Plough, Osaka, Japan) were used. Briefly, 133 µg of MINO, 1.6 mg of CLDM, and 53 µg of GM were intratracheally administrated on day 1. Thereafter, the same dose of each antibiotic was injected intraperitoneally every day for 1 week before immunization, followed by intraperitoneal injections three times per week. Twenty µg of prednisolone sodium succinate (Shionogi, Osaka, Japan) was intraperitoneally injected three times per week as a positive control. Short-term treatment with MINO and CLDM were started at the day of the third immunization. Throughout the experiment, the mice were given water containing each antibiotic at the indicated dose.
Delayed-Type Hypersensitivity Reaction Test
Mice were immunized with 50 µg of M. tuberculosis (Difco) included in CFA via their footpads. Two weeks after immunization, 10 µg of purified protein derivative from M. tuberculosis (BCG, Tokyo, Japan) was injected into the ear of each mouse, along with injections of PBS into the opposite ear. Measurements of ear swelling were performed 48 hours after the purified protein derivative challenge.
Statistical Analysis
Differences were evaluated using two-factor factorial analysis of variance and Fisher’s protected least significant difference. P values <0.05 were considered to be statistically significant.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
If there is a pre-existing immune response to P. acnes, it should be possible to detect this bacterium in the healthy lung. We therefore performed an immunohistochemical test for P. acnes on fresh frozen lung sections from C57BL/6 mice. We observed positive staining in groups of two to five round granules. All of these had been phagocytosed by lung cells, most of which lay adjacent to the alveolar space (Figure 1, A and B)
. Double immunostaining revealed that the P. acnes-positive cells expressed F4/80 (Figure 1C)
, which is a known macrophage marker, but not the dendritic cell markers CD11c (Figure 1D)
and DEC205 (Figure 1E)
.21,22
In addition, RT-PCR analyses of normal lungs with removal of the trachea and the main-stem bronchus, showed a graded distribution of P. acnes genomes (Figure 1F)
. By contrast, the genomes of M. tuberculosis and P. granulosum, which are mucosal commensal bacteria similar to P. acnes, were not detected in the normal lungs (Figure 1G)
.
|
Peripheral APCs transport antigens to draining LNs for presentation,22,23
even in the steady state. Our results in Figure 1
suggest that P. acnes exists indigenously on the normal alveolar surfaces as well as the skin and mucosal surfaces of the oral cavity and intestine.7
To test for a P. acnes-specific immune response in normal pulmonary LNs, we first established the presence of P. acnes-genomes in normal pulmonary LNs as well as other LNs by RT-PCR (Figure 2A)
. This result indicated the steady-state transport of indigenous P. acnes from the periphery to the regional LNs of the lung. Therefore, we subsequently tested whether a specific immune response to P. acnes had been established in these LNs. As expected, we found that normal peribronchial LN lymphocytes proliferated specifically in response to P. acnes, as did cells in the inguinal, hepatic, and pancreatic LNs (Figure 2B)
.
|
Interactions between antigen-bearing APC and Th cells are essential in the development of granuloma, which is enhanced by an influx of circulating antigen-primed T cells.13,24
We next determined whether the introduction of circulating P. acnes-primed T cells resulted in granuloma formation in the normal lung. We obtained P. acnes-sensitized CD4+ T cells from the draining LN of a footpad that had been repeatedly immunized with P. acnes using CFA, and injected them into the tail veins of normal mice. Two weeks after the transfer of 2 x 106 T cells, we observed granulomatous changes, in the form of aggregations of epithelioid and mononuclear cells, in the lung and the liver (Figure 3A)
, whereas the adoptive transfer of CFA-primed T cells produced no such effects (Figure 3B)
. Furthermore, the mice treated with P. acnes-primed T cells showed increased numbers of total cells and CD4+ T cell counts of both BALs and peribronchial LNs compared with the control group, which supported the histological findings (Figure 3, C and D)
.
|
As a practical application of this transfer model, and to stimulate chronic pulmonary granulomatosis, we induced the continuous extrapulmonary expansion of P. acnes-primed T cells by repeated immunization via the footpad using CFA to expand the antigen-specific T cells efficiently. Distinct granulomas formed predominantly in the subpleural and peribronchovascular areas of the lung in mice that were treated in this way (Figure 4A)
. Immunohistochemical analyses showed that the typical granulomas consisted of central APCs, with CD4+ T cells on the periphery12,13
(Figure 4, B and C)
. Furthermore, these CD4+ cells expressed IFN- but not interleukin-4, indicating that the granulomas were of the Th1-type (Figure 4D)
.
|
. These results were also consistent with the histological findings (data not shown), even though the proportion of CD4/8 cells was greatest in the group injected twice, the lungs of which showed signs of the beginning of granuloma formation (Figure 5B)
. In addition, these increases in the numbers of BAL cells were much higher than those observed in the group that received CFA alone (Figure 5C)
.
|
and abnormal accumulations of CD4+ T cells in the red pulp of the spleen (Figure 6B)
in the mice that were frequently immunized.
|
If the presence of indigenous P. acnes in the healthy lung leads to pulmonary granuloma, their numbers should influence the extent of the lesions. To test this hypothesis, we preloaded live P. acnes into healthy murine lungs before immunization. To exclude the possibility that the intratracheal injection alone induced granulomas, we ascertained that there were no granulomas in the control lungs from nonimmunized mice at either the initial stage or the end-point of the experiment (data not shown). Total counts of cells in the BAL, which were collected after three immunizations, depended on the dose of preloaded P. acnes (Figure 7A)
; the result of the histological examination of the lesions was consistent with this observation (Figure 7B)
. By contrast, there were no increases in the numbers of BAL cells in the control experiments using L. gasseri in place of P. acnes (Figure 7C)
.
|
To evaluate further the importance of pre-existing P. acnes, we reduced their numbers before immunization with the antimicrobial reagents MINO and CLDM, both of which are known to be effective drugs for acne vulgaris induced by P. acnes.26,27
GM, which is inhibitory to M. tuberculosis but not to P. acnes, was used as negative control.7
Two weeks after the third immunization, the MINO- and CLDM-treated mice, and even the short-term CLDM-treated mice, showed a marked reduction in total BAL cell counts (Figure 8A)
: the CD4+ BAL cell counts in these three groups were reduced by 53.5%, 42.1%, and 74.2%, respectively (Figure 8B)
. This effect was not observed in the mice treated with GM (Figure 8, A and B)
. Moreover, the doses of P. acnes genomes were altered in response to these antimicrobial effects (Figure 8C)
. Histological examinations also showed consistent improvement of granulomatous lesions in the mice treated with effective antibiotics (Figure 8E)
. Although immunomodulatory properties of antibiotics, especially MINO, have been described previously,28
the tests for delayed-type hypersensitivity showed no nonspecific immune effects, at least in our experimental model using C57BL/6 mice (Figure 8D)
.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
In the present study, we identified P. acnes in normal murine alveolar cells by immunostaining (Figure 1, A and B)
. All of these P. acnes were taken up by lung cells, and the P. acnes-bearing cells expressed F4/80 rather than CD11c or DEC205 (Figure 1; C, D, and E)
, consistent with the known ability of macrophages to phagocytose and deliver antigens to dendritic cells in the lung.30,31
As far as the end of the airway, airborne organisms are impacted and eliminated by mechanical defenses including mucociliary clearance and coughing.32
Nevertheless, a small number of P. acnes might escape from this system and reside on the alveolar surface. The existence of P. acnes-bearing APCs in the healthy lung prompted us to examine whether there is an immune response against P. acnes in the pulmonary LNs of normal mice. Indeed, P. acnes genomes were detected in normal pulmonary LNs as well as the lungs (Figures 1F and 2A)
, and lymphocytes from LNs showed P. acnes-specific proliferation (Figure 2B)
, suggesting that these cells had already been exposed to P. acnes by lung-derived APCs and had established a memory response. Additionally, these results indicate that P. acnes were continuously transported to pulmonary regional LN in the steady state. Because of this constant delivery of antigens to the pulmonary LNs for a long period, the small number of indigenous P. acnes in the normal lung would be enough for a specific immune response, but not for the formation of the steady-state granuloma. Although mycobacterial, atypical mycobacterial, and other propionibacterial antigens were potential candidates as endogenous microorganisms triggering pulmonary granuloma formation, genomic analyses revealed an absence of these organisms in the lungs of specific pathogen-free C57BL/6 mice (Figure 1G
and data not shown).
Because of this steady-state memory response against P. acnes in the lung, we subsequently hypothesized that an influx of P. acnes-sensitized T lymphocytes could cause lung inflammation even without artificial antigen-anchoring. The adoptive transfer of P. acnes-sensitized LN CD4+ T cells into naïve mice resulted in granulomatous changes in the lung (Figure 3A)
. This showed that extrapulmonary LN CD4+ T cells primed with P. acnes could interact with pulmonary resident cells via circulation and induce granuloma formation in the normal lung. We therefore hypothesized that a continuous supply of P. acnes-sensitized T cells should lead to chronic pulmonary granuloma formation, and consequently performed continuous remote sensitization of normal mice with P. acnes. These mice exhibited distinct pulmonary granulomas, distributed in lymph-rich spaces33
such as the subpleural, peribronchial, and perivascular areas (Figure 4A)
, and showed the typical cellular components of granuloma13
(Figure 4, B and C)
and preferential Th1 cytokine expression (Figure 4D)
. These features are similar to those of pulmonary sarcoidosis.1,3,25
In addition, an elevated ratio of CD4 to CD8 BAL lymphocytes (Figure 5B)
was observed in the group that were immunized twice, which was accompanied by increased serum calcium levels (data not shown). These observations are consistent with those of previous studies demonstrating an influence of activated macrophages on calcium metabolism34
and a positive correlation between serum calcium levels and the BAL CD4/CD8 ratios in sarcoid patients.35
Moreover, we found extrapulmonary lesions in the liver and spleen, both of which are frequently affected in sarcoidosis (Figure 6, A and B)
.1,8,25
Thus, this repeated P. acnes-immunization model, without any direct exposure of antigen to the lung, showed several similarities to the characteristics of sarcoid patients.
Because an influx of P. acnes-sensitized T cells into the lung can trigger granuloma formation (Figure 3A)
, we considered that interactions between indigenous P. acnes-loaded lung APCs and LN T cells would be essential for the development of granulomas. To confirm this, we examined whether changes in the number of pre-existing P. acnes cells in the lung had an effect on pulmonary granuloma formation. As expected, preloading of P. acnes exacerbated the pulmonary disorders (Figure 7, A and B)
, whereas reduction of the P. acnes population by antimicrobial treatment reduced the pulmonary lesions (Figure 8)
. These results suggest not only a pivotal role of normally localized P. acnes in the formation of pulmonary granuloma by extrapulmonary P. acnes sensitization, but also the potential clinical usefulness of antimicrobial eradication targeting lung-indigenous P. acnes for the treatment of pulmonary granulomatosis induced by similar pathogenesis.
The etiology of sarcoidosis remains to be resolved. Immunosuppressive, mainly corticosteroidal, therapy has been used for more than 50 years for this condition, but the long-term effects of steroidal treatment in chronic pulmonary sarcoidosis are still disputed,8 and the high relapse rate after treatment and the side effects of long-term use are often a clinical challenge.36 However, the successful clinical report of cutaneous sarcoidosis treated with minocycline,37 for example, has highlighted the usefulness of such an alternative therapy.
In this study, we have produced a novel murine pulmonary granuloma model with several features similar to those of pulmonary sarcoidosis. Sarcoidosis is a well-described clinical condition, and clearly the whole pathogenesis cannot be explained by this experimental model alone. However, if, as we have shown in mice, P. acnes also exists in the healthy human lung, people with a unique genetic background, as is reported in sarcoid patients,1,8,25 might readily form pulmonary lesions after excessive sensitization with P. acnes even at extrapulmonary sites, such as loci of acne vulgaris. We suggest that this new view of pulmonary sarcoidosis deserves further investigation and might provide the basis for novel therapeutic strategies.
|
Footnotes |
|---|
Supported in part by grants from Solution Oriented Research for Science and Technology and the Ministry of Health, Labor, and Welfare of Japan.
Accepted for publication May 7, 2004.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. Kettritz, U. Goebel, A. Fiebeler, W. Schneider, and F. Luft The protean face of sarcoidosis revisited Nephrol. Dial. Transplant., October 1, 2006; 21(10): 2690 - 2694. [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. McCaskill, K. D. Chason, X. Hua, I. P. Neuringer, A. J. Ghio, W. K. Funkhouser, and S. L. Tilley Pulmonary Immune Responses to Propionibacterium acnes in C57BL/6 and BALB/c Mice Am. J. Respir. Cell Mol. Biol., September 1, 2006; 35(3): 347 - 356. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Mukherjee Is Propionibacterium acnes the cause of sarcoidosis? Thorax, November 1, 2004; 59(11): 942 - 942. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
spleen weight x 100/whole body weight. n = 5. Data are means ± SEM. Data shown are from one representative of more than three independent experiments. E: Histological findings by H&E staining. Scale bars, 100 µm.