- Email: [email protected]
Shooting at Skin Propionibacterium acnes: To Be or Not to Be on Target
OA Alexeyev and CC Zouboulis Skin Propionibacterium acnes
The effect of TLR-2 blockade not only on secretion of IL-12 p40 and IL-6 but also on IFN-g might again reflect the IL-12 dependence of a-NAC-induced IFN-g secretion. Additionally, in a-NAC-stimulated monocytes, p44/42 (Erk) and NF-kB were phosphorylated, both of which are downstream of TLR-2. A similar phosphorylation pattern and TLR-2 dependence has been shown for bacterial HSP-60 and human monocytes (Zhao et al., 2007). More evidence comes from the receptor itself: surface levels of TLR-2 were higher in a-NACtreated cells than in controls, suggesting that a-NAC upregulates its own receptor as it has been shown for other TLR-2 ligands (Chamorro et al., 2009; Niebuhr et al., 2009). These data suggest that TLR-2 is a monocyte receptor for a-NAC. Nevertheless, TLR-2 may act in concert with other receptors, as blockade of TLR-2 did not completely block the effects of a-NAC. This is the first report that describes a nonenzymatic autoantigen capable of mediating specific IgE and T cell responses relevant in AD, which also possesses intrinsic immunomodulatory potential. Our results clearly demonstrate that a-NAC is a potent inducer of a Th1 response in human PBMCs via the induction of IL-12, and that the monocyte is the primary target cell for a-NAC. This monocyte activation seems to depend on the interaction of a-NAC and TLR-2. Moreover, our data suggest that a-NAC-activated monocytes could be particularly efficient in priming
and maintaining an exaggerated Th1 response.
Heratizadeh A, Mittermann I, Balaji H et al. (2011) The role of T-cell reactivity towards the autoantigen alpha-NAC in atopic dermatitis. Br J Dermatol 164:316–24
CONFLICT OF INTEREST
Mittermann I, Reininger R, Zimmermann M et al. (2008) The IgE-reactive autoantigen Hom S 2 induces damage of respiratory epithelial cells and keratinocytes via induction of IFNgamma. J Invest Dermatol 128:1451–9
The authors state no conflict of interest.
ACKNOWLEDGMENTS We thank Gabriele Begemann, Petra Kienlin, and Ute Staar for their excellent technical assistance. This study was supported by a grant from Deutsche Forschungsgemeinschaft (DFG) to GRK 1441/1 and DFG-KliFo 250 WE1289/8-1 and by the Austrian Science Fund (FWF). 1,3
1,3
Susanne Hradetzky , Hari Balaji , Lennart M. Roesner1, Annice Heratizadeh1, Irene Mittermann2, Rudolf Valenta2 and Thomas Werfel1 1
Division of Immunodermatology and Allergy Research, Department of Dermatology and Allergy, Hannover Medical School, Hannover, Germany and 2Center for Pathophysiology, Infectiology and Immunology, Division of Immunopathology, Department of Pathophysiology and Allergy Research, Medical University of Vienna, Vienna, Austria E-mail: [email protected] 3 These authors contributed equally to this work. SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at http://www.nature.com/jid
REFERENCES Bieber T (2008) Atopic dermatitis. N Engl J Med 358:1483–94 Chamorro S, Garcia-Vallejo JJ, Unger WW et al. (2009) Tlr triggering on tolerogenic dendritic cells results in TLR2 up-regulation and a reduced proinflammatory immune program. J Immunol 183:2984–94
Mossabeb R, Seiberler S, Mittermann I et al. (2002) Characterization of a novel isoform of alphanascent polypeptide-associated complex as IgE-defined autoantigen. J Invest Dermatol 119:820–9 Natter S, Seiberler S, Hufnagl P et al. (1998) Isolation of cDNA clones coding for IgE autoantigens with serum IgE from atopic dermatitis patients. FASEB J 12:1559–69 Niebuhr M, Lutat C, Sigel S et al. (2009) Impaired TLR-2 expression and TLR-2-mediated cytokine secretion in macrophages from patients with atopic dermatitis. Allergy 64:1580–7 Schmid-Grendelmeier P, Fluckiger S, Disch R et al. (2005) IgE-mediated and T cell-mediated autoimmunity against manganese superoxide dismutase in atopic dermatitis. J Allergy Clin Immunol 115:1068–75 Sloane JA, Blitz D, Margolin Z et al. (2010) A clear and present danger: endogenous ligands of Toll-like receptors. Neuromolecular Med 12:149–63 Werfel T (2009) The role of leukocytes, keratinocytes, and allergen-specific IgE in the development of atopic dermatitis. J Invest Dermatol 129:1878–91 Zeller S, Rhyner C, Meyer N et al. (2009) Exploring the repertoire of IgE-binding self-antigens associated with atopic eczema. J Allergy Clin Immunol 124:278–85, 285. e1–7 Zhao Y, Yokota K, Ayada K et al. (2007) Helicobacter Pylori heat-shock protein 60 induces interleukin-8 via a toll-like receptor (TLR)2 and mitogen-activated protein (Map) kinase pathway in human monocytes. J Med Microbiol 56(Part 2):154–64
Shooting at Skin Propionibacterium acnes: To Be or Not to Be on Target Journal of Investigative Dermatology (2013) 133, 2292–2294; doi:10.1038/jid.2013.116; published online 4 April 2013
TO THE EDITOR We have read with interest the article by Fitz-Gibbon et al. (2013), which reports that different Propionibacterium acnes strain populations exist in acne in comparison with normal skin. Previous
studies have, indeed, suggested that globally disseminated clonal lineages within the P. acnes type IA1 division may have an etiological role in acne, whereas other lineages may be associated with health (Lomholt and Kilian,
Accepted article preview online 8 March 2013; published online 4 April 2013
2292 Journal of Investigative Dermatology (2013), Volume 133
2010; McDowell et al., 2012). If confirmed, this observation may lead to a better understanding of acne pathogenesis and, potentially, new therapeutic strategies. There are, however, major limitations in their study design related to: (i) skin sampling methodology, (ii) choice of anatomical site for
OA Alexeyev and CC Zouboulis Skin Propionibacterium acnes
sampling, and (iii) adoption of 16S rDNA–based ribotyping for the identification of P. acnes lineages NEGLECTED BIAS IN P. ACNES SAMPLING METHODOLOGY Although genomic approaches for the analysis of microbial communities provide much greater resolution than traditional bacterial culture (Grice and Segre, 2011), sequence-based technologies still require molecular guidelines for establishing tissue-sequence correlates (Fredericks and Relman, 1996); specific microbial sequences must be demonstrated in areas of tissue pathology. In acne vulgaris, tissue pathology is confined to the pilocebaceous unit, and it is imperative that anatomical considerations are taken into account when interpreting microbiological data. It is regrettable that Fitz-Gibbon et al. failed to consider recent studies reporting direct visualization of anatomically distinct P. acnes skin populations (Alexeyev et al., 2012; Jahns et al., 2012), and the biases associated with different sampling techniques (Alexeyev and Jahns, 2012). Four distinct anatomical populations of P. acnes have been reported: superficial, intrastratum corneum, infundibular, and follicular. Fitz-Gibbon et al. purported to present the microbiota of pilosebaceous units after collecting skin ‘microcomedone’ samples using the Biore Deep Cleansing Pore strips (essentially representing a modification of the well-known tape-stripping technique). As no further details on this crucial step are provided (except that the manufacturer’s instructions were followed), the reader may assume that no skin disinfectant was used before applying the strips. This would lead to sampling of superficial P. acnes skin populations, i.e., populations residing on the stratum corneum. The tape-stripping technique represents a classic approach for removing complete cell layers from the superficial part of the human stratum corneum (Pinkus, 1951). Approximately 50 mg of tissue is removed from 1.0 cm2 after one tape-stripping procedure (Mohammed et al., 2012). This indicates that intrastratum corneum P. acnes populations are also targeted by this sampling methodology. Furthermore, the amount of stratum corneum retrieved is influenced by the force used to remove
the stripping tape and the level of pressure applied during strip application, thus making standardization of the procedure very difficult (Breternitz et al., 2007). Direct visualization of hair follicles from acne patients and controls has shown that P. acnes colonization levels range from single cells to large macrocolonies (Alexeyev et al., 2012; Jahns et al., 2012). These populations are often located in the anaerobic part of the hair follicle, which is inaccessible by the tape-stripping technique. Should P. acnes, as suggested, exist in hair follicles in the form of biofilms (Jahns et al., 2012), which are often irreversibly attached to the surface (Hall-Stoodley et al., 2004), then the utility of the tapestripping technique for collecting bacteria from a hair follicle is further diminished. To provide a meaningful quantitative analysis of bacterial sequence distribution in the human tissue, the same quantity of biological material from acne and control subjects must be tested. Unfortunately, quantitative estimates of the microcomedone numbers analyzed were not provided. Finally, detaching microcomedones from silica as part of the Biore Deep Cleansing Pore strip approach is volatile yielding highly variable amounts of material. THE NOSE IS A NONREPRESENTATIVE ANATOMICAL SITE FOR SAMPLING ACNE SKIN Another deep concern is that subjects were sampled from the nose, which is not commonly affected by acne. On this basis, we believe the nose is a very poor site for sampling. Indeed, a significant proportion of the acne patients sampled did not appear to have acne affecting the nose, suggesting that in these cases normal follicles would have been analyzed. Furthermore, even in cases where acne did affect the nose, the use of a pore strip will result in multiple follicles being sampled, including those that are normal, which we believe will further compromise the results. SINGLE LOCUS TYPING BASED ON 16S rDNA LACKS RESOLUTION The data reported by Fitz-Gibbon et al. is fundamentally different from that provided by the Aarhus and Belfast groups,
which were on the basis of population genetic analysis of isolates by Multilocus Sequence Typing (MLST) (Kilian et al., 2012; McDowell et al., 2012). Although the latter groups based their conclusions on the analysis of bacterial isolates, i.e., viable cells, Fitz-Gibbon et al. targeted 16S rDNA. The clinical significance of organisms identified solely by 16S rDNA gene sequencing is, however, unclear. In addition to viable cells, dormant and dead bacteria, as well as extracellular DNA, will be targeted. Furthermore, the 16S rRNA gene is highly conserved, and although it may be an obvious choice to identify organisms to the spp. level, it lacks sufficient resolution for useful typing of a clonal organism such as P. acnes. Previously described typing schemes for P. acnes based on MLST used multiple protein-encoding genes, which are more polymorphic; 16S rDNA genes are rarely used in MLST schemes. Unfortunately, Fitz-Gibbon et al. do not describe how their ribotypes relate to the previously described phylogenetic groupings (IA1, IA2, IB, IC, II, and III) or sequence types of P. acnes identified by MLST, some of which appear acne-associated (IA1 and IC) (Kilian et al., 2012; McDowell et al., 2012). In conclusion, evaluation of the pathogenic role of P. acnes in acne vulgaris requires careful consideration of biases associated with skin sampling. Continued failure to do so may lead to the erroneous assumption that only a follicular P. acnes population is targeted, whereas in reality, epidermal and, arguably, a minority of the follicular population are in fact being analyzed. The shortcomings associated with 16S rRNA–based typing for the identification of P. acnes lineages should also be clearly highlighted. CONFLICT OF INTEREST The authors state no conflict of interest.
Oleg A. Alexeyev1 and Christos C. Zouboulis2 1
Department of Medical Biosciences/Pathology, Umea˚ University, Umea˚, Sweden and Departments of Dermatology, Venereology, Allergology and Immunology, Dessau Medical Centre, Dessau, Germany E-mail: [email protected]
2
www.jidonline.org 2293
EA Eady and AM Layton A Distinct Acne Microbiome: Fact or Fiction?
REFERENCES Alexeyev OA, Jahns AC (2012) Sampling and detection of skin Propionibacterium acnes: current status. Anaerobe 18:479–83 Alexeyev OA, Lundskog B, Ganceviciene R et al. (2012) Pattern of tissue invasion by Propionibacterium acnes in acne vulgaris. J Dermatol Sci 67:63–6 Breternitz M, Flach M, Prassler J et al. (2007) Acute barrier disruption by adhesive tapes is influenced by pressure, time and anatomical location: integrity and cohesion assessed by sequential tape stripping. A randomized, controlled study. Br J Dermatol 156:231–40 Fitz-Gibbon S, Tomida S, Chiu BH et al. (2013) Propionibacterium acnes strain populations in the human skin microbiome associated with acne. J Invest Dermatol 133: 2152–60
Fredericks DN, Relman DA (1996) Sequencebased identification of microbial pathogens: a reconsideration of Koch’s postulates. Clin Microbiol Rev 9:18–33
Lomholt HB, Kilian M (2010) Population genetic analysis of Propionibacterium acnes identifies a subpopulation and epidemic clones associated with acne. PloS One 5:e12277
Grice EA, Segre JA (2011) The skin microbiome. Nat Rev Microbiol 9:244–53
McDowell A, Barnard E, Nagy I et al. (2012) An expanded multilocus sequence typing scheme for Propionibacterium acnes: investigation of ’pathogenic’, ’commensal’ and antibiotic resistant strains. PloS One 7:e41480
Hall-Stoodley L, Costerton JW, Stoodley P (2004) Bacterial biofilms: from the natural environment to infectious diseases. Nat Rev Microbiol 2:95–108 Jahns AC, Lundskog B, Ganceviciene R et al. (2012) An increased incidence of Propionibacterium acnes biofilms in acne vulgaris: a case-control study. Br J Dermatol 167: 50–8 Kilian M, Scholz CF, Lomholt HB (2012) Multilocus sequence typing and phylogenetic analysis of Propionibacterium acnes. J Clin Microbiol 50:1158–65
Mohammed D, Yang Q, Guy RH et al. (2012) Comparison of gravimetric and spectroscopic approaches to quantify stratum corneum removed by tape-stripping. Eur J Pharm Biopharm 82:171–4 Pinkus H (1951) Examination of the epidermis by the strip method of removing horny layers. I. Observations on thickness of the horny layer, and on mitotic activity after stripping. J Invest Dermatol 16:383–6
A Distinct Acne Microbiome: Fact or Fiction? Journal of Investigative Dermatology (2013) 133, 2294–2295; doi:10.1038/jid.2013.259; published online 11 July 2013
TO THE EDITOR We are pleased to see that Fitz-Gibbon et al. (2013) set out to conduct a prospective comparison of the microbiome of acne-prone and healthy skin using elegant molecular methods. They report that the skin of acne patients is enriched by particular strains, more specifically 16S ribotypes, with a very low prevalence in healthy subjects. The authors tell us that their study ‘demonstrates a previously unreported paradigm of commensal strain populations that could explain the pathogenesis of human diseases’, a powerful claim that requires robust evidence to support it. The role of Propionibacterium acnes in the disease after which it was named has never been conclusively proven. It is not hard to see why, given that it is a ubiquitous resident of human skin from adrenarche to old age (Leyden et al., 1975) and persists long after the acneprone years. It is the dominant bacterial resident of healthy pilosebaceous follicles (Puhvel et al., 1975). Recently, it has become possible to differentiate between different strains of P. acnes using techniques such as pulsed-field gel electrophoresis and multilocus sequence typing (Oprica et al., 2004; Lomholt and Kilian,
2010; Kilian et al., 2012; McDowell et al., 2012). Although no typing study has been specifically designed to look for differences between strains from acneprone and healthy skin, some data have been interpreted to suggest that differences may exist (Lomholt and Kilian 2010; McDowell et al., 2012). Fitz-Gibbon et al. (2013) sought to specifically address this important question. Detailed appraisal raises doubts about the authors’ methods and the conclusions drawn from them. The sample population is stated to comprise 101 subjects, 49 acne patients and 52 healthy controls, all from South California. The reader is referred to the dbGaP website (http://www.ncbi.nlm.nih.gov/ projects/gap/cgi-bin/study.cgi?study_id= phs000263.v1.p1) for further details, which show that the study sample initially comprised 148 subjects: 72 cases and 76 controls. No attempt has been made to match the age ranges of the patients and controls, and thus the oldest patient was 38 years old and the oldest control subject was 80 years old; eight control subjects were aged 42 years and above. Seven control subjects had been treated for acne in the past; it is unclear whether these had been excluded from the cohort
Abbreviation: RT, ribotype Accepted article preview online 12 June 2013; published online 11 July 2013
2294 Journal of Investigative Dermatology (2013), Volume 133
in the published article. Antibiotic therapy is known to perturb the microflora, and hence knowledge of previous treatment is pertinent; in 18 patients, treatment histories, even for the last 12 months, were not available. Among the cases, all had facial acne, but 48 (67%) did not have acne on the nose. In acne, the nose is often spared, as is the case here. Sampling from this largely unaffected site is therefore not advisable or likely to yield the evidence required to answer the question posed. Scant details are given of how the samples were collected and processed, but enough to show that they were not ideal for comparing acne versus control subjects. Pore strips were used to obtain pooled follicular casts from the nose. In both acne and non-acne, predominantly normal follicles will have been sampled by this procedure, as they always vastly outnumber lesional ones. It is doubtful whether this method will adequately sample inflamed lesions even in the minority of patients in whom they may have been present. The text refers to the pore tape samples as microcomedones and goes on to say that individual microcomedones were isolated using forceps. Acne is characterized by comedones of several types, of which whiteheads (closed comedones), blackheads (open comedones), and macrocomedones are the most common types (Cunliffe et al., 2000). Microcomedones
The effect of TLR-2 blockade not only on secretion of IL-12 p40 and IL-6 but also on IFN-g might again reflect the IL-12 dependence of a-NAC-induced IFN-g secretion. Additionally, in a-NAC-stimulated monocytes, p44/42 (Erk) and NF-kB were phosphorylated, both of which are downstream of TLR-2. A similar phosphorylation pattern and TLR-2 dependence has been shown for bacterial HSP-60 and human monocytes (Zhao et al., 2007). More evidence comes from the receptor itself: surface levels of TLR-2 were higher in a-NACtreated cells than in controls, suggesting that a-NAC upregulates its own receptor as it has been shown for other TLR-2 ligands (Chamorro et al., 2009; Niebuhr et al., 2009). These data suggest that TLR-2 is a monocyte receptor for a-NAC. Nevertheless, TLR-2 may act in concert with other receptors, as blockade of TLR-2 did not completely block the effects of a-NAC. This is the first report that describes a nonenzymatic autoantigen capable of mediating specific IgE and T cell responses relevant in AD, which also possesses intrinsic immunomodulatory potential. Our results clearly demonstrate that a-NAC is a potent inducer of a Th1 response in human PBMCs via the induction of IL-12, and that the monocyte is the primary target cell for a-NAC. This monocyte activation seems to depend on the interaction of a-NAC and TLR-2. Moreover, our data suggest that a-NAC-activated monocytes could be particularly efficient in priming
and maintaining an exaggerated Th1 response.
Heratizadeh A, Mittermann I, Balaji H et al. (2011) The role of T-cell reactivity towards the autoantigen alpha-NAC in atopic dermatitis. Br J Dermatol 164:316–24
CONFLICT OF INTEREST
Mittermann I, Reininger R, Zimmermann M et al. (2008) The IgE-reactive autoantigen Hom S 2 induces damage of respiratory epithelial cells and keratinocytes via induction of IFNgamma. J Invest Dermatol 128:1451–9
The authors state no conflict of interest.
ACKNOWLEDGMENTS We thank Gabriele Begemann, Petra Kienlin, and Ute Staar for their excellent technical assistance. This study was supported by a grant from Deutsche Forschungsgemeinschaft (DFG) to GRK 1441/1 and DFG-KliFo 250 WE1289/8-1 and by the Austrian Science Fund (FWF). 1,3
1,3
Susanne Hradetzky , Hari Balaji , Lennart M. Roesner1, Annice Heratizadeh1, Irene Mittermann2, Rudolf Valenta2 and Thomas Werfel1 1
Division of Immunodermatology and Allergy Research, Department of Dermatology and Allergy, Hannover Medical School, Hannover, Germany and 2Center for Pathophysiology, Infectiology and Immunology, Division of Immunopathology, Department of Pathophysiology and Allergy Research, Medical University of Vienna, Vienna, Austria E-mail: [email protected] 3 These authors contributed equally to this work. SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at http://www.nature.com/jid
REFERENCES Bieber T (2008) Atopic dermatitis. N Engl J Med 358:1483–94 Chamorro S, Garcia-Vallejo JJ, Unger WW et al. (2009) Tlr triggering on tolerogenic dendritic cells results in TLR2 up-regulation and a reduced proinflammatory immune program. J Immunol 183:2984–94
Mossabeb R, Seiberler S, Mittermann I et al. (2002) Characterization of a novel isoform of alphanascent polypeptide-associated complex as IgE-defined autoantigen. J Invest Dermatol 119:820–9 Natter S, Seiberler S, Hufnagl P et al. (1998) Isolation of cDNA clones coding for IgE autoantigens with serum IgE from atopic dermatitis patients. FASEB J 12:1559–69 Niebuhr M, Lutat C, Sigel S et al. (2009) Impaired TLR-2 expression and TLR-2-mediated cytokine secretion in macrophages from patients with atopic dermatitis. Allergy 64:1580–7 Schmid-Grendelmeier P, Fluckiger S, Disch R et al. (2005) IgE-mediated and T cell-mediated autoimmunity against manganese superoxide dismutase in atopic dermatitis. J Allergy Clin Immunol 115:1068–75 Sloane JA, Blitz D, Margolin Z et al. (2010) A clear and present danger: endogenous ligands of Toll-like receptors. Neuromolecular Med 12:149–63 Werfel T (2009) The role of leukocytes, keratinocytes, and allergen-specific IgE in the development of atopic dermatitis. J Invest Dermatol 129:1878–91 Zeller S, Rhyner C, Meyer N et al. (2009) Exploring the repertoire of IgE-binding self-antigens associated with atopic eczema. J Allergy Clin Immunol 124:278–85, 285. e1–7 Zhao Y, Yokota K, Ayada K et al. (2007) Helicobacter Pylori heat-shock protein 60 induces interleukin-8 via a toll-like receptor (TLR)2 and mitogen-activated protein (Map) kinase pathway in human monocytes. J Med Microbiol 56(Part 2):154–64
Shooting at Skin Propionibacterium acnes: To Be or Not to Be on Target Journal of Investigative Dermatology (2013) 133, 2292–2294; doi:10.1038/jid.2013.116; published online 4 April 2013
TO THE EDITOR We have read with interest the article by Fitz-Gibbon et al. (2013), which reports that different Propionibacterium acnes strain populations exist in acne in comparison with normal skin. Previous
studies have, indeed, suggested that globally disseminated clonal lineages within the P. acnes type IA1 division may have an etiological role in acne, whereas other lineages may be associated with health (Lomholt and Kilian,
Accepted article preview online 8 March 2013; published online 4 April 2013
2292 Journal of Investigative Dermatology (2013), Volume 133
2010; McDowell et al., 2012). If confirmed, this observation may lead to a better understanding of acne pathogenesis and, potentially, new therapeutic strategies. There are, however, major limitations in their study design related to: (i) skin sampling methodology, (ii) choice of anatomical site for
OA Alexeyev and CC Zouboulis Skin Propionibacterium acnes
sampling, and (iii) adoption of 16S rDNA–based ribotyping for the identification of P. acnes lineages NEGLECTED BIAS IN P. ACNES SAMPLING METHODOLOGY Although genomic approaches for the analysis of microbial communities provide much greater resolution than traditional bacterial culture (Grice and Segre, 2011), sequence-based technologies still require molecular guidelines for establishing tissue-sequence correlates (Fredericks and Relman, 1996); specific microbial sequences must be demonstrated in areas of tissue pathology. In acne vulgaris, tissue pathology is confined to the pilocebaceous unit, and it is imperative that anatomical considerations are taken into account when interpreting microbiological data. It is regrettable that Fitz-Gibbon et al. failed to consider recent studies reporting direct visualization of anatomically distinct P. acnes skin populations (Alexeyev et al., 2012; Jahns et al., 2012), and the biases associated with different sampling techniques (Alexeyev and Jahns, 2012). Four distinct anatomical populations of P. acnes have been reported: superficial, intrastratum corneum, infundibular, and follicular. Fitz-Gibbon et al. purported to present the microbiota of pilosebaceous units after collecting skin ‘microcomedone’ samples using the Biore Deep Cleansing Pore strips (essentially representing a modification of the well-known tape-stripping technique). As no further details on this crucial step are provided (except that the manufacturer’s instructions were followed), the reader may assume that no skin disinfectant was used before applying the strips. This would lead to sampling of superficial P. acnes skin populations, i.e., populations residing on the stratum corneum. The tape-stripping technique represents a classic approach for removing complete cell layers from the superficial part of the human stratum corneum (Pinkus, 1951). Approximately 50 mg of tissue is removed from 1.0 cm2 after one tape-stripping procedure (Mohammed et al., 2012). This indicates that intrastratum corneum P. acnes populations are also targeted by this sampling methodology. Furthermore, the amount of stratum corneum retrieved is influenced by the force used to remove
the stripping tape and the level of pressure applied during strip application, thus making standardization of the procedure very difficult (Breternitz et al., 2007). Direct visualization of hair follicles from acne patients and controls has shown that P. acnes colonization levels range from single cells to large macrocolonies (Alexeyev et al., 2012; Jahns et al., 2012). These populations are often located in the anaerobic part of the hair follicle, which is inaccessible by the tape-stripping technique. Should P. acnes, as suggested, exist in hair follicles in the form of biofilms (Jahns et al., 2012), which are often irreversibly attached to the surface (Hall-Stoodley et al., 2004), then the utility of the tapestripping technique for collecting bacteria from a hair follicle is further diminished. To provide a meaningful quantitative analysis of bacterial sequence distribution in the human tissue, the same quantity of biological material from acne and control subjects must be tested. Unfortunately, quantitative estimates of the microcomedone numbers analyzed were not provided. Finally, detaching microcomedones from silica as part of the Biore Deep Cleansing Pore strip approach is volatile yielding highly variable amounts of material. THE NOSE IS A NONREPRESENTATIVE ANATOMICAL SITE FOR SAMPLING ACNE SKIN Another deep concern is that subjects were sampled from the nose, which is not commonly affected by acne. On this basis, we believe the nose is a very poor site for sampling. Indeed, a significant proportion of the acne patients sampled did not appear to have acne affecting the nose, suggesting that in these cases normal follicles would have been analyzed. Furthermore, even in cases where acne did affect the nose, the use of a pore strip will result in multiple follicles being sampled, including those that are normal, which we believe will further compromise the results. SINGLE LOCUS TYPING BASED ON 16S rDNA LACKS RESOLUTION The data reported by Fitz-Gibbon et al. is fundamentally different from that provided by the Aarhus and Belfast groups,
which were on the basis of population genetic analysis of isolates by Multilocus Sequence Typing (MLST) (Kilian et al., 2012; McDowell et al., 2012). Although the latter groups based their conclusions on the analysis of bacterial isolates, i.e., viable cells, Fitz-Gibbon et al. targeted 16S rDNA. The clinical significance of organisms identified solely by 16S rDNA gene sequencing is, however, unclear. In addition to viable cells, dormant and dead bacteria, as well as extracellular DNA, will be targeted. Furthermore, the 16S rRNA gene is highly conserved, and although it may be an obvious choice to identify organisms to the spp. level, it lacks sufficient resolution for useful typing of a clonal organism such as P. acnes. Previously described typing schemes for P. acnes based on MLST used multiple protein-encoding genes, which are more polymorphic; 16S rDNA genes are rarely used in MLST schemes. Unfortunately, Fitz-Gibbon et al. do not describe how their ribotypes relate to the previously described phylogenetic groupings (IA1, IA2, IB, IC, II, and III) or sequence types of P. acnes identified by MLST, some of which appear acne-associated (IA1 and IC) (Kilian et al., 2012; McDowell et al., 2012). In conclusion, evaluation of the pathogenic role of P. acnes in acne vulgaris requires careful consideration of biases associated with skin sampling. Continued failure to do so may lead to the erroneous assumption that only a follicular P. acnes population is targeted, whereas in reality, epidermal and, arguably, a minority of the follicular population are in fact being analyzed. The shortcomings associated with 16S rRNA–based typing for the identification of P. acnes lineages should also be clearly highlighted. CONFLICT OF INTEREST The authors state no conflict of interest.
Oleg A. Alexeyev1 and Christos C. Zouboulis2 1
Department of Medical Biosciences/Pathology, Umea˚ University, Umea˚, Sweden and Departments of Dermatology, Venereology, Allergology and Immunology, Dessau Medical Centre, Dessau, Germany E-mail: [email protected]
2
www.jidonline.org 2293
EA Eady and AM Layton A Distinct Acne Microbiome: Fact or Fiction?
REFERENCES Alexeyev OA, Jahns AC (2012) Sampling and detection of skin Propionibacterium acnes: current status. Anaerobe 18:479–83 Alexeyev OA, Lundskog B, Ganceviciene R et al. (2012) Pattern of tissue invasion by Propionibacterium acnes in acne vulgaris. J Dermatol Sci 67:63–6 Breternitz M, Flach M, Prassler J et al. (2007) Acute barrier disruption by adhesive tapes is influenced by pressure, time and anatomical location: integrity and cohesion assessed by sequential tape stripping. A randomized, controlled study. Br J Dermatol 156:231–40 Fitz-Gibbon S, Tomida S, Chiu BH et al. (2013) Propionibacterium acnes strain populations in the human skin microbiome associated with acne. J Invest Dermatol 133: 2152–60
Fredericks DN, Relman DA (1996) Sequencebased identification of microbial pathogens: a reconsideration of Koch’s postulates. Clin Microbiol Rev 9:18–33
Lomholt HB, Kilian M (2010) Population genetic analysis of Propionibacterium acnes identifies a subpopulation and epidemic clones associated with acne. PloS One 5:e12277
Grice EA, Segre JA (2011) The skin microbiome. Nat Rev Microbiol 9:244–53
McDowell A, Barnard E, Nagy I et al. (2012) An expanded multilocus sequence typing scheme for Propionibacterium acnes: investigation of ’pathogenic’, ’commensal’ and antibiotic resistant strains. PloS One 7:e41480
Hall-Stoodley L, Costerton JW, Stoodley P (2004) Bacterial biofilms: from the natural environment to infectious diseases. Nat Rev Microbiol 2:95–108 Jahns AC, Lundskog B, Ganceviciene R et al. (2012) An increased incidence of Propionibacterium acnes biofilms in acne vulgaris: a case-control study. Br J Dermatol 167: 50–8 Kilian M, Scholz CF, Lomholt HB (2012) Multilocus sequence typing and phylogenetic analysis of Propionibacterium acnes. J Clin Microbiol 50:1158–65
Mohammed D, Yang Q, Guy RH et al. (2012) Comparison of gravimetric and spectroscopic approaches to quantify stratum corneum removed by tape-stripping. Eur J Pharm Biopharm 82:171–4 Pinkus H (1951) Examination of the epidermis by the strip method of removing horny layers. I. Observations on thickness of the horny layer, and on mitotic activity after stripping. J Invest Dermatol 16:383–6
A Distinct Acne Microbiome: Fact or Fiction? Journal of Investigative Dermatology (2013) 133, 2294–2295; doi:10.1038/jid.2013.259; published online 11 July 2013
TO THE EDITOR We are pleased to see that Fitz-Gibbon et al. (2013) set out to conduct a prospective comparison of the microbiome of acne-prone and healthy skin using elegant molecular methods. They report that the skin of acne patients is enriched by particular strains, more specifically 16S ribotypes, with a very low prevalence in healthy subjects. The authors tell us that their study ‘demonstrates a previously unreported paradigm of commensal strain populations that could explain the pathogenesis of human diseases’, a powerful claim that requires robust evidence to support it. The role of Propionibacterium acnes in the disease after which it was named has never been conclusively proven. It is not hard to see why, given that it is a ubiquitous resident of human skin from adrenarche to old age (Leyden et al., 1975) and persists long after the acneprone years. It is the dominant bacterial resident of healthy pilosebaceous follicles (Puhvel et al., 1975). Recently, it has become possible to differentiate between different strains of P. acnes using techniques such as pulsed-field gel electrophoresis and multilocus sequence typing (Oprica et al., 2004; Lomholt and Kilian,
2010; Kilian et al., 2012; McDowell et al., 2012). Although no typing study has been specifically designed to look for differences between strains from acneprone and healthy skin, some data have been interpreted to suggest that differences may exist (Lomholt and Kilian 2010; McDowell et al., 2012). Fitz-Gibbon et al. (2013) sought to specifically address this important question. Detailed appraisal raises doubts about the authors’ methods and the conclusions drawn from them. The sample population is stated to comprise 101 subjects, 49 acne patients and 52 healthy controls, all from South California. The reader is referred to the dbGaP website (http://www.ncbi.nlm.nih.gov/ projects/gap/cgi-bin/study.cgi?study_id= phs000263.v1.p1) for further details, which show that the study sample initially comprised 148 subjects: 72 cases and 76 controls. No attempt has been made to match the age ranges of the patients and controls, and thus the oldest patient was 38 years old and the oldest control subject was 80 years old; eight control subjects were aged 42 years and above. Seven control subjects had been treated for acne in the past; it is unclear whether these had been excluded from the cohort
Abbreviation: RT, ribotype Accepted article preview online 12 June 2013; published online 11 July 2013
2294 Journal of Investigative Dermatology (2013), Volume 133
in the published article. Antibiotic therapy is known to perturb the microflora, and hence knowledge of previous treatment is pertinent; in 18 patients, treatment histories, even for the last 12 months, were not available. Among the cases, all had facial acne, but 48 (67%) did not have acne on the nose. In acne, the nose is often spared, as is the case here. Sampling from this largely unaffected site is therefore not advisable or likely to yield the evidence required to answer the question posed. Scant details are given of how the samples were collected and processed, but enough to show that they were not ideal for comparing acne versus control subjects. Pore strips were used to obtain pooled follicular casts from the nose. In both acne and non-acne, predominantly normal follicles will have been sampled by this procedure, as they always vastly outnumber lesional ones. It is doubtful whether this method will adequately sample inflamed lesions even in the minority of patients in whom they may have been present. The text refers to the pore tape samples as microcomedones and goes on to say that individual microcomedones were isolated using forceps. Acne is characterized by comedones of several types, of which whiteheads (closed comedones), blackheads (open comedones), and macrocomedones are the most common types (Cunliffe et al., 2000). Microcomedones