- Email: [email protected]
Progressive and destructive hair follicle infections in a murine cutaneous anthrax model
ARTICLE IN PRESS
MICROBIAL PATHOGENESIS Microbial Pathogenesis 44 (2008) 363–369 www.elsevier.com/locate/micpath
Progressive and destructive hair follicle infections in a murine cutaneous anthrax model Christopher J. Wattsa,b, Beth L. Hahna,b, Peter G. Sohnlea,b, a
Division of Infectious Diseases, Department of Medicine, Medical College of Wisconsin, Milwaukee, WI 53226, USA b Consultant Care Division and Research Service, VA Medical Center, Milwaukee, WI 53295, USA Received 26 July 2007; received in revised form 18 October 2007; accepted 23 October 2007 Available online 1 November 2007
Abstract Hair follicles may allow pathogen entry because they represent potential barrier defects and because there is immunological privilege within actively growing follicles. Experimental cutaneous Bacillus anthracis infections in mice have previously shown prominent organism invasion and proliferation within hair follicles. For the present study, C57BL/6 mice were inoculated with B. anthracis (Sterne) spores onto abraded skin with either anagen (actively growing) or telogen (inactive) hair follicles; skin samples were evaluated by histologic methods and electron microscopy. The infections were found to progress similarly in either anagen or telogen hair follicles, with bacilli occasionally invading deeper sites in anagen hair follicles. The infections progressed from the surface inward, rather than growing outward from within the follicles. Infecting bacilli destroyed the hair follicle keratinocytes and were initially not contacted by inflammatory cells within the follicles. However, at 3–4 days after inoculation, inflammatory cells did contact and disperse the massed follicle bacilli and led to apparent resolution of the follicle infections. Therefore, in this model system B. anthracis initially attacks superficial sites in active or inactive hair follicles and then progresses inward, producing destructive infections of the hair follicles; these infections clear when the massed bacilli are eventually contacted and dispersed by inflammatory cells. r 2007 Elsevier Ltd. All rights reserved. Keywords: Cutaneous anthrax; Bacillus anthracis; Hair follicles; Keratinocytes; Anagen; Telogen
1. Introduction Anthrax in humans usually develops through contact with infected animals or products made from them, and in developed countries the disease is approximately 95% cutaneous and 5% respiratory or gastrointestinal [1,2]. The United States mail system cases in the fall of 2001 were approximately half inhalational and half cutaneous in form [3]. Despite the high percentage of cutaneous infections in humans, very little is known regarding the pathogenesis and early pathology of these lesions, except that cutaneous anthrax is thought to begin with inoculation of spores into a cut or abrasion in the skin [1,2].
Corresponding author. Research Service/151, VA Medical Center, Milwaukee, WI 53295, USA. Tel.: +1 414 384 2000x42878; fax: +1 414 383 8010. E-mail address: [email protected] (P.G. Sohnle).
0882-4010/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.micpath.2007.10.011
We have used experimental epicutaneous inoculation of Bacillus anthracis (Stern) onto intact or abraded skin of mice to investigate the interactions between the organisms and the epidermis, and these studies have demonstrated that the keratinocytes of hair follicles in inoculated abraded skin were a major site of proliferation for B. antrhacis bacilli [4]. Hair follicles represent a potential break in the stratum corneum barrier, which is an important mechanism of defense for the skin [5–7]. Furthermore, hair follicles in their actively growing state appear to be a site of immune privilege [8,9] that may also contribute to the susceptibility of these structures to cutaneous infections. Hair follicles are dynamic, with their states of growth termed anagen (actively growing), catagen (degenerative), and telogen (resting); anagen follicles extend more deeply into the skin, but when growth stops the deeper parts of the follicles degenerate to leave behind the shallower telogen follicles [5]. In mice, as opposed to humans, anagen and telogen follicles are grouped together
ARTICLE IN PRESS 364
C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
in separate skin areas [10]. It has been found in studies of experimental dermatophytosis in mice that only the anagen skin areas are susceptible to infection [11]. Also, work with experimental cutaneous candidiasis in mice has shown that some deep foci of infection could occur in the deeper parts of the anagen hair follicles [10]. The following studies were undertaken to investigate the initiation and progression of the hair follicle infections in experimental cutaneous anthrax in mice, particularly to answer the following questions: (a) are anagen hair follicles more susceptible to B. anthracis than are telogen ones? (b) Do hair follicle infections begin deep in the follicle and grow outward, or do they start at the skin surface and grow inward? (c) What happens to the deep hair follicle foci when the infections begin to clear? 2. Results 2.1. Susceptibility of anagen and telogen skin areas to infection Foci of B. anthracis bacilli were found below the infundibular outlets of hair follicles in inoculated abraded skin of telogen as well as anagen areas (Fig. 1a and b). Quantitative assessment of bacilli invading into interfollicular skin as assessed by either the number of fields infected
or the depth of bacilli invasion was not found to be different between anagen and telogen skin (Table 1). Infections of the hair follicles themselves were also not different for anagen versus telogen, with the numbers of Table 1 Effect of hair follicle (HF) growth stagea on infection after epicutaneous inoculation of B. anthracis spores onto abraded skin of C57BL/6 mice Anagen HFs
Telogen HFs
(A) Interfollicular skin Fields infected (%) Bacilli depth (mm)
58.0736.3 20.478.1
53.8721.3 18.372.5
(B) HF infections Infected (%) # Deep foci/sectionb Foci depth (mm)
58.7734.1 12.5713.5 199.2759.2
60.2728.4 9.3711.1 173.4757.0
The studies were done on tissues taken at 24 h after inoculation of spores onto abraded skin, and represent results (expressed as mean7S.D.) from 10 mice tested in five experiments for anagen and 8 mice tested in four experiments for telogen. Note that the cutaneous infections of either interfollicular skin or HFs developed approximately equally after inoculations onto skin with either anagen or telogen HFs. a Hair follicle growth stage ¼ inoculations made onto skin with hair follicles in anagen (actively growing) or telogen (inactive) stages. b Deep foci ¼ bacilli observed in hair follicles at 4100 mm below the skin surface.
Fig. 1. B. anthracis bacilli infecting hair follicles in C57BL/6 mice 24 h after epicutaneous inoculation onto abraded skin, showing: (a) infection in a telogen hair follicle; (b) infection in an anagen hair follicle; (c) infection at a very deep site (below 400 mm) in an anagen hair follicle (tissue gram stain with photomicrographs taken at an original magnification of 400 ; bar ¼ 50 mm).
ARTICLE IN PRESS C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
Fig. 2. B. anthracis bacilli infecting a hair follicle in a C57BL/6 mouse at 24 h after epicutaneous inoculation onto abraded skin, showing replacement of the keratinocytes lining the follicle with the organisms (tissue gram stain with photomicrograph taken at an original magnification of 1000 ; bar ¼ 20 mm).
follicles infected, the number of deep (below 100 mm) foci in hair follicles, or the mean foci depth of the hair follicle infections not being different between the two skin areas (Table 1). We also examined the lower portions of the anagen follicles for bacilli and found only five infections occurring 4400 mm below the skin surface (from 10 animals studied) as compared to 125 found from 100 to 400 mm. These foci were generally small and confined within the follicular structure (Fig. 1c). For comparison, measurements of anagen and telogen hair follicle depth from the (uninfected) skin surface were done in five animals each, and found to be 876.47153.6 mm in anagen versus 230.2755.2 in telogen (Po.01 by the Mann–Whitney U-test). In most cases the hair follicle infections located below the infundibular outlets consisted of massed bacilli that appeared by light microscopy to be proliferating within and replacing the keratinocytes lining the follicles (Fig. 2). In most cases very few inflammatory cells were found to be associated with the organisms inside hair follicles at early time points (6–24 h). Electron microscopy revealed that within areas of organism invasion into the keratinocytes of hair follicles, there was clearing of the usual cellular structure at this location (Fig. 3). 2.2. Progression of the cutaneous and hair follicle infections Inoculated skin was examined over the first 48 h after inoculation to determine if the infections began with spore germination within the follicles and progressed outward, or if they began at the surface and progressed inward. In these experimental infections, progression inward was found to be greater than that occurring laterally. Examination of interfollicular skin over 6–48 h revealed significant correlation between bacilli depth in micrometers during this
365
Fig. 3. B. anthracis bacilli in an infected hair follicle of a C57Bl/6 mouse at 24 h after epicutaneous inoculation onto abraded skin, showing invasion and clearing (area around point of the arrow) of the keratinocytes lining the hair follicle (electron micrograph taken at an original magnification of 4000 ; bar ¼ 2 mm).
Table 2 Progression of infection in interfollicular skin or hair follicles (HFs) after epicutaneous inoculation of B. anthracis spores onto abraded skin of C57BL/6 mice 6h
12 h
24 h
48 h
(A) Interfollicular skin Fields infected (%) 37.5715.1 51.7726.8 57.5734.3 63.3749.4 Bacilli depth (mm) 7.474.0 10.776.8 16.2717.2 68.5729.9 (B) HF infections Infected (%) 50.1711.2 53.8721.3 58.0734.0 63.0749.0 2.371.9 11.2718.9 33.7731.4 # Deep foci/sectiona 0.770.8 Depth of foci (mm) 8.370.6 10.272.0 16.275.1 23.378.2 The studies were done on tissues taken at 6, 12, 24, and 28 h after inoculation of spores onto abraded skin, and represent results (expressed as mean7S.D.) from six mice tested in three experiments for each point. Note that significant correlations (by the Pearson r correlation coefficients) versus time after inoculation were found for bacilli depth in interfollicular skin (r ¼ .80, Po.0001), number of deep HF foci (r ¼ .61, Po.0015), and depth of the deep foci (r ¼ .79, Po.0005); correlations for % of interfollicular skin fields infected or hair follicle outlets infected were not significant. a Deep foci ¼ bacilli observed in hair follicles at 4100 mm below the skin surface.
period (r ¼ .80, Po.001 by the Pearson r correlation coefficient), whereas the increases in percent of fields infected were not significant (Table 2). Similarly, increases in the number of deep (4100 mm) hair follicle foci and the mean depth of these foci both demonstrated significant correlations with time in these infections (r ¼ .61, Po.0015 and r ¼ .79, Po.0005, respectively). The sections from 12 and 24 h were also examined to find infected hair follicles that had been sectioned coronally, such that both the superficial and deeper portions of the follicles were visible;
ARTICLE IN PRESS 366
C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
in this evaluation it was found that all of the deep foci were also accompanied by bacilli in the upper parts of the follicle (Fig. 4). The bacilli did not appear first at deep sites within
the follicles. Therefore, it would appear that the organisms are fairly widely dispersed over the inoculated abraded skin at the earliest time point studied (6 h), and that they then invade inward into the deeper parts of the skin and hair follicles. It does not appear that the spores germinate within the hair follicles, with the resulting bacilli then beginning to grow outward from that site. 2.3. Clearance of the deep hair follicle infections
Fig. 4. Bar graph of hair follicle infections enumerated in coronally sectioned hair follicles at 12 and 24 h after inoculation of abraded skin with B. anthracis in C57BL/6 mice. Bars represent total numbers of infections seen in hair follicle outlets only, in deep sites only (4100 mm below the skin surface), or in both sites simultaneously (data from all coronally sectioned hair follicles seen on slides from six mice each at 12 and 24 h after inoculation). Note that no deep infections occurred unless the follicle infundibular outlet was also infected.
The fate of the deep hair follicle foci over time was investigated because the skin of most inoculated C57BL/6 mice will show these foci, but yet these animals will frequently clear their infections after a few days. Slides from 3 or 4 day infections were therefore examined for deep hair follicle foci to determine what was happening to them during this time period. Foci of infection during this time period generally seemed to persist in three general stages: (a) collections of massed bacilli confined to the follicles and appearing similar to those seen at earlier times; (b) foci in which inflammatory cell infiltrates consisting primarily of neutrophils had now invaded the follicle; (c) dermal collections of inflammatory cells with minimal numbers of dispersed organisms, which we postulate to represent the end stage of the hair follicle foci during the clearance process. Whereas at earlier times (such as 24 h) neutrophils were generally not found in large numbers in the hair follicle foci, later on these cells did enter and were able to contact the organisms within the follicles (Fig. 5a).
Fig. 5. Late B. anthracis hair follicle infections in C57BL/6 mice (4 days after epicutaneous inoculation onto abraded skin), showing: (a) neutrophils having invaded into an infected hair follicle and now contacting the organisms there; (b) neutrophils having dispersed masses of bacilli in a deep hair follicle infection; (c) infected hair follicle with apparent disruption of the follicular structure (tissue gram stain with photomicrograph taken at original magnification of 400 ; bar ¼ 50 mm).
ARTICLE IN PRESS C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
Dispersal of the massed bacilli by inflammatory cells could also be observed (Fig. 5b). In some cases the follicular structure itself was also seen to have become disrupted during the resolution process (Fig. 5c). 3. Discussion In this model system, it appears that B. anthracis can readily invade into skin areas containing either actively growing (anagen) or inactive (telogen) hair follicles; the invasion potential of this organism is therefore different from that of the dermatophytes, which can only invade skin areas with anagen follicles [11]. In the present system, the bacilli appear to attack the more superficial parts of the follicles, and then to grow down into the deeper parts of these structures. They produce infections that damage and replace the keratinocytes lining the follicles. These infections eventually clear when inflammatory cells invade into the follicles and contact the organisms, sometimes with disruption of the follicular structure. This latter process is similar to the fate of some hair follicle infections in dermatophytosis, in which deep nodular lesions known as Majocchi granulomas sometimes occur [12,13]. There are a number of reasons to explain why hair follicles may be more susceptible to infection with various cutaneous pathogens than is the rest of the skin. As noted above, these structures represent defects in the barrier function of the stratum corneum. Anagen hair follicles are also deeper structures (by approximately 3.8-fold as compared to telogen follicles, by our measurements in C57Bl/6 mice), perhaps allowing more access by invading organisms into deeper parts of the skin. As mentioned above, hair follicles are also a site of immune privilege [8,9,14]; this phenomenon has been related to the production of potent immunosuppressive factors such as alphaMSH and TGF-beta1 that are postulated to protect anagen-related autoantigens from recognition by autoreactive CD8+ T-cells [9]. Such reduced immunity may explain why experimental dermatophyte infections must be applied to anagen skin areas of mice in order to take [11], although we have found previously in studies of experimental cutaneous candidiasis that the infections occurred in telogen as well as anagen skin areas of mice [10]. Infections of both kinds of hair follicle sites developed in the present study as well; we did find a few foci of very deep hair follicle infections in anagen follicles, as was also noted in the previous experimental cutaneous candidiasis studies. One other factor that may relate to infection susceptibility of skin is that actively growing or proliferating keratinocytes (at either the skin surface or within hair follicles) appear to produce or contain smaller quantities of antimicrobial substances such as b-defensins and cathelicidins than do more differentiated cells [15,16], and therefore may be more susceptible to infections with various pathogens. However, the generally similar findings after inoculation onto anagen versus telogen in this model system would seem to minimize the significance of this factor.
367
It should be noted that neutrophils were generally not found in any quantity in the B. anthracis hair follicle foci examined at early points (6–24 h) in these experimental infections of C57Bl/6 mice. In contrast, previous studies of experimental cutaneous candidiasis have shown that neutrophils frequently accumulate and contact the organisms within 12–24 h after inoculation in either the epidermis or hair follicles of this mouse strain [10,17]. The finding of minimal early neutrophil infiltration in hair follicle infections by B. anthracis as compared to C. albicans could be due to the ability of anthrax lethal toxin to interfere with actin mediated neutrophil chemotaxis in the former infections [18]. Inability of neutrophils to accumulate along with the bacilli in the hair follicles may have allowed for early organism proliferation at this site. But protection of the organisms by the hair follicle structure appears to break down as the infections progress, with the organisms eventually being contacted, dispersed and (presumably) killed by inflammatory cells as the infections resolve. The site at which germination occurs in this model system appears to be exterior to inflammatory cells in the epicutaneous fluids that are induced by the skin abrasion procedure [19]. This extracellular germination process is therefore different from that in inhalational anthrax wherein germination of inhaled spores within alveolar macrophages has been clearly described [20,21]. On the other hand, hair follicles themselves could potentially represent an alternative germination site for spores that have found their way into them, either with or without the presence of epicutaneous exudate fluids from the abrasion process. However, the present study shows clearly that generation of bacilli from spores does not occur deep within the follicles in the absence of concurrent infundibular outlet infection; therefore, it seems that the infections begin at the surface and work their way down into the follicles, rather than the reverse. The infections produced are destructive, with obvious damage to the keratinocytes that line the interior of the follicles. B. anthracis lethal toxin has been shown to induce apoptosis in macrophages [22] and endothelial cells [23], and it is possible that a similar process could be causing cell death of keratinocytes. B. anthracis makes a number of proteases [24,25], and these could be involved in digestion of the keratinocyte structure. The hair follicle infections seem to resolve when the protection afforded the bacilli by the hair follicle breaks down and the organisms are contacted by inflammatory cells, primarily neutrophils. It has previously been shown that neutrophils from humans and mice can kill B. anthracis spores and/or bacilli in vitro [26,27]; encapsulated B. anthracis bacilli should be less susceptible to this process than the unencapsulated strain used in the present experiments, although in one study encapsulated bacilli of the Vollum strain were phagocytosed and killed fairly efficiently by human neutrophils through a mechanism dependent upon a-defensins [27].
ARTICLE IN PRESS 368
C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
4. Experimental/materials and methods 4.1. Organism We used B. anthracis Sterne for these experiments. This strain is toxigenic and unencapsulated; it is generally noninfectious in humans, but retains significant pathogenicity in mice, especially for certain susceptible strains [4,28,29]. The organism was obtained from the Colorado Serum Company (Denver, CO) and cultured on brain heart infusion agar plates. Spores were obtained by maintaining the plates at room temperature for 4–7 days after confluent growth at 37 1C. When the cultures were found to consist of 490% spores by microscopic examination, the organisms were removed from the plates, washed with distilled water, and treated by heating to 60 1C for 30 min to kill remaining vegetative forms. The spores were then layered onto 58% renographin (Bracco Diagnostics, Princeton, NJ), centrifuged at 3000 g to remove remaining vegetative forms, washed 3 times in saline, and then quantitated by both microscopic counts and colony counts to assure a viability of 490%. Spores were stored in saline with 10% glycerin at 20 1C. 4.2. Animals We used C57BL/6 mice for these experiments. This mouse strain is known to have an intermediate susceptibility to anthrax [29], and we have confirmed this characteristic in the epicutaneous inoculation system [28]. These mice were obtained from Charles Rivers Laboratories (Wilmington, MA), were either male or female, and were used at 8–14 weeks of age. The mice were housed in a separate BLS 2 section of the Veterinary Medical Unit at the Milwaukee VA Medical Center. The experimental procedures were approved by the appropriate committees at the Milwaukee Veterans Affairs Medical Center and the Medical College of Wisconsin.
of the animal, with a similar quantity of saline alone added to the disc on the opposite side. Both sites were covered with a 1.0 cm2 piece of plastic sheet (Handi-Wrap, Dow Chemical Co., Indianapolis, IN), which was then taped with Transpore tape and overwrapped with Nexcare waterproof tape (both from 3M, St. Paul, MN). 4.4. Microscopic studies After 6–24 h the occlusive dressings were removed and the sites washed 3 times with saline-soaked gauze pads. At various times from 6 to 96 h after inoculation the animals were killed and skin removed from the inoculated sites; paraffin sections were prepared and stained with tissue gram stains. The slides were analyzed in a blinded fashion by light microscopy at 400 using a 20 20 square ocular micrometer to enumerate vegetative bacilli and their depth of penetration into the interfollicular skin in each of 20 random fields. The outlets of each hair follicle infundibulum seen on the entire section was examined for the presence of bacilli, with the data expressed as percent of hair follicle outlets infected. Hair follicles were also examined for deep infections, which were taken as those with bacilli extending to 4100 mm below the skin surface. Anagen follicles were also examined for bacilli in their deeper parts (i.e., those that have degenerated during catagen and are not present in telogen follicles), defined as those extending 4400 mm below the skin surface. In some cases, hair follicles that had been sectioned longitudinally, so that the entire follicular structure was visible, were also examined to determine the numbers with bacilli at the outlet only, in areas deeper than 100 mm only, or both. The follicles were also examined to determine if inflammatory cells were present within the folliclular structure itself, and at later times (3 and 4 days after inoculation), to determine what happened to the bacilli within the follicles. 4.5. Electron microscopy
4.3. Epicutaneous innoculations Anagen and telogen skin areas in the C57BL/6 mice were identified by the black color of the former after shaving, as previously described [10]. The animals were examined on the day after shaving and only animals with clear skin were used. The sites to be inoculated were first washed with 10% povidine iodine solution and then with alcohol. The epidermis was prepared for epicutaneous inoculation under anesthesia by scraping with a surgical scalpel blade until a non-bloody glistening skin layer resulted, representing damage to the stratum corneum water barrier. For the study of anagen versus telogen skin susceptibility, animals with these areas predominantly present on the flanks were chosen and inoculated with B. anthracis spores. An inoculum of 107 spores was applied in .025 ml of saline to 4 mm filter paper discs (GB002, Schleicher & Schuell, Keene, NH) placed onto an abraded area on the left flank
In some cases inoculated skin specimens were fixed in 2% glutaraldehyde, post-fixed in 1% osmium, dehydrated in a series of graded alcohols, and then imbedded into epon. Thin sections were prepared for electron microscopy and stained with uranyl acetate and lead citrate. The thin sections were examined with an Hitachi H-600 transmission electron microscope at 75 kV. 4.6. Statistics Data were expressed as mean7S.D. The Pearson r correlation coefficient was used to determine correlations of the findings versus time. Comparisons of mean values were made using the nonparametric Kruskal–Wallis and Dunn’s multiple comparison tests, or Mann–Whitney U-test for single comparisons. Statistical significance was taken as Po.05.
ARTICLE IN PRESS C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
Acknowledgments This work was supported by the United States Department of Veterans Affairs. The authors wish to thank Clive Wells of the Facility for Electron Microscopy at the Medical College of Wisconsin for his excellent electron microscopy work. References [1] Dixon TC, Meselson M, Guillemin J, Hanna PC. Anthrax. N Engl J Med 1999;341:815–6. [2] Swartz MN. Recognition and management of anthrax—an update. N Engl J Med 2001;345:1621–6. [3] CDC. Update-investigation of bioterrorism-related anthrax—Connecticut, 2001. MMWR 2001;50:1077–9. [4] Hahn BL, Sharma S, Sohnle PG. Analysis of epidermal entry in experimental cutaneous Bacillus anthracis infections in mice. J Lab Clin Med 2005;146:95–102. [5] Lavker RM, Bertolino AP, Sun TT. Biology of hair follicles. In: Freedberg IM, Eisen AZ, Wolff K, Austen KF, Goldsmith LA, Katz SI, editors. Fitzpatrick’s dermatology in general medicine. 6th ed. New York: McGraw-Hill; 2003. p. 148–59. [6] Tsuruta D, Green KJ, Getsios S, Jones JCR. The barrier function of skinhow to keep a tight lid on water loss. Trends Cell Biol 2002;12:355–7. [7] Wagner DK, Sohnle PG. Cutaneous defense mechanisms against dermatophytes and yeasts. Clin Microbiol Rev 1995;8:317–35. [8] Niederkorn JY. Mechanisms of immune privilege in the eye and hair follicle. J Invest Dermatol Symp Proc 2003;8:168–72. [9] Paus R, Ito N, Takigawa M, Ito T. The hair follicle and immune privilege. J Invest Dermatol Symp Proc 2003;8:188–94. [10] Sohnle PG, Hahn B, Collins-Lech C. The effect of hair growth cycles on experimental cutaneous candidiasis. J Invest Dermatol 1986;86:556–9. [11] Kligman AM. Pathophysiology of ringworm infections in animals with skin cycles. J Invest Dermatol 1956;27:171–85. [12] Schwartz RA, Ceburkovas O, Janniger CK. Majocchi granuloma. /http://www.emedicine.com/derm/topic849.htmS; 2007. [13] Wilson JW, Plunkett OA, Gregersen A. Nodular granulomatous perifolliculitis of the legs caused by Trichophyton rubrum. AMA Arch Derm Syphilol 1954;63:258–77. [14] Christoph T, Muller-Rover S, Audring H, Tobin DJ, Hermes B, Cotsarelis G, et al. The human hair follicle immune system-cellular composition and immune privilege. Br J Dermatol 2000;142:862–73.
369
[15] Liu L, Wang L, Jia HP, Zhao C, Heng HHQ, Schutte BC, et al. Structure and mapping of the human b-defensin HBD-2 gene and its expression at sites of inflammation. Gene 1998;222:237–44. [16] Frohm M, Agerberth B, Ahangari G, Stahle-Backdahl M, Liden S, Wigzell H, et al. The expression of the gene coding for the antibacterial peptide LL-37 is induced in human keratinocytes during inflammatory disorders. J Biol Chem 1997;272:15258–63. [17] Sohnle PG, Hahn BL. Epidermal proliferation and the neutrophilic infiltrates of experimental cutaneous candidiasis in mice. Arch Derm Res 1989;281:279–83. [18] During RL, Li W, Hao B, Koenig JM, Stephens DS, Quinn CP, et al. Anthrax lethal toxin paralyzes neutrophil actin-based motility. J Infect Dis 2005;192:837–45. [19] Bischof TS, Hahn BL, Sohnle PG. Characteristics of spore germination in a mouse model of cutaneous anthrax. J Infect Dis 2007;195:888–94. [20] Guidi-Rontani C, Weber-Levy M, Labruyere E, Mock M. Germination of Bacillus anthracis spores within alveolar macrophages. Mol Microbiol 1999;31:9–17. [21] Hanna PC, Ireland JAW. Understanding Bacillus anthracis pathogenesis. Trends Microbiol 1999;7:180–2. [22] Park JM, Greten FR, Li W, Karin M. Macrophage apoptosis by anthrax lethal factor through p38 MAP kinase inhibition. Science 2002;297:2048–51. [23] Kirby JE. Anthrax lethal toxin induces human endothelial cell apoptosis. Infect Immun 2004;72:430–9. [24] Aronson AI, Bell C, Fulroth B. Plasmid-encoded regulator of extracellular proteases in Bacillus anthracis. J Biol Chem 2005;187: 3133–8. [25] Chung MC, Popova TG, Millis BA, Mukherjee DV, Zhou W, Liotta LA, et al. Secreted neutral metalloproteases of Bacillus anthracis as candidate pathogenic factors. J Biol Chem 2006;281: 31408–18. [26] Welkos SL, Trotter RW, Becker DM, Nelson GO. Resistance to the Sterne strain of B. anthracis—phagocytic cell responses of resistant and susceptible mice. Microb Pathog 1989;7:15–35. [27] Mayer-Scholl A, Hurwitz R, Brinkmann V, Schmid M, Jungblut P, Weinrauch Y, et al. Human neutrophils kill Bacillus anthracis. PLoS Pathog 2005;1:179–86. [28] Bischof TS, Hahn BL, Sohnle PG. Experimental cutaneous Bacillus anthracis infections in hairless HRS/J mice. Int J Exp Pathol 2007;88:75–84. [29] Welkos SL, Friedlander AM. Pathogenesis and genetic control of resistance to the Sterne strain of Bacillus anthracis. Microb Pathog 1988;4:53–69.
MICROBIAL PATHOGENESIS Microbial Pathogenesis 44 (2008) 363–369 www.elsevier.com/locate/micpath
Progressive and destructive hair follicle infections in a murine cutaneous anthrax model Christopher J. Wattsa,b, Beth L. Hahna,b, Peter G. Sohnlea,b, a
Division of Infectious Diseases, Department of Medicine, Medical College of Wisconsin, Milwaukee, WI 53226, USA b Consultant Care Division and Research Service, VA Medical Center, Milwaukee, WI 53295, USA Received 26 July 2007; received in revised form 18 October 2007; accepted 23 October 2007 Available online 1 November 2007
Abstract Hair follicles may allow pathogen entry because they represent potential barrier defects and because there is immunological privilege within actively growing follicles. Experimental cutaneous Bacillus anthracis infections in mice have previously shown prominent organism invasion and proliferation within hair follicles. For the present study, C57BL/6 mice were inoculated with B. anthracis (Sterne) spores onto abraded skin with either anagen (actively growing) or telogen (inactive) hair follicles; skin samples were evaluated by histologic methods and electron microscopy. The infections were found to progress similarly in either anagen or telogen hair follicles, with bacilli occasionally invading deeper sites in anagen hair follicles. The infections progressed from the surface inward, rather than growing outward from within the follicles. Infecting bacilli destroyed the hair follicle keratinocytes and were initially not contacted by inflammatory cells within the follicles. However, at 3–4 days after inoculation, inflammatory cells did contact and disperse the massed follicle bacilli and led to apparent resolution of the follicle infections. Therefore, in this model system B. anthracis initially attacks superficial sites in active or inactive hair follicles and then progresses inward, producing destructive infections of the hair follicles; these infections clear when the massed bacilli are eventually contacted and dispersed by inflammatory cells. r 2007 Elsevier Ltd. All rights reserved. Keywords: Cutaneous anthrax; Bacillus anthracis; Hair follicles; Keratinocytes; Anagen; Telogen
1. Introduction Anthrax in humans usually develops through contact with infected animals or products made from them, and in developed countries the disease is approximately 95% cutaneous and 5% respiratory or gastrointestinal [1,2]. The United States mail system cases in the fall of 2001 were approximately half inhalational and half cutaneous in form [3]. Despite the high percentage of cutaneous infections in humans, very little is known regarding the pathogenesis and early pathology of these lesions, except that cutaneous anthrax is thought to begin with inoculation of spores into a cut or abrasion in the skin [1,2].
Corresponding author. Research Service/151, VA Medical Center, Milwaukee, WI 53295, USA. Tel.: +1 414 384 2000x42878; fax: +1 414 383 8010. E-mail address: [email protected] (P.G. Sohnle).
0882-4010/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.micpath.2007.10.011
We have used experimental epicutaneous inoculation of Bacillus anthracis (Stern) onto intact or abraded skin of mice to investigate the interactions between the organisms and the epidermis, and these studies have demonstrated that the keratinocytes of hair follicles in inoculated abraded skin were a major site of proliferation for B. antrhacis bacilli [4]. Hair follicles represent a potential break in the stratum corneum barrier, which is an important mechanism of defense for the skin [5–7]. Furthermore, hair follicles in their actively growing state appear to be a site of immune privilege [8,9] that may also contribute to the susceptibility of these structures to cutaneous infections. Hair follicles are dynamic, with their states of growth termed anagen (actively growing), catagen (degenerative), and telogen (resting); anagen follicles extend more deeply into the skin, but when growth stops the deeper parts of the follicles degenerate to leave behind the shallower telogen follicles [5]. In mice, as opposed to humans, anagen and telogen follicles are grouped together
ARTICLE IN PRESS 364
C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
in separate skin areas [10]. It has been found in studies of experimental dermatophytosis in mice that only the anagen skin areas are susceptible to infection [11]. Also, work with experimental cutaneous candidiasis in mice has shown that some deep foci of infection could occur in the deeper parts of the anagen hair follicles [10]. The following studies were undertaken to investigate the initiation and progression of the hair follicle infections in experimental cutaneous anthrax in mice, particularly to answer the following questions: (a) are anagen hair follicles more susceptible to B. anthracis than are telogen ones? (b) Do hair follicle infections begin deep in the follicle and grow outward, or do they start at the skin surface and grow inward? (c) What happens to the deep hair follicle foci when the infections begin to clear? 2. Results 2.1. Susceptibility of anagen and telogen skin areas to infection Foci of B. anthracis bacilli were found below the infundibular outlets of hair follicles in inoculated abraded skin of telogen as well as anagen areas (Fig. 1a and b). Quantitative assessment of bacilli invading into interfollicular skin as assessed by either the number of fields infected
or the depth of bacilli invasion was not found to be different between anagen and telogen skin (Table 1). Infections of the hair follicles themselves were also not different for anagen versus telogen, with the numbers of Table 1 Effect of hair follicle (HF) growth stagea on infection after epicutaneous inoculation of B. anthracis spores onto abraded skin of C57BL/6 mice Anagen HFs
Telogen HFs
(A) Interfollicular skin Fields infected (%) Bacilli depth (mm)
58.0736.3 20.478.1
53.8721.3 18.372.5
(B) HF infections Infected (%) # Deep foci/sectionb Foci depth (mm)
58.7734.1 12.5713.5 199.2759.2
60.2728.4 9.3711.1 173.4757.0
The studies were done on tissues taken at 24 h after inoculation of spores onto abraded skin, and represent results (expressed as mean7S.D.) from 10 mice tested in five experiments for anagen and 8 mice tested in four experiments for telogen. Note that the cutaneous infections of either interfollicular skin or HFs developed approximately equally after inoculations onto skin with either anagen or telogen HFs. a Hair follicle growth stage ¼ inoculations made onto skin with hair follicles in anagen (actively growing) or telogen (inactive) stages. b Deep foci ¼ bacilli observed in hair follicles at 4100 mm below the skin surface.
Fig. 1. B. anthracis bacilli infecting hair follicles in C57BL/6 mice 24 h after epicutaneous inoculation onto abraded skin, showing: (a) infection in a telogen hair follicle; (b) infection in an anagen hair follicle; (c) infection at a very deep site (below 400 mm) in an anagen hair follicle (tissue gram stain with photomicrographs taken at an original magnification of 400 ; bar ¼ 50 mm).
ARTICLE IN PRESS C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
Fig. 2. B. anthracis bacilli infecting a hair follicle in a C57BL/6 mouse at 24 h after epicutaneous inoculation onto abraded skin, showing replacement of the keratinocytes lining the follicle with the organisms (tissue gram stain with photomicrograph taken at an original magnification of 1000 ; bar ¼ 20 mm).
follicles infected, the number of deep (below 100 mm) foci in hair follicles, or the mean foci depth of the hair follicle infections not being different between the two skin areas (Table 1). We also examined the lower portions of the anagen follicles for bacilli and found only five infections occurring 4400 mm below the skin surface (from 10 animals studied) as compared to 125 found from 100 to 400 mm. These foci were generally small and confined within the follicular structure (Fig. 1c). For comparison, measurements of anagen and telogen hair follicle depth from the (uninfected) skin surface were done in five animals each, and found to be 876.47153.6 mm in anagen versus 230.2755.2 in telogen (Po.01 by the Mann–Whitney U-test). In most cases the hair follicle infections located below the infundibular outlets consisted of massed bacilli that appeared by light microscopy to be proliferating within and replacing the keratinocytes lining the follicles (Fig. 2). In most cases very few inflammatory cells were found to be associated with the organisms inside hair follicles at early time points (6–24 h). Electron microscopy revealed that within areas of organism invasion into the keratinocytes of hair follicles, there was clearing of the usual cellular structure at this location (Fig. 3). 2.2. Progression of the cutaneous and hair follicle infections Inoculated skin was examined over the first 48 h after inoculation to determine if the infections began with spore germination within the follicles and progressed outward, or if they began at the surface and progressed inward. In these experimental infections, progression inward was found to be greater than that occurring laterally. Examination of interfollicular skin over 6–48 h revealed significant correlation between bacilli depth in micrometers during this
365
Fig. 3. B. anthracis bacilli in an infected hair follicle of a C57Bl/6 mouse at 24 h after epicutaneous inoculation onto abraded skin, showing invasion and clearing (area around point of the arrow) of the keratinocytes lining the hair follicle (electron micrograph taken at an original magnification of 4000 ; bar ¼ 2 mm).
Table 2 Progression of infection in interfollicular skin or hair follicles (HFs) after epicutaneous inoculation of B. anthracis spores onto abraded skin of C57BL/6 mice 6h
12 h
24 h
48 h
(A) Interfollicular skin Fields infected (%) 37.5715.1 51.7726.8 57.5734.3 63.3749.4 Bacilli depth (mm) 7.474.0 10.776.8 16.2717.2 68.5729.9 (B) HF infections Infected (%) 50.1711.2 53.8721.3 58.0734.0 63.0749.0 2.371.9 11.2718.9 33.7731.4 # Deep foci/sectiona 0.770.8 Depth of foci (mm) 8.370.6 10.272.0 16.275.1 23.378.2 The studies were done on tissues taken at 6, 12, 24, and 28 h after inoculation of spores onto abraded skin, and represent results (expressed as mean7S.D.) from six mice tested in three experiments for each point. Note that significant correlations (by the Pearson r correlation coefficients) versus time after inoculation were found for bacilli depth in interfollicular skin (r ¼ .80, Po.0001), number of deep HF foci (r ¼ .61, Po.0015), and depth of the deep foci (r ¼ .79, Po.0005); correlations for % of interfollicular skin fields infected or hair follicle outlets infected were not significant. a Deep foci ¼ bacilli observed in hair follicles at 4100 mm below the skin surface.
period (r ¼ .80, Po.001 by the Pearson r correlation coefficient), whereas the increases in percent of fields infected were not significant (Table 2). Similarly, increases in the number of deep (4100 mm) hair follicle foci and the mean depth of these foci both demonstrated significant correlations with time in these infections (r ¼ .61, Po.0015 and r ¼ .79, Po.0005, respectively). The sections from 12 and 24 h were also examined to find infected hair follicles that had been sectioned coronally, such that both the superficial and deeper portions of the follicles were visible;
ARTICLE IN PRESS 366
C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
in this evaluation it was found that all of the deep foci were also accompanied by bacilli in the upper parts of the follicle (Fig. 4). The bacilli did not appear first at deep sites within
the follicles. Therefore, it would appear that the organisms are fairly widely dispersed over the inoculated abraded skin at the earliest time point studied (6 h), and that they then invade inward into the deeper parts of the skin and hair follicles. It does not appear that the spores germinate within the hair follicles, with the resulting bacilli then beginning to grow outward from that site. 2.3. Clearance of the deep hair follicle infections
Fig. 4. Bar graph of hair follicle infections enumerated in coronally sectioned hair follicles at 12 and 24 h after inoculation of abraded skin with B. anthracis in C57BL/6 mice. Bars represent total numbers of infections seen in hair follicle outlets only, in deep sites only (4100 mm below the skin surface), or in both sites simultaneously (data from all coronally sectioned hair follicles seen on slides from six mice each at 12 and 24 h after inoculation). Note that no deep infections occurred unless the follicle infundibular outlet was also infected.
The fate of the deep hair follicle foci over time was investigated because the skin of most inoculated C57BL/6 mice will show these foci, but yet these animals will frequently clear their infections after a few days. Slides from 3 or 4 day infections were therefore examined for deep hair follicle foci to determine what was happening to them during this time period. Foci of infection during this time period generally seemed to persist in three general stages: (a) collections of massed bacilli confined to the follicles and appearing similar to those seen at earlier times; (b) foci in which inflammatory cell infiltrates consisting primarily of neutrophils had now invaded the follicle; (c) dermal collections of inflammatory cells with minimal numbers of dispersed organisms, which we postulate to represent the end stage of the hair follicle foci during the clearance process. Whereas at earlier times (such as 24 h) neutrophils were generally not found in large numbers in the hair follicle foci, later on these cells did enter and were able to contact the organisms within the follicles (Fig. 5a).
Fig. 5. Late B. anthracis hair follicle infections in C57BL/6 mice (4 days after epicutaneous inoculation onto abraded skin), showing: (a) neutrophils having invaded into an infected hair follicle and now contacting the organisms there; (b) neutrophils having dispersed masses of bacilli in a deep hair follicle infection; (c) infected hair follicle with apparent disruption of the follicular structure (tissue gram stain with photomicrograph taken at original magnification of 400 ; bar ¼ 50 mm).
ARTICLE IN PRESS C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
Dispersal of the massed bacilli by inflammatory cells could also be observed (Fig. 5b). In some cases the follicular structure itself was also seen to have become disrupted during the resolution process (Fig. 5c). 3. Discussion In this model system, it appears that B. anthracis can readily invade into skin areas containing either actively growing (anagen) or inactive (telogen) hair follicles; the invasion potential of this organism is therefore different from that of the dermatophytes, which can only invade skin areas with anagen follicles [11]. In the present system, the bacilli appear to attack the more superficial parts of the follicles, and then to grow down into the deeper parts of these structures. They produce infections that damage and replace the keratinocytes lining the follicles. These infections eventually clear when inflammatory cells invade into the follicles and contact the organisms, sometimes with disruption of the follicular structure. This latter process is similar to the fate of some hair follicle infections in dermatophytosis, in which deep nodular lesions known as Majocchi granulomas sometimes occur [12,13]. There are a number of reasons to explain why hair follicles may be more susceptible to infection with various cutaneous pathogens than is the rest of the skin. As noted above, these structures represent defects in the barrier function of the stratum corneum. Anagen hair follicles are also deeper structures (by approximately 3.8-fold as compared to telogen follicles, by our measurements in C57Bl/6 mice), perhaps allowing more access by invading organisms into deeper parts of the skin. As mentioned above, hair follicles are also a site of immune privilege [8,9,14]; this phenomenon has been related to the production of potent immunosuppressive factors such as alphaMSH and TGF-beta1 that are postulated to protect anagen-related autoantigens from recognition by autoreactive CD8+ T-cells [9]. Such reduced immunity may explain why experimental dermatophyte infections must be applied to anagen skin areas of mice in order to take [11], although we have found previously in studies of experimental cutaneous candidiasis that the infections occurred in telogen as well as anagen skin areas of mice [10]. Infections of both kinds of hair follicle sites developed in the present study as well; we did find a few foci of very deep hair follicle infections in anagen follicles, as was also noted in the previous experimental cutaneous candidiasis studies. One other factor that may relate to infection susceptibility of skin is that actively growing or proliferating keratinocytes (at either the skin surface or within hair follicles) appear to produce or contain smaller quantities of antimicrobial substances such as b-defensins and cathelicidins than do more differentiated cells [15,16], and therefore may be more susceptible to infections with various pathogens. However, the generally similar findings after inoculation onto anagen versus telogen in this model system would seem to minimize the significance of this factor.
367
It should be noted that neutrophils were generally not found in any quantity in the B. anthracis hair follicle foci examined at early points (6–24 h) in these experimental infections of C57Bl/6 mice. In contrast, previous studies of experimental cutaneous candidiasis have shown that neutrophils frequently accumulate and contact the organisms within 12–24 h after inoculation in either the epidermis or hair follicles of this mouse strain [10,17]. The finding of minimal early neutrophil infiltration in hair follicle infections by B. anthracis as compared to C. albicans could be due to the ability of anthrax lethal toxin to interfere with actin mediated neutrophil chemotaxis in the former infections [18]. Inability of neutrophils to accumulate along with the bacilli in the hair follicles may have allowed for early organism proliferation at this site. But protection of the organisms by the hair follicle structure appears to break down as the infections progress, with the organisms eventually being contacted, dispersed and (presumably) killed by inflammatory cells as the infections resolve. The site at which germination occurs in this model system appears to be exterior to inflammatory cells in the epicutaneous fluids that are induced by the skin abrasion procedure [19]. This extracellular germination process is therefore different from that in inhalational anthrax wherein germination of inhaled spores within alveolar macrophages has been clearly described [20,21]. On the other hand, hair follicles themselves could potentially represent an alternative germination site for spores that have found their way into them, either with or without the presence of epicutaneous exudate fluids from the abrasion process. However, the present study shows clearly that generation of bacilli from spores does not occur deep within the follicles in the absence of concurrent infundibular outlet infection; therefore, it seems that the infections begin at the surface and work their way down into the follicles, rather than the reverse. The infections produced are destructive, with obvious damage to the keratinocytes that line the interior of the follicles. B. anthracis lethal toxin has been shown to induce apoptosis in macrophages [22] and endothelial cells [23], and it is possible that a similar process could be causing cell death of keratinocytes. B. anthracis makes a number of proteases [24,25], and these could be involved in digestion of the keratinocyte structure. The hair follicle infections seem to resolve when the protection afforded the bacilli by the hair follicle breaks down and the organisms are contacted by inflammatory cells, primarily neutrophils. It has previously been shown that neutrophils from humans and mice can kill B. anthracis spores and/or bacilli in vitro [26,27]; encapsulated B. anthracis bacilli should be less susceptible to this process than the unencapsulated strain used in the present experiments, although in one study encapsulated bacilli of the Vollum strain were phagocytosed and killed fairly efficiently by human neutrophils through a mechanism dependent upon a-defensins [27].
ARTICLE IN PRESS 368
C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
4. Experimental/materials and methods 4.1. Organism We used B. anthracis Sterne for these experiments. This strain is toxigenic and unencapsulated; it is generally noninfectious in humans, but retains significant pathogenicity in mice, especially for certain susceptible strains [4,28,29]. The organism was obtained from the Colorado Serum Company (Denver, CO) and cultured on brain heart infusion agar plates. Spores were obtained by maintaining the plates at room temperature for 4–7 days after confluent growth at 37 1C. When the cultures were found to consist of 490% spores by microscopic examination, the organisms were removed from the plates, washed with distilled water, and treated by heating to 60 1C for 30 min to kill remaining vegetative forms. The spores were then layered onto 58% renographin (Bracco Diagnostics, Princeton, NJ), centrifuged at 3000 g to remove remaining vegetative forms, washed 3 times in saline, and then quantitated by both microscopic counts and colony counts to assure a viability of 490%. Spores were stored in saline with 10% glycerin at 20 1C. 4.2. Animals We used C57BL/6 mice for these experiments. This mouse strain is known to have an intermediate susceptibility to anthrax [29], and we have confirmed this characteristic in the epicutaneous inoculation system [28]. These mice were obtained from Charles Rivers Laboratories (Wilmington, MA), were either male or female, and were used at 8–14 weeks of age. The mice were housed in a separate BLS 2 section of the Veterinary Medical Unit at the Milwaukee VA Medical Center. The experimental procedures were approved by the appropriate committees at the Milwaukee Veterans Affairs Medical Center and the Medical College of Wisconsin.
of the animal, with a similar quantity of saline alone added to the disc on the opposite side. Both sites were covered with a 1.0 cm2 piece of plastic sheet (Handi-Wrap, Dow Chemical Co., Indianapolis, IN), which was then taped with Transpore tape and overwrapped with Nexcare waterproof tape (both from 3M, St. Paul, MN). 4.4. Microscopic studies After 6–24 h the occlusive dressings were removed and the sites washed 3 times with saline-soaked gauze pads. At various times from 6 to 96 h after inoculation the animals were killed and skin removed from the inoculated sites; paraffin sections were prepared and stained with tissue gram stains. The slides were analyzed in a blinded fashion by light microscopy at 400 using a 20 20 square ocular micrometer to enumerate vegetative bacilli and their depth of penetration into the interfollicular skin in each of 20 random fields. The outlets of each hair follicle infundibulum seen on the entire section was examined for the presence of bacilli, with the data expressed as percent of hair follicle outlets infected. Hair follicles were also examined for deep infections, which were taken as those with bacilli extending to 4100 mm below the skin surface. Anagen follicles were also examined for bacilli in their deeper parts (i.e., those that have degenerated during catagen and are not present in telogen follicles), defined as those extending 4400 mm below the skin surface. In some cases, hair follicles that had been sectioned longitudinally, so that the entire follicular structure was visible, were also examined to determine the numbers with bacilli at the outlet only, in areas deeper than 100 mm only, or both. The follicles were also examined to determine if inflammatory cells were present within the folliclular structure itself, and at later times (3 and 4 days after inoculation), to determine what happened to the bacilli within the follicles. 4.5. Electron microscopy
4.3. Epicutaneous innoculations Anagen and telogen skin areas in the C57BL/6 mice were identified by the black color of the former after shaving, as previously described [10]. The animals were examined on the day after shaving and only animals with clear skin were used. The sites to be inoculated were first washed with 10% povidine iodine solution and then with alcohol. The epidermis was prepared for epicutaneous inoculation under anesthesia by scraping with a surgical scalpel blade until a non-bloody glistening skin layer resulted, representing damage to the stratum corneum water barrier. For the study of anagen versus telogen skin susceptibility, animals with these areas predominantly present on the flanks were chosen and inoculated with B. anthracis spores. An inoculum of 107 spores was applied in .025 ml of saline to 4 mm filter paper discs (GB002, Schleicher & Schuell, Keene, NH) placed onto an abraded area on the left flank
In some cases inoculated skin specimens were fixed in 2% glutaraldehyde, post-fixed in 1% osmium, dehydrated in a series of graded alcohols, and then imbedded into epon. Thin sections were prepared for electron microscopy and stained with uranyl acetate and lead citrate. The thin sections were examined with an Hitachi H-600 transmission electron microscope at 75 kV. 4.6. Statistics Data were expressed as mean7S.D. The Pearson r correlation coefficient was used to determine correlations of the findings versus time. Comparisons of mean values were made using the nonparametric Kruskal–Wallis and Dunn’s multiple comparison tests, or Mann–Whitney U-test for single comparisons. Statistical significance was taken as Po.05.
ARTICLE IN PRESS C.J. Watts et al. / Microbial Pathogenesis 44 (2008) 363–369
Acknowledgments This work was supported by the United States Department of Veterans Affairs. The authors wish to thank Clive Wells of the Facility for Electron Microscopy at the Medical College of Wisconsin for his excellent electron microscopy work. References [1] Dixon TC, Meselson M, Guillemin J, Hanna PC. Anthrax. N Engl J Med 1999;341:815–6. [2] Swartz MN. Recognition and management of anthrax—an update. N Engl J Med 2001;345:1621–6. [3] CDC. Update-investigation of bioterrorism-related anthrax—Connecticut, 2001. MMWR 2001;50:1077–9. [4] Hahn BL, Sharma S, Sohnle PG. Analysis of epidermal entry in experimental cutaneous Bacillus anthracis infections in mice. J Lab Clin Med 2005;146:95–102. [5] Lavker RM, Bertolino AP, Sun TT. Biology of hair follicles. In: Freedberg IM, Eisen AZ, Wolff K, Austen KF, Goldsmith LA, Katz SI, editors. Fitzpatrick’s dermatology in general medicine. 6th ed. New York: McGraw-Hill; 2003. p. 148–59. [6] Tsuruta D, Green KJ, Getsios S, Jones JCR. The barrier function of skinhow to keep a tight lid on water loss. Trends Cell Biol 2002;12:355–7. [7] Wagner DK, Sohnle PG. Cutaneous defense mechanisms against dermatophytes and yeasts. Clin Microbiol Rev 1995;8:317–35. [8] Niederkorn JY. Mechanisms of immune privilege in the eye and hair follicle. J Invest Dermatol Symp Proc 2003;8:168–72. [9] Paus R, Ito N, Takigawa M, Ito T. The hair follicle and immune privilege. J Invest Dermatol Symp Proc 2003;8:188–94. [10] Sohnle PG, Hahn B, Collins-Lech C. The effect of hair growth cycles on experimental cutaneous candidiasis. J Invest Dermatol 1986;86:556–9. [11] Kligman AM. Pathophysiology of ringworm infections in animals with skin cycles. J Invest Dermatol 1956;27:171–85. [12] Schwartz RA, Ceburkovas O, Janniger CK. Majocchi granuloma. /http://www.emedicine.com/derm/topic849.htmS; 2007. [13] Wilson JW, Plunkett OA, Gregersen A. Nodular granulomatous perifolliculitis of the legs caused by Trichophyton rubrum. AMA Arch Derm Syphilol 1954;63:258–77. [14] Christoph T, Muller-Rover S, Audring H, Tobin DJ, Hermes B, Cotsarelis G, et al. The human hair follicle immune system-cellular composition and immune privilege. Br J Dermatol 2000;142:862–73.
369
[15] Liu L, Wang L, Jia HP, Zhao C, Heng HHQ, Schutte BC, et al. Structure and mapping of the human b-defensin HBD-2 gene and its expression at sites of inflammation. Gene 1998;222:237–44. [16] Frohm M, Agerberth B, Ahangari G, Stahle-Backdahl M, Liden S, Wigzell H, et al. The expression of the gene coding for the antibacterial peptide LL-37 is induced in human keratinocytes during inflammatory disorders. J Biol Chem 1997;272:15258–63. [17] Sohnle PG, Hahn BL. Epidermal proliferation and the neutrophilic infiltrates of experimental cutaneous candidiasis in mice. Arch Derm Res 1989;281:279–83. [18] During RL, Li W, Hao B, Koenig JM, Stephens DS, Quinn CP, et al. Anthrax lethal toxin paralyzes neutrophil actin-based motility. J Infect Dis 2005;192:837–45. [19] Bischof TS, Hahn BL, Sohnle PG. Characteristics of spore germination in a mouse model of cutaneous anthrax. J Infect Dis 2007;195:888–94. [20] Guidi-Rontani C, Weber-Levy M, Labruyere E, Mock M. Germination of Bacillus anthracis spores within alveolar macrophages. Mol Microbiol 1999;31:9–17. [21] Hanna PC, Ireland JAW. Understanding Bacillus anthracis pathogenesis. Trends Microbiol 1999;7:180–2. [22] Park JM, Greten FR, Li W, Karin M. Macrophage apoptosis by anthrax lethal factor through p38 MAP kinase inhibition. Science 2002;297:2048–51. [23] Kirby JE. Anthrax lethal toxin induces human endothelial cell apoptosis. Infect Immun 2004;72:430–9. [24] Aronson AI, Bell C, Fulroth B. Plasmid-encoded regulator of extracellular proteases in Bacillus anthracis. J Biol Chem 2005;187: 3133–8. [25] Chung MC, Popova TG, Millis BA, Mukherjee DV, Zhou W, Liotta LA, et al. Secreted neutral metalloproteases of Bacillus anthracis as candidate pathogenic factors. J Biol Chem 2006;281: 31408–18. [26] Welkos SL, Trotter RW, Becker DM, Nelson GO. Resistance to the Sterne strain of B. anthracis—phagocytic cell responses of resistant and susceptible mice. Microb Pathog 1989;7:15–35. [27] Mayer-Scholl A, Hurwitz R, Brinkmann V, Schmid M, Jungblut P, Weinrauch Y, et al. Human neutrophils kill Bacillus anthracis. PLoS Pathog 2005;1:179–86. [28] Bischof TS, Hahn BL, Sohnle PG. Experimental cutaneous Bacillus anthracis infections in hairless HRS/J mice. Int J Exp Pathol 2007;88:75–84. [29] Welkos SL, Friedlander AM. Pathogenesis and genetic control of resistance to the Sterne strain of Bacillus anthracis. Microb Pathog 1988;4:53–69.