CLINICAL AND LABORATORY INVESTIGATIONS (cid:9)
`
`DOI 10.1111h.1365-2133.2007.08028.x
`Mite-related bacterial antigens stimulate inflammatory
`cells in rosacea
`N. Lacey, S. Delaney, K. Kavanagh and F.C. Powell*
`
`Department of Biology, National University of Ireland, Maynooth, Co. Kildare, Ireland
`*Regional Centre of Dermatology, Mater Misericordiae University Hospital, Eccles Street, Dublin 7, Ireland
`Summary
`
`Correspondence
`Frank C. Powell.
`E-mail: fpowell@eircom.net
`
`Accepted for publication
`13 February 2007
`
`Key words
`antigenic proteins, Bacillus, Demodex, rosacea
`
`Conflicts of interest
`None declared.
`
`Background Patients with papulopustular rosacea have a higher density of Demodex
`folliculorum mites on their faces than normal subjects but the role, if any, of their
`mites in initiating inflammation is disputed. Selective antibiotics are effective in
`reducing the inflammatory changes of papulopustular rosacea, but their mode of
`action is unknown.
`Objectives To investigate whether a D. folliculorum-related bacterium was capable of
`expressing antigens that could stimulate an inflammatory immune response in
`patients with rosacea.
`Methods A bacterium (Bacillus oleronius) was isolated from a D. folliculorum mite extrac-
`ted from the face of a patient with papulopustular rosacea, and was investigated
`further.
`Results This bacterium produced antigens capable of stimulating peripheral blood
`mononuclear cells proliferation in 16 of 22 (73%) patients with rosacea but only
`five of 17 (29%) control subjects (P = 0.0105). This antigenic preparation was
`fractionated into 70 subfractions and the proteins in each fraction were visualized
`by sodium dodecyl sulphate—polyacrylamide gel electrophoresis. Western blot
`analysis revealed the presence of two antigenic proteins of size 62 and 83 kDa
`in fractions when probing with sera from patients with rosacea. No immuno-
`reactivity to these proteins was recorded when probing with sera from control
`patients. Two-dimensional electrophoretic separation was used to isolate these
`proteins and matrix-assisted laser desorption/ionization time-of-flight analysis
`was employed to identify the relevant peptides. The 62-kDa immunoreactive pro-
`tein shared amino add sequence homology with an enzyme involved in carbo-
`hydrate metabolism and signal transduction while the 83-kDa protein was similar
`to bacterial heat shock proteins.
`Conclusions Antigenic proteins related to a bacterium (B. oleronius), isolated from a
`D. folliculorum mite, have the potential to stimulate an inflammatory response in
`patients with papulopustular rosacea.
`
`Papulopustular rosacea is a chronic inflammatory dermatosis
`of the convexities of the central face characterized by the
`presence of multiple small dome-shaped erythematous
`papules and papulopustules arising on a background of
`fixed inflammatory erythema.' The diagnostic criteria, classi-
`fication and grading of rosacea have recently been out-
`lined,7 '3 but its aetiology and pathogenesis are poorly
`understood.4 The disease usually appears after the age of
`30 years, is commoner in individuals with fair skin, and a
`positive family history of rosacea is present in up to 30%
`
`of patients. s Histopathological study of biopsies taken of pap-
`ules and pustules from skin of patients with rosacea often
`shows a follicular-oriented inflammatory response, 6 and
`Demodex folliculomm mites are frequently seen in the follicles of
`biopsies from patients with rosacea. 7 Using the skin sur-
`face biopsy technique which extracts mites from follicular
`canals, several investigators have shown a significantly increased
`density of D. folliculorum mites in the facial skin of patients with
`rosacea when compared with control subjects," but the rele-
`vance of this finding to the pathogenesis of the condition is
`
`474 (cid:9)
`
`© 2007 The Authors
`Journal Compilation © 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp474-481
`
`1
`
`Galderma Laboratories, Inc. Ex 2010
`Dr. Reddy's Labs v. Galderma Labs., Inc.
`IPR2015-01777
`
`

`
`Bacterial antigens stimulate inflammatory cells in rosacea, N. Lacey et al. 475
`
`disputed. Papulopustular rosacea consistently responds to select-
`ed antibiotic therapy with clearing of inflammatory lesions, but
`the mechanism of action of these antibiotics is unknown, and
`the condition usually relapses within months of the antibiotic
`being discontinued.
`In an attempt to reconcile the apparently disparate findings
`of increased numbers of D. folliculorum mites, perifollicular
`inflammation, and response of papulopustular rosacea to
`selective antibiotic therapy, we decided to investigate whether
`a D. folliculorum-related bacterium could be identified which
`might express stimulatory antigens capable of initiating an
`inflammatory response in patients with papulopustular rosacea.
`A bacterium (Bacillus oleronius) cultured from a mite prepar-
`ation in this study had been previously isolated from the hind-
`gut of a termite." This microorganism was investigated
`further.
`
`U.S.A.) and washed with phosphate-buffered saline (PBS, pH
`7.2). Growth medium conditioned by the culture of B. oleronius
`('supernatant') was diluted 1/10 with PBS and filter-sterilized
`for use in peripheral blood mononuclear cells (PBMC) prolifer-
`ation assays. Bacillus oleronius cells were washed twice with
`PBS (pH 7.2) before resuspension in 2 mL of Break's
`buffer [10 mmol KC1, 3 mmol NaC1, 4 mmol
`MgC12 , 10 mmol piperazine-N,N'-bis(2-ethanesulphonic
`acid); pH 7.2] plus Triton-X 100 (10% w/v) to solubilize outer
`membrane proteins. Protease inhibitors (10 ttg leupeptin,
`pepstatin A, aprotinin and N-ot-p-tosyl-t-lysine chloromethyl-
`ketone hydrochloride) were added. Cell suspensions (5 mL)
`were sonicated for 10 s using a Branson Soniprobe type 7532B
`(Dawe Instruments, London, U.K.) to dislodge antigens associ-
`ated with the bacterial cell wall. Cells were filtered off and the
`'sonicate' preparation was stored at —20 °C.
`
`Materials and methods
`
`Extraction and dissection of mites
`
`Following informed consent, D. follicular= mites were extrac-
`ted from designated facial sites of 40 patients with papulopus-
`tular rosacea using the skin surface biopsy technique with
`cyanoacrylate glue and glass slides as previously described. 8
`The samples were studied microscopically at standard magnifi-
`cations (x 40, x 100, x 400). One live mite from each patient
`sample was removed from the microscope slide and microdis-
`sected on nutrient agar using sterile fine forceps and dissecting
`tweezers.
`
`Isolation of bacteria from facial skin swabs
`
`Each patient with papulopustular rosacea had skin surface
`swabs were taken from the same designated facial areas as
`those from which Donator mites were extracted. The swabs
`were streaked on to separate nutrient agar plates and incubated
`at 30 °C for 2 days to evaluate the microflora.
`
`Isolation of bacteria from Demodex folliculorum
`preparations
`
`The nutrient agar plates containing the microdissected mites
`were incubated at 30 °C for 2 days to achieve confluent
`bacterial growth. All cultures were sent to the National
`Collections of Industrial and Marine Bacteria (NCIMB),
`Aberdeen, U.K. for identification. The Kirby—Bauer test
`was used to test for antiobiotic sensitivity of cultured
`microorganisms.
`
`Preparation of antigen
`
`Bacillus oleronius was cultured in 250 mL nutrient broth (Oxoid,
`Basingstoke, U.K.) for 48 h at 30 °C and 180 r.p.m. Stationary
`phase cells were harvested by centrifugation (4000 g in a
`Beckman GS-6 centrifuge; Beckman Instruments, Palo Alto, CA,
`
`Peripheral blood mononuclear cells proliferation assay
`
`Peripheral blood (5 mL) was obtained from 22 patients with
`rosacea (nine women and 13 men, age range 32-64 years)
`and age- and sex-matched control volunteers. PBMC were pre-
`pared for the proliferation assay as described previously. 11.12
`A proliferation index > 2 was considered positive in accord-
`ance with the literature." - " All assays were performed on
`three independent occasions and the data were statistically
`analysed using the Sigma-Stat Statistics Package using Fisher's
`exact test. 14
`
`Antigen fractionation by Q -Sepharose charge separation
`
`The proteins in the 'Triton-X' preparation were purified and
`chromatographed by Q-Sepharose charge separation using an
`AKTA Purifier 100 system (Amersham Biosciences, Amersham,
`U.K.). The starting material was loaded on to Q-Sepharose
`(1.5 x 2 cm, 1 mL 1 mL fractions collected, eluted
`with a 30 mL linear gradient of 0.5 mol L -1 NaC1 in Break's
`buffer). Peak fractions containing the antigens of interest were
`identified by immunoreactivity against sera of patients with
`rosacea.
`
`Western blot analysis
`
`Sodium dodecyl sulphate—polyacrylamide gel electrophoresis
`(SDS-PAGE) was performed using a 12-5% (w/v) polyacryl-
`amide gel in a discontinuous buffer system and subject to
`electrophoresis at 200 V for 8 h. For Western blot analysis
`the primary antibody consisted of pooled sera from six
`patients with rosacea or controls, diluted 1/150 with anti-
`body-diluting buffer [3% (w/v) bovine serum albumin,
`10% (w/v) nonfat dried milk dissolved in TBS-T (78-8 g
`Tris-HC1, 87.65 g NaC1 and 500 IA Tween-20 in 1 L H20)]
`prior to use. The secondary antibody was antihuman IgG-
`horseradish peroxidase-linked whole antibody (Sigma Aldrich
`Chemical Co. Ltd, Poole, U.K.) which was diluted 1/1000
`with the antibody-diluting buffer prior to addition to the
`
`C) 2007 The Authors
`Journal Compilation C) 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp474-48 I
`
`2
`
`Galderma Laboratories, Inc. Ex 2010
`Dr. Reddy's Labs v. Galderma Labs., Inc.
`IPR2015-01777
`
`

`
`476 Bacterial antigens stimulate inflammatory cells in rosacea, N. Lacey et al.
`
`membrane. In all cases duplicate gels [SDS-PAGE and
`two-dimensional (21))] were prepared under identical
`conditions, one for immunoblotting and the other for
`Coomassie staining.
`
`Two-dimensional gel electrophoresis and immunoblot
`analysis of antigenic fractions to IgG
`
`Isoelectric focusing and 21) electrophoresis of fractionated
`antigenic protehi samples was performed as described previ-
`ously." Western blot analysis was carried out as above.
`
`Matrix-assisted laser desorption/ionization time-of-flight
`mass spectrometry
`
`To carry out peptide analysis, protein spots of interest were
`excised from 21) gel (Coomassie stained) and diced finely.
`Gel pieces were prepared for matrix-assisted laser desorp-
`tion/ionization time-of-flight (MALDI-ToF) analysis which
`was performed as described.' s Mass spectra were recorded
`using an Ettan MALDI-ToF spectrometer (Amersham Bio-
`sciences, Freiburg, Germany), analysed using MALDI evalu-
`ation software (Amersham Biosciences), and proteins were
`identified using the PMF Profound search engine for peptide
`mass fingerprints.
`
`Ethical approval
`
`Ethical approval for this study was granted by the Research
`Ethics Committee of the Mater Misericordiae University Flos-
`pital, Dublin, Ireland (ref. 1/378/638) and informed consent
`was required from each participant.
`
`Results
`
`Extraction and dissection of mites
`
`Mites were extracted from the following skin areas: the fore-
`head, cheek, chin or nose of 40 patients with papulopustular
`rosacea using the standardized skin surface biopsy. 8 After 48 h
`of incubation, bacterial colonies were observed growing
`around the site of the microdissected mites. Only bacterial col-
`onies with different phenotypes were extracted aseptically
`from each agar plate, streaked on to new plates and incubated
`again at 30 °C for 2 days.
`
`Isolation of bacteria from facial skin swabs
`
`After 48 h of incubation, bacterial colonies were observed
`growing where the facial swab had been streaked on the
`plates. The phenotypically different colonies were inocula-
`ted on to new agar plates and sent to the NCIMB for iden-
`tification. Bacteria isolated from facial swabs included
`Micrococcus agilis, M. luteus, Micromonospora halophytica, Staphylococcus
`simulans, S. capitis, S. epidermis, Haloanella gallinarum, Pantoea agglom-
`erans and Dietzia maris.
`
`Isolation of bacteria from Demodex folliculorum
`preparations
`
`One bacterial species, cultured as an orange colony from one
`microdissected Demodex mite from a patient with papulopustu-
`lar rosacea, appeared distinctive. This bacterium was identified
`to species level at the NCIMB by analysis of the 16s rDNA
`sequence, as an endospore-forming bacterium, B. oleronius,
`which had previously been found in association with the mid-
`gut of termites and identified as Gram negative. ° Antibiotic
`sensitivity testing showed that B. oleronius was sensitive to tetra-
`cycline, doxycycline and minocycline. Bacillus oleronius was thus
`selected for further investigation.
`
`Preparation of antigen
`
`Bacillus oleronius supernatant and sonicate antigen preparations
`were produced as described above and subsequently used in
`the experiments outlined below. Although this Bacillus could
`be cultured at both 37 °C and 30 °C, culturing of this bacter-
`ium at 30 °C rather than 37 °C induced the expression of
`increased levels of the stimulatory antigens that were the focus
`of this work.
`
`Peripheral blood mononuclear cells proliferation assay
`
`Bacillus oleronius 'sonicate' and 'supernatant' preparations were
`applied to PBMC from patients with papulopustular rosacea or
`controls. The results (Table 1) demonstrate that the B. oleronius
`'sonicate' preparation activated PBMC proliferation in 16 of
`22 (73%) patients with rosacea and five of 17 (29%) controls.
`Using Fisher's exact test," the results show statistical signifi-
`cance with P = 0.0105 (P < 0.05) (d.f. = 1). hi contrast, the
`B. oleronius 'supernatant' preparation activated PBMC in one of
`nine (11%) patients with papulopustular rosacea and six of 17
`(35%) controls — a result which was not statistically significant
`(P = 0.357) (Table I).
`
`Antigen fractionation by Q-Sepharose charge separation
`
`In order to identify the specific antigen(s) responsible for the
`differential reactivity observed above, fractions collected
`corresponding to the protein peaks on the chromatograph
`(Fig. 1 a) were analysed by SDS-PAGE, and Western blot analy-
`sis was performed. Fractions 12-56 were analysed by SDS-
`PAGE and fractions 21-37 showed the greatest number and
`intensity of bands (Fig. 1 b, c). Fractions 21-29 showed an
`increase in the number of protein bands ranging in size from
`16-5 to 83 kDa (Fig. 1 b). A decrease in the number of protein
`bands was observed in fractions 30-37 (Fig. I c), with a con-
`sistent decrease across fractions 38-54 (data not presented).
`
`Western blot analysis
`
`Immunoblot analysis of the above SDS-PAGE gels, using pooled
`sera from six patients with papulopustular rosacea as the
`
`(D 2007 The Authors
`Journal Compilation © 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp474-481
`
`3
`
`Galderma Laboratories, Inc. Ex 2010
`Dr. Reddy's Labs v. Galderma Labs., Inc.
`IPR2015-01777
`
`

`
`Bacterial antigens stimulate inflammatory cells in rosacea, N. Lacey et al. 477
`
`Table 1 Response of peripheral blood
`mononuclear cells from patients with rosacca
`and controls following challenge with Bacillus
`oleronius antigen preparations
`
`Antigen preparation
`
`Positivity in patients
`with rosacea, n (%)
`
`Positivity in
`controls, a (%)
`
`P-value
`
`B. oleronius sonicate
`B. oleronius supernatant
`
`16/22 (73%)
`1/9 (11%)
`
`5/17 (29%)
`6/17 (35%)
`
`0.0105 (P <
`0.3574 (P >
`
`0.05)
`0.05)
`
`Q Sepharose (cid:9)
`
`L1.0
`
`0-5
`
`500 -
`
`400 -
`
`300 -
`
`200 -
`
`100 -
`
`(a)
`
`Absorbance (280 nm)
`
`10 (cid:9)
`
`20
`
`30 (cid:9)
`
`40
`
`30 31 32 33 34 35 36 37
`
`II (cid:9)
`
`I
`
`21 22 23 24 25 26 27 28 29 (C) (cid:9)
`kDa (cid:9)
`
`_
`
`. (cid:9)
`•••• - •
`tr•-n (cid:9)
`aiim (cid:9)
`Ogg 011110 MO Mae ••••
`III NM (cid:9)
`ff,"ri ',Aril (cid:9)
`
`_•
`
`"••••! (cid:9)
`
`175*
`83*
`62*
`
`47.5*
`
`32.5*
`
`254-
`
`1.1" (cid:9) " (cid:9)
`
`."1" ""kg gt4i9linvo
`
`1
`
`,.•••
`
`7%05 ....2••••n•
`
`(a)
`
`62+
`
`at
`
`•avii
`
`(b)
`kDa
`
`175+
`83 4-
`62+
`
`47.5*
`
`32.54-
`
`25 4-
`
`16.5
`
`(d)
`
`83*
`
`Fig 1. (a) Chromatograph of Bacillus oleronius
`'sonicate preparation. Protein peaks correlate
`with proteins in fractions. (b) Sodium
`dodecyl sulphate—polyacrylamide gel
`electrophoresis (SDS-PAGE) analysis
`of fractions 21-29. (c) SDS-PAGE analysis of
`fractions 30-37. (d) Western blot analysis of
`fractions 21-29, probing with pooled
`rosacea patient sera. Reactivity was visualized
`in fraction 26 to the 83-kDa protein.
`(e) Western blot analysis of fractions 30-37,
`probing with pooled rosacea patient sera.
`Reactivity was visualized in fractions 30 and
`31 to the 62-kDa protein.
`
`primary antibody and antihuman IgG-horseradish peroxidase-
`linked whole antibody as the secondary antibody, confirmed
`that sera from patients with rosacea contained antibodies to pro-
`
`teins of size 83 kDa present in fraction 26 (Fig. Id) and 62 kDa
`present in fractions 30 and 31 (Fig. I e). No other immunoreac-
`tive antigens from B. oleronius were visualized in these blots.
`
`© 2007 The Authors
`Journal Compilation © 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp474-48I
`
`4
`
`Galderma Laboratories, Inc. Ex 2010
`Dr. Reddy's Labs v. Galderma Labs., Inc.
`IPR2015-01777
`
`(cid:9)
`

`
`478 Bacterial antigens stimulate inflammatory cells in rosacea, N. Lacey et al.
`
`Two-dimensional gel electrophoresis immunoblot
`analysis of antigenic fractions to IgG
`
`2D gel electrophoresis was performed on fractions 26, 30 and
`31 to isolate the proteins of interest prior to further character-
`ization by mass spectrometry. The samples were primarily
`separated according to their isoelectric point, across a pH
`gradient, followed by separation due to molecular size on an
`SDS-PAGE gel. 2D gels showed more than 30 reproducible
`protein spots from fraction 26 (Fig. 2a), 25 from fraction 30
`(Fig. 2b) and 18 from fraction 31 (Fig. 2c). The 83-kDa pro-
`tein from fraction 26 (Fig. 2a) was shown to consist of two
`proteins of different pI values (5.35 and 5.5). Following
`immunoblot analysis of this 21) gel (as previously described),
`sera from patients with rosacea showed reactivity only towards
`one of these proteins at pI 5.5 (Fig. 2a). The 62-kDa protein
`was shown to consist of one large protein (pi 5-0), to which
`the sera of patients with rosacea was strongly reactive
`(Fig. 2b, c).
`
`Matrix-assisted laser desorption/ionization time-of-flight
`mass spectrometry
`
`The spots visualized by Coomassie staining (Fig. 2a—c), corres-
`ponding to those that were immunoreactive with rosacea
`patient antibodies by Western blot analysis, were prepared for
`MALDI-ToF analysis. From the MALDI-ToF spectra of spot 1
`(83 kDa, pI 5.5), spot 2 (62 kDa, pi 5.0) and spot 3 (62 kDa,
`
`pi 5-0), peptide masses were used for searches in the National
`Center for Biotechnology Information (Bethesda, MD, U.S.A.)
`database. The parameters were as follows: enzyme, trypsin;
`maximum missed cleavage, 1; peptide mass tolerance, I Da;
`monoisotopic masses tolerance, 1 Da. In the tryptic digest of
`the 83-kDa protein spot that evoked reactivity from patient
`sera, protein coverage with other candidates on the database
`ranged from 8% to 18%. Results include a 9% protein cover-
`age with an 85-kDa heat shock-like protein (Z score = 1.54)
`and an 18% coverage with an NADP-dependent malic enzyme
`(Z score = 1.29) (Table 2). The 62-kDa spots (spots 2 and 3)
`also produced an antibody response, and revealed protein
`similarity with other candidates on the database ranging from
`9% to 21%. Candidate functions included phosphorylation,
`protease activity, enzyme activity and surface adhesion
`protein.
`
`Discussion
`
`Danodex mites are inhabitants of normal adult human skin
`whose life-cycle lasts about 14 days. 16 While they were first
`described as early as 1842 their role in the biology of the
`skin has received little attention from dermatologists. As the
`vast majority of individuals experience no adverse reaction
`from the presence of these mites, it is likely that they either
`avoid exposure to host immune defences or have the ability
`to downregulate host immunity so that they can survive in
`the cutaneous environment of their human host. Dernodex
`
`kDa
`(a)
`175 44
`
`83
`
`62 4.
`
`47.5 40
`
`321 41
`
`25
`
`•
`
`0
`
`p1
`
`10
`
`kDa
`(b)
`
`•
`
`175 41
`
`3
`
`pI
`
`10
`
`83 4.
`
`62 1
`
`47.5
`
`32.5 4'
`
`25
`
`•
`
`10
`
`kDa
`(c)
`
`83
`
`62
`
`47.5
`
`32.5
`
`25
`
`62
`
`4--
`
`0
`
`Fig 2. Two-dimensional gel and Western blot analysis of (a) antigenic fraction 26 (83-kDa protein), (b) antigenic fraction 30 (62-kDa protein)
`and (c) antigenic fraction 31 (62-kDa protein).
`
`© 2007 The Authors
`Journal Compilation © 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp474-481
`
`5
`
`Galderma Laboratories, Inc. Ex 2010
`Dr. Reddy's Labs v. Galderma Labs., Inc.
`IPR2015-01777
`
`

`
`Table 2 Comparison of 62-kDa and 83-kDa antigens with proteins in the National Center for Biotechnology Information (NCB!) database
`
`Bacterial antigens stimulate inflammatory cells in rosacea, N. Lacey et al. 479
`
`Function
`62-kDa antigen
`Rapid utilization of
`carbohydrates/signal transduction
`Catalyses the synthesis of
`major membrane phospholipid
`Hydrolysis and inactivation
`of penicillin
`Protease activity
`Enzyme
`Gene encoding class of
`surface proteins (lipoproteins)
`83-kDa antigen
`Molecular chaperone
`(stress response)
`Metabolic enzyme/adherence
`molecule
`Molecular chaperone
`(stress response)
`
`Identified proteins
`
`Spot
`number
`
`NCBI accession
`number
`
`% protein
`coverage
`
`pI value
`
`Z score
`
`Phosphoenolpyruvate-protein
`phosphotransferase
`Putative cardiolipin
`synthetase
`Metallo-beta-lactamase
`superfamily protein
`Hypothetical protein
`Major catalase in spores
`Adhesin pMGA1.4
`
`Heat shock-like 85-kDa
`protein
`NADP-dependent malic
`enzyme
`Small heat shock-like
`protein Hsp3
`
`2
`
`3
`
`2
`
`2
`3
`3
`
`1
`
`1
`
`1
`
`NP_975272.1
`
`NP_269350.1
`
`AA082612.1
`
`CAG20611.1
`NP_391742.1
`AAF91415.1
`
`HS85_TRYCR
`
`NP_422343.1
`
`NP_280056.1
`
`9
`
`16
`
`21
`
`13
`10
`18
`
`9
`
`18
`
`8
`
`5-4
`
`6.1
`
`5-8
`
`6.0
`5.5
`5.3
`
`5-1
`
`5.2.
`
`5.5
`
`1.79
`
`1.39
`
`1.34
`
`P32
`0.53
`0.49
`
`1-54
`
`P29
`
`0.30
`
`follicular= is usually found in the upper canal of the pilo-
`sebaceous follicle and often there are several mites in a single
`follicle, with mouthparts directed towards the fundus." The
`role of Demodex mites in the production of cutaneous inflam-
`mation is disputed, but Demodex mites have been previously
`shown to act as vectors for the transportation of microbes
`on the face" and some authors attribute specific facial erup-
`tions to these organisms." Patients with rosacea have a
`higher density of D. folliculorum mites in the areas of facial
`skin affected by the condition compared with subjects with
`normal skin, as demonstrated by the skin surface biopsy
`technique which extracts these mites from the follicular canal
`and allows their quantification."
`The mechanism by which antibiotics successfully clear the
`inflammatory lesions of papulopustular rosacea is unknown. It
`has been suggested that antibiotics work through anti-inflam-
`matory mechanisms, but other potent anti-inflammatory drugs
`are ineffective and immunosuppressive agents such as steroids
`(administered topically or systemically) and tacrolimus can
`make the inflammatory eruption of rosacea worse. 2°
`The fact that only selective antibiotics are effective in rosa-
`cea suggested to us the possibility that a bacterial agent may
`be involved in the pathogenesis of the disease. A D. folliculorum-
`related bacterium sensitive to the antibiotics used in the treat-
`ment of rosacea could explain the induction of inflammatory
`changes in papulopustular rosacea. Such changes would be
`centred on the centrofacial pilosebaceous follicles which are
`highly populated with these mites in papulopustular rosacea.
`Appropriate antibiotic treatment could reduce the bacterial-
`induced inflammatory response and possibly affect the viabil-
`ity of mites, with clinical relapse occurring either when the
`bacterial antigen load increased again after cessation of treat-
`ment or when the follicle again became distended with a
`
`repopulation of multiple mites causing 'leakage' of bacterial
`antigens through follicular walls.
`Following analysis of the microorganisms cultured from
`swabs and skin surface biopsies, we decided to study in detail
`B. oleronius because this microorganism has previously been
`reported as a member of the hindgut flora of the termite Red-
`culitemies santonensis' ° and is not a recognized human cutaneous
`commensal nor a likely contaminant of samples. This Bacillus is
`an endospore-forming bacterium previously identified as
`being Gram negative, a finding supported by our Gram stain
`method. However, this bacterium has contradictory character-
`istics, sharing Gram-positive cell wall components like all Bacil-
`lus species." Bacillus oleronius demonstrated sensitivity to several
`antibiotics which are effective in the treatment of rosacea (tetra-
`cycline, doxycycline and minocycline). Exposure of PBMC
`from patients with papulopustular rosacea to the 'sonicate'
`preparation fractions containing antigenic components of
`B. okronius resulted in significantly more PBMC stimulation
`(73%) than PBMC from control patients (29%) which suggests
`prior sensitization of PBMC from patients with rosacea to
`these surface bacterial antigens. Characterization of the bacter-
`ial antigens by SDS-PAGE and Western blot analysis revealed
`two specific antigenic proteins (83 kDa and 62 kDa) which
`were immunoreactive. Further characterization by MALDI-ToF
`analysis showed the 83-kDa protein to have homology with
`heat shock proteins, while the 62-kDa protein shared amino
`acid sequence homology with a protease enzyme involved in
`carbohydrate metabolism and signal transduction. Both these
`types of antigenic proteins have the potential to be involved
`in the immune-based inflammatory cascade.
`The significance of the finding of B. oleronius in associ-
`ation with D. folliculorum mites and its role, if any, in biological
`functioning of these mites needs to be further defined. The
`
`© 2007 The Authors
`Journal Compilation © 2007 British Association of Dermatologists • British Journal of Dermawlogy 2007 157, pp474-481
`
`6
`
`Galderma Laboratories, Inc. Ex 2010
`Dr. Reddy's Labs v. Galderma Labs., Inc.
`IPR2015-01777
`
`

`
`480 Bacterial antigens stimulate inflammatory cells in rosacea, N. Lacey et al.
`
`possible presence of other mite-related bacteria also needs to
`be explored as this Bacillus may represent only one of several
`different or even closely related species or subspecies present
`in these mites. It has been shown that several such related
`bacteria can be identified sharing a niche by multiple culture
`conditions and by 16s rDNA analysis. 2 ' Bacillus oleronius was
`isolated from one microdissected D. folliculorum mite from one
`patient with rosacea. Our failure to isolate this Bacillus from
`mites extracted from other patients with rosacea is puzzling,
`and may be related to previous antibiotic treatment of some
`of the patients with rosacea prior to sampling which may have
`eradicated or inhibited the growth of the bacterium. Another
`possible reason for our inability to culture this Bacillus from
`mites extracted from other patients is the questionable reliabil-
`ity of the traditional culture techniques we used. Other stud-
`ies21 ' 22 have shown that these culture techniques do not select
`for all bacteria. By culturing our primary facial swab samples
`and microdissected mites at 30 °C we may have reduced the
`diversity of microflora retrieved. Subculturing B. oleronius at
`both 30 °C and 37 °C showed that the Bacillus grew more
`actively at the latter temperature. However, B. oleronius grown
`at 30 °C was used in the experiments described as this condi-
`tion favoured expression of the antigens investigated. Future
`investigations of the flora of D. folliculorum-related bacteria
`should incorporate cultivation at 30 °C and 37 °C in both
`aerobic and anaerobic conditions as well as 16s rDNA analysis
`on whole mites.
`Bacillus oleronius may have a symbiotic relationship with Demo-
`dex mites as is suggested by the findings of this Bacillus as a
`member of the hindgut flora of the termite R. santonensis. Such
`symbiotic associations between microorganisms and higher
`eukaryotes are common, and range from mutualistic (bene-
`ficial) to commensal (neutral) and parasitic (harmful)." Such
`symbiotic microorganisms can have clinical importance as
`therapeutic targets or by exposure of the host to infection. A
`recent report has shown that doxycycline treatment of patients
`with filariasis resulted in a marked reduction in the worm
`population through the antibiotic effect on its endosymbiotic
`Wolbachia bacteria, 24 while suppurative cutaneous reactions
`have been reported to gut-related symbiotic bacteria of leeches
`(Hirudo medicinalis). 25 It is possible that B. oleronius has a sym-
`biotic relationship with Demodex mites similar to Wolbachia
`endosymbionts which facilitate embryogenesis and fertility in
`filarial worms. It has been repeatedly shown that patients with
`rosacea have increased facial Demodex mite populations. The
`large numbers of dying mites in the follicles of patients with
`rosacea may increase released bacterial antigen load to critical
`levels. The presence of multiple mites (both dead and alive)
`in individual follicles could distend or damage the follicular
`integrity to an extent that allows diffusion of mite-related
`bacterial antigens through the follicular wall, triggering a peri-
`follicular host immune response.
`Selective antibiotics are effective in clearing papulopustular
`rosacea but their mode of action is unknown. This work
`shows that consideration must be given to their potential to
`affect microorganisms such as B. oleronius or other follicular or
`
`mite-related bacteria. By reducing the antigenic load of
`these microorganisms or altering the numbers or biological
`fimctioning of D. folliculorum mites they may consequently
`affect the inflammatory changes which form the characteristic
`clinical features of the disease we recognize as papulopustular
`rosacea.
`
`Acknowledgments
`
`This work was supported by a grant from the National
`Rosacea Society (U.S.A.).
`
`References
`
`1 Powell FC. Rosacca. N Engl J Mad 2005; 352:793-803.
`2 Wilkin J, Dahl M, Detmar M et al. Standard classification of rosacea:
`Report of the National Rosacea Society Expert Committee on the
`Classification and Staging of Rosacea. J Am Acad Demiatol 2002;
`46:584-7.
`3 Wilkin J, Dahl M, Detmar M et al. Standard grading system for ros-
`acea: Report of the National Rosacea Society Expert Committee on
`the Classification and Staging of Rosacea. J Am Acad Demunol 2004;
`50:907-12.
`4 Powell FC. What's going on in rosacea? J Eur Acad Dermatol Venereal
`2001; 14:351-5.
`5 Rebora A. The red face: rosacea. Clin Dermatol 1993; 11:225-34.
`6 Powell FC. Rosacea and the pilosebaceous follicle. Cutis 2004;
`74:9-12.
`7 Aroni K, Tsagroni F. Lazaris AC et al. Rosacea: a clinicopathological
`approach. Dermatology 2004; 209:177-82.
`8 Bomar E, Eustace P. Powell FC. The Demodex mite population in
`rosacea. J Am Acad Dermatol 1993; 28:443-8.
`9 Forton F. Seys B. Density of Demodex folliculorum in rosacea: a case
`control study using standardized skin surface biopsy. Br J Dermatol
`1993; 128:650-9.
`10 Kuhnigk T, Borst EM, Breunig A et al. Bacillus oleronius sp. nov., a
`member of the hind-gut flora of the termite Reticulitermes santonensis
`(Feytaud). Can J Microbiol 1995; 41:699-706.
`11 von Poblotzki A, Gerdes C, Reischl U et al. Lymphoproliferative
`responses after infection with human parvovirus B19. I Viral 1996;
`70:7327-33.
`12 Yokote R, Yagi H, Furukawa F, Takigawa M. Regulation of periph-
`eral blood mononuclear cell response to Dermatophagoides farinae by
`substance P in patients with atopic dermatitis. Arch Dermatol Iles
`1998; 4:191-7.
`13 Sanni LA, Alsopp CE, Reubasaet L et al. Cellular response to Plasmo-
`dium falciparum erythrocyte membrane protein-1: use of relatively
`conserved synthetic peptide pools to determine CD4 T-cell
`responses in malaria-exposed individuals in Benin, West Africa.
`Malaria J 2002; 26:7-12.
`14 Wardlaw AL. Practical Statistics for Experimental Biologists. Chichester:
`Wiley, 2000.
`15 Bergin D, Reeves EP, Renwick J et al. Superoxide production in
`haemocytes of Galle