Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Acute skin irritant effects of cyanobacteria (blue‐green algae) in healthy volunteers

Acute skin irritant effects of cyanobacteria (blue‐green algae) in healthy volunteers Abstract Objective: To assess the skin irritant potential of a range of laboratory grown cyanobacterial species using skin-patch testing on human volunteers. Methods: Cell suspensions and extracts of cyanobacterial cultures of Microcystis aeruginosa (non-toxic strain), Anabaena circinalis and Nodularia spumigena were applied to 64 volunteers in one trial, and Microcystis aeruginosa (toxic strain), Apanocapsa incerta and Cylindrospermopsis raciborskii were applied to 50 volunteers in a second trial. Six cell concentrations of each organism in the range from less than 5,000 to greater than 200,000 cells/mL were applied in random order using adhesive skin patches (Finn Chambers). In addition, the applications included two treatments of each cyanobacterial species, involving whole and lysed cells, and positive (sodium lauryl sulphate) and negative (culture media) controls. Patches were removed after 24 hours and assessment of erythema was made by a dermatologist blinded to the species, cell type and concentration. Results: On average, between 20% and 24% of individuals with 95% confidence interval ±8% reacted across the concentration range tested for these cyanobacterial species. The reaction rates were lower (11% to 15%) among the subset of subjects not reacting to negative controls. The reaction was mostly mild, and in all cases was resolved without treatment. This was the case for both whole and lysed cells with little difference in reaction rates between these two treatments. There was also no dose-response across the concentration range for any of the cyanobacterial species tested. Conclusion: A small proportion of healthy people (around 20%) may develop a skin reaction to cyanobacteria in the course of normal water recreation, but the reaction is mild and resolved without treatment. (Aust N Z J Public Health 2004; 28: 220-4) Louis Pilotto Flinders Centre for Epidemiology & Biostatistics and the Department of General Practice, Flinders University, South Australia Peter Hobson, Michael D. Burch Cooperative Research Centre for Water Quality and Treatment, South Australia Geetha Ranmuthugala National Centre for Epidemiology and Population Health, Australian National University, Australian Capital Territory Robyn Attewell Covance Pty Ltd, Australian Capital Territory Warren Weightman Department of Dermatology, Queen Elizabeth Hospital, South Australia signif icant proportion of cyanobacteria produces one or more of a range of potent toxins. Cyanotoxins belong to a chemically rather diverse group of substances, each of which shows specific toxic mechanisms in animals. Some cyanotoxins are strong neurotoxins (anatoxin-a, anatoxin-a(s), and saxitoxins). Others are primarily toxic to the liver (microcystins, nodularin and cylindrospermopsin), and yet other potentially toxic bioactive compounds (e.g. lipopolysaccharides) may cause health impairments such as gastroenteritis or dermal effects, which are poorly understood.1 In relation to recreational activity in water contaminated with cyanobacteria, dermal contact with cyanobacteria is an important exposure route. The nature of the reports, however, for both dermal and allergic reactions are sporadic and largely anecdotal. Dermal effects from cyanobacterial exposure have been reported at Lakes Alexandrina and Albert in South Australia involving Nodularia blooms2 and in the Murray River, South Australia involving the genera Anabaena, Aphanizomenon, and Oscillatoria.3 Pilotto et al.4 carried out a comprehensive survey of health effects in people undertaking normal water recreation in waters contaminated by cyanobacteria at several sites across south-eastern Australia. The participants recorded skin irritations as one of a class of symptoms to quite low cell densities of cyanobacteria. Dermal toxicity of cyanobacteria has been attributed to lipopolysaccharides (LPSs), which are produced by all cyanobacteria.5 LPSs are an integral component of the cell wall of all Gram-negative bacteria, including cyanobacteria. LPSs can elicit irritant and allergenic responses in human and animal tissues that come in contact with the compounds.1 However, cyanobacterial LPS has been shown to be less potent than LPSs from pathogenic gram-negative bacteria such as Salmonella.6,7 The few studies linking cyanobacteria to Submitted: August 2003 Revision requested: October 2003 Accepted: December 2003 Correspondence to: Professor Louis Pilotto, Flinders Centre for Epidemiology & Biostatistics, Department of General Practice, Flinders University, GPO Box 2100, Adelaide, South Australia 5001. Fax: (08) 8276 3305; e-mail: louis.pilotto@flinders.edu.au AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2004 VOL. 28 NO. 3 Viruses and Bacteria Acute skin irritant effects of cyanobacteria skin contact irritant effects are observational, with limited4 or no measurement of personal levels of exposure, and therefore do not provide a strong basis for the establishment of guidelines for recreational activities involving water contact in cyanobacterial contaminated water. The purpose of this study was to quantitatively assess skin irritant properties of a range of cyanobacterial species with an experimental study design that would assess possible threshold and dose-response effects. Methods Following an e-mail and invitations at lectures for Year 1 medical students, and invitations among administrative staff from the School of Medicine at the Flinders University of South Australia, a total of 114 volunteers were recruited for three rounds of skinpatch testing conducted in May and October 2001 and in May 2002. The first (24 volunteers) and second (40 volunteers) rounds were used to investigate Microcystis aeruginosa (non-toxic strain), Anabaena circinalis and Nodularia spumigena. The third round (50 volunteers) was used to investigate Microcystis aeruginosa (toxic strain), Apanocapsa incerta and Cylindrospermopsis raciborskii. Cultures were grown under laboratory conditions (25oC and 15 µmol m2 s1, continuous illumination) and harvested in log (active growth) phase on the same day of application to the volunteers. This avoided cell lysis and loss in cell integrity. The cultures were mono-cyanobacterial but not axenic (i.e. not free from other contaminant bacteria). Solutions of known cell density of each cyanobacterial culture were prepared by serial dilution and applied to the skin using adhesive patches with 10 individual chambers (8 mm diameter), each containing a small filter pad (Finn Chambers). Each chamber held a 20 µL volume of culture, which was administered to the filter pad by micro-pipette. Four patches (i.e. 40 chambers) were applied in a random sequence on the volunteers’ upper backs. Each volunteer was exposed to three different cyanobacterial species at six cell concentrations per species. Table 1 gives the dose-concentrations for each species tested. These concentrations were consistent with those found in a previous study that examined water bodies containing cyanobacteria that were used for recreational water activities.4 Both whole and lysed preparations of each cyanobacterial species were applied on each volunteer. It was hypothesised that lysed cell material would allow contact with intra-cellular components that may display irritant properties. Two positive controls (1 and 5% solutions of sodium lauryl sulphate (SLS)) and two negative controls (culture media and an empty patch) were also included. The 1% solution was introduced in round 2 due to the strong reaction to the 5% solution in round 1. SLS is a well-recognised irritant that works by disorganisation of the lipid layers leading to impairment of barrier function. This effect is time and concentration dependent.8 SLS was used to provide information on how irritant the algae were in comparison to a known irritant. Participants were advised not to get the patch wet (i.e. not to shower or work up a sweat) until the patches were removed. Volunteers completed a questionnaire outlining their medical and family history of atopy. Individuals who had been diagnosed as having eczema, hay fever, asthma or an atopic condition were not excluded from the study. However, severe recurrent skin conditions and pregnancy were exclusion criteria. Patches were removed after 24 hours and a dermatologist, using the following clinical grades, made an assessment for the presence of erythematous reactions: 0 No reaction or erythema 1+ Minimal or very weak spotty erythema 2+ Mild diffuse erythema 3+ Moderate diffuse erythema 4+ Severe diffuse erythema with oedema A one-hour interval was allowed after removing patches and before assessment to allow for any reaction induced by the adhesive to subside. Assessments were made for each of the 40 patches, as well as for normal skin. Erythema readings were repeated 48 hours later in round 1 (i.e. a total of 72 hours after the patches were applied). The second assessment was not included in rounds 2 and 3 due to the marked reduction in response after 72 hours, compared with round 1. The rounds were carried out as double-blind experiments. Both Table 1: Cyanobacterial species and cell densities used in rounds 1, 2 and 3 of the skin irritation study. Species and strains tested Microcystis aeruginosa strain 338 (non-toxic) 5,260 20,800 55,200 103,000 204,000 289,000 Anabaena circinalis strain 118AR (neurotoxic) 3,792 13,490 56,700 82,100 163,000 287,000 Nodularia spumigena strain 001E (hepatotoxic) 4,884 15,800 59,700 79,500 121,000 214,000 Microcystis aeruginosa strain 309(1) CA (hepatotoxic) 3 5,450 22,750 70,667 109,500 203,000 300,500 Cylindrospermopsis raciborskii strain 031C (toxic) 4,670 14,458 51,015 81,795 123,675 150,792 Aphanocapsa incerta strain 001 (non-toxic) 82,578 107,583 225,944 200,667 311,567 351,900 Round Cell density (cells/mL) 8,708 35,571 100,600 155,273 295,333 364,800 3,854 20,416 38,976 86,700 232,202 324,360 5,582 24,385 61,104 86,156 186,750 304,560 2004 VOL. 28 NO. 3 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Pilotto et al. Article Table 2: Odds ratios (OR) and 95% confidence intervals (CI) for skin reactions (grades 1 to 4) estimated in logistic regression models for whole and lysed cells from different species (MA338, AC, NS in rounds 1 and 2; MA309, CR, AI in round 3). Rounds 1 & 2 (n=64) Term ORa MA338 Whole Lysed AC Whole Lysed NS Whole Lysed 1.76 1.53 1.98 1.79 2.04 1.51 1.98 1.77 2.18 the project officer applying the patches and the dermatologist assessing the reactions were not aware of the type and cell concentration of the cyanobacterial preparations applied to each individual. Statistical methods 95% CI 1.08-2.86 0.90-2.61 1.21-3.23 1.13-2.84 1.17-3.56 0.95-2.41 1.21-3.23 1.04-3.03 1.33-3.59 Round 3 (n=50) Term MA309 Whole Lysed CR Whole Lysed AI Whole Lysed ORa 95% CI 1.33-4.35 1.35-5.14 1.27-3.74 1.79-4.21 1.42-3.21 2.00-5.84 1.93-4.20 2.13-5.87 1.55-3.21 0.004 0.005 0.005 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 Notes: (a) OR is odds ratio for skin reactions to active patches vs. negative control patches. MA=Microcystis aeruginosa; AC=Anabaena circinalis; NS=Nodularia spumigena; CR=Cylindrospermopsis raciborskii; AI=Aphanocapsa incerta Table 3: Mean percentages and 95% confidence intervals (95% CI) for skin reactions (grades 1 to 4) to an active (non-control) patch by species, overall (Total) and for the subset of subjects not reacting to negative controls. Rounds 1 and 2 Total subjects (n=64) Species Mean 95% CI MA338 AC NS 22% 22% 24% 16-31% 16-29% 18-33% The distribution of clinical gradings by patch type (control/ active), species, cell type and cell concentration was assessed. Due to the relatively small number of high-level gradings, each observation was dichotomised into no reaction (grade 0) and a positive reaction (1+, 2+, 3+ or 4+) prior to modelling. Two sets of logistic regression models were then fitted. First, reactions to active patches were compared with reactions to negative control patches producing odds ratios (and 95% confidence intervals) for reactions for exposure to an active versus a negative control patch. The average percentage of reactions to an active patch (and 95% confidence interval) was also calculated based on estimates from this model for each species. Second, dose-response models were fitted based on the cell counts of the active patches estimating the odds ratio for reactions per 10,000 cells/mL. Separate regression models were fitted for each species (whole and lysed cells combined and whole and lysed cells separately) for data from each round and combined. Dose-response modelling was also repeated excluding the subjects who reacted to the negative control patches. The repeated nature of the data (multiple observations per subject) was taken into account using the Generalised Estimating Equation (GEE) approach as implemented in the STATA Version 7 (Stata Corporation, 4905 Lakeway Drive, College Station, TX 77845, USA) procedure ‘xtlogit’ using robust standard error estimation. The results were also disaggregated by atopy status (defined as self report of ever being diagnosed with asthma, eczema, hayfever or atopy). Tests with significance levels below 0.05 are interpreted as statistically significant. Results The 114 volunteers ranged in age from 17 to 54 years (mean 28, standard deviation 7, median 26 years). There were 44 men and 70 women. Half were atopic (50%). Most subjects (88%, 100 out of 114) had an erythematous reaction to the positive control patches. However, reactions were also recorded for the negative control or blank patches in almost one-quarter (23%, 26 out of 114) of the subjects. Subjects were more likely to react to the active patches than to the negative control patches for all species, and for both whole and lysed cells (see Table 2). Almost all results are statistically significant (p<0.05) and the odds ratios for active versus control patches range between 1.5 and 3.5. The reaction rates estimated for the six cyanobacterial species lie in the range 20% to 24% with 95% confidence intervals approximately ± 8% (see Table 3). If the volunteers who exhibited reactions to the negative 2004 VOL. 28 NO. 3 Subjects not reacting to negative controls (n=49) Mean 95% CI 13% 13% 15% 8-21% 8-19% 10-23% Round 3 Total subjects (n=50) Mean 95% CI 20% 22% 23% 13-29% 15-31% 16-32% Species MA309 CR AI Subjects not reacting to negative controls (n=39) Mean 95% CI 13% 11% 11% 7-22% 6-18% 7-17% Note: MA=Microcystis aeruginosa; AC=Anabaena circinalis; NS= Nodularia spumigena; CR = Cylindrospermopsis raciborskii; AI= Aphanocapsa incerta. AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Viruses and Bacteria Acute skin irritant effects of cyanobacteria control patches are excluded from the analysis, the reaction rates decrease to between 11% and 15% (± 7%). There is no evidence of a consistent increasing dose-response relationship between positive gradings and increasing cell concentrations for the majority of the cyanobacterial species tested whether they are whole or lysed (see Table 4). Statistical significance was reached for two species (see Table 4), however, in one case the relationship was inverse (i.e. reduction in clinical response with increasing cell concentration). Furthermore, there is no evidence for a threshold effect, i.e. a particular concentration above which there were frequent or strong reactions. The percentage of subjects who reacted to the active patches is not uniformly consistently higher or lower for atopic versus nonatopic subjects for any species or concentration. Table 4: Odds ratios (OR) and 95% confidence intervals (CI) for skin reactions (grades 1 to 4) estimated in logistic regression dose-response models with cell counts for whole and lysed cells from different species (MA338, AC, NS in rounds 1 and 2; MA309, CR, AI in round 3). Rounds 1 & 2 (n=64) Term ORa MA338 Whole Lysed AC Whole Lysed NS Whole Lysed 1.005 1.011 1.000 1.000 1.003 0.997 0.999 1.000 1.005 95% CI 0.997-1.014 0.997-1.026 0.984-1.016 0.988-1.012 0.985-1.021 0.985-1.009 0.983-1.014 0.981-1.019 0.997-1.014 Discussion In summary, results have shown that between 20% to 24% of both atopic and non-atopic individuals reacted to the six cyanobacterial species studied across the concentration range tested. These reaction rates may be overstated by 10% since a significant number of subjects also reacted to the negative control patches. Both whole and lysed cells displayed the same reaction rates. There was considerable variation in the pattern of reactions across the cell concentrations and no consistently sustained increasing dose-response relationship was identified for any of the cyanobacterial species tested. The results from this study are in agreement with a range of earlier investigations that have demonstrated variable positive reactions to skin exposure to cyanobacteria and algae. Heise9 observed allergenic reactions in persons given intradermal injections of glycerosaline extracts of dried Microcystis and Oscillatoriaceae species. Ten persons were found to react to both groups of cyanobacteria. However, another 50 people showed no reaction to either of the test organisms. They concluded that both Microcystis and Oscillatoriaceae species contained similar antigens and that only certain individuals would show an allergenic response. McElhenney et al.10 performed intradermal skin tests using four different green algae species on 140 children, of which 20 were non-allergenic and 120 had pollen and/or other inhalant sensitivities. None of the non-allergenic group showed a reaction. Of the 120 allergenic children, 98 showed positive reactions to one or more of the algal species while 22 showed no reaction. Mittal et al.11 carried out a study investigating the association between allergy sufferers and irritant effects of algae. Results were presented for 4,000 intradermal skin tests performed on 400 participants suffering from nasal-bronchial allergy and 300 skin tests on 30 healthy persons with 10 common algae isolated from Delhi, India. Genera studied included Lyngbya, Phormidium, Anabaena, Scytonema, Chlorella, Westiellopsis, Anabaenopsis, Oscillatoria, Nostoc and Chlorococcum. Response rates ranged from 25.7% for Lyngbya to 1.7% for Oscillatoria in allergic volunteers. In non-allergic volunteers no positive skin reactions were identified. 2004 VOL. 28 NO. 3 Round 3 (n=50) Term MA309 Whole Lysed CR Whole Lysed AI Whole Lysed ORa 95% CI 0.987-1.025 1.002-1.031 0.962-1.025 0.956-1.013 0.936-0.997 0.96-1.037 0.995-1.027 0.985-1.02 0.995-1.05 Notes: (a) OR is odds ratio for skin reactions per 10,000 cells/mL. MA=Microcystis aeruginosa; AC=Anabaena circinalis; NS=Nodularia spumigena; CR=Cylindrospermopsis raciborskii; AI=Aphanocapsa incerta While Heise 9 suggested that certain individuals were specifically allergic to cyanobacteria, McElhenney et al.10 and Mittal et al.11 showed a direct link between those people who suffered from nasal-bronchial allergies and skin irritation due to exposure to cyanobacteria. In contrast, the present study showed that individuals who were atopic, i.e. people who suffered from eczema, hay fever, asthma or who had been diagnosed as atopic, did not show a statistically significant increase in response rate compared with non-atopic individuals. It should be noted that the studies by Heise,9 McElhenney et al.10 and Mittal et al.11 all used intra-dermal applications of cyanobacteria. It may be expected that a more significant reaction would be observed for sub-cutaneous application compared with exposure of unbroken skin. In general, bathing or recreational activities would not result in exposure to cyanobacteria via broken skin. Therefore, the use of skin patches in the current study is more representative of normal exposure to cyanobacteria in water recreation situations. However, it is often reported anecdotally that irritation can occur around mucous membranes or broken skin and in situations where the cyanobacterial cells are rubbed into the skin (e.g. underneath swimsuits and wetsuits). Exposure in these situations where the epidermis is abraded may be more comparable with intra-dermal applications of the earlier studies.9-11 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Pilotto et al. Article The interpretation of the results from this study for the protection of individuals from adverse effects during recreation, i.e. for guideline development, is problematic. The demonstration of irritation in a small proportion of the population at low to moderate cell densities confirms that there is an adverse health hazard from exposure. The severity of the irritation in the affected individuals did not appear severe or debilitating, was self-limiting and was resolved within a short period (24-72 hours). The absence of a dose-dependent response and a threshold makes it impossible to recommend a quantitative protective guideline based upon our study. The issue of the significance of the characterised toxins produced by cyanobacteria (i.e. hepato- and neurotoxins) in this study requires comment. There was no difference in skin irritation reaction for strains that produce known toxins (M. aeruginosa, C. raciborskii, N spumigena, and A circinalis) compared with nontoxic strains (A. incerta). Based upon knowledge of the mode of action of these toxins (microcystins, nodularin, cylindrospermopsin and saxitoxins), it is expected that they would not be absorbed and exert a toxic systemic effect in unbroken skin. It follows that this route of exposure (dermal absorption) in water recreational activity is less significant when compared with the potential effects from ingestion, which definitely needs to be addressed in the setting of guidelines for recreational water use. ence skin irritation reactions or possibly allergic reactions associated with cyanobacteria. In any case, the minor nature of the potential irritant dermatitis from this type of exposure would be less important than the need to protect from toxic effects from known characterised toxins (hepato- and neurotoxins) primarily via bathing water guidelines for these toxins. Acknowledgements Financial support for this study was provided by the Cooperative Research Centre for Water Quality and Treatment, Eraring Energy Pty Ltd, the New South Wales Health Department, and the Sydney Catchment Authority. Many thanks to the students and staff of the School of Medicine, and the Department of General Practice, Flinders University, for participating as volunteers in this study. Special thanks to Raelene Burnley, Caroline Fazekas, Cecilia Freeman and Leon Linden for assistance with volunteer trials. Thanks also to Peter Baker for supplying the cyanobacterial cultures used for this work. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian and New Zealand Journal of Public Health Wiley

Acute skin irritant effects of cyanobacteria (blue‐green algae) in healthy volunteers

Download PDF
5 pages

Loading next page...
 
/lp/wiley/acute-skin-irritant-effects-of-cyanobacteria-blue-green-algae-in-B16ubnDj1S

References (14)

Publisher
Wiley
Copyright
Copyright © 2004 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1326-0200
eISSN
1753-6405
DOI
10.1111/j.1467-842X.2004.tb00699.x
Publisher site
See Article on Publisher Site

Abstract

Abstract Objective: To assess the skin irritant potential of a range of laboratory grown cyanobacterial species using skin-patch testing on human volunteers. Methods: Cell suspensions and extracts of cyanobacterial cultures of Microcystis aeruginosa (non-toxic strain), Anabaena circinalis and Nodularia spumigena were applied to 64 volunteers in one trial, and Microcystis aeruginosa (toxic strain), Apanocapsa incerta and Cylindrospermopsis raciborskii were applied to 50 volunteers in a second trial. Six cell concentrations of each organism in the range from less than 5,000 to greater than 200,000 cells/mL were applied in random order using adhesive skin patches (Finn Chambers). In addition, the applications included two treatments of each cyanobacterial species, involving whole and lysed cells, and positive (sodium lauryl sulphate) and negative (culture media) controls. Patches were removed after 24 hours and assessment of erythema was made by a dermatologist blinded to the species, cell type and concentration. Results: On average, between 20% and 24% of individuals with 95% confidence interval ±8% reacted across the concentration range tested for these cyanobacterial species. The reaction rates were lower (11% to 15%) among the subset of subjects not reacting to negative controls. The reaction was mostly mild, and in all cases was resolved without treatment. This was the case for both whole and lysed cells with little difference in reaction rates between these two treatments. There was also no dose-response across the concentration range for any of the cyanobacterial species tested. Conclusion: A small proportion of healthy people (around 20%) may develop a skin reaction to cyanobacteria in the course of normal water recreation, but the reaction is mild and resolved without treatment. (Aust N Z J Public Health 2004; 28: 220-4) Louis Pilotto Flinders Centre for Epidemiology & Biostatistics and the Department of General Practice, Flinders University, South Australia Peter Hobson, Michael D. Burch Cooperative Research Centre for Water Quality and Treatment, South Australia Geetha Ranmuthugala National Centre for Epidemiology and Population Health, Australian National University, Australian Capital Territory Robyn Attewell Covance Pty Ltd, Australian Capital Territory Warren Weightman Department of Dermatology, Queen Elizabeth Hospital, South Australia signif icant proportion of cyanobacteria produces one or more of a range of potent toxins. Cyanotoxins belong to a chemically rather diverse group of substances, each of which shows specific toxic mechanisms in animals. Some cyanotoxins are strong neurotoxins (anatoxin-a, anatoxin-a(s), and saxitoxins). Others are primarily toxic to the liver (microcystins, nodularin and cylindrospermopsin), and yet other potentially toxic bioactive compounds (e.g. lipopolysaccharides) may cause health impairments such as gastroenteritis or dermal effects, which are poorly understood.1 In relation to recreational activity in water contaminated with cyanobacteria, dermal contact with cyanobacteria is an important exposure route. The nature of the reports, however, for both dermal and allergic reactions are sporadic and largely anecdotal. Dermal effects from cyanobacterial exposure have been reported at Lakes Alexandrina and Albert in South Australia involving Nodularia blooms2 and in the Murray River, South Australia involving the genera Anabaena, Aphanizomenon, and Oscillatoria.3 Pilotto et al.4 carried out a comprehensive survey of health effects in people undertaking normal water recreation in waters contaminated by cyanobacteria at several sites across south-eastern Australia. The participants recorded skin irritations as one of a class of symptoms to quite low cell densities of cyanobacteria. Dermal toxicity of cyanobacteria has been attributed to lipopolysaccharides (LPSs), which are produced by all cyanobacteria.5 LPSs are an integral component of the cell wall of all Gram-negative bacteria, including cyanobacteria. LPSs can elicit irritant and allergenic responses in human and animal tissues that come in contact with the compounds.1 However, cyanobacterial LPS has been shown to be less potent than LPSs from pathogenic gram-negative bacteria such as Salmonella.6,7 The few studies linking cyanobacteria to Submitted: August 2003 Revision requested: October 2003 Accepted: December 2003 Correspondence to: Professor Louis Pilotto, Flinders Centre for Epidemiology & Biostatistics, Department of General Practice, Flinders University, GPO Box 2100, Adelaide, South Australia 5001. Fax: (08) 8276 3305; e-mail: louis.pilotto@flinders.edu.au AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 2004 VOL. 28 NO. 3 Viruses and Bacteria Acute skin irritant effects of cyanobacteria skin contact irritant effects are observational, with limited4 or no measurement of personal levels of exposure, and therefore do not provide a strong basis for the establishment of guidelines for recreational activities involving water contact in cyanobacterial contaminated water. The purpose of this study was to quantitatively assess skin irritant properties of a range of cyanobacterial species with an experimental study design that would assess possible threshold and dose-response effects. Methods Following an e-mail and invitations at lectures for Year 1 medical students, and invitations among administrative staff from the School of Medicine at the Flinders University of South Australia, a total of 114 volunteers were recruited for three rounds of skinpatch testing conducted in May and October 2001 and in May 2002. The first (24 volunteers) and second (40 volunteers) rounds were used to investigate Microcystis aeruginosa (non-toxic strain), Anabaena circinalis and Nodularia spumigena. The third round (50 volunteers) was used to investigate Microcystis aeruginosa (toxic strain), Apanocapsa incerta and Cylindrospermopsis raciborskii. Cultures were grown under laboratory conditions (25oC and 15 µmol m2 s1, continuous illumination) and harvested in log (active growth) phase on the same day of application to the volunteers. This avoided cell lysis and loss in cell integrity. The cultures were mono-cyanobacterial but not axenic (i.e. not free from other contaminant bacteria). Solutions of known cell density of each cyanobacterial culture were prepared by serial dilution and applied to the skin using adhesive patches with 10 individual chambers (8 mm diameter), each containing a small filter pad (Finn Chambers). Each chamber held a 20 µL volume of culture, which was administered to the filter pad by micro-pipette. Four patches (i.e. 40 chambers) were applied in a random sequence on the volunteers’ upper backs. Each volunteer was exposed to three different cyanobacterial species at six cell concentrations per species. Table 1 gives the dose-concentrations for each species tested. These concentrations were consistent with those found in a previous study that examined water bodies containing cyanobacteria that were used for recreational water activities.4 Both whole and lysed preparations of each cyanobacterial species were applied on each volunteer. It was hypothesised that lysed cell material would allow contact with intra-cellular components that may display irritant properties. Two positive controls (1 and 5% solutions of sodium lauryl sulphate (SLS)) and two negative controls (culture media and an empty patch) were also included. The 1% solution was introduced in round 2 due to the strong reaction to the 5% solution in round 1. SLS is a well-recognised irritant that works by disorganisation of the lipid layers leading to impairment of barrier function. This effect is time and concentration dependent.8 SLS was used to provide information on how irritant the algae were in comparison to a known irritant. Participants were advised not to get the patch wet (i.e. not to shower or work up a sweat) until the patches were removed. Volunteers completed a questionnaire outlining their medical and family history of atopy. Individuals who had been diagnosed as having eczema, hay fever, asthma or an atopic condition were not excluded from the study. However, severe recurrent skin conditions and pregnancy were exclusion criteria. Patches were removed after 24 hours and a dermatologist, using the following clinical grades, made an assessment for the presence of erythematous reactions: 0 No reaction or erythema 1+ Minimal or very weak spotty erythema 2+ Mild diffuse erythema 3+ Moderate diffuse erythema 4+ Severe diffuse erythema with oedema A one-hour interval was allowed after removing patches and before assessment to allow for any reaction induced by the adhesive to subside. Assessments were made for each of the 40 patches, as well as for normal skin. Erythema readings were repeated 48 hours later in round 1 (i.e. a total of 72 hours after the patches were applied). The second assessment was not included in rounds 2 and 3 due to the marked reduction in response after 72 hours, compared with round 1. The rounds were carried out as double-blind experiments. Both Table 1: Cyanobacterial species and cell densities used in rounds 1, 2 and 3 of the skin irritation study. Species and strains tested Microcystis aeruginosa strain 338 (non-toxic) 5,260 20,800 55,200 103,000 204,000 289,000 Anabaena circinalis strain 118AR (neurotoxic) 3,792 13,490 56,700 82,100 163,000 287,000 Nodularia spumigena strain 001E (hepatotoxic) 4,884 15,800 59,700 79,500 121,000 214,000 Microcystis aeruginosa strain 309(1) CA (hepatotoxic) 3 5,450 22,750 70,667 109,500 203,000 300,500 Cylindrospermopsis raciborskii strain 031C (toxic) 4,670 14,458 51,015 81,795 123,675 150,792 Aphanocapsa incerta strain 001 (non-toxic) 82,578 107,583 225,944 200,667 311,567 351,900 Round Cell density (cells/mL) 8,708 35,571 100,600 155,273 295,333 364,800 3,854 20,416 38,976 86,700 232,202 324,360 5,582 24,385 61,104 86,156 186,750 304,560 2004 VOL. 28 NO. 3 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Pilotto et al. Article Table 2: Odds ratios (OR) and 95% confidence intervals (CI) for skin reactions (grades 1 to 4) estimated in logistic regression models for whole and lysed cells from different species (MA338, AC, NS in rounds 1 and 2; MA309, CR, AI in round 3). Rounds 1 & 2 (n=64) Term ORa MA338 Whole Lysed AC Whole Lysed NS Whole Lysed 1.76 1.53 1.98 1.79 2.04 1.51 1.98 1.77 2.18 the project officer applying the patches and the dermatologist assessing the reactions were not aware of the type and cell concentration of the cyanobacterial preparations applied to each individual. Statistical methods 95% CI 1.08-2.86 0.90-2.61 1.21-3.23 1.13-2.84 1.17-3.56 0.95-2.41 1.21-3.23 1.04-3.03 1.33-3.59 Round 3 (n=50) Term MA309 Whole Lysed CR Whole Lysed AI Whole Lysed ORa 95% CI 1.33-4.35 1.35-5.14 1.27-3.74 1.79-4.21 1.42-3.21 2.00-5.84 1.93-4.20 2.13-5.87 1.55-3.21 0.004 0.005 0.005 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 Notes: (a) OR is odds ratio for skin reactions to active patches vs. negative control patches. MA=Microcystis aeruginosa; AC=Anabaena circinalis; NS=Nodularia spumigena; CR=Cylindrospermopsis raciborskii; AI=Aphanocapsa incerta Table 3: Mean percentages and 95% confidence intervals (95% CI) for skin reactions (grades 1 to 4) to an active (non-control) patch by species, overall (Total) and for the subset of subjects not reacting to negative controls. Rounds 1 and 2 Total subjects (n=64) Species Mean 95% CI MA338 AC NS 22% 22% 24% 16-31% 16-29% 18-33% The distribution of clinical gradings by patch type (control/ active), species, cell type and cell concentration was assessed. Due to the relatively small number of high-level gradings, each observation was dichotomised into no reaction (grade 0) and a positive reaction (1+, 2+, 3+ or 4+) prior to modelling. Two sets of logistic regression models were then fitted. First, reactions to active patches were compared with reactions to negative control patches producing odds ratios (and 95% confidence intervals) for reactions for exposure to an active versus a negative control patch. The average percentage of reactions to an active patch (and 95% confidence interval) was also calculated based on estimates from this model for each species. Second, dose-response models were fitted based on the cell counts of the active patches estimating the odds ratio for reactions per 10,000 cells/mL. Separate regression models were fitted for each species (whole and lysed cells combined and whole and lysed cells separately) for data from each round and combined. Dose-response modelling was also repeated excluding the subjects who reacted to the negative control patches. The repeated nature of the data (multiple observations per subject) was taken into account using the Generalised Estimating Equation (GEE) approach as implemented in the STATA Version 7 (Stata Corporation, 4905 Lakeway Drive, College Station, TX 77845, USA) procedure ‘xtlogit’ using robust standard error estimation. The results were also disaggregated by atopy status (defined as self report of ever being diagnosed with asthma, eczema, hayfever or atopy). Tests with significance levels below 0.05 are interpreted as statistically significant. Results The 114 volunteers ranged in age from 17 to 54 years (mean 28, standard deviation 7, median 26 years). There were 44 men and 70 women. Half were atopic (50%). Most subjects (88%, 100 out of 114) had an erythematous reaction to the positive control patches. However, reactions were also recorded for the negative control or blank patches in almost one-quarter (23%, 26 out of 114) of the subjects. Subjects were more likely to react to the active patches than to the negative control patches for all species, and for both whole and lysed cells (see Table 2). Almost all results are statistically significant (p<0.05) and the odds ratios for active versus control patches range between 1.5 and 3.5. The reaction rates estimated for the six cyanobacterial species lie in the range 20% to 24% with 95% confidence intervals approximately ± 8% (see Table 3). If the volunteers who exhibited reactions to the negative 2004 VOL. 28 NO. 3 Subjects not reacting to negative controls (n=49) Mean 95% CI 13% 13% 15% 8-21% 8-19% 10-23% Round 3 Total subjects (n=50) Mean 95% CI 20% 22% 23% 13-29% 15-31% 16-32% Species MA309 CR AI Subjects not reacting to negative controls (n=39) Mean 95% CI 13% 11% 11% 7-22% 6-18% 7-17% Note: MA=Microcystis aeruginosa; AC=Anabaena circinalis; NS= Nodularia spumigena; CR = Cylindrospermopsis raciborskii; AI= Aphanocapsa incerta. AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Viruses and Bacteria Acute skin irritant effects of cyanobacteria control patches are excluded from the analysis, the reaction rates decrease to between 11% and 15% (± 7%). There is no evidence of a consistent increasing dose-response relationship between positive gradings and increasing cell concentrations for the majority of the cyanobacterial species tested whether they are whole or lysed (see Table 4). Statistical significance was reached for two species (see Table 4), however, in one case the relationship was inverse (i.e. reduction in clinical response with increasing cell concentration). Furthermore, there is no evidence for a threshold effect, i.e. a particular concentration above which there were frequent or strong reactions. The percentage of subjects who reacted to the active patches is not uniformly consistently higher or lower for atopic versus nonatopic subjects for any species or concentration. Table 4: Odds ratios (OR) and 95% confidence intervals (CI) for skin reactions (grades 1 to 4) estimated in logistic regression dose-response models with cell counts for whole and lysed cells from different species (MA338, AC, NS in rounds 1 and 2; MA309, CR, AI in round 3). Rounds 1 & 2 (n=64) Term ORa MA338 Whole Lysed AC Whole Lysed NS Whole Lysed 1.005 1.011 1.000 1.000 1.003 0.997 0.999 1.000 1.005 95% CI 0.997-1.014 0.997-1.026 0.984-1.016 0.988-1.012 0.985-1.021 0.985-1.009 0.983-1.014 0.981-1.019 0.997-1.014 Discussion In summary, results have shown that between 20% to 24% of both atopic and non-atopic individuals reacted to the six cyanobacterial species studied across the concentration range tested. These reaction rates may be overstated by 10% since a significant number of subjects also reacted to the negative control patches. Both whole and lysed cells displayed the same reaction rates. There was considerable variation in the pattern of reactions across the cell concentrations and no consistently sustained increasing dose-response relationship was identified for any of the cyanobacterial species tested. The results from this study are in agreement with a range of earlier investigations that have demonstrated variable positive reactions to skin exposure to cyanobacteria and algae. Heise9 observed allergenic reactions in persons given intradermal injections of glycerosaline extracts of dried Microcystis and Oscillatoriaceae species. Ten persons were found to react to both groups of cyanobacteria. However, another 50 people showed no reaction to either of the test organisms. They concluded that both Microcystis and Oscillatoriaceae species contained similar antigens and that only certain individuals would show an allergenic response. McElhenney et al.10 performed intradermal skin tests using four different green algae species on 140 children, of which 20 were non-allergenic and 120 had pollen and/or other inhalant sensitivities. None of the non-allergenic group showed a reaction. Of the 120 allergenic children, 98 showed positive reactions to one or more of the algal species while 22 showed no reaction. Mittal et al.11 carried out a study investigating the association between allergy sufferers and irritant effects of algae. Results were presented for 4,000 intradermal skin tests performed on 400 participants suffering from nasal-bronchial allergy and 300 skin tests on 30 healthy persons with 10 common algae isolated from Delhi, India. Genera studied included Lyngbya, Phormidium, Anabaena, Scytonema, Chlorella, Westiellopsis, Anabaenopsis, Oscillatoria, Nostoc and Chlorococcum. Response rates ranged from 25.7% for Lyngbya to 1.7% for Oscillatoria in allergic volunteers. In non-allergic volunteers no positive skin reactions were identified. 2004 VOL. 28 NO. 3 Round 3 (n=50) Term MA309 Whole Lysed CR Whole Lysed AI Whole Lysed ORa 95% CI 0.987-1.025 1.002-1.031 0.962-1.025 0.956-1.013 0.936-0.997 0.96-1.037 0.995-1.027 0.985-1.02 0.995-1.05 Notes: (a) OR is odds ratio for skin reactions per 10,000 cells/mL. MA=Microcystis aeruginosa; AC=Anabaena circinalis; NS=Nodularia spumigena; CR=Cylindrospermopsis raciborskii; AI=Aphanocapsa incerta While Heise 9 suggested that certain individuals were specifically allergic to cyanobacteria, McElhenney et al.10 and Mittal et al.11 showed a direct link between those people who suffered from nasal-bronchial allergies and skin irritation due to exposure to cyanobacteria. In contrast, the present study showed that individuals who were atopic, i.e. people who suffered from eczema, hay fever, asthma or who had been diagnosed as atopic, did not show a statistically significant increase in response rate compared with non-atopic individuals. It should be noted that the studies by Heise,9 McElhenney et al.10 and Mittal et al.11 all used intra-dermal applications of cyanobacteria. It may be expected that a more significant reaction would be observed for sub-cutaneous application compared with exposure of unbroken skin. In general, bathing or recreational activities would not result in exposure to cyanobacteria via broken skin. Therefore, the use of skin patches in the current study is more representative of normal exposure to cyanobacteria in water recreation situations. However, it is often reported anecdotally that irritation can occur around mucous membranes or broken skin and in situations where the cyanobacterial cells are rubbed into the skin (e.g. underneath swimsuits and wetsuits). Exposure in these situations where the epidermis is abraded may be more comparable with intra-dermal applications of the earlier studies.9-11 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Pilotto et al. Article The interpretation of the results from this study for the protection of individuals from adverse effects during recreation, i.e. for guideline development, is problematic. The demonstration of irritation in a small proportion of the population at low to moderate cell densities confirms that there is an adverse health hazard from exposure. The severity of the irritation in the affected individuals did not appear severe or debilitating, was self-limiting and was resolved within a short period (24-72 hours). The absence of a dose-dependent response and a threshold makes it impossible to recommend a quantitative protective guideline based upon our study. The issue of the significance of the characterised toxins produced by cyanobacteria (i.e. hepato- and neurotoxins) in this study requires comment. There was no difference in skin irritation reaction for strains that produce known toxins (M. aeruginosa, C. raciborskii, N spumigena, and A circinalis) compared with nontoxic strains (A. incerta). Based upon knowledge of the mode of action of these toxins (microcystins, nodularin, cylindrospermopsin and saxitoxins), it is expected that they would not be absorbed and exert a toxic systemic effect in unbroken skin. It follows that this route of exposure (dermal absorption) in water recreational activity is less significant when compared with the potential effects from ingestion, which definitely needs to be addressed in the setting of guidelines for recreational water use. ence skin irritation reactions or possibly allergic reactions associated with cyanobacteria. In any case, the minor nature of the potential irritant dermatitis from this type of exposure would be less important than the need to protect from toxic effects from known characterised toxins (hepato- and neurotoxins) primarily via bathing water guidelines for these toxins. Acknowledgements Financial support for this study was provided by the Cooperative Research Centre for Water Quality and Treatment, Eraring Energy Pty Ltd, the New South Wales Health Department, and the Sydney Catchment Authority. Many thanks to the students and staff of the School of Medicine, and the Department of General Practice, Flinders University, for participating as volunteers in this study. Special thanks to Raelene Burnley, Caroline Fazekas, Cecilia Freeman and Leon Linden for assistance with volunteer trials. Thanks also to Peter Baker for supplying the cyanobacterial cultures used for this work.

Journal

Australian and New Zealand Journal of Public Health – Wiley

Published: Jun 1, 2004

There are no references for this article.