University of Cincinnati

UNIVERSITY OF CINCINNATI

Date:___11-13-2006______

I, ______Yulia Yulianova Iossifova______, hereby submit this work as part of the requirements for the degree of: DOCTOR OF PHILOSOPHY in:

ENVIRONMENTAL HEALTH It is entitled: (1-3)-ß-D-GLUCANS IN INDOOR ENVIRONMENTS - LABORATORY ANALYSIS AND WHEEZE IN INFANTS

This work and its defense approved by:

Chair: Tiina Reponen, PhD Linda Levin, PhD Grace Lemasters, PhD Gurjit Khurana-Hershey, MD, PhD Vincent Castranova, PhD

(1-3)-ß-D-GLUCANS IN INDOOR ENVIRONMENTS -

LABORATORY ANALYSIS AND WHEEZE IN INFANTS

A dissertation submitted to the PhD committee members

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSPOHY

in the Department of Environmental Health

of the College of Medicine

November, 2006

Yulia Yulianova Iossifova

M.D. Medical University of Varna, Bulgaria, 2003

Committee Chair: Tiina Reponen, Ph.D.

ABSTRACT Indoor exposure to fungi has been associated with respiratory symptoms, often attributed to their major cell wall component, (1-3)-b-D-glucan. Currently there are two methods available for the analysis of (1-3)-b-D-glucan: the Limulus Amebocyte Lysate assay (LAL) and the inhibition Enzyme Immunoassay (EIA). (1-3)-b-D-glucan is a fungal cell wall component, suspected to cause respiratory and general symptoms in adults. However, very little is known on the possible health effects of (1-3)-b-D-glucan during infancy.

The first aim of this research was to compare the specificity of the LAL vs. EIA methods in detecting eight alpha and beta-glucan standards, and their sensitivity for the analysis of (1-3)-b-D-glucan content of common indoor fungal species and indoor dust samples. It was also examined which indoor mold species predict (1-3)-b-D-glucan concentration in field dust samples, and thus whether (1-3)-b-D-glucan can be used as a surrogate for mold exposure. The second aim was to asses the association between (1-3)- b-D-glucan exposure and the prevalence of allergen sensitization and wheezing during the first year of life in a birth cohort of 574 infants born to atopic parents.

Common indoor fungal species were cultured from pure ATCC strains on agar media, and analyzed for (1-3)-b-D-glucan content by both the LAL and EIA. (1-3)-b-D- glucan exposure was also measured in settled dust collected from infants’ primary activity rooms using LAL. The primary outcomes at approximately age one included parental reports of recurrent wheezing and allergen sensitization evaluated by skin prick testing to a panel of 15 aeroallergens.

iii

This study revealed that most prevalent species in indoor environments, such as

Cladosporium and Aspergillus species were the main (1-3)-b-D-glucan contributors followed by Epicoccum nigrum, Wallemia sebi and Penicillium brevicompactum. In contrast, Alternaria alternata did not contribute much to the (1-3)-b-D-glucan load.

Exposure to high (1-3)-b-D-glucan concentration was associated with reduced likelihood of both recurrent wheezing and recurrent wheezing combined with allergen sensitization.

Similar trends were found between (1-3)-b-D-glucan concentrations and allergen sensitization.

EXECUTIVE SUMMARY

Indoor exposure to fungi has been associated with respiratory symptoms, often attributed to their major cell wall component, (1-3)-b-D-glucan. This and the ease and low cost of performing (1-3)-b-D-glucan analysis rather than cultivation or microscopic counting of mold spores, has prompted many to use (1-3)-b-D-glucan as a surrogate for mold exposure. Therefore, (1-3)-b-D-glucans, which comprise up to 50% of the fungal cell wall, may be a better predictor for health risk than the commonly used determination of viable fungal spores. (1-3)-b-D-glucan is suspected to cause respiratory and general symptoms in adults. However, very little is known on the possible health effects of (1-3)- b-D-glucan during infancy. Currently there are two methods available for the analysis of

(1-3)-b-D-glucan. One method is based upon the bioactivity of this molecule in the factor-G-mediated Limulus coagulation pathway - the Limulus Amebocyte Lysate assay

(LAL). The other method is based on (1-3)-b-D-glucan antigen-antibody reaction - the inhibition Enzyme Immunoassay (EIA).

The first objective of this dissertation was to compare the sensitivity, specificity and accuracy of these two methods, LAL and EIA, in measuring fungal (1-3)-ß-D-glucan, as determined against known types and quantities of glucan standards and mold spores, as well as in real field samples.

The second objective was to investigate whether (1-3)-ß-D-glucan concentrations in indoor dust are associated with the concentrations of common indoor fungal species, as well as with the development of allergic sensitisation and wheeze in infants at age one.

In the first objective, the specificity, sensitivity and accuracy of these two methods in detecting three alpha-glucan (linear and branched), two linear (1-3)-b-D- glucan and three (1-3)(1-6)-b-D-glucan standards of increasing branching were tested. In addition, twelve fungal species (two Cladosporium species, five Aspergillus species,

Aureobasidium pullulans, Penicillium brevicompactum, Epiccocum nigrum, Wallemia sebi, and Stachybotrys chartarum) were cultured from pure ATCC strains on agar media.

Seventy dust and 40 air samples were also collected from residential buildings. All samples were analyzed for (1-3)-b-D-glucan content by both the LAL (GlucatellTM,

Associates of Cape Cod, East Falmouth, MA) and the EIA assays (antibody: mouse IgG, kappa light; Biosupplies Australia, Parkville Victoria, Australia).

It was found that the LAL assay is more sensitive, specific and accurate in measuring both linear and branched (1-3)-b-D-glucans than the EIA assay. Of the indoor fungal species tested, E. nigrum was the species of the greatest (1-3)-ß-D-glucan content per spore (241 pg/spore), mainly due to having also the largest spore size (28 mm).

Although several samples were below the detection limit of the EIA assay, the biomass- normalized (1-3)-ß-D-glucan content measured by both assays was within similar range

(LAL: 0.003 to 146.33 pg/mm2, 0.22-240.54 pg/mm3; EIA: 0.04 – 197.00 pg/mm2, 0.03-

300 pg/mm3). Although both assays determined different fungal species as major contributors to (1-3)-b-D-glucan in field samples, there was a strong significant correlation between LAL-and EIA-analyzed (1-3)-b-D-glucan concentrations in both indoor dust and air samples. Furthermore, the (1-3)-b-D-glucan content per spore measured by the LAL correlated with the fungal spore size and the respective prevalence

and concentration in indoor dust samples. Therefore, LAL-analyzed (1-3)-b-D-glucan concentration in dust samples may be used as a measure of total mold.

Based on the above results, the second objective was to examine which indoor mold species predict LAL-analyzed (1-3)-b-D-glucan concentration in field dust samples, and thus whether or not (1-3)-b-D-glucan can be used as a surrogate for total mold or specific mold genera exposure. The PCR (polymerase chain reaction) method was used to analyze 36 indoor fungal species in 297 indoor dust samples, which were also simultaneously analyzed for (1-3)-b-D-glucan concentration using the LAL assay. Linear regression analysis, followed by factor analysis and structural equation modeling, were utilized in order to identify fungal species that mostly contribute to the (1-3)-b-D-glucan concentrations in field dust samples.

The study revealed that Cladosporium and Aspergillus species were the main (1-

3)-b-D-glucan contributors. In addition, Epicoccum nigrum, Wallemia sebi and

Penicillium brevicompactum, also contributed substantially to the (1-3)-ß-D-glucan concentration in dust samples. Another finding of the study is that the species that contributed most to the (1-3)-ß-D-glucan concentration were also the ones that were most prevalent in indoor environments. However, Alternaria alternata, the third most common fungal species in indoor dust, did not seem to be a significant source of (1-3)-ß-D-glucan.

It can be speculated whether or not this was due to low extraction efficiency, or that allergic effects of Alternaria are not determined by the (1-3)-ß-D-glucan content.

Finally, the association between LAL-analyzed (1-3)-b-D-glucan exposure and the prevalence of allergen sensitization and wheezing during the first year of life in a birth cohort of 574 infants born to atopic parents was examined. For this purpose, (1-3)-

vii

b-D-glucan and endotoxin concentrations were measured in settled dust collected from the infants’ primary activity rooms. The primary outcomes at approximately age one included parental reports of recurrent wheezing and allergen sensitization evaluated by skin prick testing to a panel of 15 aeroallergens as well as milk and egg white.

Exposure to high (1-3)-b-D-glucan concentration (within 4th quartile) was found to be associated with reduced likelihood of both recurrent wheezing [adjusted Odds Ratio

(aOR) = 0.39; 95% CI=0.16-0.93] and recurrent wheezing combined with allergen sensitization (aOR=0.13; 95% CI=0.03-0.61). Similar trends were found between (1-3)- b-D-glucan concentrations and allergen sensitization (aOR=0.57, 95%CI 0.30-1.10). In contrast, recurrent wheezing with or without allergen sensitization was positively associated with low (1-3)-b-D-glucan exposure (within 1st quartile aOR=3.04, 95%CI

1.25-7.38; aOR=4.89, 95%CI 1.02-23.57). There were no significant associations between endotoxin exposure and any of the studied health outcomes.

The analysis of variance of three categorical levels of visible mold exposure (no, low, and high) showed no significant overall differences for the (1-3)-b-D-glucan exposure (respective geometric mean values for concentrations: 53.2, 57.4, and 49.7 mg/g, and for loadings: 17.2, 19.0, and 25.7 mg/m2). The test of linear trend between increasing levels of (1-3)-b-D-glucan and visible mold was not significant either (both units). (1-3)-b-D-glucan concentrations more likely reflect exposure from multiple environmental sources of (1-3)-b-D-glucan, including mold, pollen, plants, and their fragments. Therefore, the concentration of (1-3)-b-D-glucan seems to be a measure of biological exposure that is independent from observed visible mold.

viii

In conclusion, the LAL assay is more sensitive and specific in detecting both linear and branched (1-3)-b-D-glucans than the EIA assay. It is also more accurate than the EIA assay. Although the (1-3)-b-D-glucan concentrations in field samples measured by the LAL and EIA assay correlated, data shall be analyzed with caution, as assays give different weight to different fungal species. In addition, the strong significant associations between the LAL-analyzed (1-3)-b-D-glucan and fungal spore size, and indoor dust fungal concentration, indicates that LAL-analyzed (1-3)-b-D-glucan could be used to estimate the total fungal load in indoor samples.

This is the first study to report that indoor exposures to high levels of (1-3)-b-D- glucan (=61µg/g, =19µg/m2), but not endotoxin, are associated with decreased risk for recurrent wheezing among infants at age one. By following this cohort, it will be determined if (1-3)-b-D-glucan is the strongest predictor for the Hygiene Hypothesis, stating that early exposure to microbial components have a protective effect against atopy development.

ACKNOWLEDGEMENTS

First and foremost, I would like to express my sincere gratitude to Dr. Tiina Reponen, for not only being a self-giving mentor, teacher, and advisor, but above all, for being as a mother and friend to me. She was always there when needed for help, advice, or just to listen to me. Not to mention what a great example of an outstanding researcher and person she is. Completing my PhD research within such a short term was not going to be possible without her.

Second, I would like to thank Dr. and Mrs. Clark and Dr. Grinshpun, for all the moral support, kindness, and advices in areas of both professional development and personal adjustment to life in a new country and surviving the PhD education.

My committee members were all the most helpful and kind people anyone can wish to have on their committee. Dr. Linda Levin was wonderful to spend so much time on helping with the statistical analysis. She is also a great teacher, and I can honestly say that all the statistics I know today is thanks to her and Dr Paul Succop. Dr. Gurjit Hershey and

Dr. Mike Daines, from the Cincinnati Children Hospital, were extremely generous to provide me not only with the resources to conduct the immuno-assay, but were also very kind and patient teachers. This PhD research was not going to be possible without Dr.

LeMasters, who not only let me take part in the CCAPPS study, but also kept me focused on the practical implications of the research.

I would also like to thank the numerous students that helped me through my PhD endeavor. Students and fellows, like Heidi Sucharew, Paloma Campo, Carlos Crawford,

Heidi Lenhart, Haoyue Lee, who dedicated their time to teaching me and helping me with my research. My classmates, Danius Martuzevicius, Shu-an Lee, Susan Kotowski,

Kristen Hahn, Kris Adkins, Nancy Hopf, Rob Eninger, Seung-Hyun Cho, and many others, with whom I have lots of great memories.

Most of all, I would like to dedicate this research to my family, back in Bulgaria and

Turkey, and my host family in the USA, who were the people always to cheer me up and not let me give up.

And last, but not least, I would like to thank all the families that participated in the

CCAAPS study, all researchers, students and master-minds behind this project. If this research has added even one drop to the vast ocean of knowledge that can change our lives to better, it would be the greatest accomplishment in one’s life.

"Personal self satisfaction is the death of the scientist. Collective self-satisfaction is the death of the research. It is restlessness, anxiety, dissatisfaction, agony of mind that nourish science."

Jacques Monod (Nobel Laurete in Medicine or Physiology)

xii

LIST OF PEER REVIEWED PUBLICATIONS

I. Iossifova Y, Reponen T, Daines M, Crawford C, Hershey GK. Comparison of

EIA and LAL analytical methods for detecting (1-3)-ß-D-glucan in pure fungal cultures.

Appl Environ Microb (to be submitted in November, 2006) [Specific aims 1 and 2]

II. Iossifova Y, Sucharew H, Succop P, Vesper S, Reponen T. Use of (1-3)-b-D- glucan concentrations in dust as a surrogate method for estimating specific mold exposures. Indoor Air (submitted) [Specific aim 3]

III. Iossifova Y, Reponen T, Bernstein D, Levin L, Kalra H, Campo P, Zeigler H,

Villareal M, Lockey J, Khurana-Hershey G, LeMasters G. House dust (1-3)-b-D-glucan and wheezing in infants? Allergy (revision submitted) [Specific aim 4]

The full texts of peer-reviewed publications are attached in Appendices A1 through A3.

xiii

Table of Contents

ABSTRACT...... iii

EXECUTIVE SUMMARY...... v

ACKNOWLEDGEMENTS ...... xi

LIST OF PEER REVIEWED PUBLICATIONS...... xiii

LIST OF FIGURES ...... xix

LIST OF TABLES ...... xx

HYPOTHESES AND SPECIFIC AIMS...... 1

BACKGROUND ...... 3

General Overview:...... 3 Composition and sources of (1-3)-b-D-glucan...... 3 Health effects ...... 5 Analysis of (1-3)-ß-D-glucan vs. mold...... 6 Rationale for specific aims 1 and 2:...... 7 Rationale for specific aim 3:...... 10 Rationale for specific aim 4:...... 12

SPECIFIC AIM 1...... 15

COMPARE LAL AND EIA ANALYZED (1-3)-ß-D-GLUCAN

CONCENTRATIONS IN PURIFIED GLUCAN STANDARDS AND IN REAL

FIELD SAMPLES ...... 15

1.1. INTRODUCTION ...... 15 1.2. MATERIALS AND METHODS...... 16 1.2.1. Laboratory analysis of (1-3)-ß-D-glucan...... 16 1.2.2. Assays specificity in detecting branched (1-3)-b-D-glucans ...... 17 1.2.3. Field samples ...... 18 1.2.4. Data analysis ...... 19 1.3. RESULTS ...... 20 1.3.1. Specificity and Accuracy of LAL and EIA in measuring glucan of different linkage and branching...... 20 1.3.2. Comparison of LAL vs. EIA analyzed (1-3)-ß-D-glucan in dust and air samples...... 22 1.4. DISCUSSION...... 23

xiv

1.5. CONCLUSIONS FOR SPECIFIC AIM 1...... 25

SPECIFIC AIM 2...... 26

COMPARE LAL- AND EIA- ANALYZED (1-3)-ß-D-GLUCAN CONTENT OF

COMMON INDOOR MOLD SPECIES...... 26

2.1. INTRODUCTION ...... 26 2.2. MATERIALS AND METHODS...... 26 2.2.1. Laboratory analysis of (1-3)-ß-D-glucan...... 26 2.2.2. Fungal species selection...... 27 2.2.3. Preparation of pure fungal species...... 27 2.2.4. Field samples ...... 29 2.2.5. MSQPCR analysis of dust ...... 29 2.2.6. Data analysis ...... 30 2.3. RESULTS ...... 31 2.3.1. Spore characteristics ...... 31 2.3.2. (1-3)-ß-D-glucan content in fungal spores...... 31 2.3.3. Associations between (1-3)-ß-D-glucan content in fungal spores and concentrations in field dust samples ...... 32 2.3.4. Estimated vs. predicted (1-3)-ß-D-glucan concentrations in field samples.... 33 2.4. DISCUSSION...... 35 2.5. CONCLUSIONS FOR SPECIFIC AIM 2...... 40

SPECIFIC AIM 3...... 41

COMPARE LAL-ANALYZED INDOOR (1-3)-ß-D-GLUCAN

CONCENTRATIONS WITH MOLD SPORE CONCENTRATIONS AS

ANALYZED BY THE PCR METHOD IN DUST SAMPLES...... 41

3.1. INTRODUCTION ...... 41 3.2. MATERIALS AND METHODS...... 42 3.2.1. On-site home visit and exposure assessment ...... 42 3.2.2. MSQPCR analysis of dust ...... 42 3.2.3. (1-3)-b-D-glucan analysis in dust...... 42 3.2.4. Linear Regression Analysis ...... 42 3.2.5. Factor analysis and Structural Equation Modeling...... 43 3.3. RESULTS ...... 45 3.3.1. Linear Regression Analysis ...... 45 3.3.2. Factor analysis and Structural Equation Modeling...... 47 3.4. DISCUSSION...... 47 3.5. CONCLUSIONS FOR SPECIFIC AIM 3...... 51

SPECIFIC AIM 4...... 52

STUDY THE MAGNITUDE AND DIRECTION OF THE ASSOCIATION

BETWEEN (1-3)-ß-D-GLUCAN AND THE DEVELOPMENT OF WHEEZE AND

ALLERGEN SENSITIZATION IN INFANTS...... 52

4.1. INTRODUCTION ...... 52 4.2. MATERIALS AND METHODS...... 53 4.2.1.Recruitment ...... 53 4.2.2.Exposure assessment...... 54 4.2.3. (1-3)-b-D-glucan and Endotoxin Analysis ...... 54 4.2.4. Medical evaluation of infants...... 55 4.2.5. Data analysis ...... 56 4.3. RESULTS ...... 58 4.3.1. Exposure and subject characteristics...... 58 4.3.2. Infantile wheezing...... 59 4.3.3. Allergen sensitization...... 63 4.4. DISCUSSION...... 64 4.5. CONCLUSIONS FOR SPECIFIC AIM 4...... 71 OVERALL CONCLUSIONS (PRACTICAL IMPLICATIONS)...... 72

FUTURE DIRECTIONS...... 74

REFERENCES: ...... 76

LIST OF APPENDICES ...... 118

APPENDIX A: COPIES OF PEER-REVIEWED PUBLICATIONS RESULTED

FROM THE PhD STUDY...... 120

A1: Comparison of EIA and LAL analytical methods for detecting (1-3)-ß-D-glucan in pure fungal cultures and in home dust samples ...... 121 A2: Use of (1-3)-b-D-glucan concentrations in dust as a surrogate method for estimating specific mold exposures ...... 162 A3: House dues (1-3)-b-D-glucan and wheezing in infants ...... 186

APPENDIX B: LIST OF OTHER PUBLICATIONS (NOT INCLUDED IN THE

PhD DISSERTATION) AUTHORED/CO-AUTHORED BY Ms. YULIA

xvi

IOSSIFOVA DURING HER GRADUATE STUDY IN THE UNIVERSITY OF

CINCINNATI...... 226

B1: Peer-Reviewed Publications: ...... 227 B2: Conference Proceedings and Abstracts:...... 228

C. LAL PROTOCOLS FOR (1-3)-b-D-GLUCAN ANALYSIS ...... 229

C1: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – DUST SAMPLES – ENDPOINT ASSAY ...... 230 C2: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – DUST SAMPLES – KINETIC ASSAY ...... 235 C3: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – AIR SAMPLES.. 240 C4: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – PURE FUNGAL SPORES SUSPENSIONS ...... 244

APPENDIX D: EIA PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS...... 249

xvii

xviii

LIST OF FIGURES Fig. 1-1 Comparison of eight purified glucans and their reactivity as measured by the kinetic LAL (A) and endpoint EIA (B) assays.

Fig. 1-2 LAL vs. EIA-analyzed (1-3)-ß-D-glucan in 70 dust samples (CCAAPS study) in mg/g.

Fig. 1-3 LAL vs. EIA-analyzed (1-3)-ß-D-glucan in 70 dust samples (CCAAPS study) in mg/m2.

Fig. 2-1 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding spore size (geometric mean).

Fig. 2-2 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding frequency (%) in indoor dust samples

(PCR analysis).

Fig. 2-3 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding concentration (geometric mean of cells/mg) in indoor dust samples (PCR analysis).

Fig. 4-1. Smoothed plot of the adjusted prevalence rates of recurrent wheezing in relation to the log-transformed (1-3)-b-D-glucan concentration (solid lines). Dotted lines represent ± 1 standard error (SE).

xix

LIST OF TABLES Table 1-1. Comparison of eight purified glucans and their relative reactivity as measured by the LAL and EIA assays. The standard against which the concentrations are measured is Pachyman for the LAL assay, and Laminarin for the EIA assay.

Table 2-1. Fungal species selected for this study and their frequency and concentration (geometric mean), as measured by PCR analysis of 297 dust samples

(Cincinnati Childhood Allergy and Air Pollution Study).

Table 2-2. Characteristics of fungal species – spore size, spore surface area and spore volume (average of n=30 spores for each fungal species).

Table 2-3. Average (1-3)-ß-D-glucan contents of twelve common indoor fungal species as measured by the LAL and EIA assay (picograms per spore, picograms per spore surface area, and picograms per spore volume).

Table 2-4. Contribution of predicted fungal (1-3)-ß-D-glucan (pg/mg dust) based on LAL and EIA-analyses and concentration of fungal species (total in all 297 dust samples, cells/mg dust).

Table 3-1. Frequency and concentration of 36 indoor fungal species analyzed with the PCR method and correlations between the concentration of each species and the cumulative (1-3)-b-D-glucan concentration (n = 297 dust samples).

Table 3-2. Fungi associated with the indoor (1-3)-ß-D-glucan concentration.

Table 3-3. Final 6-factor SEM model with Loadings. Only Factor 1 and 3 were significant, and thus are reported.

Table 4-1. Geometric mean (GM), geometric standard deviation (GSD) and interquartile (IQ) range of (1-3)-ß-D-glucan and endotoxin concentration (mg/g, EU/mg)

and loading (mg/m2, EU/m2), measured in homes of 574 infants.

Table 4-2. Characteristics of predictor variables and prevalence and percent of infants reporting health outcome* by levels of (1-3)-b-D-glucan, endotoxin, demographic characteristics. Numbers in brackets for each predictor variable represents the number of infants that fall in each class of that predictor variable. Cell entries in the health outcome columns are number of subjects reporting outcome (% of column total).

xxi

HYPOTHESES AND SPECIFIC AIMS

Indoor exposure to fungi has been associated with adverse respiratory effects in adults, often attributed to their major cell wall component, (1-3)-b-D-glucan. This and the ease and low cost of performing (1-3)-b-D-glucan analysis rather than cultivation or microscopic counting of mold spores, has prompted many to use (1-3)-b-D-glucan as a surrogate for mold exposure. Currently, there are two methods available for the analysis of (1-3)-b-D-glucan: the Limulus Amebocyte Lysate assay (LAL) and the inhibition

Enzyme Immunoassay (EIA). (1-3)-b-D-glucan is suspected of causing respiratory and general symptoms in adults. Very little is known on the possible health effects of (1-3)-b-

D-glucan during infancy. Infancy is a time when immune responses may be modified by exposure to microbial products, such as endotoxins. In addition, two recent European studies found increased dust-borne (1-3)-b-D-glucan concentrations to have a slight protective effect on atopic wheeze in school children, and both asthma and persistent wheeze in children at age 1-4 (see Background).

The current PhD thesis studies two important aspects of (1-3)-ß-D-glucan in indoor environments – their laboratory analysis and respiratory health effects in infants.

The first hypothesis is that the LAL assay is a more sensitive, specific, and accurate method for estimating (1-3)-ß-D-glucan mold exposure than the EIA, as measured against known types and quantities of glucan standards and mold spores, as well as in field samples.

The following specific aims were performed in order to test the first hypothesis:

Specific aim 1: Compare LAL and EIA analyzed (1-3)-ß-D-glucan concentrations in purified (1-3)-ß-D-glucan standards {linear (1-3)-ß-D-glucan, and branched [(1-3)(1-6)]-ß-D-glucan standards} and in real field samples.

Specific aim 2: Compare paired differences and measures of association between

LAL and EIA analyzed (1-3)-ß-D-glucan contents of common indoor mold species.

The second hypothesis is that (1-3)-ß-D-glucan concentrations in indoor dust are positively associated with concentrations of common indoor fungal species and negatively associated with the development of allergic sensitisation and wheeze in infants at age one. Based on the findings of the specific aims 1 and 2, the assay found to be more sensitive, specific and accurate, will be used for the (1-3)-ß-D-glucan analysis in specific aims 3 and 4.

The following specific aims were performed in order to test the second hypothesis:

Specific aim 3: Compare indoor (1-3)-ß-D-glucans concentrations with mold spore concentrations (both the total spore concentration and the concentration of individual fungal species) as analyzed by the PCR method in dust samples collected from children’s homes.

Specific aim 4: Study the magnitude and direction of the association between (1-

3)-ß-D-glucan and the development of wheeze and allergic sensitization in infants.

BACKGROUND

General Overview:

Composition and sources of (1-3)-b-D-glucan

Moisture damage is common around the world and in all kinds of buildings such as homes, schools, offices, and hospitals. Exposure to fungi in occupational and residential environments is associated with respiratory (nose and throat irritation, cough) and general (lethargy and headache) symptoms, allergic reactions and organic dust toxic syndrome (Storey et al., 2004; IOM, 2004). Similar general and respiratory symptoms and airways inflammation are reported in occupational and other indoor exposure to (1-

3)-b-D-glucan, polyglucose component of fungi, pollen and some bacteria (Rylander and

Lin, 2000; Douwes et al., 2005).

(1-3)-b-D-glucan is a biologically active polyglucose molecule comprising up to

60% of the cell wall of mold, some soil bacteria and plants. It retains its toxicity after the death of the organism (Stone and Clark, 1992) and has been used in several studies as a surrogate of mold exposure (Young and Castranova, 2005; Rylander R., 2000).

In the fungal cell wall, glucans comprise a three-dimensional network of (1-3)- and (1-6)-ß-D-linked anhydroglucose repeat units that are connected to other carbohydrates, proteins and lipids. The backbone of the (1-3)-ß-D-glucan polymer is composed of glucose subunits connected by intra-chain glycosidic (1-3)-ß-linkages. In a branched (1-3)-ß-D-glucan polymer, the branches are connected by (1-6)-ß-linkages (as shown in the Fig. 1 on next page). In the figure, the side-chain branches are single glucose subunits. (1-3)-ß-D-glucan polymers can exist as a single polymer strand with a

helical conformation (single helix) or as a stable complex of three polymer strands forming a triple helix. The triple helical form is generally considered to be the preferred form in nature (Young and Castranova 2005).

A) Linear (non-branched) fungal (1-3)-ß-D-glucan polymer

CH2OH CH2OH CH2OH O O O

O O

OH OH OH n

OH OH OH

B) Branched fungal (1-3)(1-6)-ß-D-glucan polymer

CH OH 2 O

OH O

CH OH CH2OH CH2 2 OH O O O O O

OH OH OH

Fig.1. Primary structure of a single, non-branched and a single branched (1-3)-ß-D- glucan polymer.

(1-3)-ß-D-glucans are non-allergenic water-insoluble structural cell wall components of most fungi, some bacteria, most higher plants and many lower plants, where they have storage, structural or protective roles (Stone and Clark, 1992). The (1-3)-

ß-D-glucan-containing bacteria are soil inhabitants (Alcaligenes faecalis, Agrobacterium rhizogenes, A. radiobacter), Streptococcus pneumoniae and Anabaena cylindrica. These bacteria are not representative of the indoor bacteria and thus are not expected to interfere as a source of indoor (1-3)-ß-D-glucans (Stone and Clark, 1992). Lichens and higher plants are another source of (1-3)-ß-D-glucans where the main linkages can be either (1-

3) or (1-4) depending on the species. Among the non-fungal components listed above, only pollen is found in significant concentrations in air. Pollen, however, has (1-4)- linkage of the main chain (Rylander et al., 1999e). Glucans containing (1-3)- and (1-6)-ß- glucosidic linkages are important components of cell walls and secretions of fungi in classes within the Mastigomycotine, Ascomycotina, Basidiomycotina and

Deuteromycotina. (1-3)-ß-D-glucans are also storage carbohydrates of some fungi

(Mastigomycotina and Basidomycotina) (Stone and Clark, 1992). The (1-3)(1-6)-ß- glucans are usually highly branched.

Health effects

Exposure to indoor molds has long been associated with induction of asthma and allergic rhinitis through IgE-mediated mechanisms, hypersensitivity pneumonitis, and life-threatening primary and secondary infections in immunocompromised patients

(Edmondson et al., 2005; Gent et al., 2002; Belanger et al., 2003). The risk of reported wheezing was increased up to five-fold in homes with visible mold or water damage

(Belanger et al., 2003; Cho et al., 2005). Similar general and respiratory symptoms are also reported for (1-3)-b-D-glucan exposure in adults. The association between (1-3)-b-

D-glucan exposure in occupational and residential environments and respiratory symptoms (nose and throat irritation, cough), airways inflammation and decrease in lung function in adults is well established (Douwes et al., 2005).

The (1-3)-ß-D-glucans are potentially useful as immune potentiators as these are non-toxic, non-antigenic and non-virulent. One study suggested that (1-3)-ß-D-glucan was associated with an increased risk of atopy in people (Thorn and Rylander, 1998), similar as has been observed in some animal studies (Douwes, 2005). This was, however, not confirmed in a smaller study (Rylander, 1999a). An association between (1-3)-ß-D- glucan exposure and an increased T helper 1 cell (Th1) immune response was suggested

(Beijer et al., 2003). The latest appears to be contradictive with the finding of a higher prevalence of atopy in subjects with high (1-3)-ß-D-glucan exposure (atopy is a Th2 driven immune response). An up-regulation of Th1 immune activity was, however, only shown in non-atopic subjects. Finally, some studies suggested that associations between

(1-3)-ß-D-glucan exposure and symptoms and lung function changes were stronger in atopics than in non-atopics (Douwes et al., 2000a, Rylander, 1998b).

Analysis of (1-3)-ß-D-glucan vs. mold

A primary benefit of investigating fungal (1-3)-ß-D-glucan besides its biological properties is the ease and cost of assessment. Accurate assessment of mold exposure is difficult with currently available sampling and analysis methods. No single measurement technique is entirely suitable, and sampling should never be conducted alone but in

conjunction with inspection. A measure of culturable molds (as colony-forming units) represents only a small fraction of the total number of mold spores present in air, dust, or bulk samples (Rylander and Etzel, 1999b). On the other hand, (1-3)-ß-D-glucan retains its toxicity after the death of the organism (Fogelmark and Rylander, 1997), which makes it a more sensitive indicator of mold exposure than the traditional culturable count. Spore trap or filter sampling with microscopy counting of spores can also provide a measure of fungal concentration but is laborious and requires considerable experience (Rylander and

Etzel, 1999b). Also in large population-based studies analysis by culturable-count and microscopic count is not feasible, because of the considerable time and labor required.

Another advantages of studying (1-3)-ß-D-glucan exposure rather than mold exposure are that it is a less time consuming analysis (Douwes et al., 2003a).

Rationale for specific aims 1 and 2:

Currently there are two methods available for the analysis of (1-3)-ß-D-glucan, as described in Table 1.

One method is based upon the bioactivity of this molecule in the factor-G- mediated Limulus coagulation pathway - the Limulus Amebocyte Lysate assay (LAL)

(Obayashi et al., 1985). The other method is based on (1-3)-?ß-D-glucan antigen-antibody reaction - the inhibition Enzyme Immunoassay (EIA) (Douwes et al., 1996).

Table 1. Comparison of LAL vs. EIA

Assay LAL-assay EIA-assay

Type of reaction Activation of Factor G Antibody-antigen inhibition

coagulation enzyme reaction

Samples analyzed in previous Air samples Air and dust samples studies

Lower Limit of Detection 1 pg/ml 40 ng/ml

Glucan linkages detected Detects both linear & Detects only (1-3)-ß-D-glucan

branched (1-3)-and (1-6)-ß-D- linkages

glucan linkages

Both methods are used for the analysis of air samples, but only EIA assay has previously been used for dust samples. The LAL method is much more sensitive with a

Lower Limit of Detection (LOD) = 1 pg/ml compared to EIA, which has LOD = 40 ng/ml, which limits the EIA assay to settled dust and air samples collected from high- exposure environments only.

The EIA assay developed by Douwes et al. (1996) is reported to be highly specific to linear (1-3)-ß-D-glucan only (Douwes et al., 1996). Other modifications of the

EIA assay were developed by Adachi et al. (1994) and Milton et al. (2001), which are highly specific to the measurement of the branched (1-3)(1-6)-b-D-glucans only. On the other hand, the LAL is extremely sensitive (LOD=1pg/ml) assay suggested to recognize both linear and branched (1-3)-(1-6)-ß-D-glucan found in fungi (Tanaka et al., 1991;

Thorne et al., 2004). Its ability to detect both linear and branched (1-3)-b-D-glucans, as

well as -mannan in previous studies (Tanaka et al., 1991) were viewed as disadvantage indicating low specificity. However, studies have shown that both linear and branched (1-

3)-b-D-glucans are recognized by the toll like receptors and are important for the activation of macrophages (Kataoka et al., 2002; Brown and Gordon, 2005; Sakurai et al.,

1997).

While there are some data on the content of (1-3)-b-D-glucans in spores of the indoor fungal species of Penicillium, Aspergillus, Cladosporium and Stachybotrys

(Fogelmark and Rylander, 1997; Foto et al., 2004), analyzed by the LAL assay, very little is known on the EIA-analyzed (1-3)-b-D-glucan of spores from different species.

LAL-analyzed (1-3)-b-D-glucan is a recognized indicator of mold biomass based on health effects and correlation with total fungal count (Alwis et al., 1999; Rylander et al., 1999cde, Wan and Li, 1999; Mandryk et al., 2000). While EIA-analyzed (1-3)-b-D- glucan in settled dust has also been used as an indicator of mold biomass (Douwes et al.,

1996; Chew et al., 2001; Gehring et al., 2001), there is very little data on any correlation between EIA analyzed air samples and mold counts due to its low sensitivity (i.e., (1-3)- b-D-glucan concentration in most air samples is expected to be below the lower limit of detection). As the LAL assay has been used predominantly for the analysis of (1-3)-b-D- glucan concentration of air samples, while the EIA has been used for dust samples, there is a need for a comparison of these two methods by using unified source of environmental samples. Due to the low sensitivity of the EIA method, such a unified source for analysis by both assays is indoor dust samples (not air samples).

Thus the proposed study will compare for the first time the ability of LAL vs. EIA methods to assess (1-3)-ß-D-glucan content in common indoor mold spores and

concentration in field samples. It will also provide data on the specific branching structure of (1-3)-ß-D-glucans that each assay particularly detects.

Rationale for specific aim 3:

LAL-analyzed air (1-3)-ß-D-glucan is a recognized indicator of mold biomass based on health effects and correlation with total fungal count (Rylander 1999a, Alwis et al., 1999, Wan and Li 1999, Mandryk et al. 2000). Associations between EIA-analyzed

(1-3)-ß-D-glucan in settled dust and the total culturable mold count were found by

Douwes et al. (1996, 1998), Chew et al. (2001) and Gehring et al. (2001). Chew et al.

(2001) found out that (1-3)-ß-D-glucan concentrations were associated with the concentration of total culturable fungi (CFU/g dust, CFU=Colony forming units), with an increase in CFU/g from the 25th to the 75th percentile associated with a 1.3 (1.1-1.6)-fold increase in (1-3)-ß-D-glucan. Levels of (1-3)-ß-D-glucan were also positively associated with CFU of Aspergillus versicolor, Eurotium repens and Wallemia spp., and with the combined total amount of Penicillium and Aspergillus. In addition, Douwes et al. (1996) found (1-3)-ß-D-glucan levels to be positively associated with the CFU of Alternaria.

Gehring et al. (2001) found increased (1-3)-b-D-glucan loading in homes with visible mold compared to homes without. Interestingly, when concentrations were expressed as micrograms per gram of dust instead of micrograms per square meter, no statistically significant correlation was found between (1-3)-b-D-glucan levels and concentrations of total CFU or CFU of individual mold genera. These findings of positive associations

between (1-3)-ß-D-glucan and total culturable fungi in house dust suggest that both LAL and EIA-analyzed (1-3)-b-D-glucan may be used as a marker for indoor fungal exposure.

Previous studies have shown that (1-3)-b-D-glucan content varies between mold species of different genera, as well as within same genus (Foto et al., 2004; Fogelmark and Rylander, 1997). The variability of the (1-3)-b-D-glucan content among fungal species may lead to variance of health outcomes by fungal genera. A recent study

(Osborne et al., 2006) has found inverse association between the concentration of airborne Cladosporium and allergen sensitization (p < 0.05), but positive associations between Penicillium/Aspergillus (p < 0.01), or Alternaria (p<0.01) and allergen sensitization. The above finding may be due to different (1-3)-b-D-glucan content. For example, it is reported that patients with deep mycosis such as candidasis, aspergillosis, trichosporosis and carinii pneumonia but not mucormycosis or cryptococcosis, have serum (1-3)-b-D-glucan concentrations positively correlated with clinical symptoms and pathological changes (Yoshida et al., 1997a; Miyazaki et al., 1995; Yuasa et al., 1996).

As the biological properties of (1-3)-b-D-glucan are not dependent on viability, it is also important to identify and quantify both viable and non-viable indoor fungi. A promising approach for this is the Polymerase Chain Reaction (PCR), which unlike the

CFU count detects both viable and non-viable fungi, and unlike microscopic counting is less time consuming and more precise in identifying fungi up to species level.

The proposed investigation will examine which indoor mold species contribute most to the (1-3)-b-D-glucan content in field dust samples, and thus whether (1-3)-b-D- glucan can be used as a surrogate for either total mold exposure or whether it rather predicts the exposure to specific mold species.

Rationale for specific aim 4:

Animal data on the effects of exposure to (1-3)-b-D-glucan on respiratory responses are controversial. Experimental exposures of mice, guinea pigs and adult human subjects to particulate (1-3)-b-D-glucan, the most common form of environmental

(1-3)-b-D-glucan, failed to elicit inflammatory responses in the nasal or bronchial airways (Fogelmark et al., 1994; Korpi et al., 2003; Rylander, 1996). On the other hand, soluble (1-3)-b-D-glucan has been shown to enhance allergen induced airway inflammation by increasing eosinophil infiltration and specific IgE in guinea pigs and mice sensitized to ovalbumin (Rylander and Holt, 1998a; Ormstad et al., 2000).

Population studies have shown a positive association between occupational exposure to (1-3)-b-D-glucan and general (fatigue, headache) and respiratory symptoms

(nose and throat irritation, cough), airways inflammation and lung function in the following workplaces: organic dusts in wood-work and paper industry (0.6-97.7 ng/m3)

(Alwis et al., 1999; Mandryk et al., 2000; Rylander et al., 1999d); household waste collection (9.2-52 ng/m3) (Thorn, 2001; Heldal et al., 2003), poultry rearing (0.01-70 ng/m3) (Rylander et al., 2006), composting (0.36-4.85 mg/m3) (Douwes et al., 2000a), and sewage treatment (4.8-40 ng/m3) (Rylander, 1999c; Gladding et al., 2003). Even though indoor exposures are lower, similar symptoms have been found in adult populations exposed to indoor environments with elevated (1-3)-b-D-glucan (0.2-15.3 ng/m3)

(Rylander, 1997; Wan and Li, 1999; Wouters et al., 2002; Thorn and Rylander, 1998).

endotoxin during infancy may have a protective effect on subsequent aeroallergen sensitization in childhood (von Mutius et al., 2000; Gereda et al., 2000; Belanger et al.,

2003). Exposure to (1-3)-b-D-glucan in children may have a similar protective effect

(Schram-Bijkerk et al., 2005). Although no previous studies have linked mold exposure to protective effects (Douwes and Pearce, 2003b), a recent study has shown an inverse relationship between exposure to Cladosporium and allergen sensitization to any allergen

(p < 0.05), as well as to aeroallergens (p < 0.05) in infants (Osborne et al., 2006). There is limited epidemiological data on health effects of (1-3)-b-D-glucan exposure in children. Rylander et al. (1998b) reported that exposure to increased airborne (1-3)-b-D- glucan (>15.3 ng/m3) positively correlated with upper airway symptoms in atopic school children (6-13 years old). Two recent European studies found increased dust-borne (1-3)- b-D-glucan concentrations to have a slight protective effect on atopic wheeze (1.2 fold decrease; p<0.10) in school children (5-13 years old) (Schram-Bijkerk et al., 2005), and both asthma (aOR 0.70, 95% CI 0.30-1.60) and persistent wheeze (aOR 0.43, 95% CI

0.15-1.21) in children at age 1-4 (Douwes et al., 2006).

As described above, existing data on health effects of (1-3)-b-D-glucan are contradictory. No studies todate have assessed (1-3)-b-D-glucan exposure on clinical outcomes in an infant birth cohort. The hypothesis of aim 4 is that the prevalence of wheezing and allergen sensitization in high-risk infants is inversely related to exposure to high indoor concentrations of (1-3)-b-D-glucan.

Overall, very little si known how the LAL results compare to the EIA results.

Also, there is no standard procedure developed for the LAL test in settled dust samples.

Currently it is difficult to conclude on the reliability of (1-3)-ß-D-glucan as a mold

surrogate as data presented in the literature deals primarily with the correlation with the concentrations of culturable count. Furthermore, there is very little data on the effect of

(1-3)-?ß-D-glucan exposure during infancy as a trigger of or protector against respiratory atopy in early lifetime. The current PhD thesis provides answers to these questions.

SPECIFIC AIM 1

COMPARE LAL AND EIA ANALYZED (1-3)-ß-D-GLUCAN CONCENTRATIONS IN PURIFIED GLUCAN STANDARDS AND IN REAL FIELD SAMPLES

1.1. INTRODUCTION

Currently there are two methods available for the analysis of (1-3)-b-D-glucan.

As described in the Background section of the thesis, these are the LAL and EIA assays.

Data on health effects and mold exposures associated with environmental (1-3)-b-D- glucan has been accumulated. However, no study has used both assays simultaneously, and no published study so far has assessed the comparability of these two methods.

Through comparison of sensitivity, specificity and accuracy of these two assays, we can determine which of these two assays is more accurate and useful in future exposure assessments studies. For this purpose, the first aim of the thesis was to compare the specificity, sensitivity and accuracy of LAL and EIA methods through the analysis of

(1-3)-b-D-glucan concentration in purified glucan standards and in real field (dust and air) samples.

1.2. MATERIALS AND METHODS

1.2.1. Laboratory analysis of (1-3)-ß-D-glucan

The LAL test is a quantitative direct method for the detection of (1-3)-b-D- glucans that uses (1-3)-b-D-glucan-sensitive factor G (Obayashi et al., 1985). The kinetic chromogenic Limulus Amebocyte lysate assay {GlucatellTM, Associates of Cape Cod,

East Falmouth, MA} was performed, using laboratory materials (pipette tips, tubes, etc.) certified free of contaminating glucans by the manufacturer (Associates of Cape Cod).

From each sample, 0.5 ml aliquot (or 25 mg of dust) was extracted with 0.5 ml (or 1ml to dust) of 0.6 M NaOH by shaking for 1 hour at room temperature, to unwind the triple- helix structure of (1-3)-b-D-glucan and make it water-soluble. Twenty-five ml of

Glucatell reagent was added to each well of serially diluted (from 1:1 to 1:1011) extract and a control standard (1-3)-b-D-glucan (Pachyman), placed in a 96-well, flat-bottom microplate. The prepared (expected) concentrations of the glucan standards for the LAL assay were: 3.125, 12.50, 50 and 100 pg/ml. The optical density (OD) was read at 405 nm, recorded at time of onset at OD = 0.03. All samples of (1-3)-b-D-glucan were above the lower limit of detection (LOD) of the Glucatell assay protocol (3.125 pg/ml). The median coefficient of variation (CV) was 9% for the intra-plate variability and 27% for the inter-plate variability.

The EIA test was performed as described by Douwes et al. (1996). The primary monoclonal antibody to (1-3)-b-D-glucan was mouse IgG, kappa light (Biosupplies

Australia, Parkville Victoria, Australia). The secondary antibody was peroxidase- conjugated affinipure sheep anti-mouse IgG (H+L) (Jackson ImmunoResearch, West

Grove, PA). The sample extraction was accomplished by heat extraction in an autoclave

at 120oC for 1 hour. The prepared (expected) concentrations of the glucan standards for the EIA assay were 250, 1000, 2500, and 5000 ng/ml. LOD values (250 ng/ml) were divided by the square root of two for the data analyses. For the EIA assay, the median CV was 13.6 % for the intra-plate variability and 24.2 % for the inter-plate variability.

The results of both assays were expressed as pg/ml for the glucan standards, but converted to pg/spore for the spore suspensions. After measuring the spore size and calculating the spore surface area and volume, the results were also converted to pg/mm2 and pg/mm3. Results of the (1-3)-b-D-glucan concentration in dust samples were reported as mg/g of dust and mg/m2of floor area.

1.2.2. Assays specificity in detecting branched (1-3)-b-D-glucans

The following (1-3)-b-D-glucan standards to test the specificity of the LAL and

EIA tests in detecting linear vs. branched (1-3)-D-glucans were used: 1) Pachyman {99% linear (1-3)-b-D-glucan}, 2) Curdlan {linear (1-3)-b-D-glucan}, 3) Laminarin {primarily linear, with some (1-6)- b - interstrand linkages and branch points}, 4) Schizophyllan

{(1-6)-b-branching at every third linkage}, 5) MG-glucan (MacroGardR is extract form baker’s yeast containing 97% (1-3),(1-6)- b -D-glucan), 6) Mannan {(1-6)-a-D-

Mannose}, 7) Dextran {(1-3),(1-6)- a -D-glucose}, and 8) Pullulan{(1-3),(1-4)- a -D- glucose}. The glucan standards were purchased from Sigma Chemical Co. (St. Louis,

MO) except for Pachyman and MG-glucan, which were obtained from Megazyme

International Ireland Ltd (Bray, Ireland) and Nutritional Scientific Corporation (Liberty,

TX), respectively.

1.2.3. Field samples

Field samples were obtained through the Cincinnati Childhood Allergy and Air Pollution

Study (CCAAPS). The CCAAPS is a prospective birth cohort study aimed at investigating the role of aeroallergens and diesel exhaust particles in the development of atopy and atopic respiratory disorders (Ryan et al., 2005; Biagini et al., 2006). As part of the CCAAPS study, when infants reached an average age of 8 months, families were visited at their homes to administer a detailed questionnaire to the parents on home characteristics, including observations of visible mold/water damage. Dust samples were vacuumed from the baby’s primary activity room floor in order to assess exposure to indoor aeroallergens and mold (Cho et al., 2006b). The baby’s primary activity room was defined as the one where the infant spends most of his/her time, and in >90% of cases, it was the living room or the family room. Thus bias due to selection of different types of rooms was minimized. In brief, for carpeted floor, samples were collected from area of 2 m2 at a vacuuming rate of 2 min/m2. For non-carpeted floor only one sample was collected from the entire room at a rate of 1 min/m2. The home dust sample was sieved

(355 µm sieve), and the fine dust was divided into sub-samples and stored at -20°C before analyses.

A set of 50 mg and 40 mg subsamples (n=70) were analyzed for (1-3)-ß-D-glucan concentration in dust samples by the LAL- and EIA assays, respectively. Taking into account the dust amount used for the analysis of each sample and the surface area from where the dust was sampled, the (1-3)-b-D-glucan concentrations in dust samples were reported as mg/g and mg/m2.

Air samples were collected using a Button Personal Inhalable Aerosol Sampler

(SKC Inc., PA, USA), as described in detail by Osborne et al. (2006). The sampler collects inhalable particles onto a 25 mm polycarbonate membrane filter of 3 mm pore size (GE Osmonics Inc., Minnetonka, MN, USA). Forty samples were collected from the primary activity room at a flow rate of 4 l/min of 48 hours. Each filter was extracted into

2 ml of phosphate buffer, of which 0.7 ml aliquot was used for (1-3)-ß-D-glucan assay

(0.5 ml for the LAL and 0.2 ml for the EIA). The (1-3)-ß-D-glucan concentrations in air samples were reported as ng/m3.

1.2.4. Data analysis

Specificity answers the question whether an assay uniquely measures the analyte of interest, i.e., whether other substances interfere with assay or cross-react significantly

(Diamandis and Christopoulos, 1996). The fungal (1-3)-b-D-glucan consists of both linear (1-3) and branched (1-3)(1-6)-b-D-glucans, both of similar importance for fungal survival and health effects of fungi (Kataoka et al., 2002; Brown and Gordon, 2005;

Sakurai et al., 1997). Therefore, for the purpose of this thesis, detection of both linear (1-

3)-b-D-glucan and branched (1-3)(1-6)-b-D-glucan was viewed as specific detection of fungal (1-3)-b-D-glucans.

The lowest detectable analyte concentration or detection limit, also known as sensitivity of an assay, is defined as the analyte concentration that gives a response which has a statistically significant difference from the response of the zero analyte sample (the blank) (Diamandis and Christopoulos, 1996).

Accuracy of an assay is the degree to which it actually measures what it is intended to measure. It was assessed by comparison of linear and branched a and b- glucans with their reference standards.

(1-3)-b-D-glucan concentrations in the subset of dust (n=70) and air (n=40) samples collected from CCAAPS homes were tested for normality of distribution. The

EIA-analyzed (1-3)-b-D-glucan concentration from indoor dust and air samples did not follow the Gaussian distribution, probably due to the small sample size (n=70, n=40, respectively). Thus, non-parametric statistics were used to test for correlation and difference between the LAL- and EIA-analyzed (1-3)-b-D-glucan concentrations in field samples.

1.3. RESULTS

1.3.1. Specificity and Accuracy of LAL and EIA in measuring glucan of different linkage and branching.

The reactivities of LAL and EIA assays to a- and ß-glucans of various degree of branching are shown in Fig. 1-1. As the LAL assay (Fig. 1-1 A) is a kinetic assay measuring the onset of time at OD = 0.03, the later the reaction occurs (mean onset time), the lesser the specificity for the particular purified glucan at that concentration in comparison to the other glucan standards. In the endpoint EIA assay (Fig. 1-1 B), the best curve is curvilinear, with a rapid straight decrease in the absorbance units with the increase of the concentration till saturation of the curve is reached (i.e., minimal concentration detected). As seen in Fig. 1-1, both LAL and EIA assays were specific to

linear (1-3)-ß-D-glucans (Curdlan and Pachyman) and non-reactive to alpha-glucans

(presented by straight horizontal lines). The reactivity of the LAL assay slightly decreased with the increase in the degree of branching (Fig. 1-1.A). Although the EIA assay was also sensitive in recognizing close to linear structures (Laminarin), its sensitivity to branched structures was negligible {(Schizophyllan, branched (1-3)(1-6)-b-

D-glucan)} (Fig. 1-1.B).

As the LAL assay measures the availability of reactive (1-3)-b-D-glucan molecules in the sample that can activate the factor G enzyme, it detects any structures that have the single helix of (1-3)-b-D-glucan. This and the fact that it uses the linear (1-

3)-b-D-glucan Pachyman as a standard against which the rest of the glucans are measured, assures for the detection of any glucans containing (1-3)-b-D-glucan as part of their molecule. The EIA assay is an antigen-antibody test that uses antibodies against

Laminarin, a linear (1-3)-b-D-glucan with some (1-6)-branching. Thus it is expected to be highly reactive against epitopes containing close to linear (1-3)-b-D-glucan.

The glucan standard concentrations were calculated based on a standard curve of

Pachyman for the LAL assay (2 runs) and Laminarin for the EIA assay (1 run). These measured values, expressed as % of the expected concentrations, are presented in Table

1-1. For each standard concentration the LAL was more accurate in measuring concentrations of (1-3)-b-D-glucan standards than the EIA as demonstrated by the narrower range of the % expected concentration and the smaller median value of this %

(Table 1-1).

1.3.2. Comparison of LAL vs. EIA analyzed (1-3)-ß-D-glucan in dust and air samples

The EIA-measured (1-3)-ß-D-glucan levels {mg/g: geometric mean (GM) =4.20; mg/m2: GM=3.58) in dust samples were higher than that the LAL-measured (mg/g:

GM=1.60; mg/m2: GM=0.97). However, the variability in (1-3)-ß-D-glucan levels was greater as measured by the LAL {mg/g: Geometric standard deviation (GSD) GSD=1.70; mg/m2: GSD=2.00} than the EIA assay (mg/g: GSD=1.45; mg/m2: GSD=1.81). The same trend was observed in the air samples, too (ng/m3) (EIA: GM=12.8, GSD=10.00; LAL:

GM=6.4, GSD=2.4). There was no correlation between LAL- and EIA- analyzed (1-3)-ß-

D-glucan in dust samples, when concentrations were expressed per gram, but strong and significant when expressed per square meter (Figs. 1-2 and 1-3). In both units, the difference between the LAL- and EIA-analyzed (1-3)-ß-D-glucan levels was significant, the LAL measured values being about seven times lower than the ones measured by the

EIA (mg/g: z-value = -7.21, p<0.001; mg/m2: z-value = -7.26, p<0.001). Similarly, in air samples there was a strong correlation between the LAL- and EIA-analyzed (1-3)-ß-D- glucan concentrations (r=0.44, p=0.04). Due to the low sensitivity of the EIA assay, (1-

3)-ß-D-glucan concentrations in 18 out of the 40 air samples were below the LOD of the

EIA method. Thus the difference between assays in air samples was not significant (z- value=-1.54, p=0.12).

1.4. DISCUSSION

(1-3)-b-D-glucans are polyglucose compounds consisting of a glucose chain united by b-1,3-linked linkages, and branched with variable amounts of b-1,6- and b-1,4- linked glucose side chains (Shematek et al., 1980). Environmental exposure to (1-3)-b-D- glucan has been measured by the LAL or the EIA assays in previous studies (Rylander,

2000; Douwes, 2005). This study confirmed that the LAL assay recognizes both linear and branched (1-3)-b-D-glucans, as previously reported by Tanaka et al. (1991) and

Thorne et al. (2004). By using 0.6 M NaOH in the extraction process, the triple helix conformer of (1-3)-b-D-glucans was unwounded into the more reactive single-helix, thus increasing the sensitivity of the LAL assay in detecting conformers of (1-3)-b-D-glucans.

The current research confirmed the data reported by Douwes et al. (1996) that the EIA immunoassay reacts with linear (1-3)-b-D-glucans. Other modifications of the EIA assay were developed by Adachi et al. (1999) and Milton et al. (2001), which are highly specific to and measure the content of branched (1-3)(1-6)-b-D-glucans only. It has been reported that both linear and branched type (1-3)-b-D-glucans are ubiquitous in the cell wall of fungi (Perez and Ribas, 2004), and both linear and branched (1-3)-b-D-glucans are important for in vivo priming of macrophages (Ohno et al., 1995). Thus measurement of both linear and branched (1-3)-b-D-glucans is equally important. This makes both the

LAL and inhibition EIA assays more advantageous to use in comparison to the new

ELISA methods, as the latest cannot detect both the linear and branched forms. However, the LAL assay showed greater sensitivity and specificity (LOD=1pg/ml, detects both linear and branched (1-3)-b-D-glucans with comparable sensitivity) in comparison to the

EIA assay (LOD=40 ng/ml, detects preferably the linear (1-3)-b-D-glucans). In addition,

the LAL was more accurate in measuring concentrations of (1-3)-b-D-glucan standards than the EIA as demonstrated by the narrower range of the % expected concentration and the smaller median value of this %

To my knowledge, this is the first report on LAL-analyzed (1-3)-ß-D-glucan concentration in dust samples. The EIA assay measured much lower indoor dust (1-3)-ß-

D-glucan (mg/g: GM=4.20; mg/m2: GM=3.58), as compared to other studies. However, the smaller sample size can contribute to this as the range of geometric means reported in literature varies widely – from 35.1 mg/g (GSD=1.80, n=20; Milton et al. 2001) to 1,711 mg/g (GSD=1.9, n=395; Gehring et al. 2001), and from 90 mg/m2 (no GSD reported, n=508; Douwes et al. 2006) to 1,197mg/m2 (GSD=2.5, n=395; Gehring et al. 2001). In addition, the source of primary and secondary antibodies used in the EIA assay is different between the studies. Interestingly, the GM values of (1-3)-ß-D-glucan concentrations (ng/m3) measured in air samples in this study (LAL: GM=6.4; EIA:

GM=12.8) are comparable to the ones reported earlier in indoor residential buildings for the LAL assay (0.5-6.00) (Beijer et al., 2003; Rylander and Lin, 2000), but higher for what is reported for the EIA assay (GM range 0.2-6.6) (Alwis et al., 1999; Douwes et al.,

1996). This is probably due to the large % (>75%) of air samples in these studies for which the (1-3)-b-D-glucan concentration is below the LOD, and thus results are inconclusive.

The strong correlations between LAL-and EIA-analyzed (1-3)-b-D-glucan concentration in indoor dust samples per square meter and the lack of correlation, when levels were expressed per gram of dust were due to smaller variation in the values expressed per square meter. The variance in levels per square meter was largely

determined by the amount of dust sampled.

1.5. CONCLUSIONS FOR SPECIFIC AIM 1

In conclusion, the LAL assay is more specific, sensitive and accurate than the EIA assay in detecting both linear and branched (1-3)-b-D-glucans.

SPECIFIC AIM 2

COMPARE LAL- AND EIA- ANALYZED (1-3)-ß-D-GLUCAN CONTENT OF COMMON INDOOR MOLD SPECIES.

2.1. INTRODUCTION

As already described in the background section, it is suspected that (1-3)-b-D- glucan exposures are the cause of common general and respiratory symptoms associated with fungi. In addition, it is known from literature that different fungal species consist of different amounts of (1-3)-b-D-glucan. As the two assays (LAL vs. EIA) differ in their specificity and sensitivity to detect linear and branched fungal (1-3)-b-D-glucans (see results of specific aim 1), it is expected that they will also yield different estimates of (1-

3)-b-D-glucan content in different fungal species. As the amount and type of (1-3)-b-D- glucan is important regarding health effects associated with exposure, it is essential to determine the content of (1-3)-b-D-glucan in common indoor fungal species.

Thus, the second aim of this thesis was to compare the specificity and sensitivity of LAL and EIA methods through the analysis of (1-3)-b-D-glucan content in common indoor fungal species. This will also help determine which assay, if any, is a better surrogate of total fungal exposure.

2.2. MATERIALS AND METHODS

2.2.1. Laboratory analysis of (1-3)-ß-D-glucan

As described in specific aim 1 (section 1.2.1).

2.2.2. Fungal species selection

Twelve fungal species were selected based on their prevalence in field samples, genus variability, and public health concerns: two Cladosporium species, five Aspergillus species, Aureobasidium pullulans, Penicillium brevicompactum, Epiccocum nigrum,

Wallemia sebi, and Stachybotrys chartarum (Table 2-1). Results from an ongoing field study, {Cincinnati Childhood Allergy and Air Pollution Study (CCAAPS), see section

Field Samples below} were used to identify species that are commonly found in homes.

Based on Polymerase Chain Reaction (PCR) analysis of dust samples from 297 homes

(Vesper et al., 2006), eight fungal species that were most commonly found (>90% frequency) and had highest median concentrations were selected for this study (Table 2-

1). This list included two Aspergillus species. Three additional Aspergillus species were included in order to study the within species variability of (1-3)-b-D-glucan. P. brevicompactum and S. chartarum were chosen to represent the species of medium frequency and concentration. In addition, S. chartarum was included due to reasons of being of health concern in indoor environments (Lai, 2005; Hossain et al., 2004). A non- toxic strain was generously provided by NIOSH, Morgantown, WV (research collection,

No: 29-51-05). All the other species were purchased from the American Type Culture

Collection (ATCC, Manassas, VA).

2.2.3. Preparation of pure fungal species

penicillioides), and potato dextrose agar (the rest of the species). Spores from one-week old pure cultures were harvested from the agar surface by using micro-beads (Schmechel et al., 2006), and transferred into 5 ml sterile tube, containing 0.02% Tween solution in pyrogen and glucan-free reagent water. Serial dilutions of 10-0 to 10-6 were prepared for each fungal spore suspension, which was used for determining (1-3)-b-D-glucan concentration (as described above), spore concentration, and spore size.

In order to determine the spore concentration (spores/ml), one milliliter of spore suspension was first filtered through a mixed cellulose ester (mixture of cellulose acetate and cellulose nitrate, pore size 1.2 µm, diameter 13 mm; Millipore Corporation, Bedford,

MA, USA) membrane filter and cleared by a modified acetone vaporizing unit (Model:

Quixfix; Environmental Monitoring System, Charleston, SC, USA). The filter was stained with glycerin jelly (gelatin 20 g, phenol crystals 2.4 g, glycerol 60 ml, water 70 ml). Spores were counted under a microscope as described by Adhikari et al. (2003).

For determining the spore size, about 50 ml of the pure spore suspension was placed on a microscopic slide, covered by a cover slide and the spore size of 30 spores of each species was immediately measured at 1000X magnification oil immersion using an optical microscope samples digitally imaged by a color video camera (SPOT advanced software, version 3.4, Diagnostic Instruments Inc, Sterling Heights, MI, USA). Based on the average spore size (diameter for spherical; width and length for ellipsoidal spores), the spore surface area and volume were calculated. Spores of Cladosporium species, S. chartarum and A. pullulans are ellipsoidal, and therefore, equations for a prolate spheroid were used. The surface area and volume for the spores of the other fungal species were calculated according to the formulas for a sphere.

2.2.4. Field samples

The collection of dust samples is described in specific aim 1 –Field samples

(section 1.2.3).

Based on the estimated (1-3)-ß-D-glucan content in common indoor fungal species (LAL and EIA), the predicted total (1-3)-ß-D-glucan concentration in dust samples was calculated. The predicted (1-3)-ß-D-glucan concentration of individual fungal species in dust samples was calculated by multiplying the measured concentration

(cells/mg dust) (MSQPCR) with the LAL- and EIA-estimated (1-3)-ß-D-glucan content of the respective fungal species. The total predicted (1-3)-ß-D-glucan concentration was the sum of the predicted (1-3)-ß-D-glucan concentration contributed by each individual fungal species from all dust samples, assuming that the main contribution of (1-3)-ß-D- glucan came from the 12 fungal species studied.

2.2.5. MSQPCR analysis of dust

Methods have been reported previously for preparing conidial suspensions from dust samples, extracting DNA, performing Mold Specific Quantitative Polymerase Chain

Reaction (MSQPCR) analyses and preparing standard calibration curves (Vesper et al.

2004, Vesper et al., 2006). All primer and probe sequences, as well as known species comprising the assay groups, were published at the website:

http://www.epa.gov/microbes/moldtech.htm.

Primers and probes were synthesized commercially (Applied Biosystems, Foster

City, CA; Integrated DNA Technologies, Coralville IA; Sigma Genosys, Woodlands,

TX).

A 5 mg sub-sample of the fine dust from 297 samples was used for the PCR analysis (Vesper et al., 2006) in order to identify and quantify 36 fungal species. The results were reported as cells/mg dust. Mold concentration data having a minimum detection limit of 1 cell mg-1 dust were treated as left-censored data with appropriate statistical methods applied (Helsel, 2005). Non-detections were set as 1.

2.2.6. Data analysis

(1-3)-b-D-glucan content of the pure fungal spores was tested for normality of distribution. As the (1-3)-b-D-glucan content of fungal spores, as well as their respective spore size, surface areas and volumes, were not normally distributed even after log- transformation, the correlations with and between the LAL- and EIA-analyzed spore (1-

3)-b-D-glucan contents were tested with the non-parametric Spearman correlation, and the difference of means with the Wilcoxon statistics, using non-log transformed data.

(1-3)-b-D-glucan concentration in 297 dust samples collected from the CCAAPS homes and the fungal cell concentrations analyzed with PCR followed the normal

2.3. RESULTS

2.3.1. Spore characteristics

The spore size, spore surface area, and spore volume of the twelve pure fungal species analyzed are provided in Table 2-2. Among the fungal species, E. nigrum had the largest spore size (28 mm), and thus the biggest surface area (2,463 mm2) and volume

(11,494 mm3). The Aspergillus species were small in size, and Aspergillus versicolor was the smallest among them (size 2.96 mm, surface area 27.53 mm2, and volume 13.58 mm3).

The measured spore sizes and surface areas were within the range reported earlier

(Reponen et al., 2001; Bauer et al., 2002; Foto et al., 2004).

2.3.2. (1-3)-ß-D-glucan content in fungal spores

Based on the spore characteristics reported in Table 2-2, the (1-3)-ß-D-glucan content per spore, spore surface area, and spore volume were calculated, and presented in

Table 2-3. Although E. nigrum was the species of greatest (1-3)-ß-D-glucan content per spore (241 pg/spore), this was mainly due to having also the largest spore size (28 mm).

Other fungi of high (1-3)-ß-D-glucan content per spore measured by the LAL assay were

C. herbarum and A. pullulans. The EIA assay revealed W. sebi and C. cladosporioides as the species of greatest (1-3)-ß-D-glucan content per spore after E. nigrum. However, the biomass-normalized (1-3)-ß-D-glucan content (pg per surface area and pg per spore volume) measured by both assays was within similar range (LAL: 0.003 to 146.33 pg/mm2, 0.22-240.54 pg/mm3; EIA: 0.04 – 197.00 pg/mm2, 0.03-300 pg/mm3). The LAL assay determined C. herbarum, followed by E. nigrum and P. brevicompactum, as the fungi of highest (1-3)-ß-D-glucan content per surface area and volume. The EIA assay

ranked W. sebi as leading in the (1-3)-ß-D-glucan content, followed by E. nigrum, P. brevicompactum and C. cladosporioides (Table 2-3). Both assays recognized Aspergillus species and S. chartarum to be the lowest in the (1-3)-ß-D-glucan content.

The variation between all the tested 12 fungal species and within Aspergillus genus (5 species), measured by both assays, was further analyzed. Both assays showed lower variation of (1-3)-b-D-glucan content within the Aspergillus genus than between the 12 fungal species (LAL: Coefficients of variation = 45% and 57%, respectively; EIA:

Coefficients of variation = 22% and 42%, respectively).

As (1-3)-ß-D-glucan can contribute up to 50% of the composition of the fungal cell wall, it may be expected the spore size and surface area to affect the fungal (1-3)-ß-

D-glucan content. Indeed, this study found a close to significant correlation between

LAL-analyzed (1-3)-ß-D-glucan content (pg/spore) and spore size (Fig. 2-1), but not with the spore surface area (r=0.39, p=0.22), or volume (r=0.36, p=0.26). The lack of correlation was even more obvious between EIA-analyzed (1-3)-ß-D-glucan content per fungal spore and the corresponding spore size (r=0.22, p=0.53), surface area (r=0.27; p=0.40), or volume (r=-0.23, p=0.47). Furthermore, the correlation between LAL and

EIA-analyzed (1-3)-ß-D-glucan content per spore was close to significant (r=0.48, p=0.11), but not per surface area (r=0.24, p=0.45) or per volume (r=0.20, p=0.53).

2.3.3. Associations between (1-3)-ß-D-glucan content in fungal spores and concentrations in field dust samples

There was a strong association between LAL-analyzed (1-3)-ß-D-glucan content in pure fungal spores and the field data on respective fungal species. (1-3)-ß-D-glucan

content per spore correlated with both the frequency and concentration of the respective fungal species in the 297 dust samples analyzed using the PCR in the CCAAPS study

(Figs. 2-2 and 2-3). The EIA analyzed (1-3)-ß-D-glucan content failed to show such correlations (frequency: r=0.27, p=0.40; concentration: r=0.27, p=0.39). Spore size also correlated with the (1-3)-ß-D-glucan content (r=0.46, p=0.13), frequency (r=0.64, p=0.03), and concentration (r=0.53, p=0.08). Therefore, a test was performed to distinguish whether it is the (1-3)-ß-D-glucan content (pg/spore) or the spore size that drives the correlation between (1-3)-ß-D-glucan content per spore and the fungal frequency/concentration. For this purpose, the (1-3)-ß-D-glucan content vs. the values of frequency and concentration divided by the spore size, were compared. The analysis showed that the (1-3)-ß-D-glucan content is strongly correlated with the concentration of fungi in indoor dust samples (r=0.78, p=0.003), but it is the spore size that correlated with the fungal frequency rather than the (1-3)-ß-D-glucan content (correlation between (1-3)-

ß-D-glucan content and fungal concentration divided by spore size: r=0.05, p=0.88). In addition there was a significant correlation between indoor fungal species frequency and concentration (r=0.95, p<0.001).

2.3.4. Estimated vs. predicted (1-3)-ß-D-glucan concentrations in field samples

as measured in 297 dust samples was compared versus the predicted total (1-3)-ß-D- glucan assuming that the main contribution of (1-3)-ß-D-glucan comes from these 12 fungal species. Of these, E. nigrum, Cladosporium species, A. chevalieri and W. sebi emerged as the major sources of (1-3)-b-D-glucan as identified by both assays, contributing more than 85 % of the total (1-3)-b-D-glucan concentration in the dust samples. The predicted total (1-3)-ß-D-glucan was five times greater than the measured one with the LAL assay in per gram units, and six times greater in per square meter units.

No correlation between the predicted and measured (1-3)-ß-D-glucan concentration was observed (mg/g: r=0.03, p=0.57; mg/m2: r=0.01, p=0.87). The measured (1-3)-ß-D-glucan concentration was significantly lower than the predicted (1-3)-ß-D-glucan concentration

(mg/g: t-test= -7.37, p<0.001; mg/m2: t-test= -16.03, p<0.001). Interestingly, the predicted values of (1-3)-ß-D-glucan as calculated based on the LAL and EIA-determined fungal

(1-3)-ß-D-glucan content in pure spores, were strongly correlated (mg/g: r=0.94, p<0.001; mg/m2: r=0.82, p<0.001). Therefore, regardless of the different contributions of W. sebi,

A. pullulans and two Cladosporium species to the total (1-3)-ß-D-glucan concentration in dust samples when measured by the LAL vs. the EIA assay, if all these six commonly found indoors species (C. herbarum, C. cladosporioides, E. nigrum, W. sebi, P. brevicompactum, A. pullulans) grow together, the LAL and EIA assays will show comparable results. Still, both assays are significantly different from each other, with the

LAL-determined fungal (1-3)-ß-D-glucan content being lower than the EIA determined

(mg/g: t-test= -2.38, p=0.02; mg/m2: t-test= -12.18, p<0.001).

2.4. DISCUSSION

The utility of the LAL assay in measuring serum fungal (1-3)-b-D-glucans has been evaluated in numerous laboratory and clinical studies (Yoshida et al., 1997b; Yuasa et al., 1996; Odabasi et al., 2006), and is currently routinely used in Japan and Europe for the detection of invasive fungal infections (Odabasi et al., 2006; Stevens et al., 2002).

However, there is very little data on the utility of LAL and EIA assays in the analysis of fungal (1-3)-b-D-glucan concentration in environmental samples. Thus the current study is the first one to directly compare the specificity and sensitivity of LAL versus EIA assays in detecting fungal (1-3)-b-D-glucans.

From previous studies using the LAL assay on serum and culture supernatants of clinical mold isolates, it is known that there is a wide range of difference in the (1-3)-ß-

D-glucan content of different fungal species (Odabasi et al., 2006; Yoshida et al., 1997ab;

Yuasa et al., 1996). This study also showed a wide range of (1-3)-b-D-glucan content between species and within Aspergillus genus, as demonstrated by the large coefficients of variation measured in this study. The (1-3)-b-D-glucans content varied considerably between the five Aspergillus spp. spores unlike what was observed by Odabasi et al.

(2006) (range in 12 Aspergillus spp.: 1,311-2,480 pg/ml, GM= 1,769). However, as the latest study presented (1-3)-b-D-glucan content per ml of supernatant of culture broth, but not biomass normalized per spore, results cannot be directly compared.

A major disadvantage of the EIA assay is its low sensitivity. Thus the low content of EIA-reactive (1-3)-b-D-glucan antigens in the pure fungal suspensions of C. herbarum, A. pullulans and S. chartarum is a limitation of the study. This difference in the detection sensitivity and specificity between assays (see results from specific aim 1),

may explain the difference of (1-3)-b-D-glucans content in the same fungus species when determined by LAL vs. EIA.

In this study, Aspergillus species and S. chartarum, had the lowest (1-3)-b-D- glucan contents, but the Aspergillus species had higher content than S. chartarum.

Fogelmark and Rylander (1997) have reported that the Stachybotrys atra (=S. chartarum)

(median 3.9 pg/spore, range 0.863-39.330 pg/spore) has 1,000 times higher LAL- analyzed (1-3)-b-D-glucan than Penicillium aurantiogriseum (median: 0.09 pg/spore, range 0.005-1.8 pg/spore) and Aspergillus fumigatus (median: 0.11 pg/spore, range

0.008-0.7 pg/spore). The much lower (1-3)-b-D-glucan content in the Stachybotrys spores in this study may be due to the fact it was a non-toxic strain. The last is based on the assumption that fungal toxicity may be related to pathogenicity, as low fungal (1-3)- glucan content is associated with lower pathogenicity (Hogan et al., 1994; Rementeria et al., 2005; Rappleye et al., 2006). Also, the growth medium can affect the content of fungal (1-3)-b-D-glucans (Foto 2004). The ATCC mold species in this study were grown on agar media, while the ones in Rylander’s study were grown on rice and wood. In addition, a study by Foto et al. (2004), showed that the LAL-analyzed (1-3)-b-D-glucan content in S. chartarum (mean 0.012 pg/mm2) is lower than that of A. versicolor (mean

0.022 pg/mm2) and C. cladosporioides (mean 0.060 pg/mm2), all grown on 2% malt extract agar. Although the (1-3)-b-D-glucan contents measured by the LAL in this study are higher than those reported by Foto et al. (2004), the same trend among species was observed.

The order of EIA analyzed fungal (1-3)-b-D-glucan content in this study was different from what was reported by the monoclonal IgM ELISA assay in a mixture of

fungal mycelia and spores (Milton et al., 2001). Aspergillus flavus isolated from stored urine and Aspergillus ochraceus isolated from outdoor air samples had higher (1-3)-b-D- glucan content than Cladosporium spp. isolated from bedroom air and Wallemia spp. isolated from outdoor air. This may be due to the different sources from which species were isolated, as well as that both spores and mycelia contribute to the measured (1-3)-b-

D-glucan content. The IgM ELISA measures the content of branched (1-3)(1-6)-b-D- glucan only, and thus may underestimate the (1-3)-b-D-glucan content of Wallemia and

Cladosporium spp., as they have been reported to contain predominantly linear glucans

(San-Blas et al., 1996).

Due to the different spatial distribution of (1-3)-b-D-glucan in spores of different fungal species (Perez and Ribas, 2004; Schmid et al., 2001; Ishibashi et al., 2004), as well as during morphological development (Steele et al., 2005), it is important to measure the

(1-3)-b-D-glucan content not only per spore, but also per surface area and volume. This data set supports this by the fact that the order of fungi in relation to their (1-3)-b-D- glucan glucan content changed when data was presented per spore compared to when it was presented per surface area/volume. The order, however, was relatively the same whether (1-3)-b-D-glucan content was expressed per surface area or per volume.

Foto et al (2004) have reported that LAL-analyzed (1-3)-b-D-glucan content was proportional to spore size, while such data is not available on the EIA assay. In supplement to Foto’s (2004) study, this study also found a positive correlation between the LAL-analyzed (1-3)-b-D-glucan content per spore and the spore size. The lack of correlation with surface area and volume, may be due to higher extraction efficiency of

(1-3)-b-D-glucan from larger cell surfaces, i.e., larger spore size, but lower efficiency

from thicker cell walls such as of some Aspergillus species (Reijula, 1991). However, as the surface structures of some species, such as E. nigrum is very rough (uneven) and, the actual surface area is much larger than the one estimated here for a smooth surface, thus the calculated (1-3)-b-D-glucan content per surface area may be higher than the actual.

In addition, E. nigrum produces a high yield of extracellular polysaccharides (Schmid et al., 2001), some of which may have been adhered to the microbeads during the preparation of spore suspensions, and thus yield higher than expected (1-3)-b-D-glucan content.

The strong and statistically significant correlation between LAL-analyzed fungal

(1-3)-b-D-glucan content and the concentration of the corresponding fungi in indoor dust samples may attribute to the ecological advantage to survival of fungal species of high (1-

3)-b-D-glucan content. The strong correlation between fungal frequencies and concentrations indicates that most common indoor fungal species are also the ones that dominate in concentration in individual dust samples. However, it was the fungi with greater spore size, rather than with greater (1-3)-b-D-glucan content that were of higher frequency indoors. This indicates that floor dust samples contain predominantly fungi of greater spore size, as they settle down due to gravity, while smaller size spores remain in air longer and may deposit to vertical surfaces to greater extent than larger spores.

Nevertheless, the strong correlation of LAL-measured (1-3)-b-D-glucan content per spore with the fungal spore size and with the prevalence of indoor fungi indicates that the LAL- measured (1-3)-b-D-glucan may be used as a measure of total mold.

The current study have also found that species that are more frequently found in indoor environments are also the ones that contribute most to the total (1-3)-b-D-glucan

concentration in indoor samples. Of these, E. nigrum, Cladosporium species, A. chevalieri and W. sebi are the major sources of (1-3)-b-D-glucan as identified by both assays, contributing more than 85 % of the total (1-3)-b-D-glucan concentration.

Although W. sebi, A. pullulans and two Cladosporium species showed different contributions as measured by the LAL vs. the EIA assay, if all these six commonly found indoors species (C. herbarum, C. cladosporioides, E. nigrum, W. sebi, P. brevicompactum, A. pullulans) grow together, the LAL and EIA assays will show comparable results. However, as different climates, occupational and indoor environments, favor the growth of one species over other, these two assays are expected to show different (1-3)-b-D-glucan concentration in dust samples from different environments.

There was a lack of correlation between the predicted and measured (1-3)-ß-D- glucan concentration, with the measured (1-3)-ß-D-glucan concentration significantly lower than the predicted. The lower measured (1-3)-b-D-glucan concentration may be explained by the mixture of spores and mycelia that environmental samples contain, while the predicted (1-3)-b-D-glucan concentration is estimated on the assumption that samples contain only spores. In addition, unfavorable growth conditions, as well as, competition among fungal species in field samples are associated with less (1-3)-ß-D- glucan fungal content (Stone and Clark, 1992; Walker et al., 1984; de la Cruz et al.,

1995).

2.5. CONCLUSIONS FOR SPECIFIC AIM 2

Although the (1-3)-b-D-glucan concentration in this field samples measured by the LAL and EIA assay correlated, data shall be analyzed with caution, as assays give different weight to different fungal species. In addition, the strong significant associations of the LAL-analyzed (1-3)-b-D-glucan with fungal spore size and with indoor dust fungal concentration, indicates that LAL-analyzed (1-3)-b-D-glucan could be used to estimate the total fungal load in indoor samples.

Therefore, based on the results of specific aims 1 and 2, the LAL assay is more sensitive, specific and accurate. In addition, unlike the EIA assay, it may be used as estimate of the total fungal load in dust samples. Thus, the LAL assay was chosen for the conduction of the (1-3)-b-D-glucan analysis in specific aims 3 and 4.

SPECIFIC AIM 3

COMPARE LAL-ANALYZED INDOOR (1-3)-ß-D-GLUCAN CONCENTRATIONS WITH MOLD SPORE CONCENTRATIONS AS ANALYZED BY THE PCR METHOD IN DUST SAMPLES

3.1. INTRODUCTION

As stated earlier in the background section, many studies have used (1-3)-b-D- glucan assessment as a surrogate for mold exposure. Although, the analysis of (1-3)-b-D- glucan analysis has many advantages over methods that count mold spores and culturable colony units, it was of practical interest to prove or disprove the utility of (1-3)-b-D- glucan analysis as such a surrogate.

Based on the results from specific aims 1 and 2, it was shown that the LAL assay is more sensitive, specific and accurate than the EIA assay. The LAL assay was chosen to measure (1-3)-b-D-glucan concentrations in a larger number (n=297) of field dust samples and compare them against the mold concentration in the same dust samples.

From many available methods to assess mold exposure, the Polymerase Chain Reaction

(PCR) is accurate, cheap and quick (Vesper et al., 2004), as well as it measures both viable and non-viable fungi, and provides data up to fungal species level.

Specific aim 3 examined which indoor mold species, assessed by the PCR analysis, contribute most to the (1-3)-b-D-glucan content in field dust samples, and thus whether (1-3)-b-D-glucan can be used as a surrogate for total mold exposure or whether it rather predicts the exposure to specific mold species.

3.2. MATERIALS AND METHODS

3.2.1. On-site home visit and exposure assessment

The on-site home visit and collection of dust samples is described in details in specific aim 1- Field samples (1.2.3).

Altogether 297 were selected for the study that included PCR analysis of a 5 mg sub-sample of the fine dust (Vesper et al., 2006). Another 50 mg sub-sample was analyzed for (1-3)-ß-D-glucan content by the LAL assay.

3.2.2. MSQPCR analysis of dust

As described in specific aim 2 (section 2.2.5)

3.2.3. (1-3)-b-D-glucan analysis in dust

The (1-3)-b-D-glucan content in 297 dust samples was determined by the endpoint chromogenic Limulus Amebocyte lysate assay (Associates of Cape Cod, East

Falmouth, MA). The protocol is described in detail in specific aim 4 (section 4.2.3.).

Results were reported as mg/g.

3.2.4. Linear Regression Analysis

Data were visualized as box-plots, histograms and scatter diagrams of each fungal species against the independent variable of (1-3)-b-D-glucan content. This revealed non-

Gaussian, skewed to the left distributions, which are usually observed for environmental concentration and loading data. Thus, we performed a log-transformation of all data

{fungal and (1-3)-b-D-glucan concentrations}. The log-transformed data had Gaussian distributions.

The distributions of the concentrations of individual mold species were skewed to the left (tests for normality: Shapiro-Willks and Kolmogorov-Smirnov. Diagnostics on the data were also performed for co-linearity because mold grows in clusters of different genera, so this would be expected. The co-linearity for all variables (n=36) was evaluated using the Pearson correlation. A linear regression model was developed for 36 mold species in 297 homes in Cincinnati. Statistical analyses were performed using Statistical

Analysis System (SAS Institute Inc., Cary, NC).

To test the first null hypothesis that one or more of the fungi will predict the (1-

3)-b-D-glucan content, the best regression model for the log-transformed data was fit.

Many fungal species grow together. Thus a second hypothesis was that the 10 most common indoor fungal species combined would better predict the (1-3)-b-D-glucan content than individually. To test this hypothesis these ten fungal species were included using a backward elimination approach with a=0.05. Then, the same independent variables were included using a forward selection approach, a=0.05.

Due to the simultaneous fungal growth, there is a potential for interaction effect between the independent variables. A third approach was to test all possible 45 two- variable interactions among the ten most common indoor fungi.

3.2.5. Factor analysis and Structural Equation Modeling

Further analysis was conducted using the multivariate methods of factor analysis and structural equation modeling (SEM) to determine the relationships that may exist

among the 36 mold species and to determine how these groups of related mold species predict (1-3)-b-D-glucan content. Several preliminary models were run in steps, which led to the final SEM model. These preliminary steps included exploratory factor analysis

(EFA), confirmatory factor analysis (CFA), deriving a base SEM, and then the final

SEM. The log-transformed data were used in all analyses.

The SAS procedure Factor with maximum likelihood method specified was first used to determine the number of factors needed and to obtain a preliminary description of the mold groupings that determined these factors; a cutoff of > 0.3 loading was used. The

SAS procedure Calis was then used to perform the CFA. The initial CFA was based on the final EFA. In a stepwise fashion, mold species were added to the factors based on the largest Lagrange multipliers that were statistically significant with a positive association.

Similarly, residual correlations were added based on the largest Lagrange multipliers that were statistically significant. This procedure continued until all estimated loadings and all residual correlations were statistically significant at the 0.05 level. The initial SEM was based on the final CFA with the additional model predicting (1-3)-b-D-glucan with the factors determined in the final CFA. Factors that were not statistically significant at the 0.05 level were removed from the (1-3)-b-D-glucan model in a stepwise fashion until only significant factors remained in the model.

3.3. RESULTS

3.3.1. Linear Regression Analysis

Frequency of and geometric mean concentrations in dust samples for the 36 fungal species studied are given in Table 3-1. There were no strong correlations between any mold species concentration and its corresponding (1-3)-b-D-glucan content (Table 3-1).

However, there were cross-correlations among specific fungal species, of which 12 had a coefficient of correlation r2 >0.5: Alternaria alternata - Aureobasidium pullulans, A. alternata - Cladosporium cladosporioides, Aspergillus penicillioides - Aspergillus chevalieri, A. penicillioides – Wallemia sebi, Aspergillus unguis - Aspergillus ustus, A. pullulans - Cladosporium herbarum, C. cladosporioides - C. herbarum, C. cladosporioides - Cladosporium sphaerospermum, C. cladosporioides - Epiccocum nigrum, C. sphaerospermum - W. sebi, and A. chevalieri - W. sebi. The co-linearity was also diagnosed for the final model with two variables with VIF (VIF=1). The VIF was not larger using the rule of thumb for VIF<10. No outliers were observed in the final model.

The following linear regression models were estimated:

1) First, a backward elimination approach including all variables was used with a=0.05, which gave a linear regression, which included Aspergillus flavus only. Next, a forward inclusion selection approach was used at the same significance level, which gave a model including A. flavus and C. herbarum. The above models were presented by the following equations:

Backward elimination model:

(1-3)-b-D-glucan = 1.637 + 0.115(A. flavus) (1)

Forward inclusion model:

(1-3)-b-D-glucan =1.420 + 0.118(s. flavus) + 0.095(C. herbarum) (2)

2) Using only the ten most common fungal species in a backward elimination and forward selection approaches (a=0.05) gave same regression model, reported below

(equation 3). The only significant species in the model were C. cladosporioides and A. flavus.

(1-3)-b-D-glucan = 1.338 + 0.151(C. cladosporioides) + 0.140(A. flavus) (3)

The same above-described procedures were performed by changing the alpha levels to a =0.20, a =0.15 and a=0.10. The models did not change.

3) All possible 45 interaction terms between the ten most common indoor fungi were included in the forward and backward approach with all other 36 variables, a =0.05.

Both gave the same model, which included six interaction effects and only three main effects. Therefore, we forced all main effects that were involved in the interactions in a forward selection and backward elimination approach.

In conclusion, the regression model revealed C. cladosporioides, P. brevicompactum, A. flavus and A. unguis as significant contributors to the (1-3)-b-D- glucan content in indoor dust samples (results summarized in Table 3-2).

Combining species of the same genus in one variable (e.g. all Cladosporium species were combined in a new group called Cladosporium, all Aspergillus species were combined in a group called Aspergillus, etc.) gave the most parsimonious model. This

model was estimated by forward inclusion and revealed two statistically significant variables: combined Cladosporium species and E. nigrum (Table 3-2).

3.3.2. Factor analysis and Structural Equation Modeling

Six factors were obtained in the final EFA and were then used in the initial CFA.

The final 6-factor CFA fit the data well (c2 (278) = 235, p-value = 0.97). The mold species that form these 6 factors are shown in Table 3-3. In the final SEM, factors 1 and

3 were significant for predicting B-glucan with 0.116 loading for factor 1 and 0.114 for factor 3. The final SEM also fit the data well with c2 (313) = 272, p-value = 0.95.

The confirmatory factor analysis revealed the following model as significant:

(1-3)-b-D-glucan = 0.1155*factor1 + 0.1141*factor3 + 1.0000 e0 (4)

Std Err 0.0639 f1_0 0.0634 f3_0

t Value 1.8068 1.7986

3.4. DISCUSSION

Most of the biological functions related to its survival, such as pathogenicity and toxicity, reside in the fungal cell wall, which, being the outermost part of the cell, mediates the host-fungus interplay (Chaffin et al., 1998). The microfibrillar polymers {(1-

3) and (1-6)-ß-D-glucans and chitin} represent the structural components of the wall.

These form a rigid skeleton that provides strong physical properties to the cell. From a quantitative point of view, b-D-glucans are the main constituent, accounting for 47 to

60% by weight of the cell wall (Young and Castranova, 2005). Therefore, it will be

expected that the indoor species of highest prevalence will contribute most to the (1-3)-b-

D-glucan content. Although the species identified by this study certainly are of high frequency and concentration, it shall be emphasized that non-toxic, non-pathogenic species also tend to synthesize more cell wall glucans in favorable environmental conditions (Santos et al., 1979). The factor analysis first combined the species based on the correlations among the species. Thus, species of coincident growth, concentration and frequency would be grouped in one factor. Therefore, species of high frequency and concentration were combined in one factor group, while Aspergillus species of low frequency and concentration were combined in another factor group. Thus, the combined

(1-3)-b-D-glucan contribution of the genera of Aspergillus and Cladosporium is more important than the individual contributions.

Furthermore, this study also suggests that (1-3)-b-D-glucan content varies between species, demonstrated by the different (1-3)-b-D-glucan loadings contributed by different fungal species in the above analyses. The latest has been demonstrated in both serum and culture supernatants of clinical mold isolates (Odabasi et al., 2006; Yoshida et al., 1997ab; Yuasa et al., 1996), as well as environmental mold species (Foto et al., 2004;

Fogelmark and Rylander, 1997; see also results from specific aim 2).

Another interesting finding is that the analysis failed to recognize A. alternata, allergenic fungus of high prevalence and concentration in indoor dust samples, as a contributor to the (1-3)-b-D-glucan content. Although immuno-modulating effects of (1-

3)-b-D-glucan are suspected (Rylander and Holt, 1998a; Ormstad et al., 2000), the scientific data are not affirmative on the (1-3)-b-D-glucan being an allergenic factor

(Douwes et al., 2005). Therefore, it may be only speculated whether (1-3)-b-D-glucan

content can modify the allergic effects of fungi. In addition, there is a possibility that the

(1-3)-b-D-glucan in Alternaria spores is not readily available for extraction and analysis by the LAL assay. Ultimately, as this is a field sample, other fungal species, such as

Cladosporium and Aspergillus species, could have introduced glucanases and made the environment unfavorable for A. alternata. In such cases, it may be speculated that although A. alternata sporulates, its (1-3)-b-D-glucan can decrease.

The advantage of this study is that determined fungal species contributing to the

(1-3)-b-D-glucan from real field samples, rather than individual spores or colony forming units isolates. The current study showed that certain species (such as Cladosporium and

Aspergillus) tend to cluster together. Therefore, the synthesized cell wall (1-3)-b-D- glucan by fungal species in indoor environments can be different from the content of their spores or cultures in laboratory studies.

This study has shown that the most common Cladosporium species (C. cladosporioides, C. herbarum and C. sphaerospermum) contribute to the (1-3)-b-D- glucan loading in dust samples. A recent study of the same cohort (Osborne et al, 2006) have found Cladosporium genus to be inversely associated with allergen sensitization. In the same infant cohort, an inverse association between exposure to high (1-3)-b-D-glucan and recurrent wheeze, recurrent wheeze with allergen sensitization, and allergen sensitization was observed (see results to specific aim 4). Thus, it can be expected that fungal species of high (1-3)-b-D-glucan content as determined in specific aim 3, will be associated with lower allergen sensitization, too. Some species from the Aspergillus and

Penicillium genera contribute to high (1-3)-b-D-glucan loading in the environment (A. flavus, A. unguis, A. penicillioides, A. chevalieri, A. versicolor, P. brevicompactum).

However, other species of these genera did not reveal such contributions. In addition, the study on the (1-3)-b-D-glucan content of pure fungal spores (specific aim 2), has shown that Aspergillus species have the lowest (1-3)-b-D-glucan content. Aspergillus and

Penicillium, as well as A. alternata are recognized as allergenic fungi due to their allergenic antigens. Thus, it can be speculated that a combination of low (1-3)-b-D- glucan content in these species, and their allergenic antigens may cause increase in the prevalence of allergenic sensitization among infants. This might explain the different direction of associations found between health outcomes and the concentrations of

Cladosporium vs. Aspergillus, Penicillium and Alternaria (Osborne et al. 2006).

As part of the CCAAPS study, the same database of mold species and indoor dust samples was used in establishing a quantitative index of moldiness, called an EPA

Relative Moldiness Index© (ERMI©). The ERMI index was calculated from mold specific quantitative PCR (MSQPCR) measurements on the concentration of 36 species of molds in floor dust samples called. The ERMI index may be used to predict the occurrence of illness in homes as certain groups of mold species were found to be more dominant in the homes associated with wheezing (Vesper et al., 2006). The fungal species identified as greatest contributors of (1-3)-b-D-glucan, were also identified as of significantly higher concentration in “more moldy homes” than in “less moldy homes”

(shown in Table 3-1) (Vesper et al. 2006). Ultimately, this data shows us that (1-3)-b-D- glucan content is representative of most prevalent fungal species, as well as those associated with moldy homes as determined by the ERMI index. Therefore, total (1-3)-b-

D-glucan may be used as a measure of total mold.

3.5. CONCLUSIONS FOR SPECIFIC AIM 3

The present study utilized advanced statistical methods in order to identify the group of indoor fungi most contributing to the (1-3)-ß-D-glucan load in residential dust samples. It revealed these species to be of the Cladosporium and Aspergillus genera. In addition E. nigrum, W. sebi, and P. brevicompactum also can contribute substantially to the (1-3)-ß-D-glucan concentration in dust samples. Another interesting finding of the study is that fungi that contribute most to the (1-3)-ß-D-glucan concentration are also the ones that are most prevalent in indoor environments. The latest supports the findings of specific aim 2, where it was demonstrated that LAL-determined (1-3)-ß-D-glucan content of fungal spores correlate with the concentration and frequency of these fungal species in dust samples.

SPECIFIC AIM 4

STUDY THE MAGNITUDE AND DIRECTION OF THE ASSOCIATION BETWEEN (1-3)-ß-D-GLUCAN AND THE DEVELOPMENT OF WHEEZE AND ALLERGEN SENSITIZATION IN INFANTS

4.1. INTRODUCTION

Exposure to some indoor molds during infancy has been associated with respiratory symptoms, such as increased risk for persistent cough and wheeze (Gent et al.,

2002; Belanger et al., 2003). On the other hand, exposure to Cladosporium has been inversely associated with allergen sensitization (Osborne et al., 2006). However, whether such effects are attributable also to the fungal component (1-3)-b-D-glucan is unclear, as data on (1-3)-b-D-glucan exposure and association with respiratory symptoms and atopy in infancy is scarce.

Exposure occurring in early years of life to microbial products, such as endotoxins, may have a protective effect on subsequent aeroallergen sensitization in childhood (von Mutius et al., 2000; Gereda et al., 2000; Belanger et al., 2003). As bacteria and fungi grow together, exposure to (1-3)-b-D-glucan may have a similar protective effect (Schram-Bijkerk et al., 2005; Douwes et al., 2006). However, existing data on health effects of (1-3)-b-D-glucan in children are inconclusive and no studies to date have assessed LAL-analyzed (1-3)-b-D-glucan exposure on clinical outcomes in an infant birth cohort.

Thus, the objective of this aim is to determine whether the prevalence of wheezing and allergen sensitization in a large birth cohort of high-risk infants is inversely

related to exposure to high indoor concentrations of both (1-3)-b-D-glucan and endotoxin.

4.2. MATERIALS AND METHODS

4.2.1.Recruitment

Infants were identified from birth certificate records in the Greater Cincinnati area from October 2001 through July 2003. Parents were interviewed for allergic symptoms and those reporting one or more symptoms were invited to be skin prick tested (SPT) for

15 aeroallergens. Infants with at least one parent having positive SPT [SPT(+)] were eligible for enrollment in the Cincinnati Childhood Allergy and Air Pollution study

(CCAAPS) as described previously (Ryan et al., 2005; LeMasters et al., 2006; Cho et al.,

2006; Campo et al., 2006). One purpose of CCAAPS is to examine gene-environmental interactions; thus, parental atopy was a critical criterion to obtain the most genetically at risk group. There were 1879 families with at least one parent reporting allergy symptoms.

Of these, 1152 parents agreed to participate in the SPT, and 881 had at least one parent with a SPT(+). Thus, the infants of the 881 families were eligible and 758 agreed to participate. The participating children (n=758) were 20.1% African-American, reflecting the greater Cincinnati, Ohio area distribution of 23.4%. The University of Cincinnati

Institutional Review Board approved the study.

4.2.2.Exposure assessment

See specific aim 1 – Field samples (section 1.2.3). Based on a visual inspection, the homes were classified into three groups: 1) no mold or water damage, 2) low visible mold (area<0.2m2), and 3) high visible mold (area =0.2 m2) (Cho et al., 2006b).

4.2.3. (1-3)-b-D-glucan and Endotoxin Analysis

As the exposure to (1-3)-b-D-glucan was not included in the CCAAPS study, only dust samples with sufficient amount of dust (at least 25 mg) were analyzed, after analyses for all other allergens was completed. Thus from the initially collected 758 samples, only 574 were available for the additional analysis. The (1-3)-b-D-glucan and endotoxin activities in dust samples were determined by the endpoint chromogenic

Limulus Amebocyte lysate assay (LAL; Associates of Cape Cod, East Falmouth, MA).

Two separate modifications of the assay were used, the Glucatell assay for (1-3)-b-D- glucan analysis, and the Pyrochrome assay for endotoxin. Each modification used a unique enzyme: Factor G in the Glucatell assay, and Factor C in the Pyrochrome assay.

Thus false-positive results were avoided. Endotoxin analysis was done as described by

Campo et al. (2006).

For the (1-3)-b-D-glucan, 50 mg of each dust sample was extracted in 2 ml of 0.6

M NaOH and shaken for 1 hour at -4oC. Twenty-five ml of Glucatell reagent was added to each well of serially diluted (1:100,000 and 1:1,000,000) dust extract and a control standard (1-3)-b-D-glucan (Pachyman), placed in a 96-well, flat-bottomed microplate.

After 30 minutes incubation at 37oC, diazo-reagents were added to stop the reaction. The optical density was recorded at 540 nm. All samples of (1-3)-b-D-glucan were above the

lower limit of detection (LOD) of the Glucatell assay (5 pg/ml). Thirty-five of the 574 endotoxin samples were below the LOD of 0.0625 EU/ml and were recorded as LOD.

LOD values were divided by the square root of two for the data analyses.

As currently it is not clear whether the concentration (mg/g) or loading (mg/m2) unit better represents the actual exposure (Douwes et al., 2000a), results were reported in both measures - expressed as µg/g and mg/m2 for (1-3)-ß-D-glucan and EU/mg and

EU/m2 for endotoxin exposures.

4.2.4. Medical evaluation of infants

The medical evaluation of infants was performed during physician office visits at the average age of 13 months (range 11-18 months, of these 95% were less than 15 months old). Sensitization to both food and aeroallergens before age one is an important risk factor for development of persistent wheeze symptoms and asthma in children born to atopic parents (Rhodes et al., 2002). Thus, infants were tested for allergen sensitization by SPT to a panel of food (milk, egg) and 15 common indoor and outdoor aeroallergens

(7 pollen, 4 mold, cat, dog, German cockroach, house dust mite). An SPT(+) to at least one allergen was defined as a wheal >=3 mm larger than the saline control after 15 minutes (LeMasters et al., 2006; Cho et al., 2006; Campo et al., 2006). SPT positivity to at least one allergen (regardless whether food or aeroallergen) was defined as SPT(any), while SPT positivity to at least one aeroallergen was defined as SPT(aero). At clinic visits, the parents were interviewed regarding wheezing using questions adapted from the

ISAAC questionnaire for 4-5 years old (Beasley et al., 2006). The following outcome variables were used: recurrent wheezing (>= 2 episodes in the past 12 months), recurrent

wheezing combined with SPT(+), and allergen sensitization (a positive SPT to at least one aeroallergen and/or food antigen). The reference group for recurrent wheezing consisted of infants with <=1 wheezing episodes in the past 12 months regardless of the

SPT status. Two reference groups for the recurrent wheezing with allergen sensitization were used: the 1st reference group consisted of infants with <= 1 wheezing episodes and

SPT(-) status; and the 2nd reference group consisted of infants with <=1 wheezing episodes but SPT(+) status.

4.2.5. Data analysis

The associations between (1-3)-b-D-glucan and each health outcome were investigated for 574 infants. Histograms and quantile-quantile plots showed that (1-3)-b-

D-glucan levels were approximately log-normally distributed.

Univariate logistic regressions were initially performed to evaluate associations between wheezing outcomes and predictor variables believed, a priori, to be related to wheezing in infants (Gold et al., 1999). Predictor variables that were significant at the

20% level in the univariate analyses were initially included in the multivariate logistic regression analyses. This significance level was chosen to include covariates that were moderately correlated to wheeze outcomes. Variables, which maintained significance levels approximately equal to 5% in at least one wheeze model, and/or changed the regression coefficient (or standard error) of another variable by at least 15% when dropped from the model, were kept in the final model. (1-3)-b-D-glucan remained, a priori, in all models. Predictor variables that were evaluated, but not included in the final model, were loge endotoxin, interaction between loge endotoxin and (1-3)-b-D-glucan,

breastfeeding duration (<1, 1-24, >25 weeks), and number of dogs and/or cats in the home.

After univariate analyses, allergen sensitization (SPT+), recurrent wheeze, and recurrent wheeze with allergen sensitization were analyzed by multiple logistic regression analyses, in which loge (1-3)-b-D-glucan, mother’s smoking (average number of cigarettes per day smoked in the last 12 months), and number of siblings in the same household, were continuously modeled. Categorically coded day-care attendance (yes, no), either parent asthma (yes, no), gender, race (African-American, non-African-

American), visible mold in home (none, low, high), lower respiratory condition (none, at least one of whooping cough, croup, viral infections, bronchitis/ bronchiolitis, flu, pneumonia), and upper respiratory condition (none, at least one of cold, ear infection, sinus infection, strep throat, tonsillitis, colored drainage) were also modeled. Covariates were chosen using a backward elimination technique.

Odds ratios and 95% confidence intervals for continuous variables were obtained to estimate the odds of each health outcome for an infant having a high (75th percentile) versus low (25th percentile) value of the variable. These percentiles represent the interquartile range of the predictor variable. Thus, the reference value was the lower endpoint of the interquartile range. Odds ratios and 95% confidence intervals were obtained for categorical predictor variables, where the reference category was chosen to facilitate the interpretation of results.

Graphical interpretations of wheeze prevalence (both wheeze outcomes) versus

(1-3)-b-D-glucan showed a nonlinear relationship, which was modeled by dividing the range of (1-3)-b-D-glucan into 4 non-overlapping intervals, approximately equal to (1-3)-

b-D-glucan quartiles. On each interval, a third degree function of (1-3)-b-D-glucan was fitted to wheeze prevalence. Fitted curves were smooth at the points of connection and were constrained to be linear in the tails. This transformation is known as a restricted cubic spline function. It allowed parameter estimates to be obtained in order to estimate the effect of (1-3)-b-D-glucan on wheeze outcomes over the interquartile range of (1-3)- b-D-glucan values. Endotoxin was modeled in a similar manner.

Other analyses that were performed outside of the regression model included correlations between (1-3)-b-D-glucan and endotoxin by levels of visible mold and analysis of variance testing differences among means of log-transformed (1-3)-b-D- glucan levels in homes with visible mold levels 0, 1, and 2. The latter analysis was followed by a test of linear trend between increasing levels of (1-3)-b-D-glucan levels and visible mold. The analyses were performed using S-Plus software (Insightful Corp.,

Seattle, Washington, 2000).

4.3. RESULTS

4.3.1. Exposure and subject characteristics

The descriptive statistics for (1-3)-b-D-glucan and endotoxin levels are presented in Table 4-1. (1-3)-b-D-glucan levels in concentration units correlated significantly with those in loading units (Spearman correlation: r=0.69, p<0.001). The correlation between endotoxin and (1-3)-b-D-glucan was significant in loading units (r=0.51, p<0.001), and non-existent in concentration units (r=0.08, p=0.052). These values of correlation between endotoxin and (1-3)-b-D-glucan are similar to what has been reported before

(Douwes et al., 2006; Gehring et al., 2001), and are due to the variance in the amount of dust sampled.

The analysis of variance of three categorical levels of visible mold exposure (no, low, and high) showed no significant overall differences for the (1-3)-b-D-glucan exposure (respective geometric mean values for concentrations: 53.2, 57.4, and 49.7 mg/g, and for loadings: 17.2, 19.0, and 25.7 mg/m2; results not shown in table). The test of linear trend between increasing levels of (1-3)-b-D-glucan and visible mold was not significant either (both units). In order to check for seasonal variations of the (1-3)-b-D- glucan concentration as increase in such was reported in warm months (Thorn, 2001), an

ANOVA test of the (1-3)-b-D-glucan concentrations among four seasons was performed.

No seasonal variation was found (F=1.82, p=0.14).

Among the infants, 169 (29.4%) were sensitized to at least one allergen (food and/or aeroallergen), 114 (19.9%) had recurrent wheezing (defined as two or more episodes in the last 12 months) and 41 (7.1%) had recurrent wheezing combined with a positive SPT to at least one allergen (Table 4-2). Day-care attendance, parents without asthma, African-American race, and no siblings were associated with lower prevalence of recurrent wheeze and/or recurrent wheeze with allergen sensitization. As expected, those who experience lower and/or upper respiratory condition, also tend to wheeze more.

4.3.2. Infantile wheezing

First, the association between the wheezing outcomes and (1-3)-b-D-glucan quartile exposure in concentration unit (mg/g) was investigated (Table 4-3). This analysis showed strong significant inverse associations between recurrent wheezing and (1-3)-b-

D-glucan exposure. Recurrent wheezing was significantly less likely among infants with very high (1-3)-b-D-glucan exposure levels (=61µg/g). In contrast, recurrent wheezing with or without allergen sensitization was positively associated with (1-3)-b-D-glucan exposure in the 1st quartile. The recurrent wheezing combined with allergen sensitization

[SPT(any)] was also significantly less likely in infants exposed to high (1-3)-b-D-glucan concentrations (=61µg/g). Similar trend was observed also for recurrent wheezing combined with sensitization to aeroallergens only [SPT(aero)}, but was not statistically significant (data not shown).

Data were also analyzed using loading unit (mg/m2) for (1-3)-b-D-glucan exposure (Table 4-4). Similar to the data in concentration units, an increase in wheezing outcomes when (1-3)-b-D-glucan exposure levels were low (1st and 2nd interquartile range), and a decrease when (1-3)-b-D-glucan exposure levels were high (=19µg/m2) (3rd and 4th interquartile range) was observed (Table 4-4). These associations however, were mostly non-significant. Only the inverse association between recurrent wheezing combined with allergen sensitization and exposure to high (1-3)-b-D-glucan loading

(within 4th quartile) was statistically significant.

After stratification by sensitization group, the inverse association between (1-3)- b-D-glucan exposure and recurrent wheeze with allergen sensitization became stronger in the group of sensitized wheezers compared to sensitized non-wheezers (Tables 4-3 and 4-

4, last columns), than those compared to the non-sensitized non-wheezers (µg/g unit only).

Visually, the association between continuously measured log-transformed (1-3)- b-D-glucan and wheezing is presented in Fig.4-1. A curvilinear relationship was found

which was approximated by fitting a restricted cubic spline (rcs) function to the wheezing data and the levels of (1-3)-b-D-glucan. Points where contiguous curves meet are called knots. The spline functions that best fit the data had three knots. The cubic spline function was modeled allowing for 3 turning points (knots) in the curvilinear relationship between (1-3)-b-D-glucan and log(odds of wheeze). Knots were located at the 5%, 50%, and 95% percentiles of (1-3)-b-D-glucan.

The visual presentation of data prompted us to perform an additional analysis of the odds of wheeze when (1-3)-b-D-glucan was equal to the value at which the predicted value of wheeze turns around (approximately the midpoint of the range: 60 mg/g,

19mg/m2) compared to the minimum value. The same analysis was also done when (1-3)- b-D-glucan was equal to the maximum value (900 mg/g, 2966 m? g/m2) compared to the midpoint. As expected from Fig. 4-1, the logistic regression analyses showed that (1-3)- b-D-glucan levels below the turning point were associated with increase in the wheeze outcomes (3-60 mg/g: recurrent wheeze aOR 3.90, 95%CI 1.26-12.09, recurrent wheeze with allergen sensitization: aOR 6.05, 95% CI 0.84-43.79), (0.2-19 mg/m2: recurrent wheeze aOR 1.41, 95%CI 0.41-4.87, recurrent wheeze with allergen sensitization aOR

9.23, 95% CI 0.85-99.99). (1-3)-b-D-glucan levels above the turning point were associated with decrease in the wheeze outcomes (60-900 mg/g: recurrent wheeze aOR

0.32, 95%CI 0.11-0.97, recurrent wheeze with allergen sensitization aOR 0.08, 95% CI

0.01-0.59), (19-2966 mg/m2: recurrent wheeze aOR 0.80, 95%CI 0.21-3.08, recurrent wheeze with allergen sensitization aOR 0.08, 95% CI 0.01-0.97). Therefore, the conclusions of inverse association between increase in (1-3)-b-D-glucan levels above the midpoint and wheeze outcomes (Tables 4-3 and 4-4) were confirmed.

Several studies have pointed out the potential modifying effects of co-exposure to

(1-3)-b-D-glucan and endotoxin (Fogelmark et al., 1994; Young et al., 2002), and how both may influence the wheeze outcome in children (Belanger et al., 2003; Schram-

Bijkerk et al., 2005; Schram et al., 2005; von Mutius et al., 2000; Gereda et al., 2000;

Rylander et al., 1998b). Thus the interaction effect between (1-3)-b-D-glucan and endotoxin was tested for, and as such was not found in the wheezing outcomes, endotoxin was included as a confounder in the wheezing analyses. Two types of analyses were performed – with and without endotoxin included in the final logistic models. The results were similar, and endotoxin showed no effect on recurrent wheezing or allergen sensitization. In order to test whether it is endotoxin, rather than (1-3)-b-D-glucan responsible for the observed results, endotoxin was also modeled in quartiles, while the

(1-3)-b-D-glucan level was held fixed (25th-75th percentile). No associations between endotoxin and any health outcomes were present. Due to the scientific interest in endotoxin, the data are reported with endotoxin in the model.

As literature suggests that (1-3)-b-D-glucan may be related to visible mold

(Gehring et al., 2001), the models with no visible mold as a confounder, were also run, in order to check for over-adjustment. Similar results were obtained, the trend of significant inverse association was preserved, however the 95% CI became wider. As in this study no correlation between (1-3)-b-D-glucan and visible mold was found, and the role of (1-

3)-b-D-glucan as a surrogate of mold exposure is uncertain due to various sources of the

(1-3)-b-D-glucan, such as pollen and plants (Rylander et al., 1999e), the full models including visible mold as a confounder are reported.

Among the other covariates, high visible mold, mother’s smoking, parent’s asthma, Afro-American race, siblings, as well as other lower and upper respiratory conditions were risk factors for both recurrent wheezing and recurrent wheezing combined with allergen sensitization (Tables 4-3 and 4-4).

4.3.3. Allergen sensitization

Among the infants 29.4% were sensitized to any allergen (either food or aeroallergen); 4.3% were sensitized to food only and 13.7% to aeroallergens only.

Exposure to high (1-3)-b-D-glucan concentrations and loadings had borderline significant inverse association with allergen sensitization assessed by SPT(any) [mg/g: within 3rd quartile, aOR=0.89, 95%CI 0.74-1.06; within 4th quartile, aOR=0.57, 95%CI 0.30-1.10 and mg/m2: within 3rd quartile, aOR=0.90, 95% CI 0.76-1.06; within 4th quartile, aOR=0.48, 95% CI 0.18-1.26]. Interaction effect between (1-3)-b-D-glucan and endotoxin was also tested for, and such was found in the outcome including allergen sensitization to any allergens, but only in per square meter unit (p=0.02). Thus, the interaction effect in the SPT(any) model was included when the exposure was per square meter unit: within 3rd quartile, aOR=1.33, 95% CI 0.83-2.15; within 4th quartile, aOR=152.62, 95% CI 6.57-3543.18. The latest result, however, may be due to the strong and significant correlation between (1-3)-b-D-glucan and endotoxin in this unit, and thus whether the result is attributed to (1-3)-b-D-glucan or endotoxin, or both, is unclear.

None of the other covariates in the multivariate analysis were significantly associated with positive SPT.

There were no significant associations between (1-3)-b-D-glucan and allergen sensitization assessed by SPT(aero) [(mg/g: within 4th quartile, aOR=1.07; 95% CI=0.71-

1.62) and (mg/m2: within 4th quartile, aOR=1.21; 95%CI=0.77-1.89)].

4.4. DISCUSSION

This study demonstrated a significant inverse association between increasing exposure to high (1-3)-b-D-glucan concentrations (>60µg/g) and recurrent wheezing

(aOR=0.39, 95% CI 0.16-0.93) in high risk infants. Even stronger association was found for recurrent wheezing combined with allergen sensitization (aOR=0.13, 95% CI 0.03-

0.61). Although others have reported a weak inverse relationship with (1-3)-b-D-glucan and atopic wheeze in children aged 1-4 and 5-15 years (Schram-Bijkerk et al., 2005;

Douwes et al., 2006), this is the first study to demonstrate a statistically significant relationship between high (1-3)-b-D-glucan exposure and reduced wheezing in infants.

No significant differences in (1-3)-b-D-glucan concentrations or loadings between the homes in three visible mold categories were found, although increasing trend was seen in (1-3)-b-D-glucan loadings. This finding is in line with the findings of Gehring et al. (2001) who only found difference in (1-3)-b-D-glucan loading, and not in concentration, between homes in two visible mold categories. The difference in correlations when exposures are assessed in per gram versus per square meter units indicates that the variance in levels per square meter was largely determined by the amount of dust sampled. Thus after transformation from concentration to loading unit, there was a decrease in the between samples variation.

Previous studies have shown that (1-3)-b-D-glucan concentrations do not consistently correlate with total culturable mold spore counts (Gehring et al., 2001;

Douwes et al., 1998; Wan and Li, 1999), and more likely reflect exposure from multiple environmental sources of (1-3)-b-D-glucan, including mold, pollen, plants, and their fragments (Rylander et al., 1999e; Foto et al., 2004). Furthermore, (1-3)-b-D-glucan content varies between mold species (Foto et al., 2004), and this may lead to variance of health outcomes by fungal genera (Osborne et al., 2006). For example, in this cohort, an inverse association between the concentration of airborne Cladosporium and SPT(+) to any allergen (p < 0.05), and Cladosporium and SPT(+) to aeroallergens (p < 0.05) was also found. However, positive associations between Penicillium/Aspergillus and SPT(+) to any allergen (p < 0.01), and Alternaria and SPT(+) to any allergen (p < 0.01) were also observed (Osborne et al., 2006). Molds contain (1-3)-b-D-glucan but also number of other agents such as sugars and enzymes. Apart from the fungi themselves mold growth is often associated with growth of other microbes. Thus, no conclusions concerning causal relationships between (1-3)-b-D-glucan and mold can be made. However, it seems

(1-3)-b-D-glucan may be an independent measure of biologically active exposure. This may explain why visible mold exposure, unlike (1-3)-b-D-glucan, was a risk factor for recurrent wheezing in this cohort (Cho et al., 2006a), as well as allergic rhinitis and upper respiratory infections (Biagini et al., 2006).

This study demonstrated that low levels of (1-3)-b-D-glucan exposure are associated with increase in the prevalence of recurrent wheeze, while the opposite was observed for the exposure to high levels of (1-3)-b-D-glucan. Similar trend was also found for the health outcome including wheeze combined with allergen sensitization.

Children are a special group of interest as their immune system is evolving. Thus they may be more susceptible to environmental exposure factors (Holt et al., 2005). In fact, research has shown (Holt et al., 1999) that the effective priming of aeroallergen-specific memory T-cells is initiated during infancy and is consolidated by the end of preschool years in relation to the Th1-Th2 balance. An interesting observation that Th2 (allergic) priming is preferentially favored by low-dose antigen exposure, whereas higher doses favor Th1 priming, was made by Constant et al (1995) and Rogers et al (1999).

Moreover, epidemiological studies on exposure to indoor allergens have revealed a biphasic pattern in which sensitization risk increases with exposure levels until a plateau is reached, above which risk decreases with further increase in exposure (Platts-Mills et al., 2001; Cullinan et al., 2004; Holt et al., 2005). This same pattern was demonstrated in this study.

To date, knowledge on health effects of (1-3)-b-D-glucan in children is limited.

Rylander et al. (1998b) reported that both non-atopic and atopic wheezing in school children (age 6-13 years) were positively correlated with airborne (1-3)-b-D-glucan concentrations, analyzed by the LAL assay. Douwes et al. (2006) also showed that increased levels of dust-borne (1-3)-b-D-glucan were positively related to variability in peak expiratory flow in non-atopic and atopic children 7-11 years of age. In contrast,

Schram-Bijkerk et al. (2005) and Douwes et al. (2006) reported that dust-borne (1-3)-b-

D-glucan concentrations were inversely associated with atopic wheeze in 5-13 year old children (OR=0.83; 95%CI=0.71-1.01) and asthma in 1-4 years old children (OR=0.61;

95%CI=0.28-1.33). All these studies, similarly to this one, showed that the health outcome was significantly associated with exposure in the subgroup of atopic children.

Interestingly, in the latter two studies (Schram-Bijkerk et al, 2005; Douwes et al., 2006) inclusion of both (1-3)-b-D-glucan and endotoxin in the models lead to loss of significance of either the effect of endotoxin (Schram-Bijkerk et al., 2005) or (1-3)-b-D- glucan (Douwes et al., 2006). These two studies used the inhibition immuno assay (EIA) and had a large number of samples below the detection limit for (1-3)-b-D-glucan, which could have caused an exposure misclassification. This was a major limitation in these studies, as pointed out by the authors themselves. This low limit of detection and the consequent loss of samples is a major drawback of the EIA assay. So far, there are no published studies comparing the EIA and the LAL assay (used in the current study and by

Rylander and Lin (2000). As shown in specific aims 1 and 2 that these two methods give different loading of (1-3)-b-D-glucan content of same fungal species, comparison of results obtained by these two assays is based on uncertain grounds.

In addition, these two previous studies reported a strong and significant correlation between endotoxin and (1-3)-b-D-glucan. Thus it is uncertain whether the observed effects were driven by endotoxin, (1-3)-b-D-glucan or both. No association was found between the investigated health outcomes and endotoxin exposure. This finding agrees with the results of Schram-Bijkerk et al. (2005) who showed that endotoxin loses significance after adjusting for (1-3)-b-D-glucan. This finding contradicts the study results reported by Douwes et al. (2006). A separate analysis focused on the endotoxin exposure in this cohort showed that endotoxin was not independently associated with wheezing or allergen sensitization. High endotoxin exposure in the presence of multiple dog ownership was, however, associated with reduced wheezing (Campo et al., 2006).

Furthermore, the lack of correlation between (1-3)-b-D-glucan and endotoxin

concentrations in this study database strengthens the finding that the effect seen for (1-3)- b-D-glucan exposure is not confounded by the endotoxin exposure.

These findings that homes with high (1-3)-b-D-glucan have less recurrent wheezing in sensitized infants, and that (1-3)-b-D-glucan plays more important role than endotoxin in early life wheezing is consistent with the study by Schram-Bijkerk et al.

(2005) and Douwes et al. (2006). A possible explanation for the stronger inverse association found in this study may be attributable to the younger age and selection of high-risk cohort (born to atopic parents). In this cohort, an inverse association between the concentration of airborne Cladosporium and allergen sensitization was found

(Osborne et al, 2006). Indeed this data is somewhat consistent with the hygiene hypothesis, which postulates that exposure to microbial products (such as endotoxin) early in life favors modification of Th2 directed immune responses to Th1 (von Mutius et al., 2000). Studies on exposure to high endotoxin and increased or decreased frequency of wheezing during infancy have shown conflicting results (von Mutius et la., 2000; Gereda et al., 2000; Park et al., 2001). Possible explanation for this controversy may be that these studies did not concomitantly assess indoor (1-3)-b-D-glucan levels. However, several studies revealed a strong adjuvant activity of (1-3)-b-D-glucan on the systemic allergic immune response in animal models (Rylander and Holt, 1998a; Ormstad et al., 2000;

Leaderer et al., 2002). Therefore it must be emphasized that the immunologic impact of

(1-3)-b-D-glucan exposure on Th2 directed atopic disorders remains uncertain and requires further investigation.

The above findings were further supported by the inverse trend between (1-3)-b-

D-glucan exposure and allergen sensitization. This and the stronger inverse association

between (1-3)-b-D-glucan and wheezing in allergen sensitized infants compared to wheezing in all infants, even stronger after stratification by sensitization, suggests that (1-

3)-b-D-glucan could modify allergic respiratory responses in infants. It must be emphasized that it is uncertain if this early observed effect will impact the risk of later development of childhood asthma among this cohort of sensitized infants with recurrent wheezing as health outcomes associated with (1-3)-b-D-glucan exposure determined in this infant population may be transient.

The hygiene hypothesis states that infections re-enforce the physiological mechanisms of natural dominant tolerance through the expansion of naturally occurring regulatory T cells. Pro-inflammatory ligands of Toll-like receptors expressed by natural

T-regulatory cells, both of microbial (e.g., lipopolysaccharide, peptidoglycans) and endogenous (e.g., stress proteins) origin, may play a critical role in their activation and expansion (Demengeot et al., 2006). The results from tspecific aim 4 suggest the (1-3)-b-

D-glucan as one more candidate of microbial exposure that possibly increases the activity of the T-regulatory cells via the Toll-like receptors.

To my knowledge, there is only one previous study reporting the associations between (1-3)-b-D-glucan concentration versus loading units with health outcomes.

Douwes et al. showed stronger association between peak flow variability in children and dust (1-3)-b-D-glucan in the loading unit than in the concentration unit (Douwes et al.,

2003b). Similar trends between the health outcomes and (1-3)-b-D-glucan exposures in the two units were found, but the associations were stronger for the concentration unit.

These differences may be due to the different health outcome or different (1-3)-b-D-

glucan analysis methods used: Douwes et al. analyzed (1-3)-b-D-glucan content by the

EIA, while this study used the LAL.

Dust samples were collected only from the baby’s primary activity room, as it is most representative of its exposure. Despite the fact that infants spend most of their time in bed, a recent European study (Douwes et al., 2006) has shown association only between living area exposure and asthma in 4 years old children, but none with mattress exposure. This may be due to the fact that the main living room allergen levels for dust allergens were higher than that in the bedding in the infants’ homes (Leaderer et al.,

2002) and because mattresses for infants were purchased new (use of less than 3 months)

(Douwes et al., 2006). In addition, the collection of dust by trained technicians using a standardized protocol and the same type of vacuum cleaner is strength of the present study as it decreases collection bias. Further, home exposure assessment was conducted almost concurrently with health outcome assessment.

A potential limitation of the study generalizability is that the cohort consists only of infants born to atopic parents. However, the Third National Health and Nutrition

Examination Survey (NHANES III, Arbes et al., 2005), 1988-94, designed to obtain nationally representative information on the health of the population of the United States, showed that 54.3% of the U.S. population was sensitized to one or more allergens. This finding suggests that the results observed in the current study can be widely generalized to over 50% of the U.S. population. Research has shown that exposure to maternal smoking and aeroallergens (Upham and Holt, 2005) impact the development of allergen sensitization as early as in pregnancy. Thus, the absence of antenatal exposure data may be another study limitation. In addition, the low number of recurrent wheezers with

positive SPT also limits the power of the study. Unfortunately, as the CCAAPS study did not initially include the analysis for (1-3)-b-D-glucans, only samples with sufficient dust amount left were analyzed, after analyses for all other allergens were completed.

Although this introduces a selection bias, and may limit the generalizability of the results, still 76% of the samples were analyzed. In addition, the amount of dust and wheeze outcomes did not correlate (p=0.32). Therefore, dust amount is not expected to bias the results.

4.5. CONCLUSIONS FOR SPECIFIC AIM 4

In conclusion, it was found that the concentration of (1-3)-b-D-glucan is a measure of biological exposure that is independent from observed visible mold. A significant inverse association was found between high (1-3)-b-D-glucan levels and recurrent wheezing. This association was even stronger in a subgroup of allergen- sensitized infants. It seems that exposure to high (1-3)-b-D-glucan levels (>=61µg/g,

>=19µg/m2) may be conducive to reduced wheezing in infants at high risk for developing asthma (as those born to atopic parents). Long-term follow-up of this cohort will help determine how early (1-3)-b-D-glucan exposure affects the development of phenotypes such as atopy, allergic rhinitis and asthma in later childhood.

OVERALL CONCLUSIONS (PRACTICAL IMPLICATIONS)

In conclusion, the LAL assay is more specific, sensitive and accurate than the EIA assay in detecting both linear and branched (1-3)-b-D-glucans (specific aim 1).

Although the (1-3)-b-D-glucan concentration in the field samples measured by the

LAL and EIA assay correlated, data shall be analyzed with caution, as assays give different weight to different fungal species. In addition, the strong significant associations of the LAL-analyzed (1-3)-b-D-glucan with fungal spore size and with indoor dust fungal concentration, indicates that LAL-analyzed (1-3)-b-D-glucan could be used to estimate the total fungal load in indoor samples (specific aim 2). In addition, this study has shown that (1-3)-b-D-glucan content is representative of most prevalent fungal species, as well as those associated with moldy homes as determined by the ERMI index.

The hygiene hypothesis states that microbial components, such as endotoxins, flagelins, peptidoglycans, re-enforce the physiological mechanisms of natural dominant tolerance through the expansion of naturally occurring regulatory T cells. Pro- inflammatory ligands of Toll-like receptors expressed by natural T-regulatory cells, both of microbial and endogenous origin, may play a critical role in their activation and expansion. The results from this Ph.D. research suggest the (1-3)-b-D-glucan as one more candidate of microbial exposure that possibly increases the activity of the T-regulatory cells via the Toll-like receptors.

This study has shown that all three most common Cladosporium species (C. cladosporioides, C. herbarum and C. sphaerospermum) contribute to the (1-3)-b-D-

glucan loading in dust samples (specific aim 3). A recent study of the same cohort

(Osborne et al, 2006) have found Cladosporium genus to be inversely associated with allergen sensitization. In this same infant cohort, the current PhD thesis found an inverse association between exposure to high (1-3)-b-D-glucan and recurrent wheeze, recurrent wheeze with allergen sensitization, and allergen sensitization (specific aim 4). Thus, it can be speculated that fungal species of high (1-3)-b-D-glucan content, such as

Cladosporium genus, Epicoccum nigrum, Wallemia sebi, and Penicillium brevicompactum (determined in specific aims 2 and 3), may be associated with lower allergen sensitization, too. On the other hand, species that do not significantly contribute to (1-3)-ß-D-glucan in field samples, such as Alternaria alternata, and most species of the Aspergillus and Penicillium genera will not be expected to have such a protective effect on wheeze or allergen sensitization. In addition, the Aspergillus species tested in this study had the lowest (1-3)-b-D-glucan content among the 12 tested fungal species.

Aspergillus and Penicillium, as well as A. alternata are recognized as allergenic fungi due to their allergenic antigens. Thus, it can be speculated that a combination of low (1-3)-b-

D-glucan content in these species, and their allergenic antigens may cause increase in the prevalence of allergenic sensitization among infants. This might explain the different direction of associations observed between health outcomes and the concentrations of

Cladosporium vs. Aspergillus, Penicillium and Alternaria.

FUTURE DIRECTIONS

· Follow-up of the cohort in order to asses the development of atopy in relation to early

life exposure to (1-3)-ß-D-glucan

· Simultaneous assessment of the association between LAL vs. EIA-analyzed (1-3)-ß-

D-glucan exposure and wheeze/allergen sensitization in children

· ?lpha-glucans have been associated with host immune reactions. In some species,

such as Aspergillus fumigatus, the alpha-glucans consist 40% of the dry weight of the

fungal cell wall (Stone and Clark, 1992). However, the role of (1 3)- -glucan is still

controversial. On the one hand, -glucan appears to be a nonessential component of

the cell wall in Aspergillus nidulans and Schizophyllum commune (Sietsma and

Wessels, 1988; Zonneveld et al., 1973). On the other hand, Beauvais and Latgé

(2001) suggested that cell wall -glucan may have a structural role in the pathogenic

fungus. Thus analysis of a-glucans exposure and association with atopy development

in children, may answer the question whether a-glucans play a role in the immune

system development.

· Study the (1-3)-ß-D-glucan content of all fungal species that are of frequency and

concentration in indoor samples >50th percentile.

· Study both the (1-3)-ß-D-glucan content and allergenic antigens of Alternaria

alternata.

· Analysis of which fungal species contribute most to (1-3)-ß-D-glucan concentrations

in air field samples (as fungi in dust samples are also the ones of greatest spore size,

and smaller size fungal spores may not be taken into consideration).

· Estimate the (1-3)-ß-D-glucan content of mycelia vs. spores of Cladosporium species,

Epiccoccum nigrum, Wallemia sebi, Penicillium brevicompactum (and other common

indoor fungal species identified as having high (1-3)-ß-D-glucan spore content).

REFERENCES:

Adachi Y, Ohno N, Yadomae T. Preparation and antigen specificity of an anti-(1-->3)- beta-D-glucan antibody. Biol Pharm Bul 1994;17:1508-12.

Adhikari A, Martuzevicius D, Reponen T, Grinshpun S, Cho S, Sivasubramani S, Zhong,

W, Levin L, Kelly A, St. Clair H. and LeMasters G. Performance of the button personal inhalable sampler for the measurement of outdoor aeroallergens. Atmos Environ

2003;37:4723–4733.

Alwis K, Mandryk J, Hocking A. Exposure to biohazards in wood dust: bacteria, fungi, endotoxins, and (1-3)-b-D-glucans. Applied Occup Environ Hyg 1999;14:598-608.

Arbes SJ Jr, Gergen PJ, Elliott L, Zeldin DC. Prevalences of positive skin test responses to 10 common allergens in the US population: results from the third National Health and

Nutrition Examination Survey. J Allergy Clin Immunol 2005;116:377-83.

Bauer H, Kasper-Giebl A, Zibuschka F, Hitzenberger R, Kraus G, Puxbaum H.

Determination of the carbon content of airborne fungal spores. Anal Chem 2002;74:91-5.

Beasley R. The International Study of Asthma and Allergies in Childhood (ISAAC)

Steering Committee pp. 1225-1232 Worldwide variation in prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and atopic eczema: ISAAC. Lancet 1998;351:1225-

32.

Beauvais, A. and Latgé, J.-P. Membrane and cell wall targets in Aspergillus fumigatus.

Drug Resist. Updat. 2001;4:38–49.

Beijer L, Thorn J, Rylander R. Mould exposure at home relates to inflammatory markers in blood. Eur Respir J 2003;21:317-22.

Belanger K, Beckett W, Triche E, Bracken MB, Holford T, Ren P et al. Symptoms of wheeze and persistent cough in the first year of life: associations with indoor allergens, air contaminants, and maternal history of asthma. Am J Epidemiol 2003;158:195-202.

Biagini JM, LeMasters GK, Ryan PH, Levin L, Reponen T, Bernstein DI, Villareal M,

Khurana Hershey GK, Burkle J, Lockey J. Environmental risk factors of rhinitis in early infancy. Pediatr Allergy Immunol 2006;17:278-84.

Brown GD, Gordon S. Immune recognition of fungal beta-glucans. Cell Microbiol

2005;7:471-9.

Campo P, Kalra HK, Levin L, Reponen T, Olds R, Lummus ZL et al. Influence of dog ownership and high endotoxin on wheezing and atopy during infancy. J Allergy Clin

Immunol 2006 (in press).

Chaffin W, Lopez-Ribot j, Casanova M, Gozalbo D, Martinez J. Cell-wall and secreted proteins of Candida albicans: identifications, function, and expression. Microbiol Mol

Biol Rev 1998;62:130-180.

Chew G, Douwes J, Doekes G, Higgins K, Strien R, Spithoven J, Brunekreef B. Fungal extracellular polysaccharides, b(1,3)-glucans and culturable fungi in repeated sampling of house dust. Indoor Air 2001;11:171-178.

Cho S-H, Reponen T, LeMasters G, Levin L, Huang J, Meklin T et al. Mold damage in homes and wheezing in infants. Ann Allergy Asthma Immunol 2006a;97:539-45.

Cho S-H, Reponen T, Levin L, Liu X, Bernstein D, Olds R, Storrs A, Wilson K,

LeMasters G. The effect of home characteristics on dust allergen levels in homes. Sci

Total Environ 2006b;371:31-43.

Constant S, Pfeiffer C, Woodard A, Pasqualini T, Bottomly K. Extent of T cell receptor ligation can determine the functional differentiation of naive CD4+ T cells. J Exp Med

1995;182:1591-6.

Cullinan P, MacNeill S, Harris JM, Moffat S, White C, Mills P, Newman Taylor A. Early allergen exposure, skin prick responses, and atopic wheeze at age 5 in English children: a cohort study. Thorax 2004;59:855-61.

Curtis L. Adverse Health Effects of Indoor Molds. J Nutr Environ Med 2004;14:261.

Diamandis E, Christopoulos T. Immunoassay. San Diego: Academic Press Inc 1996.

De la Cruz J, Pintor-Toro J, Benitez T, Llobell A, Romero L. A novel endo-beta-1,3- glucanase, BGN13.1, involved in the mycoparasitism of Trichoderma harzianum. J

Bacteriol. 1995;177:6937-45.

Demengeot J, Zelenay S, Moraes-Fontes MF, Caramalho I, Coutinho A. Regulatory T cells in microbial infection. Springer Semin Immunopathol. 2006;28:41-50.

Douwes J, Doekes G, Heinrich J, Koch A, Bischof W, Brunekreef B. Endotoxin and ß(1-

3)-Glucan in house dust and the relation with home characteristics: a pilot study in 25

German houses. Indoor Air 1998;8:255-263.

Douwes J, Doekes G, Montijn R, Heederik D, Brunekreef B. Measurement of b(1,3)- glucans in occupational and home environments with an inhibition enzyme immunoassay. Appl Environ Microbiol 1999;62:3176-3182.

Douwes J, Wouters I, Dubbeld H, van Zwieten L, Steerenberg P, Doekes G, Heederik D.

Upper airway inflammation assessed by nasal lavage in compost workers: A relation with bio-aerosol exposure. Am J Ind Med 2000a;37:459-68.

Douwes J, Zuidhof A, Doekes G, van der Zee SC, Wouters I, Boezen MH et al. (1-->3)- beta-D-glucan and endotoxin in house dust and peak flow variability in children. Am J

Respir Crit Care Med 2000b;162:1348-54.

Douwes J, Thorne P, Pearce N, Heederick D. Bioaerosol health effects and exposure assessment: progress and prospects. Ann Occup Hyg 2003a;47:187-200.

Douwes J, Pearce N. Invited commentary: is indoor mold exposure a risk factor for asthma? Am J Epidemiol 2003b;158:203-6.

Douwes J. 1-3-b-D-glucans and respiratory health: a review of the scientific evidence.

Indoor Air 2005;15:160-169.

Douwes J, van Strien R, Doekes G, Smit J, Kerkhof M, Gerritsen J, Postma D, de Jongste

J, Travier N, Brunekreef B. Does early indoor microbial exposure reduce the risk of asthma? The Prevention and Incidence of Asthma and Mite Allergy birth cohort study.

J Allergy Clin Immunol 2006;117:1067-73.

Edmondson DA, Nordness ME, Zacharisen MC, Kurup VP, Fink JN. Allergy and "toxic mold syndrome". Ann Allergy Asthma Immunol. 2005;94:234-9.

Fogelmark B, Sjostrand M, Rylander R. Pulmonary inflammation induced by repeated inhalations of beta(1,3)-D-glucan and endotoxin. Int J Exp Pathol 1994;75:85-90.

Fogelmark B, Rylander R. (1®3)-b-D-glucan in some indoor air fungi. Indoor Built

Environ 1997;6:291-294.

Fogelmark B, Thorn J, Rylander R. Inhalation of (1-->3)-beta-D-glucan causes airway eosinophilia. Mediators Inflamm 2001;10:13-9.

Foto M, Plett J, Berghout J, Miller J. Modification of the Limulus amebocyte lysate assay for the analysis of glucan in indoor environments. Anal Bioanal Chem

2004;379:156-62.

Gent JF, Ren P, Belanger K, Triche E, Bracken MB, Holford TR et al. Levels of household mold associated with respiratory symptoms in the first year of life in a cohort at risk for asthma. Environ Health Perspect 2002;110:781-6.

Gehring U, Douwes J, Doekes G, Koch A, Bischof W, Fahlbusch B, Richter K,

Wichmann H, Heinrich J. b(1,3)-glucan in house dust of German homes: Housing characteristics, occupant behavior, and relations with endotoxins, allergens and molds.

Environ Health Perspect 2001;109:139- 144.

Gereda JE, Leung DY, Thatayatikom A, Streib JE, Price MR, Klinnert MD et al. Relation between house-dust endotoxin exposure, type 1 T-cell development, and allergen sensitization in infants at high risk of asthma. Lancet 2000;355:1680-3.

Gladding T, Thorn J, Stott D. Organic dust exposure and work-related effects among recycling workers. Am J Ind Med 2003;43:584-91.

Gold DR, Burge HA, Carey V, Milton DK, Platts-Mills T, Weiss ST. Predictors of repeated wheeze in the first year of life: the relative roles of cockroach, birth weight, acute lower respiratory illness, and maternal smoking. Am J Respir Crit Care Med

1999;160:227-36.

Heldal K, Halstensen A, Thorn J, Eduard W, Halstensen T. Airway inflammation in waste handlers exposed to bioaerosols assessed by induced sputum. Eur Respir J

2003;21:641-5.

Helsel D.R. Nondetects and Data Analysis, Statistics for Censored Environmental Data.

Wiley and Sons, Inc. NY, NY. Environmental Data. Wiley and Sons Inc. Hoboken, NJ.

2005.

Hogan L, Klein B. Altered expression of surface alpha-1,3-glucan in genetically related strains of Blastomyces dermatitidis that differ in virulence. Infect Immun 1994;62:3543-6.

Holt PG, Macaubas C, Stumbles PA, Sly PD. The role of allergy in the development of asthma. Nature 1999;402:B12-7.

Holt PG, Upham JW, Sly PD. Contemporaneous maturation of immunologic and respiratory functions during early childhood: implications for development of asthma prevention strategies. Review. J Allergy Clin Immunol 2005;116:16-24.

Hossain M, Ahmed M, Ghannoum M. Attributes of Stachybotrys chartarum and its association with human disease. J Allergy Clin Immunol 2004;113:200-208.

Institute of Medicine. Damp Indoor Spaces and Health. The National Academies Press,

Washington, D.C. USA.2004;65-66.

Ishibashi K, Miura NN, Adachi Y, Tamura H, Tanaka S, Ohno N. The solubilization and biological activities of Aspergillus beta-(1 --> 3)-D-glucan. FEMS Immunol Med

Microbiol 2004;42:155-66.

Johanning, E. Indoor moisture and mold related health problems Eur Ann Allergy Clin

Immunol 2004;36:182-185.

Kataoka K, Muta T, Yamazaki S, Takeshige K. Activation of macrophages by linear (1-

3)-beta-D-glucans. Implications for the recognition of fungi by innate immunity. J Biol

Chem 2002;277:36825-31.

Korpi A, Kasanen JP, Kosma VM, Rylander R, Pasanen AL. Slight respiratory irritation but not inflammation in mice exposed to (1-->3)-beta-D-glucan aerosols. Mediators

Inflamm 2003;12:139-46.

Lai K. Hazard identification, dose-response and environmental characteristics of stachybotryotoxins and other health-related products from Stachybotrys. Environ Technol

2006;27:329-35.

Leaderer B, Belanger K, Triche E, Holford T, Gold DR, Kim Y, Jankun T et al. Dust mite, cockroach, cat, and dog allergen concentrations in homes of asthmatic children in the northeastern United States: impact of socioeconomic factors and population density.

Environ Health Perspect 2002;110:419-25.

Lemasters GK, Wilson K, Levin L, Biagini J, Ryan P, Lockey JE, Stanforth S, Maier S,

Yang J, Burkle J, Villareal M, Khurana Hershey GK, Bernstein DI. High prevalence of aeroallergen sensitization among infants of atopic parents. J Pediatr 2006;149:505-511.

Lindsley W, Schmechel D, Chen B. A two-stage cyclone using microcentrifuge tubes for personal bioaerosol sampling. J Environ Monit 2006;8:1-7.

Mandryk J, Alwis K, Hocking A. Effects of personal exposure on pulmonary function and work-related symptoms among sawmill workers. Ann Occup Hyg 2000;44:281-289.

Milton D, Alwis U, Fisette L, and Muilenberg M. Enzyme-Linked Immunosorbant Assay

Specific for (1 6) Branched, (1 3)-b -D-Glucan Detection in Environmental Samples.

Appl Environ Microbiol 2001;67:5420-5424.

Miyazaki T, Kohno S, Mitsutake K, Maesaki S, Tanaka K, Ishikawa N, Hara K. Plasma

(1 ? 3)-beta-D -glucan and fungal antigenemiain patients with candidemia, aspergillosis, and cryptococcosis. J. Clin. Microbiol. 1995;33:3115-3118.

Obayashi T, Tamura H, Tanaka S, Ohki M, Takahashi S, Arai M, Masuda M, Kawai T. A new chromogenic endotoxin-specific assay using recombined limulus coagulation enzymes and its clinical applications. Clin Chim Acta 1985;149:55-65.

Odabasi Z, Paetznick V, Rodriguez J, Chen E, Mcginnis M, Ostrosky-Zeichner L.

Differences in beta-glucan levels in culture supernatants of a variety of fungi. Med Mycol

2006;44:267-272.

Ohno N, Miura N, Chiba N, Adachi Y, Yadomae T. Comparison of the mmunopharmacological activities of triple and single-helical Schizophyllan in mice. Biol

Pharm Bull 1995;18:1242-1247.

Ormstad H, Groeng EC, Lovik M, Hetland G. The fungal cell wall component beta-1,3- glucan has an adjuvant effect on the allergic response to ovalbumin in mice. J Toxicol

Environ Health 2000;61:55-67.

Osborne M, Reponen T, Adhikari A, Cho SH, Grinshpun SA, Levin L, Bernstein DI,

LeMasters G. Specific fungal exposures, allergic sensitization, and rhinitis in infants.

Pediatr Allergy Immunol 2006;17:450-457.

Park JH, Gold DR, Spiegelman DL, Burge HA, Milton DK. House dust endotoxin and wheeze in the first year of life. Am J Respir Crit Care Med 2001;163:322-8.

Perez P, Ribas J. Cell wall analysis. Methods 2004;33:245-251.

Platts-Mills T, Vaughan JW, Blumenthal K, Woodfolk J, Sporik R. Decreased prevalence of asthma among children with high exposure to cat allergen: relevance of the modified Th2 response. Mediators Inflamm 2001;10:288-91.

Rappleye CA, Goldman WE. Defining Virulence Genes in the Dimorphic Fungi. Annu

Rev Microbiol 2006;60:281-303.

Rementeria A, Lopez-Molina N, Ludwig A, Vivanco AB, Bikandi J, Ponton J, Garaizar J.

Genes and molecules involved in Aspergillus fumigatus virulence. Rev Iberoam Micol

2005;22(1):1-23.

Reponen T, Grinshpun S, Conwell K, Wiest J, Anderson M. Aerodynamic versus physical size of spores: measurement and implication for respiratory deposition. Grana

2001;40:119-125.

Reijula K. Two common fungi associated with farmer’s lung: fine structure of

Aspergillus fumigatus and Aspergillus umbrosus. Mycopathologia 1991;113:143-149.

Rhodes HL, Thomas P, Sporik R, Holgate ST, Cogswell JJ. A birth cohort study of subjects at risk of atopy: twenty-two-year follow-up of wheeze and atopic status. Am J

Respir Crit Care Med 2002;165:176-80.

Rogers P, Croft M. Peptide dose, affinity, and time of differentiation can contribute to the

Th1/Th2 cytokine balance. J Immunol 1999;163:1205-13.

Ryan P, LeMasters G, Biagini J, Bernstein D, Grinshpun S, Shukla R et al. Is it traffic type, volume, or distance? Wheezing in infants living near truck and bus traffic. J Allergy

Clin Immunol 2005;116:279-84.

Rylander R. Symptoms and mechanisms: Inflammation of the lung. Am J Ind Med

1994;25:19-23.

Rylander R. Airway responsiveness and chest symptoms after inhalation of endotoxin or

1,3, beta-d-glucan. Indoor Build Environ 1996;154:106-111

Rylander R. Airborne (1®3)-b-glucan and airway disease in a day-care center before and after renovation. Arch Environ Health 1997;52;281-285.

Rylander R, Holt PG. (1-->3)-beta-D-glucan and endotoxin modulate immune response to inhaled allergen. Mediators Inflamm 1998a;7:105-10.

Rylander R, Norrhall M, Engdahl U, Tunsater A, Holt PG. Airways inflammation, atopy, and (1--> 3)-beta-D-glucan exposures in two schools. Am J Respir Crit Care Med

1998b;158:1685-7.

Rylander R. Indoor air-related effects and airborne (1-3)-b-D-glucan. Environ Health

Perspect 1999a;107(S3), S501-S503

Rylander R, Etzel R. Introduction and summary workshop on children’s health and indoor mold exposure. Env Health Perspect 1999b;107 (S3)

Rylander R. Health effects among workers in sewage treatment plants. Occup Environ

Med 1999c;56:354-357.

Rylander R, Thorn J, Attefors R. Airways inflammation among workers in a paper industry. Eur Respir J 1999d;13:1151-1157.

Rylander R, Fogelmark B, McWilliam A, Currie A. (1-->3)-beta-D-glucan may contribute to pollen sensitivity. Clin Exp Immunol 1999e;115:383-4.

Rylander R, Lin RH. (1-->3)-beta-D-glucan - relationship to indoor air-related symptoms, allergy and asthma. Toxicology 2000;152:47-52.

Rylander R, Carvalheiro M. Airways inflammation among workers in poultry houses.

Int Arch Occup Environ Health 2006;79:487-90.

Sakurai T, Kaise T, Yadomae T, Matsubara C. Different role of serum components and cytokines on alveolar macrophage activation by soluble fungal (1-->3)-beta-D-glucan.

Eur J Pharmacol 1997;334:255-63.

San-Blas G, Guanipa O, Moreno B, Pekerar S, San-Blas F. Cladosporium carrionii and

Hormoconis resinae (C. resinae): cell wall and melanin studies. Curr Microbiol

1996;32:11-6.

Santos T, Sanchez M, Villanueva J, Nombela C. Derepression of b-1,3-glucanases in

Penicillium italicum: localization of the various enzymes and correlation with the cell wall glucan mobilization and autolysis. J Bacteriol 1979;137:6-12.

Schram-Bijkerk D, Doekes G, Douwes J, Boeve M, Riedler J, Ublagger E et al. Bacterial and fungal agents in house dust and wheeze in children: the PARSIFAL study. Clin Exp

Allergy 2005;35:1272-8.

Schram D, Doekes G, Boeve M, Douwes J, Riedler J, Ublagger E et al. Bacterial and fungal components in house dust of farm children, Rudolf Steiner school children and reference children--the PARSIFAL Study. Allergy 2005;60:611-8.

Schmechel D, Simpson J, Beezhold D, Lewis D. The development of species-specific immunodiagnostics for Stachybotrys chartarum: The role of cross-reactivity. J Immunol

Methods 2006;309:150-159.

Schmid F, Stone BA, McDougall BM, Bacic A, Martin KL, Brownlee RT, Chai E,

Seviour RJ. Structure of epiglucan, a highly side-chain/branched (1 --> 3;1 --> 6)-beta- glucan from the micro fungus Epicoccum nigrum. Carbohydr Res 2001;331:163-71.

Shematek E, Braatz J, Cabib E. J Biol Chem 1980;255:888-94.

Sietsma J. and Wessels J. Total inhibition of wall synthesis by 2-deoxyglucose and polyoxin D in protoplasts of Schizophyllum commune. Acta Bot. Neerl.1988;37:23–29.

Steele C, Rapaka R, Metz A, Pop S, Williams D, Gordon S, Kolls J, Brown G. The beta- glucan receptor dectin-1 recognizes specific morphologies of Aspergillus fumigatus.

PLoS Pathog 2005;1:42.

Stevens DA. Diagnosis of fungal infections: Current status. J Antimicrob Chemother

2002;49 (S1):11-19.

Stone BA, Clarke AE. Chemistry and Biology of (1-3)-ß-glucans. La Trobe University

Press, Bundoora, Australia. 1992

Storey, E., Dangman, K.H., Schenck, P., DeBernardo, R.L., Yang, C.S., Bracker, A., and

Hodgson, M.J. Guidance for Clinicians on the Recognition and Management of Health

Effects Related to Mold Exposure and Moisture Indoors. University of Connecticut

Health Center, Farmington, Connecticut, 2004. http://oehc.uchc.edu/clinser/MOLD%20GUIDE.pdf (accessed 01/31/05)

Tanaka S, Aketagawa J, Takahashi S, Shibata Y, Tsumuraya Y, Hashimoto Y. Activation of a limulus coagulation factor G by (1-3)-b-D-glucans. Carbohydr Res 1991;218:167-

174.

Thorn J, Rylander R. Airways inflammation and glucan in a row house area. Am J Respir

Crit Care Med 1998;157:1798-1803.

Thorn J. Seasonal variations in exposure to microbial cell wall components among household waste collectors. Ann Occup Hyg 2001;45:153-156.

Thorne P, Duchaine C, Douwes J, Eduard W, Gorny R, Jacobs R, Reponen T, Schierl R,

Szponar B. Working group report 4: exposure assessment for biological agents. Am J Ind

Med 2004;46:419-422.

Vesper SJ, Varma M, Wymer LJ, Dearborn DG, Sobolewski J, Haugland RA.

Quantitative polymerase chain reaction analysis of fungi in dust from homes of infants who developed idiopathic pulmonary hemorrhaging. J Occup Environ Med 2004;46:596-

601.

Vesper S, McKinstry C, Haugland R, Iossifova Y, LeMasters G, Levin L, Hershey GK,

Villareal M, Bernstein DI, Reponen T. EPA Relative Moldiness Index© as Predictor of

Childhood Respiratory Illness. J Expo Anal Environ Epi 2006 (in press)

von Mutius E, Braun-Fahrlander C, Schierl R, Riedler J, Ehlermann S, Maisch S et al.

Exposure to endotoxin or other bacterial components might protect against the development of atopy. Clin Exp Allergy 2000;30:1230-4.

Upham JW, Holt PG. Environment and development of atopy. Curr Opin Allergy Clin

Immunol 2005;5:167-72.

Yoshida M, Obayashi T, Iwama A, Ito M, Tsunoda S, Suzuki T, Muroi K, Ohta M,

Sakamoto S, Miura Y. Detection of plasma (1 ? 3)-beta-D -glucan in patients with

Fusarium, Trichosporon, Saccharomyces and Acremonium fungaemias. J. Med. Vet.

Mycol. 1997a;35:371-374.

Yoshida K, Niki Y, Ohno M, Watanabe S, Hashiguchi K, Nakajima M, Matsushima T.

Clinical significance of (1-->3)-beta-D-glucan in pleural effusion and liquor.

Kansenshogaku Zasshi 1997b;71:1210-5.

Young SH, Robinson VA, Barger M, Zeidler P, Porter DW, Frazer DG et al. Modified endotoxin responses in rats pretreated with 1-->3-beta-glucan (zymosan A). Toxicol Appl

Pharmacol 2002;178:172-9.

Young S-H and Castranova V. Toxicology of (1-3)-b-D-glucans. Glucans as a marker for fungal exposure. CRC Press Taylor&Francis group. 2005

Yuasa K, Goto H, Iguchi M, Okamura T, Ieki R. Evaluation of the diagnostic value of the measurement of (1-->3)-beta-D-glucan in patients with pulmonary aspergillosis.

Respiration 1996;63:78-83.

Walker GJ, Brown RA, Taylor C. Activity of Streptococcus mutans alpha-D- glucosyltransferases released under various growth conditions. J Dent Res. 1984;63:397-

400.

Wan, G. -H., Li C.S. Indoor endotoxin and glucan in association with airway inflammation and systemic symptoms. Arch Environ Health 1999;54:172-179.

Wouters I., Hilhorst S., Kleppe P., Doekes G., Douwes J., Peretz C., Heederik D. Upper airway inflammation and respiratory symptoms in domestic waste collectors. Occup

Environ Med 2002;59:106-112.

Zonneveld B. Inhibitory effect of 2-deoxyglucose on cell wall -1,3-glucan synthesis and cleistothecium development in Aspergillus nidulans. Dev. Biol. 1973;34:1–8.

3500 A. LAL Mannan, Dextran, Pullulan 3000

Schizophylan 2500 Laminarin 2000 Mg-glucan

1500

Mean Onset Time (min) 1000 Curdlan 500 Pachyman 0 0 20 40 60 80 100 120

1.2 B. EIA

1.0 Pullulan

0.8 Dextran Schizophylan 0.6 Mannan 0.4 Mg-glucan Pachyman 0.2 Absorbance units (450nm)

0.0 Curdlan Laminarin

0 10x106 20x106 30x106 40x106 50x106 60x106 (1-3)-b -D-glucan concentration (pg/ml) Fig. 1-1 Comparison of eight purified glucans and their reactivity as measured by (A) the kinetic LAL and (B) endpoint EIA assays.

105 g/g) m -D-glucan ( b 104

EIA -analyzed (1-3)- r=0.17, p=0.17

103 100 101 102 103 104

LAL-analyzed (1-3)-b-D-glucan (mg/g)

Fig. 1-2. LAL vs. EIA-analyzed (1-3)-ß-D-glucan in 70 dust samples (CCAAPS study) in mg/g.

106 r=0.57, p<0.001 ) 2 g/m

m 105 -D-glucan (

b 104

103 EIA-analyzed (1-3)-

102 10-1 100 101 102 103 LAL-analyzed (1-3)-b-D-glucan (mg/m2)

Fig. 1-3. LAL vs. EIA-analyzed (1-3)-ß-D-glucan in 70 dust samples (CCAAPS study) in mg/m2.

103

102

101

100 r=0.46, p=0.13

10-1 -D-glucan content (pg/spore) b

-2 (1-3)- 10

10-3 1.0 10.0 100.0 Spore size (mm)

Fig. 2-1 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding spore size.

103

102

r=0.81, p=0.02 101

100

10-1 -D-glucan content (pg/spore) b

10-2 (1-3)-

10-3 0 20 40 60 80 100 Fungal frequency (%) in indoor dust samples

Fig. 2-2 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding frequency (%) in indoor dust samples

(PCR analysis).

103

102 r=0.85, p=0.001

101

100

10-1 -D-glucan content (pg/spore) b

-2

(1-3)- 10

10-3 100 101 102 103 104 Fungal concentration (cell/mg dust) in indoor dust samples

Fig. 2-3 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding concentration (geometric mean of cells/mg of dust) in 297 indoor dust samples (PCR analysis).

100

aOR 3.90, aOR=0.32, aOR 1.41, aOR=0.80, 1 SE 0.30 1 SE

+ 95%CI 1.26-12.09 95%CI=0.11-0.97 95%CI 0.41-4.87 95%CI=0.21-3.08

0.25

0.20 0.20

0.15

0.10

0.10 0.05

0.05 0.00 Adjusted Prevalence Rates of Recurrent Wheeze Adjusted Prevalence Rates of Recurrent Wheeze 1.0 7.4 54.6 403.4 0.1 1.0 7.4 54.6 403.4 2980.9

µg/g µg/m2

(1-3)-ß-D-Glucan Concentration

Figure 4-1. Smoothed plot of the adjusted prevalence rates of recurrent wheezing in relation to the log-transformed (1-3)-b-D-glucan concentration (solid lines). Dotted lines represent ± 1 standard error (SE).

101

Table 1-1. Comparison of eight purified glucans and their relative reactivity as measured

by the LAL and EIA assays. The standard against which the concentrations are measured

is Pachyman for the LAL assay, and Laminarin for the EIA assay.

Glucan Linkages LAL – % of the expected EIA– % of the expected

concentration (%)* concentration (%)*

Range Median Range Median

value value

Curdlan (1-3)-b-D-glucan 43-100 96.4 95-150 97.0

Pachyman (1-3)-b-D-glucan 97-118 104.8 56 - 94 73.7

Laminarin (1-3)(1-6)-b-D-glucan 78-118 99.3 17-120 98.2

(some branching)

Schizophyllan (1-3)(1-6)-b-D-glucan 47-118 84.7 No response

(33% branching)

MG-glucan (1-3)(1-6)-b-D-glucan 74-240 96.0 No response

extract from baker’s yeast

Mannan (1-6)-a-D-Mannose No response No response

Dextran (1-3)(1-6)-a-D-glucose No response No response

Pullulan (1-3)(1-4)-a-D-glucose No response No response

* The prepared (expected) concentrations of the glucan standards for the LAL assay

were: 3.125 pg/ml, 12.50 pg/ml, 50 pg/ml and 100 pg/ml. The prepared (expected)

concentrations of the glucan standards for the EIA assay were 250 ng/ml, 1000ng/ml,

2500ng/ml, and 5000 ng/ml. The glucan measured concentrations were calculated based

102

on a standard curve of Pachyman for the LAL assay (2 runs) and Laminarin for the EIA assay (1 run). The measured values were then expressed as % of the expected concentrations.

103

Table 2-1. Fungal species selected for this study and their frequency and concentration (geometric mean), as measured by PCR analysis of 297 dust samples (Cincinnati Childhood Allergy and Air Pollution Study). Fungal species name ATCC number of the Frequency indoors Concentration strain used for the (1- (total number and % of (cells/mg 3)-ß-D-glucan content the total number of dust)* analysis houses in which the species occurred)* Aureobasidium pullulans 58926 297 (100%) 4365.2 Cladosporium 6721 297 (100%) 1520.6 cladosporioides Epicoccum nigrum 58875 293 (98.7%) 250.6 Cladosporium herbarum 58927 293 (98.7%) 184.1 Aspergillus chevalieri 66451 291 (98.0%) 153.5 Aspergillus penicillioides 16910 277 (93.3%) 39.4 Wallemia sebi 42694 275 (92.6%) 61.7 Penicillium brevicompactum 9056 212 (71.4%) 30.8 Stachybotrys chartarum 29-51-05, NIOSH, 153 (51.5%) 4.0 Morgantown, WV Aspergillus flavus 11489 103 (34.7%) 3.0 Aspergillus unguis 10032 86 (29.0%) 2.1 Aspergillus versicolor 52173 68 (23.0%) 4.2 *Vesper et al. 2006

104

Table 2-2. Characteristics of fungal species – spore size, spore surface area and spore volume (average of n=30 spores for each fungal species).

Fungal species Spore Size Spore surface Spore volume

(µm) area (µm2) (µm3)

C. herbarum 4.76* 59.18 36.00

E. nigrum 28.10 2463 11494

P. brevicompactum 3.05 14.39 5.13

A. pullulans 11.83* 628.45 786.1

C. cladosporioides 5.64* 83.42 59.72

A. chavalieri 4.50 63.62 47.71

W. sebi 3.95 49.02 32.27

A. flavus 4.92 76.05 62.36

A. versicolor 2.96 27.53 13.58

A. penicillioides 4.45 62.21 46.14

A. unguis 3.11 30.39 15.75

S. chartarum 7.45* 148.48 172.15

* Geometric mean of width and length for ellipsoidal spores

105

Fungal species LAL –measured (1-3)-ß-D-glucan EIA – measured (1-3)-ß-D-glucan

pg/spore pg/µm2x103 pg/µm3 x103 pg/spore pg/µm2 x103 pg/µm3 x103

C. herbarum 8.66 146.33 240.54 (0.06)* (1.01)* (1.72)*

E. nigrum 241.57 98.08 21.02 379.97 153.25 32.72

P. brevicompactum 0.21 7.02 13.81 3.39 116.10 228.39

A. pullulans 3.76 5.98 4.78 (0.02)* (0.04)* (0.03)*

C. cladosporioides 0.25 3.00 4.19 7.20 86.30 120.55

A. chavalieri 0.22 3.46 4.61 0.24 3.81 5.08

W. sebi 0.12 2.40 3.72 9.68 197.00 300.01

A. flavus 0.03 0.39 0.48 0.10 1.34 1.64

A. versicolor 0.03 0.91 1.84 0.08 3.01 6.10

A. penicillioides 0.01 0.16 0.22 0.13 2.06 2.78

A. unguis 0.005 0.16 0.32 0.02 0.57 1.10

S. chartarum 0.004 0.03 0.33 (0.58)* (3.94)* (3.40)*

* The value for this fungal (1-3)-ß-D-glucan concentration was below the lower detection limit of the EIA assay. One half of the LOD of the pg/ml value was assigned for the (1-

3)-ß-D-glucan concentration of those fungal species.

106

Table 2-4. Contribution of predicted fungal (1-3)-ß-D-glucan (pg/mg dust) based on

LAL and EIA-analyses and concentration of fungal species (total in all 297 dust

samples, cells/mg dust).

Fungal species name Total Predicted fungal content of (1-3)-ß-D-glucan, pg/mg concentration dust (pg/m2 area) (cells/mg dust) Based on LAL Based on EIA in all 297 dust pg/mg pg/m2 pg/mg pg/m2 samples Aureobasidium pullulans 3x106 12 x106 18 x106 0.06 x106 0.12 x106 Cladosporium 1 x106 0.3 x106 3.4 x106 8 x106 97 x106 cladosporioides Epicoccum nigrum 0.6 x106 140 x106 57 x106 220 x106 89 x106

Cladosporium herbarum 0.1 x106 1 x106 20 x106 0.008x106 0.1 x106 Aspergillus chevalieri 6 x106 1 x106 22 x106 1.5 x106 24 x106 Aspergillus penicillioides 0.05 x106 0.0005 x106 0.0009 x106 0.007 x106 0.110 x106 Wallemia sebi 4 x106 0.5 x106 9 x106 38 x106 768 x106 Penicillium 0.07 x106 0.01 x106 0.05 x106 0.2 x106 8 x106 brevicompactum Stachybotrys chartarum 0.005 x106 0.00002x106 0.0002 x106 0.003 x106 0.02 x106 Aspergillus flavus 0.007 x106 0.0002 x106 0.003 x106 0.0007 x106 (0.010 x106) Aspergillus unguis 0.1 x106 0.0005 x106 0.017 x106 0.002 x106 0.06 x106 Aspergillus versicolor 0.1 x106 0.003 x106 0.10 x106 0.009 x106 0.33 x106 TOTAL predicted 155 x106 131 x106 267 x106 979 x106

TOTAL measured 28 x106 20 x106

107

Fungal species name Frequency indoors (total GM (cell/mg Non-parametric p-value number and % of the total dust) correlation* number of houses) Aureobasidium pullulans† 297 (100%) 4365.2 0.08 0.15 Cladosporium cladosporioides 297 (100%) 1520.6 0.12 0.04 (Type I) † Alternaria alternata† 287 (96.63%) 400.8 0.04 0.54 Epicoccum nigrum† 293 (98.7%) 250.6 0.03 0.63 Cladosporium herbarum 293 (98.7%) 184.1 0.10 0.09 Eurotium chevalieri† 291 (98%) 153.5 0.10 0.08 Mucor racemosus† 291 (98%) 108.6 0.10 0.08 Cladosporium 283 (95.3%) 144.5 0.14 0.02 sphaerospermum† Cladosporium cladosporioides 279 (93.9%) 39.7 0.09 0.12 (Type II) † Aspergillus penicillioides† 277 (93.3%) 39.4 0.07 0.23 Wallemia sebi† 275 (92.6%) 61.7 0.10 0.09 Penicillium chrysogenum† 251 (84.5%) 77.6 0.13 0.02 Trichoderma viride† 231 (77.8%) 25.7 0.06 0.31 Aspergillus niger 226 (76.09%) 6.6 0.05 0.37 Penicillium brevicompactum† 212 (71.4%) 30.8 0.09 0.13 Paecilomyces variotii† 197 (66.3%) 7.1 0.05 0.37 Aspergillus fumigatus† 193 (64.98%) 10.0 0.10 0.09 Penicillium variabile 190 (64%) 5.4 0.04 0.51 Scopulariopsis brevicaulis† 180 (60.6%) 3.7 0.09 0.14 Aspergillus ustus 153 (51.5%) 5.0 0.04 0.55 Stachybotrys chartarum 153 (51.5%) 4.0 0.01 0.87 Chatomium globosum† 147 (49.5%) 3.4 0.04 0.51 Acremonium strictum 133 (44.78%) 4.0 0.01 0.81 Rhizopus stolonifer† 130 (43.8%) 3.0 0.10 0.08 Scopulariopsis chartarum† 129 (43.4%) 2.2 0.003 0.96 Aspergillus sclerotiorum† 104 (35%) 3.4 0.002 0.97 Aspergillus flavus† 103 (34.68%) 3.0 0.14 0.02 Aspergillus ochraceus† 93 (31.31%) 4.1 0.07 0.23 Aspergillus unguis† 86 (29%) 2.1 -0.01 0.85 Aspergillus versicolor† 68 (23%) 4.2 0.01 0.93 Penicillium spinulosum† 52 (17.5%) 2.2 0.08 0.20 Aspergillus restrictus† 44 (14.8%) 2.1 0.08 0.17 Aspergillus sydowii† 40 (13.5%) 2.0 0.01 0.85 Penicillium crustosum 40 (13.5%) 2.0 0.02 0.76 Penicillium purpurogenum† 26 (8.8%) 1.2 0.04 0.46 Penicillium corylophilum† 14 (4.7%) 1.2 -0.01 0.91 108

* Correlation between the (1-3)-b-D-glucan concentration and the cell concentration measured by the PCR

†Fungal species identified as being significantly in higher concentration in “more moldy homes” than in “non-moldy homes” (Vesper et al., 2006)

109

Table 3-2. Fungi associated with the indoor (1-3)-ß-D-glucan concentration.

Model Standardized Frequency indoors ( % coefficients of the total number of houses) No interactions Aspergillus flavus 0.12* 34.7% Cladosporium herbarum 0.10* 98.7%

Two -way interactions Cladosporium cladosporioides type II 0.44*** 93.9% Penicillium brevicompactum 0.51** 71.4% Aspergillus flavus 0.11** 34.7% Aspergillus unguis 0.67** 29% Interaction between Cladosporium 0.30** cladosporioides types I and II

Three-way interactions Combined Cladosporium species 0.42*** 97% Epicoccum nigrum 0.22 (p=0.20) 98.7% Interaction between Cladosporium -0.04* species and Epicoccum nigrum *p<0.05, **p<0.01, *** p<0.001

The standardized coefficient for Alternaria alternata (X0) was 0, and thus is not included in the equation, despite its involvement in an interaction effect.

110

Table 3-3. Final 6-factor SEM model with Loadings. Only Factor 1 and 3 were significant, and thus are reported.

Factor 1 Loading Frequency indoors ( % of the total number of houses) Log X17 Cladosporium cladosporioides (Type 1.1635 93.9% 2) Log X22 Mucor racemosus 0.8338 98.0% Log X19 Cladosporium sphaerospermum 0.7533 95.3% Log X36 Wallemia sebi 0.4114 92.6% Log X34 Stachybotrys chartarum 0.2670 51.5% Log X18 Cladosporium herbarum 0.2535 98.7% Log X14 Aureobasidium pullulans 0.1245 100%

Factor 3 Loading Log X7 Aspergillus penicillioides 0.7949 93.3% Log X21 Aspergillus chevalieri 0.7652 98.0% Log X36 Wallemia sebi 0.6295 92.6% Log X13 Aspergillus versicolor 0.5031 23.0% Log X6 Aspergillus ochraceus 0.3086 31.31% Log X8 Aspergillus restrictus 0.3037 14.8% Log X19 Cladosporium sphaerospermum 0.2878 95.3% Log X5 Aspergillus niger 0.2171 76.1% Log X10 Aspergillus sydowii 0.1531 13.5% Log X11 Aspergillus unguis 0.1298 29%

111

Table 4-1. Geometric mean (GM), geometric standard deviation (GSD) and interquartile (IQ) range of (1-3)-ß-D-glucan and endotoxin concentration (mg/g, EU/mg) and loading (mg/m2, EU/m2), measured in homes of 574 infants. n=574 GM GSD IQ*

(1-3)-ß-D-Glucan mg/g 55.1 3.7 21.9-133.5 mg/m2 18.4 5.7 5.9-57.9 Endotoxin EU/mg 70.7 3.4 39.8-165.0 EU/m2 23.7 5.6 9.4-74.5 EU: endotoxin units *Interquartile range = [25th percentile – 75th percentile]

112

Allergen sensitization Recurrent wheeze Recurrent wheeze with Characteristic of predictor variables (SPT+) n=114 (19.9% of 574) allergen sensitization n=169 (29.4% of 574) n=41 (11.0% of 373) (1-3)-b-D-glucan quartiles 2 2 2 Exposure in mg/g Exposure in mg/m2 mg/g mg/m mg/g mg/m mg/g mg/m I : 3-22 (144) I: 0.2-6 (143) 44 (26.0%) 40 (23.7%) 22 (19.3%) 27 (23.7%) 9 (22.0%) 8 (19.5%) II: 22-60 (149) II: 6-18 (144) 45 (26.6%) 51 (30.2%) 34 (29.8%) 24 (21.1%) 16 (39.0%) 11(26.8%) III: 61–134 (138) III: 19-58 (143) 43 (25.4%) 41 (24.3%) 35 (30.7%) 35 (30.1%) 10 (24.4%) 15(36.6%) IV: 134-900(143) IV: 58-2966 (144) 37 (21.9%) 37 (21.9%) 23 (20.2%) 28 (24.6%) 6 (14.6%) 7 (17.1%) p-value** 0.75 0.30 0.08 0.40 0.18 0.25 Endotoxin quartiles EU/mg EU/m2 EU/mg EU/m2 EU/mg EU/m2 EU/mg EU/m2 35 (24.3%) 45 (31.3%) 29 (20.1%) 31 (21.5%) 8 (8.6%) 8 (8.6%) I: 3-39 (144) I: 0.09-9 (144) 51 (35.7%) 36 (25%) 25 (17.5%) 25 (17.5%) 10 (10.8%) 12 (12.9%) II: 39-80 (143) II: 9-25 (143) 41 (28.7%) 58 (40.6%) 33 (23.1%) 21 (14.7%) 13 (14%) 6 (6.5%) III:80-171 (143) III:25-74 (143) 42 (29.2%) 30 (21.0%) 27 (18.8%) 37 (25.7%) 10 (10.6%) 15 (16.0%) IV:171-2800(144) IV: 74-5120(144)

0.21 0.002 0.67 0.10 0.71 0.17 p-value** Visible Mold None (255) 76 (45.0%) 43 (37.7%) 13 (31.7%) low (<0.2m2) (296) 82 (48.5%) 61 (53.5%) 22 (53.7%) high ³0. 2m2) (23) 11 (6.5%) 10 (8.8%) 6 (14.6%) p-value** 0.12 0.008 0.0009 Mother’s smoking (average number of 1.9 3.1 3.6 cigarettes per day) Daycare (yes) (50) 15 (8.9%) 15 (13.2%) 3 (7.3%) p-value** 0.93 0.06 0.74 Breastfeeding duration None (177) 53 (31.4%) 42 (36.8%) 18 (43.9%) (1-24) weeks (252) 80 (47.3%) 48 (42.1%) 17 (41.5%) (25+) weeks (145) 36 (21.3%) 24 (21.1%) 6 (14.6%) p-value** 0.34 0.25 0.11 Dog in home (yes) (219) 72 (42.6%) 41 (36.0%) 17 (41.5%) p-value** 0.33 0.84 0.56 Cat in home (yes) (114) 38 (22.5%) 23 (20.2%) 11 (26.8%) p-value** 0.51 0.91 0.45 Either parent asthma (yes) (183) 56 (33.1%) 51 (44.7%) 21 (51.2%) 0.68 0.001 0.006 p-value

113

Gender (male) (311) 93 (55.0%) 67 (58.8%) 24 (58.5%) p-value** 0.79 0.27 0.56 Race Afro-American (96) 31 (18.3%) 24 (21.0%) 11 (26.8%) All other races (478) 138 (81.7%) 90 (79.0%) 30 (73.2%) p-value** 0.50 0.17 0.07 Siblings 0 (193) 65 (38.5%) 28 (24.6%) 9 (22.0%) 1 (220) 57 (33.7%) 43 (37.7%) 16 (39.0%) >=2 (161) 47 (27.8%) 43 (37.7%) 16 (39.0%) p-value 0.22 0.02 0.16 Lower Respiratory Condition† (yes) (208) 56 (33.1%) 70 (61.4%) 26 (63.4%) p-value** 0.25 <0.0001 0.0005 Upper Respiratory Condition‡ (yes) (355) 106 (62.7%) 89 (78.1%) 33 (80.5%) p-value** 0.98 0.0003 0.02 EU: endotoxin units *Allergen sensitization (SPT+) = Positive skin prick test to any of the tested 17 allergens (n=169). Infants that were SPT(-) were used as the comparison group (n=405). Total n=169+405=574. Recurrent wheeze = 2 or more wheezing episodes in the last 12 months (n=114). Infants that had one or no wheezing episodes in the last 12 months were used as the comparison group (n=460). Total n=114+460=574. Recurrent wheeze with allergen sensitization = two or more wheezing episodes in the last 12 months and SPT(+) (n=41). Infants that had one or no wheezing episodes in the last 12 months and were SPT(-) were used as the comparison group (n=332). Total n=41+332=373. **p-value – significance of differences in numbers across the predictor variable levels, Chi-square statistics †Lower Respiratory condition includes any of the following: whooping cough, croup, viral infections, bronchitis/bronchiolitis, flu, pneumonia ‡Upper respiratory condition includes any of the following: cold, ear infection, sinus infection, Strep throat (positive culture), tonsillitis, colored drainage

114

Table 4-3. Adjusted odds ratios (aOR) and 95% confidence intervals (95% CI) for recurrent wheeze and recurrent wheeze with allergen sensitization in relation to upper vs. lower endpoints of continuously measured (1-3)-b-D-glucan quartiles (mg/g) (reference category is the value of the lower endpoint of each quartile). Recurrent Recurrent wheeze Recurrent wheeze with wheeze* with SPT(+) vs. no SPT(+) vs. no wheeze, wheeze, SPT(-)** SPT(+)*** aOR (95% CI) aOR (95% CI) aOR (95% CI) (1-3)-b-D-glucan quartile endpoints (mg/g) I.3-22 3.04 (1.25-7.38) 4.89 (1.02-23.57) 160.51 (4.85-5,311.00) II.23-60 1.29 (0.99-1.67) 1.23 (0.79-1.92) 2.54 (0.97-6.62) III.61–133 0.82 (0.65-1.05) 0.59 (0.38-0.92) 0.17 (0.05-0.57) IV.134-900 0.39 (0.16-0.93) 0.13 (0.03-0.61) 0.00 (0.00-0.07) Endotoxin fixed at continuous 0.99 (0.71-1.37) 1.17 (0.69-1.98) 1.60 (0.58-4.41) level from 39.19 to 171.17 (mg/g) Visible Mold (low vs. none) 1.18 (0.73-1.91) 1.29 (0.57-2.90) 2.64 (0.89-7.86) Visible Mold (high vs. none) 4.44 (1.63-12.05) 9.51 (2.34-38.63) 42.47 (4.70-384.14) Mother’s smoking (20 vs. 0 5.16 (2.33-11.44) 10.17 (2.58-40.09) 10.17 (2.58-40.09) cigarettes/day) Parental asthma 1.87 (1.17-3.00) 2.22 (1.05-4.71) 2.09 (0.76-5.77) Race (Afro-American vs. other) 2.08 (1.15-3.73) 3.93 (1.57-9.84) 10.04 (2.45-41.14) Siblings (1vs. 0) 1.38 (0.77-2.47) 1.84 (0.66-5.09) 8.87 (1.85-42.51) Siblings (=2 vs. 0) 1.96 (1.08-3.57) 2.46 (0.87-6.93) 7.83 (1.60-38.38) Lower respiratory condition† 3.98 (2.47-6.41) 4.63 (2.05-10.46) 9.93 (3.06-32.16) Upper respiratory condition‡ 2.15 (1.26-3.67) 2.75 (1.08-7.04) 4.47 (1.24-16.07) * Recurrent wheeze = 2 or more wheezing episodes in the last 12 months (n=114). Infants that had one or no wheezing episodes in the last 12 months were used as the comparison group (n=460).

115

**Recurrent wheeze with allergen sensitization = two or more wheezing episodes in the last 12 months and SPT(+) (n=41). Infants that had one or no wheezing episodes in the last 12 months and were SPT(-) were used as the comparison group (n=332). ***Recurrent wheeze with allergen sensitization = two or more wheezing episodes in the last 12 months and SPT(+) (n=41). Infants that had one or no wheezing episodes in the last 12 months and were SPT(+) were used as the comparison group (n=128). †Lower Respiratory condition includes any of the following: whooping cough, croup, viral infections, bronchitis/bronchiolitis, flu, pneumonia ‡Upper respiratory condition includes any of the following: cold, ear infection, sinus infection, Strep throat (positive culture), tonsillitis, colored drainage

116

Table 4-4. Adjusted odds ratios (aOR) and 95% confidence intervals (95% CI) for recurrent wheeze and recurrent wheeze with allergen sensitization in relation to upper vs. lower endpoints of continuously measured (1-3)-b-D-glucan quartiles (mg/m2) (reference category is the value of the lower endpoint of each quartile). Recurrent wheeze* Recurrent wheeze Recurrent wheeze with with SPT(+) vs. no SPT(+) vs. no wheeze, wheeze, SPT(-)** SPT(+)*** aOR (95% CI) aOR (95% CI) aOR (95% CI) (1-3)-b-D-glucan quartile endpoints (mg/m2) I.0.2-6 1.62 (0.54-4.82) 7.51 (0.96-58.70) 147.81 (1.38-15814.02) II.7-18 1.08(0.86-1.36) 1.26 (0.82-1.91) 2.16 (0.98-4.77) III.19-58 0.95 (0.76-1.19) 0.64 (0.42-0.97) 0.60 (0.22-1.64) IV.59-2966 0.65 (0.21-2.06) 0.05 (0.00-0.51) 0.01 (0.00-4.75) Endotoxin fixed at continuous 1.07 (0.77-1.49) 1.45 (0.89-2.37) 1.40 (0.58-3.42) level from 38 to 165 (mg/m2) Visible Mold (low vs. none) 1.18 (0.73-1.90) 1.33 (0.60-2.98) 2.16 (0.77-6.01) Visible Mold (high vs. none) 4.25 (1.56-11.58) 12.56 (2.96-53.35) 32.16 (4.03-256.60) Mother’s smoking (20 vs. 0 5.39 (2.43-11.92) 10.44 (2.65-41.14) 10.44 (2.65-41.14) cigarettes/day) Parental asthma 1.85 (1.16-2.95) 2.35 (1.10-5.04) 1.99 (0.74-5.37) Race (Afro-American vs. other) 1.94 (1.08-3.48) 3.96 (1.60-9.82) 4.47 (1.29-15.49) Siblings (1vs. 0) 1.46 (0.82-2.61) 1.84 (0.66-5.16) 5.74 (1.48-22.23) Siblings (=2 vs. 0) 2.11 (1.16-3.84) 2.51 (0.88-7.15) 6.55 (1.54-27.94) Lower respiratory condition† 4.10 (2.55-6.58) 5.33 (2.34-12.17) 10.75 (3.37-34.22) Upper respiratory condition‡ 2.11 (1.24-3.57) 2.42 (0.96-6.12) 3.69 (1.13-12.01) See footnotes for Table 4-3.

117

LIST OF APPENDICES

APPENDIX A: COPIES OF PEER-REVIEWED PUBLICATIONS RESULTED FROM THE PhD STUDY

A1: Iossifova Y, Sucharew H, Succop P, Vesper S, Reponen T. Use of (1-3)-b-D- glucan concentrations in dust as a surrogate method for estimating specific mold exposures. Indoor Air (submitted) [Specific aim 3]

A2: Iossifova Y, Reponen T, Daines M, Crawford C, Hershey GK. Comparison of EIA and LAL analytical methods for detecting (1-3)-ß-D-glucan in pure fungal cultures. Appl Environ Microb (submitted) [Specific aims 1 and 2]

A3: Iossifova Y, Reponen T, Bernstein D, Levin L, Kalra H, Campo P, Zeigler H,

Villareal M, Lockey J, Khurana-Hershey G, LeMasters G. House dust (1-3)-b-D-glucan and wheezing in infants? Allergy (in revision) [Specific aim 4]

APPENDIX B: ABSTRACTS OF ADDITIONAL PAPERS THAT THE AUTHOR

CO-AUTHORED AS PART OF HER PhD THESIS:

B1: Vesper S, McKinstry C, Haugland R, Iossifova Y, LeMasters G, Levin L, Hershey

GK, Villareal M, Bernstein DI, Reponen T. EPA Relative Moldiness Index© as Predictor of Childhood Respiratory Illness. J Expo Anal Environ Epi 2006 (in press)

118

B2: Lee T-K, Grinshpun S, Kim K-Y, Iossifova Y, Adhikari A, Reponen T. Relationship between indoor and outdoor airborne fungal spores, pollen and (1-3)-ß-D-glucan in homes without visible mold growth. Aerobiologia 22:227-236, 2006

B3. Seo S-C, Cho S-H, Grinshpun S, Iossifova Y, Schmechel D, Rao C, Reponen T. A new field-compatible method for the collection of fungal fragments. (submitted)

APPENDIX C LAL PROTOCOLS FOR (1-3)-b-D-GLUCAN ANALYSIS

C1: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – DUST SAMPLES – ENDPOINT ASSAY

C2: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – DUST SAMPLES – KINETIC ASSAY

C3: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – AIR SAMPLES

C4: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – PURE FUNGAL SPORES SUSPENSIONS

APPENDIX D EIA PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS

119

APPENDIX A: COPIES OF PEER-REVIEWED PUBLICATIONS RESULTED FROM THE PhD STUDY

120

A1: Comparison of EIA and LAL analytical methods for detecting (1-3)-ß-D-glucan in pure fungal cultures and in home dust samples

Y. Iossifova1, T. Reponen1†, M. Daines2, C. Crawford1, G.K. Hershey2

1 Department of Environmental Health, University of Cincinnati,

2 Division of Allergy and Immunology, Cincinnati Children’s Hospital Medical Center.

†Contact author: Tiina Reponen, Center for Health Related Aerosol Studies, Department of Environmental Health, University of Cincinnati, 3223 Eden Ave, PO Box 670056,

Cincinnati, Ohio 45267-0056, USA;

Email: [email protected], Fax: (513) 558-2263, Tel: (513) 558-0571

Short title: (1-3)-b-D-GLUCAN AND MOLD

Manuscript 11/06/06

To Be Submitted to:

APPL ENVIRON MICROB

121

ABSTRACT

Exposure to fungal (1-3)-b-D-glucan is associated with various respiratory responses. It has been speculated whether 1-3-b-D-glucan exposure can be used as a surrogate for mold exposure. Currently there are two methods available for the analysis of (1-3)-b-D-glucan: the Limulus Amebocyte Lysate assay (LAL) and the inhibition

Enzyme Immunoassay (EIA). The aim of this study was to compare the specificity of these two methods in detecting eight alpha and beta-glucan standards, and their sensitivity for the analysis of (1-3)-b-D-glucan content of common indoor fungal species and indoor dust samples. Twelve fungal species (two Cladosporium species, five

Aspergillus species, Aureobasidium pullulans, Penicillium brevicompactum, Epiccocum nigrum, Wallemia sebi, and Stachybotrys chartarum) were cultured from pure ATCC strains on agar media. All samples were analyzed for (1-3)-b-D-glucan content by both the LAL assay (GlucatellTM, Associates of Cape Cod, East Falmouth, MA) and the EIA

(antibody: mouse IgG, kappa light; Biosupplies Australia, Parkville Victoria, Australia).

We found that the LAL assay is more accurate and specific in measuring both linear and branched (1-3)-b-D-glucans than the EIA assay. Although E. nigrum was the species of the greatest (1-3)-ß-D-glucan content per spore (241 pg/spore), this was mainly due to having also the largest spore size (28 mm). Although several samples were below the detection limit of the EIA assay, the biomass-normalized (1-3)-ß-D-glucan content measured by both assays was within similar range (LAL: 0.003 to 146.33 pg/mm2, 0.22-

240.54 pg/mm3; EIA: 0.04 – 197.00 pg/mm2, 0.03-300 pg/mm3). Furthermore, the (1-3)-b-

D-glucan content per spore measured for the twelve fungal species by the LAL correlated with the respective fungal spore size and the respective species prevalence in indoor dust

122

samples. Therefore LAL-analyzed (1-3)-b-D-glucan concentration in dust samples may be used as a measure of total mold.

INTRODUCTION

Exposure to fungi in occupational and indoor environments is associated with respiratory (nose and throat irritation, cough) and general (tiredness and headache) symptoms, allergic reactions and organic dust toxic syndrome (Rylander et al., 1994;

Johanning, 2004; Curtis, 2004). Similar general and respiratory symptoms and airways inflammation are reported in occupational and indoor exposures to (1-3)-b-D-glucan, polyglucose component of fungi, pollen, and some bacteria (Rylander and Lin, 2000;

Douwes et al., 2005). The biological properties of (1-3)-b-D-glucans are not dependent on the viability and (1-3)-b-D-glucans from dead organisms may thus be equally relevant in causing potential health effects. Therefore, (1-3)-b-D-glucans, which comprise up to

50% of the fungal cell wall, may be a better predictor for health risk than the commonly used determination of viable fungal spores. In addition, (1-3)-b-D-glucan measurement is a cheaper, quicker and more sensitive and specific method than analysis of culturable or total mold count (Rylander and Lin, 2000; Douwes et al., 1996, 2003).

Currently there are two methods available for the analysis of (1-3)-b-D-glucan.

One method is based upon the bioactivity of this molecule in the factor-G-mediated

Limulus coagulation pathway - the Limulus Amebocyte Lysate assay (LAL) (Obayashi et al., 1985). The other method is based on (1-3)-b-D-glucan antigen-antibody reaction - the inhibition Enzyme Immunoassay (EIA) (Douwes et al., 1996), and other EIA

123

modifications (Adachi et al., 1994; Milton et al., 2001). The LAL method is much more sensitive with a Lower Limit of Detection (LOD) = 1pg/ml compared to EIA LOD = 40 ng/ml, which limits the EIA assay to settled dust and air samples collected from high- exposure environments only. Due to its high sensitivity, the LAL assay has predominantly been used for the analysis of air samples. The EIA assay described by

Douwes et al. (1996) reacts with both linear and branched (1-3)-b-D-glucans. Other modifications of the EIA assay were developed by Adachi et al. (1994) and Milton et al.

(2001), which are highly specific to the measurement of the branched (1-3)(1-6)-b-D- glucans only. On the other hand, the LAL is extremely sensitive (LOD=1pg/ml) assay suggested to recognize both linear and branched (1-3)-b-D-glucans (Tanaka et al., 1991;

Thorne et al., 2004). However, its ability to detect both linear and branched (1-3)-b-D- glucans, as well as yeast -D-mannan in previous studies (Tanaka et al., 1991) were viewed as disadvantage indicating low specificity.

While there are some data on the content of (1-3)-b-D-glucans in spores of the indoor fungal species of Penicillium, Aspergillus, Cladosporium and Stachybotrys

(Fogelmark and Rylander, 1997; Foto et al., 2004), analyzed by the LAL assay, very little is known on the EIA-analyzed (1-3)-b-D-glucan of spores from different species.

LAL-analyzed (1-3)-b-D-glucan is a recognized indicator of mold biomass based on health effects and correlation with total fungal count (Alwis et al., 1999; Rylander et al.,

1999, Wan and Li, 1999; Mandryk et al., 2000). While EIA-analyzed (1-3)-b-D-glucan in settled dust has also been used as an indicator of mold biomass (Douwes et al., 1996;

Chew et al., 2001; Gehring et al., 2001), there is very little data on any correlation between EIA analyzed air samples and mold counts due to its low sensitivity (i.e., (1-3)-

124

b-D-glucan concentration in most air samples is expected to be below the lower limit of detection). As the LAL assay has been used predominantly for the analysis of (1-3)-b-D- glucan concentration of air samples, while the EIA has been used for dust samples, there is a need for a comparison of these two methods by using unified source of environmental samples. Due to the low sensitivity of the EIA method, the best approach is to use dust samples for this comparison.

Thus, the aim of this study is to compare the specificity and sensitivity of LAL and EIA methods through the analysis of (1-3)-b-D-glucan concentration in purified glucan standards, common indoor fungal species and field dust samples, in order to determine which assay, if any, is a better surrogate of total fungal exposure.

MATERIALS AND METHODS

Laboratory analysis of (1-3)-ß-D-glucan

The LAL test is a quantitative direct method for the detection of (1-3)-b-D-glucans that uses (1-3)-b-D-glucan-sensitive factor G (Obayashi et al., 1985). We performed the kinetic chromogenic Limulus Amebocyte lysate assay {GlucatellTM, Associates of Cape

Cod, East Falmouth, MA}, using laboratory materials (pipette tips, tubes, etc.) certified free of contaminating glucans by the manufacturer (Associates of Cape Cod). From each sample, 0.5 ml aliquot was extracted with 0.5 ml of 0.6 M NaOH by shaking for 1 hour at room temperature, to unwind the triple-helix structure of (1-3)-b-D-glucan and make it water-soluble. Twenty-five ml of Glucatell reagent was added to each well of serially

125

diluted (from 1:1 to 1:1011) extract and a control standard (1-3)-b-D-glucan (Pachyman), placed in a 96-well, flat-bottom microplate. The optical density (OD) was read at 405 nm, recorded at time of onset at OD = 0.03. All samples of (1-3)-b-D-glucan were above the lower limit of detection (LOD) of the Glucatell assay protocol (3.125 pg/ml). The median coefficient of variation (CV) was 9% for the intra-plate variability and 27% for the inter- plate variability.

The EIA test was performed as described by Douwes et al. (1996). The primary monoclonal antibody to (1-3)-b-D-glucan was mouse IgG, kappa light (Biosupplies

Australia, Parkville Victoria, Australia). The secondary antibody was peroxidase- conjugated affinipure sheep anti-mouse IgG (H+L) (Jackson ImmunoResearch, West

Grove, PA). The sample extraction was accomplished by heat extraction in an autoclave at 120oC for 1 hour. LOD values (250,000 pg/ml) were divided by the square root of two for the data analyses. For the EIA assay, the median CV was 13.6 % for the intra-plate variability and 24.2 % for the inter-plate variability.

The results of both assays were expressed as pg/ml for the glucan standards, but converted to pg/spore for the spore suspensions. After measuring the spore size and calculating the spore surface area and volume, the results were also converted to pg/m2 and pg/m3. Results of the (1-3)-b-D-glucan concentration in dust samples were reported as mg/g of dust and mg/m2of floor area.

Assays specificity in detecting branched (1-3)-b-D-glucans

The following (1-3)-b-D-glucan standards to test the specificity of the LAL and

EIA tests in detecting linear vs. branched (1-3)-D-glucans were used: 1) Pachyman {99%

126

linear (1-3)-b-D-glucan}, 2) Curdlan {linear (1-3)-b-D-glucan}, 3) Laminarin {primarily linear, with some (1-6)-ß- interstrand linkages and branch points}, 4) Schizophyllan {(1-

6)-ß- branching at every third linkage}, 5) MG-glucan (MacroGardR is extract form baker’s yeast containing 97% (1-3),(1-6)-b-D-glucan), 6) Mannan {(1-6)- a -D-

Mannose}, 7) Dextran {(1-3),(1-6)- a -D-glucose}, and 8) Pullulan{(1-3),(1-4)- a -D- glucose}. The glucan standards were purchased from Sigma Chemical Co. (St. Louis,

MO) except for Pachyman and MG-glucan, which were obtained from Megazyme

International Ireland Ltd (Bray, Ireland) and Nutritional Scientific Corporation (Liberty,

TX), respectively.

Suspensions of pure fungal spores

Fungal species selection

Wallemia sebi, and Stachybotrys chartarum (Table 1). Results from an ongoing field study, {Cincinnati Childhood Allergy and Air Pollution Study (CCAAPS), see section

Field Samples below} were used to identify species that are commonly found in homes.

Based on Polymerase Chain Reaction (PCR) analysis of dust samples from 297 homes

(Vesper et al., 2006), eight fungal species that were most commonly found (>90% frequency) and had highest median concentrations were selected for this study (Table 1).

This list included two Aspergillus species. Three additional Aspergillus species were included in order to study the within species variability of (1-3)-b-D-glucan. P.

127

brevicompactum and S. chartarum were chosen to represent the species of medium frequency and concentration. In addition, S. chartarum was included due to reasons of being of health concern in indoor environments (Lai, 2005; Hossain et al., 2004). We used a non-toxic strain, which was generously provided by NIOSH, Morgantown, WV

(research collection, No: 29-51-05). All the other species were purchased from the

American Type Culture Collection (ATCC, Manassas, VA).

Preparation of pure fungal species

The freeze-dried pure fungal cultures were re-hydrated and prepared by following the ATCC instructions. Thus cultures were inoculated on the following media: malt extract agar (Aspergillus chevalieri), Harrold’s M40Y (Wallemia sebi, Aspergillus penicillioides), and potato dextrose agar (the rest of the species). Spores from one-week old pure cultures were harvested from the agar surface by using micro-beads (Schmechel et al. 2006), and transferred into 5 ml sterile tube, containing 0.02% Tween solution in pyrogen and glucan-free reagent water. Serial dilutions of 10-0 to 10-6 were prepared for each fungal spore suspension, which was used for determining (1-3)-b-D-glucan concentration (as described above), spore concentration, and spore size.

MA, USA) membrane filter and cleared by a modified acetone vaporizing unit (Model:

Quixfix; Environmental Monitoring System, Charleston, SC, USA). The filter was

128

stained with glycerin jelly (gelatin 20 g, phenol crystals 2.4 g, glycerol 60 ml, water 70 ml). Spores were counted under a microscope as described by Adhikari et al. (2003).

Field samples

Field samples were obtained through the CCAAPS study. The CCAAPS is a prospective birth cohort study aimed at investigating the role of aeroallergens and diesel exhaust particles in the development of atopy and atopic respiratory disorders (Ryan et al., 2005). When participating infants reached an average age of 8 months, families were visited at their homes and dust samples were vacuumed from the baby’s primary activity room floor in order to assess exposure to indoor aeroallergens and mold as described by

Cho et al. (2006). In brief, for carpeted floor, samples were collected from area of 2 m2 at a vacuuming rate of 2 min/m2. For non-carpeted floor only one sample was collected from the entire room at a rate of 1 min/m2. The home dust sample was sieved (355 µm

129

sieve), and the fine dust was divided into sub-samples and stored at -20°C before analyses. A 5 mg sub-sample of the fine dust from 297 samples was used for the PCR analysis (Vesper et al., 2006) in order to identify and quantify 36 fungal species. The results were reported as cells/mg dust. Another set of 50 mg and 40 mg subsamples

(n=70) were analyzed for (1-3)-ß-D-glucan concentration in dust samples by the LAL- and EIA assays, respectively. Taking into account the dust amount used for the analysis of each sample (50 mg) and the surface area from where the dust was sampled, the (1-3)- b-D-glucan concentration in dust samples was reported as mg/g and mg/m2.

Data analysis

(1-3)-b-D-glucan content of the pure fungal spores and concentration in the subset

(n=70) of dust samples collected from CCAAPS homes were tested for normality of distribution. As the (1-3)-b-D-glucan content of fungal spores, as well as their respective spore size, surface areas and volumes, were not normally distributed even after log- transformation, the correlations with and between the LAL- and EIA-analyzed spore (1-

3)-b-D-glucan contents were tested with the non-parametric Spearman correlation, and the difference of means with the Wilcoxon statistics. The EIA-analyzed (1-3)-b-D-glucan concentration from indoor dust samples also did not follow the Gaussian distribution, probably to the smaller sample size (n=70). Thus we used non-parametric analyses to test for correlation and difference between the LAL- and EIA-analyzed (1-3)-b-D-glucan concentrations in dust samples.

(1-3)-b-D-glucan concentration in 297 dust samples collected from the CCAAPS homes and the fungal cell concentrations analyzed with PCR followed the normal

130

Gaussian distribution after log-transformation. Therefore, the log-transformed data was used in paired samples analysis to test correlations (Pearson) and differences (t-test) between LAL-analyzed and LAL-predicted (1-3)-b-D-glucan concentration in indoor dust samples, as well as between LAL-predicted and EIA-predicted (1-3)-b-D-glucan contents. The predicted total (1-3)-ß-D-glucan concentration in dust samples was calculated based on the estimated (1-3)-ß-D-glucan content in common indoor fungal species (LAL and EIA) multiplied by the measured concentrations (cells/mg dust) of the respective fungal species, assuming that the main contribution of (1-3)-ß-D-glucan came from these 12 fungal species.

RESULTS

Specificity and Accuracy of LAL and EIA in measuring glucans of different linkage and branching.

The reactivities of LAL and EIA assays to a- and ß-glucans of various degree of branching are shown in Fig. 1. As the LAL assay (Fig.1 A) is a kinetic assay measuring the onset of time at OD = 0.03, the later the reaction occurs (mean onset time), the lesser the specificity for the particular purified glucan at that concentration in comparison to the other glucan standards. In the endpoint EIA assay, the best curve is curvilinear, with a rapid straight decrease in the absorbance units with the increase of the concentration till saturation of the curve is reached (i.e., minimal concentration detected)). As seen in Fig.

1, both LAL and EIA assay were specific to linear (1-3)-ß-D-glucans (Curdlan and

131

Pachyman) and non-reactive to alpha-glucans (presented by straight horizontal lines).

The reactivity of the LAL assay slightly decreased with the increase in the degree of branching (Fig. 1A). Although the EIA assay was also sensitive in recognizing close to linear structures (Laminarin), its sensitivity to branched structures was negligible

{(Schizophylan, branched (1-3)(1-6)-b-D-glucan)} (Fig. 1B). As the LAL assay measures the availability of reactive (1-3)-b-D-glucan molecules in the sample that can activate the factor G enzyme, it detects any structures that have the single helix of (1-3)-b-D-glucan.

This and the fact that it uses the linear (1-3)-b-D-glucan Pachyman as a standard against which the rest of the glucans are measured, assures for the detection of any glucans containing (1-3)-b-D-glucan as part of their molecule. The EIA assay is an antigen- antibody test that uses antibodies against Laminarin, a linear (1-3)-b-D-glucan with some

(1-6)-branching. Thus it is expected to be highly reactive against epitopes containing close to linear (1-3)-b-D-glucan.

The glucan standard concentrations were calculated based on a standard curve of

Pachyman for the LAL assay (2 runs) and Laminarin for the EIA assay (1 run). These values, expressed as % of the expected concentrations, are presented in Table 2. The prepared (expected) concentrations of the glucan standards for the LAL assay were:

3.125 pg/ml, 12.50 pg/ml, 50 pg/ml and 100 pg/ml. The prepared (expected) concentrations of the glucan standards for the EIA assay were 250 ng/ml, 1000ng/ml,

2500ng/ml, and 5000 ng/ml. For each standard concentration the LAL was more accurate in measuring concentrations of (1-3)-b-D-glucan standards than the EIA as demonstrated by the narrower range of the % expected concentration and the smaller median value of this % (Table 2).

132

Sensitivity of LAL and EIA in measuring (1-3)-b-D-glucan in pure fungal spores.

The spore size, spore surface area, and spore volume of the twelve pure fungal species analyzed are provided in Table 3. Among the fungal species, E. nigrum had the largest spore size (28 mm), and thus the biggest surface area (2,463 mm2) and volume

(11,494 mm3). The Aspergillus species were small in size, and Aspergillus versicolor was the smallest among them (size 2.96m? m, surface area 27.53 mm2, and volume 13.58 mm3).

The measured spore sizes and surface areas were within the range reported earlier

(Reponen et al., 2001; Bauer et al., 2002; Foto et al., 2004).

Based on the spore characteristics reported in Table 3, the (1-3)-ß-D-glucan content per spore, spore surface area, and spore volume were calculated, and presented in

Table 4. Although E. nigrum was the species of greatest (1-3)-ß-D-glucan content per spore (241 pg/spore), this was mainly due to having also the largest spore size (28 mm).

Other fungi of high (1-3)-ß-D-glucan content per spore measured by the LAL assay were

133

We further analyzed the variation between fungal species and within Aspergillus genus (5 species), measured by both assays. Both assays showed lower variation of (1-3)- b-D-glucan content within the Aspergillus genus than between the 12 fungal species

(LAL: Coefficients of variation = 45% and 57%, respectively; EIA: Coefficients of variation = 22% and 42%, respectively).

As (1-3)-ß-D-glucan can contribute up to 50% of the composition of the fungal cell wall, we expected the spore size and surface area to affect the fungal (1-3)-ß-D- glucan content. Indeed, we found a close to significant correlation between LAL- analyzed (1-3)-ß-D-glucan content (pg/spore) and spore size (Fig. 2), but not with the spore surface area (r=0.39, p=0.22), or volume (r=0.36, p=0.26). The lack of correlation was even more obvious between EIA-analyzed (1-3)-ß-D-glucan content per fungal spore and the corresponding spore size (r=0.22, p=0.53), surface area (r=0.27; p=0.40), or volume (r=-0.23, p=0.47). Furthermore, we found close to significant correlation between

LAL and EIA-analyzed (1-3)-ß-D-glucan content per spore (r=0.48, p=0.11), but not per surface area (r=0.24, p=0.45) or per volume (r=0.20, p=0.53).

In addition, there was a strong association between LAL-analyzed (1-3)-ß-D- glucan content in pure fungal spores and the field data on respective fungal species. (1-

3)-ß-D-glucan content per spore correlated with both the frequency and concentration of the respective fungal species in the 297 dust samples analyzed using the PCR in the

CCAAPS study (Figs. 3 and 4). The EIA analyzed (1-3)-ß-D-glucan content failed to show such correlations (frequency: r=0.27, p=0.40; concentration: r=0.27, p=0.39).

However, as spore size also correlated with the (1-3)-ß-D-glucan content (r=0.46, p=0.13), frequency (r=0.64, p=0.03), and concentration (r=0.53, p=0.08), we tested

134

whether it is the (1-3)-ß-D-glucan content (pg/spore) or the spore size that drives the correlation between (1-3)-ß-D-glucan content per spore and the fungal frequency / concentration. For this purpose, we compared the (1-3)-ß-D-glucan content vs. the values of frequency and concentration divided by the spore size. The analysis showed that the

(1-3)-ß-D-glucan content is strongly correlated with the concentration of fungi in indoor dust samples (r=0.78, p=0.003), but it is the spore size that correlated with the fungal frequency rather than the (1-3)-ß-D-glucan content (r=0.05, p=0.88). In addition there was a significant correlation between indoor fungal species frequency and concentration

(r=0.95, p<0.001).

Based on the estimated (1-3)-ß-D-glucan content in common indoor fungal species, measured by both the LAL and EIA assay, the predicted total (1-3)-ß-D-glucan content due to each fungal species can be calculated when the concentration of these species is known in field samples (Table 5). We compared the total (1-3)-ß-D-glucan concentration as measured in 297 dust samples versus the predicted total (1-3)-ß-D- glucan assuming that the main contribution of (1-3)-ß-D-glucan comes from these 12 fungal species. Of these, E. nigrum, Cladosporium species, A. chevalieri and W. sebi emerged as the major sources of (1-3)-b-D-glucan as identified by both assays, contributing more than 85 % of the total (1-3)-b-D-glucan concentration in the dust samples. The predicted total (1-3)-ß-D-glucan was five times greater than the measured one with the LAL assay in per gram units, and six times greater in per square meter units.

No correlation between the predicted and measured (1-3)-ß-D-glucan concentration was observed (mg/g: r=0.03, p=0.57; mg/m2: r=0.01, p=0.87). The measured (1-3)-ß-D-glucan

135

concentration was significantly lower than the predicted (1-3)-ß-D-glucan concentration

(mg/g: t-test= -7.37, p<0.001; mg/m2: t-test= -16.03, p<0.001).

Interestingly, the predicted values of (1-3)-ß-D-glucan as calculated based on the

LAL and EIA-determined fungal (1-3)-ß-D-glucan content in pure spores, were strongly correlated (mg/g: r=0.94, p<0.001; mg/m2: r=0.82, p<0.001). Therefore, regardless of the different contributions of W. sebi, A. pullulans and two Cladosporium species to the total

(1-3)-ß-D-glucan concentration in dust samples when measured by the LAL vs. the EIA assay, if all these six commonly found indoors species (C. herbarum, C. cladosporioides,

E. nigrum, W. sebi, P. brevicompactum, A. pullulans) grow together, the LAL and EIA assays will show comparable results. Still, both assays are significantly different with the

LAL-determined fungal (1-3)-ß-D-glucan content being lower than the EIA determined

(mg/g: t-test= -2.38, p=0.02; mg/m2: t-test= -12.18, p<0.001).

Comparison of LAL vs. EIA analyzed (1-3)-ß-D-glucan in dust samples

The EIA-measured (1-3)-ß-D-glucan levels {mg/g: geometric mean (GM) =4.20; mg/m2: GM=3.58) in dust samples were higher than that the LAL-measured (mg/g:

GM=1.60; mg/m2: GM=0.97). However, the variability in (1-3)-ß-D-glucan levels was greater as measured by the LAL {mg/g: Geometric standard deviation (GSD) GSD=1.70; mg/m2: GSD=2.00} than the EIA assay (mg/g: GSD=1.45; mg/m2: GSD=1.81). There was no correlation between LAL- and EIA- analyzed (1-3)-ß-D-glucan in dust samples, when concentration was expressed per gram, but strong and significant when expressed per square meter (Figs. 5 and 6). In both units, the difference between the LAL- and EIA-

136

analyzed (1-3)-ß-D-glucan levels was significant (mg/g: z-value = -7.21, p<0.001; mg/m2: z-value = -7.26, p<0.001).

DISCUSSION

The utility of the LAL assay in measuring serum fungal (1-3)-b-D-glucans has been evaluated in numerous laboratory and clinical studies (Yoshida et al., 1997; Yuasa et al., 1996; Odabasi et al., 2006), and is currently routinely used in Japan and Europe for the detection of invasive fungal infections (Odabasi et al., 2006; Stevens et al., 2002).

2000; Douwes, 2005). Our study confirmed that the LAL assay recognizes both linear and branched (1-3)-b-D-glucans, as previously reported by Tanaka et al. (1991) and

We confirmed the data reported by Douwes et al. (1996) that the EIA immunoassay

137

reacts with linear (1-3)-b-D-glucans. Other modifications of the EIA assay were developed by Adachi et al. (1999) and Milton et al. (2001), which are highly specific to and measure the content of branched (1-3)(1-6)-b-D-glucans only. It has been reported that both linear and branched type (1-3)-b-D-glucans are ubiquitous in the cell wall of fungi (Perez and Ribas, 2004), and both linear and branched (1-3)-b-D-glucans are important for in vivo priming of macrophages (Ohno et al., 1995). Thus measurement of both linear and branched (1-3)-b-D-glucans is equally important. This makes both the

LAL and inhibition EIA assays more advantageous to use in comparison to the new

EIA assay (LOD=40 ng/ml, detects preferably the linear (1-3)-b-D-glucans). In addition, the LAL was more accurate in measuring concentrations of (1-3)-b-D-glucan standards than the EIA as demonstrated by the narrower range of the % expected concentration and the smaller median value of this %

From previous studies using the LAL assay on serum and culture supernatants of clinical mold isolates, it is known that different fungal species produce a wide range of

(1-3)-ß-D-glucan content (Odabasi et al., 2006; Yoshida et al., 1997; Yuasa et al., 1996).

This study also showed a wide range of (1-3)-b-D-glucan content between species and within Aspergillus genus, as demonstrated by the large coefficients of variation measured in this study. The (1-3)-b-D-glucans content varied considerably between the five

Aspergillus spp. spores unlike what was observed by Odabasi et al. (2006) (range in 12

Aspergillus spp.: 1,311-2,480 pg/ml, GM= 1,769). However, as the latest study presented

138

(1-3)-b-D-glucan content per ml of supernatant of culture broth, but not biomass normalized per spore, results cannot be directly compared.

3)-b-D-glucans content in the same fungus species when determined by LAL vs. EIA.

In our study, Aspergillus species and S. chartarum, had the lowest (1-3)-b-D-glucan contents, but the Aspergillus species had higher content than S. chartarum. Fogelmark and Rylander (1997) have reported that the Stachybotrys atra (=S. chartarum) (median

3.9 pg/spore, range 0.863-39.330 pg/spore) has 1,000 times higher LAL-analyzed (1-3)- b-D-glucan than Penicillium aurantiogriseum (median: 0.09 pg/spore, range 0.005-1.8 pg/spore) and Aspergillus fumigatus (median: 0.11 pg/spore, range 0.008-0.7 pg/spore).

The much lower (1-3)-b-D-glucan content in the Stachybotrys spores in our study may be due to the fact it was a non-toxic strain. The last is based on the assumption that fungal toxicity may be related to pathogenicity, as low fungal (1-3)-glucan content is associated with lower pathogenicity (Hogan et al., 1994; Rementeria et al., 2005; Rappleye et al.,

2006). Also, the growth medium can affect the content of fungal (1-3)-b-D-glucans (Foto

2004). The ATCC mold species in our study were grown on agar media, while the ones in

Rylander’s study were grown on rice and wood. In addition, a study by Foto et al (2004), showed that the LAL-analyzed (1-3)-b-D-glucan content in S. chartarum (mean 0.012 pg/mm2) is lower than that of A. versicolor (mean 0.022 pg/mm2) and C. cladosporioides

(mean 0.060 pg/mm2), all grown on 2% malt extract agar. Although the (1-3)-b-D-glucan

139

contents measured by the LAL in our study are higher than those reported by Foto et al.

(2004), we observed the same trend among species. The order of EIA analyzed fungal (1-

3)-b-D-glucan content in our study was different from what was reported by the monoclonal IgM ELISA assay in a mixture of fungal mycelia and spores (Milton et al.,

2001). Aspergillus flavus isolated from stored urine and Aspergillus ochraceus isolated from outdoor air samples had higher (1-3)-b-D-glucan content than Cladosporium spp. isolated from bedroom air and Wallemia spp. isolated from outdoor air. This may be due to the different sources from which species were isolated, as well as that both spores and mycelia contribute to the measured (1-3)-b-D-glucan content. The IgM ELISA measures the content of branched (1-3)(1-6)-b-D-glucan only, and thus may underestimate the (1-

3)-b-D-glucan content of Wallemia and Cladosporium spp., as they have been reported to contain predominantly linear glucans (San-Blas et al., 1996).

(1-3)-b-D-glucan content not only per spore, but also per surface area and volume. Our data set supports this by the fact that the order of fungi in relation to their (1-3)-b-D- glucan glucan content changed when data was presented per spore compared to when it was presented per surface area/volume. The order, however, was relatively the same whether (1-3)-b-D-glucan content was expressed per surface area or per volume.

140

LAL-analyzed (1-3)-b-D-glucan content per spore and the spore size. The lack of correlation with surface area and volume, may be due to higher extraction efficiency of

(1-3)-b-D-glucan from larger cell surfaces, i.e., larger spore size, but lower efficiency from thicker cell walls such as of some Aspergillus species (Reijula, 1991). However, as the surface structures of some species, such as E. nigrum is very rough (uneven) and, the actual surface area is much larger than the one we estimated for a smooth surface, thus the calculated (1-3)-b-D-glucan content per surface area may be higher than the actual.

The strong and statistically significant correlation between LAL-analyzed fungal

(1-3)-b-D-glucan content and the concentration of the corresponding fungi in indoor dust samples may attribute to the ecological advantage to survival of fungal species of high (1-

Nevertheless, the strong correlation of LAL-measured (1-3)-b-D-glucan content per spore

14 1

with the fungal spore size and with the prevalence of indoor fungi, indicates that the

LAL-measured (1-3)-b-D-glucan may be used as a measure of total mold.

We have also found that species that are more frequently found in indoor environments are also the ones that contribute most to the total (1-3)-b-D-glucan concentration in indoor samples. Of these, E. nigrum, Cladosporium species, A. chevalieri and W. sebi are the major sources of (1-3)-b-D-glucan as identified by both assays, contributing more than 85 % of the total (1-3)-b-D-glucan concentration.

To our knowledge, this is the first report on LAL-analyzed (1-3)-ß-D-glucan concentration in dust samples. Our EIA assay measured much lower indoor dust (1-3)-ß-

D-glucan (mg/g: GM=4.20; mg/m2: GM=3.58), as compared to other studies. However,

142

the smaller sample size can contribute to this as the range of geometric means reported in literature varies widely – from 35.1 mg/g (GSD=1.80, n=20; Milton et al. 2001) to 1,711 mg/g (GSD=1.9, n=395; Gehring et al. 2001), and from 90 mg/m2 (no GSD reported, n=508; Douwes et al. 2006) to 1,197mg/m2 (GSD=2.5, n=395; Gehring et al. 2001). In addition, the source of primary and secondary antibodies used is also different between the EIA assays.

The strong correlations between LAL-and EIA-analyzed (1-3)-b-D-glucan concentration in indoor dust samples per square meter and the lack of correlation, when levels were expressed per gram of dust was due to smaller variation in the values expressed per square meter. The variance in levels per square meter was largely determined by the amount of dust sampled.

In conclusion, the LAL assay is more specific, sensitive and accurate in detecting both linear and branched (1-3)-b-D-glucans. Although the (1-3)-b-D-glucan concentration in our field samples measured by the LAL and EIA assay correlated, data shall be analyzed with caution, as assays give different weight to different fungal species.

In addition, the strong significant associations of the LAL-analyzed (1-3)-b-D-glucan with fungal spore size and with indoor dust fungal concentration, indicates that LAL- analyzed (1-3)-b-D-glucan could be used to estimate the total fungal load in indoor samples.

143

ACKNOWLEDGEMENTS

This study was supported by National Institute of Environmental Health Sciences

(NIEHS) grant ES11170 and the National Institute for Occupational Safety and Health

(NIOSH) Pilot Research Project Training Program of the University of Cincinnati

Education and Research Center Grant T42/OH008432.

REFERENCES

Adachi Y, Ohno N, Yadomae T. Preparation and antigen specificity of an anti-(1-

->3)-beta-D-glucan antibody. Biol Pharm Bul 1994;17:1508-12.

Adhikari A, Martuzevicius D, Reponen T, Grinshpun S, Cho S, Sivasubramani S,

Zhong, W, Levin L, Kelly A, St. Clair H. and LeMasters G. Performance of the button personal inhalable sampler for the measurement of outdoor aeroallergens. Atmos Environ

2003;37:4723–4733.

Alwis K, Mandryk J, Hocking A. Exposure to biohazards in wood dust: bacteria, fungi, endotoxins, and (1-3)-b-D-glucans. Applied Occup Environ Hygiene 1999;14:598-

608.

Bauer H, Kasper-Giebl A, Zibuschka F, Hitzenberger R, Kraus G, Puxbaum H.

Determination of the carbon content of airborne fungal spores. Anal Chem 2002;74:91-5.

Chew G, Douwes J, Doekes G, Higgins K, Strien R, Spithoven J, Brunekreef B.

Fungal extracellular polysaccharides, b(1,3)-glucans and culturable fungi in repeated sampling of house dust. Indoor Air 2001;11:171-178.

144

Cho S-H, Reponen T, LeMasters G, Levin L, Huang J, Meklin T et al. Mold damage in homes and wheezing in infants. Ann Allergy Asthma Immunol 2006;27:1-7.

Curtis L. Adverse Health Effects of Indoor Molds. J Nutr Environ Med

2004;14:261.

Douwes J, Doekes G, Montijn R, Heederik D, Brunekreef B. Measurement of b(1,3)-glucans in occupational and home environments with an inhibition enzyme immunoassay. Appl Environ Microbiol 1999;62:3176-3182.

Douwes J, Thorne P, Pearce N, Heederick D. Bioaerosol health effects and exposure assessment: progress and prospects. Ann Occup Hyg 2003;47:187-200.

Douwes J. 1-3-b-D-glucans and respiratory health: a review of the scientific evidence. Indoor Air 2005;15:160-169.

Douwes J, van Strien R, Doekes G, Smit J, Kerkhof M, Gerritsen J, Postma D, de

Jongste J, Travier N, Brunekreef B. Does early indoor microbial exposure reduce the risk of asthma? The Prevention and Incidence of Asthma and Mite Allergy birth cohort study.

J Allergy Clin Immunol 2006;117:1067-73.

Fogelmark B, Rylander R. (1®3)-b-D-glucan in some indoor air fungi. Indoor

Built Environ 1997;6:291-294.

Foto M, Plett J, Berghout J, Miller J. Modification of the Limulus amebocyte lysate assay for the analysis of glucan in indoor environments. Anal Bioanal Chem

2004;379:156-62.

Gehring U, Douwes J, Doekes G, Koch A, Bischof W, Fahlbusch B, Richter K,

Wichmann H, Heinrich J. b(1,3)-glucan in house dust of German homes: Housing

145

characteristics, occupant behavior, and relations with endotoxins, allergens and molds.

Environ Health Perspect 2001;109:139- 144.

Hossain M, Ahmed M, Ghannoum M. Attributes of Stachybotrys chartarum and its association with human disease. J Allergy Clin Immunol 2004;113:200-208.

Iossifova Y, Reponen T, Bernstein D, Levin L, Kalra H, Campo P, Zeigler H,

Villareal M, Lockey J, Khurana-Hershey G, LeMasters G. House dust (1-3)-b-D-glucan and wheezing in infants? Allergy (submitted) 2006

Ishibashi K, Miura NN, Adachi Y, Tamura H, Tanaka S, Ohno N. The solubilization and biological activities of Aspergillus beta-(1 --> 3)-D-glucan. FEMS

Immunol Med Microbiol 2004;42:155-66.

Johanning, E. Indoor moisture and mold related health problems Eur Ann Allergy

Clin Immunol 2004;36:182-185.

Lai K. Hazard identification, dose-response and environmental characteristics of stachybotryotoxins and other health-related products from Stachybotrys. Environ Technol

2006;27:329-35.

Lindsley W, Schmechel D, Chen B. A two-stage cyclone using microcentrifuge tubes for personal bioaerosol sampling. J Environ Monit 2006;8:1-7.

Mandryk J, Alwis K, Hocking A. Effects of personal exposure on pulmonary function and work-related symptoms among sawmill workers. Ann Occup Hyg

2000;44:281-289.

Milton D, Alwis U, Fisette L, and Muilenberg M. Enzyme-Linked

Immunosorbent Assay Specific for (1 6) Branched, (1 3)-b -D-Glucan Detection in

Environmental Samples. Appl Environ Microbiol 2001;67:5420-5424.

146

Obayashi T, Tamura H, Tanaka S, Ohki M, Takahashi S, Arai M, Masuda M,

Kawai T. A new chromogenic endotoxin-specific assay using recombined limulus coagulation enzymes and its clinical applications. Clin Chim Acta 1985;149:55-65.

Odabasi Z, Paetznick V, Rodriguez J, Chen E, Mcginnis M, Ostrosky-Zeichner L.

Differences in beta-glucan levels in culture supernatants of a variety of fungi. Med Mycol

2006;44:267-272.

Ohno N, Miura N, Chiba N, Adachi Y, Yadomae T. Comparison of the mmunopharmacological activities of triple and single-helical Schizophyllan in mice. Biol

Pharm Bull 1995;18:1242-1247.

Perez P, Ribas J. Cell wall analysis. Methods 2004;33:245-251.

Reponen T, Grinshpun S, Conwell K, Wiest J, Anderson M. Aerodynamic versus physical size of spores: measurement and implication for respiratory deposition. Grana

2001;40:119-125.

Reijula K. Two common fungi associated with farmer’s lung: fine structure of

Aspergillus fumigatus and Aspergillus umbrosus. Mycopathologia 1991;113:143-149.

Ryan P, LeMasters G, Biagini J, Bernstein D, Grinshpun S, Shukla R et al. Is it traffic type, volume, or distance? Wheezing in infants living near truck and bus traffic. J

Allergy Clin Immunol 2005;116:279-84.

Rylander R. Symptoms and mechanisms: Inflammation of the lung. Am J Ind Med

1994;25:19-23.

Rylander R, Thorn J, Attefors R. Airways inflammation among workers in a paper industry. Eur Respir J 1999;13:1151-1157.

147

Rylander R, Lin R. (1-->3)-beta-D-glucan - relationship to indoor air-related symptoms, allergy and asthma. Toxicology 2000;152:47-52.

San-Blas G, Guanipa O, Moreno B, Pekerar S, San-Blas F. Cladosporium carrionii and Hormoconis resinae (C. resinae): cell wall and melanin studies. Curr

Microbiol 1996;32:11-6.

Schmechel D, Simpson J, Beezhold D, Lewis D. The development of species- specific immunodiagnostics for Stachybotrys chartarum: The role of cross-reactivity. J

Immunol Methods 2006;309:150-159.

Schmid F, Stone BA, McDougall BM, Bacic A, Martin KL, Brownlee RT, Chai

E, Seviour RJ. Structure of epiglucan, a highly side-chain/branched (1 --> 3;1 --> 6)-beta- glucan from the micro fungus Epicoccum nigrum. Carbohydr Res 2001;331:163-71.

Shematek E, Braatz J, Cabib E. J Biol Chem 1980;255:888-94.

Steele C, Rapaka R, Metz A, Pop S, Williams D, Gordon S, Kolls J, Brown G.

The beta-glucan receptor dectin-1 recognizes specific morphologies of Aspergillus fumigatus. PLoS Pathog 2005;1:42.

Stevens DA. Diagnosis of fungal infections: Current status. J Antimicrob

Chemother 2002;49 (S1):11-19.

Tanaka S, Aketagawa J, Takahashi S, Shibata Y, Tsumuraya Y, Hashimoto Y.

Activation of a limulus coagulation factor G by (1-3)-b-D-glucans. Carbohydr Res

1991;218:167-174.

Thorne P, Duchaine C, Douwes J, Eduard W, Gorny R, Jacobs R, Reponen T,

Schierl R, Szponar B. Working group report 4: exposure assessment for biological agents. Am J Ind Med 2004;46:419-422.

148

Vesper S, McKinstry C, Haugland R, Iossifova Y, LeMasters G, Levin L,

Hershey GK, Villareal M, Bernstein DI, Reponen T. EPA Relative Moldiness Index© as

Predictor of Childhood Respiratory Illness. J Expo Anal Environ Epi 2006 (in press)

Yoshida K, Niki Y, Ohno M, Watanabe S, Hashiguchi K, Nakajima M,

Matsushima T. Clinical significance of (1-->3)-beta-D-glucan in pleural effusion and liquor. Kansenshogaku Zasshi 1997;71:1210-5.

Yuasa K, Goto H, Iguchi M, Okamura T, Ieki R. Evaluation of the diagnostic value of the measurement of (1-->3)-beta-D-glucan in patients with pulmonary aspergillosis. Respiration 1996;63:78-83.

Wan, G. -H., Li C.S. Indoor endotoxin and glucan in association with airway inflammation and systemic symptoms. Arch Environ Health 1999;54:172-179.

149

FIGURE TITLES:

Fig. 1 Comparison of eight purified glucans and their reactivity as measured by the kinetic LAL (A) and endpoint EIA (B) assays.

Fig. 2 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding spore size (geometric mean).

Fig. 3 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding frequency (%) in indoor dust samples (PCR analysis).

Fig. 4 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding concentration (geometric mean of cells/mg) in indoor dust samples (PCR analysis).

Fig. 5 LAL vs. EIA-analyzed (1-3)-ß-D-glucan in 70 dust samples (CCAAPS study) in mg/g.

Fig. 6 LAL vs. EIA-analyzed (1-3)-ß-D-glucan in 70 dust samples (CCAAPS study) in mg/m2.

150

3500 A. LAL Mannan, Dextran, Pullulan 3000

Schizophylan 2500 Laminarin 2000 MG-glucan

1500

Mean Onset Time (min) 1000 Curdlan 500 Pachyman 0 0 20 40 60 80 100 120

1.2 B. EIA

1.0 Pullulan

0.8 Dextran Schizophylan 0.6 Mannan 0.4 MG-glucan Pachyman 0.2 Absorbance units (450nm)

0.0 Curdlan Laminarin

0 10x106 20x106 30x106 40x106 50x106 60x106

(1-3)-b -D-glucan concentration (pg/ml) Fig. 1 Comparison of eight purified glucans and their reactivity as measured by (A) the kinetic LAL and (B) endpoint EIA assays.

151

103

102

101

100 r=0.46, p=0.13

10-1 -D-glucan content (pg/spore) b

-2 (1-3)- 10

10-3 0 5 10 15 20 25 30 Spore size (mm)

Fig. 2 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding spore size.

152

103

102

r=0.81, p=0.02 101

100

10-1 -D-glucan content (pg/spore) b

10-2 (1-3)-

10-3 0 20 40 60 80 100 Fungal frequency (%) in indoor dust samples

Fig. 3 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding frequency (%) in indoor dust samples (PCR analysis).

153

103

102 r=0.85, p=0.001

101

100

10-1 -D-glucan content (pg/spore) b

-2

(1-3)- 10

10-3 100 101 102 103 104 Fungal concentration (cell/mg dust) in indoor dust samples

Fig. 4 Correlation between LAL-analyzed (1-3)-b-D-glucan content (pg/spore) of twelve fungal species and their corresponding concentration (geometric mean of cells/mg of dust) in 297 indoor dust samples (PCR analysis).

154

105 g/g) m -D-glucan ( b 104

EIA -analyzed (1-3)- r=0.17, p=0.17

103 100 101 102 103 104 LAL-analyzed (1-3)-b-D-glucan (mg/g)

Fig. 5 LAL vs. EIA-analyzed (1-3)-ß-D-glucan in 70 dust samples (CCAAPS study) in mg/g.

155

106 r=0.57, p<0.001 ) 2 g/m

m 105 -D-glucan (

b 104

103 EIA-analyzed (1-3)-

102 10-1 100 101 102 103 LAL-analyzed (1-3)-b-D-glucan (mg/m2)

Fig. 6 LAL vs. EIA-analyzed (1-3)-ß-D-glucan in 70 dust samples (CCAAPS study) in mg/m2.

156

Table 1. Fungal species selected for this study and their frequency and concentration

(geometric mean), as measured by PCR analysis of 297 dust samples (Cincinnati

Childhood Allergy and Air Pollution Study).

Fungal species name ATCC number of the strain Frequency indoors (total Concentration used for the (1-3)-ß-D- number and % of the (cells/mg dust)* glucan content analysis total number of houses in which the species occurred)* Aureobasidium pullulans 58926 297 (100%) 4365.2 Cladosporium 6721 297 (100%) 1520.6 cladosporioides Epicoccum nigrum 58875 293 (98.7%) 250.6 Cladosporium herbarum 58927 293 (98.7%) 184.1 Aspergillus chevalieri 66451 291 (98.0%) 153.5 Aspergillus penicillioides 16910 277 (93.3%) 39.4 Wallemia sebi 42694 275 (92.6%) 61.7 Penicillium brevicompactum 9056 212 (71.4%) 30.8 Stachybotrys chartarum 29-51-05, NIOSH, 153 (51.5%) 4.0 Morgantown, WV Aspergillus flavus 11489 103 (34.7%) 3.0 Aspergillus unguis 10032 86 (29.0%) 2.1 Aspergillus versicolor 52173 68 (23.0%) 4.2 *Vesper et al. 2006

157

Table 2. Comparison of eight purified glucans and their relative reactivity as measured by

the LAL and EIA assays. The standard against which the concentrations are measured is

Pachyman for the LAL assay, and Laminarin for the EIA assay.

Glucan Linkages LAL – % of the expected EIA– % of the expected concentration (%)* concentration (%)*

Range Median Range Median value value Curdlan (1-3)-b-D-glucan 43-100 96.4 95-150 97.0 Pachyman (1-3)-b-D-glucan 97-118 104.8 56 - 94 73.7 Laminarin (1-3)(1-6)-b-D-glucan 78-118 99.3 17-120 98.2 (some branching) Schizophylan (1-3)(1-6)-b-D-glucan 47-118 84.7 No response (33% branching) MG-glucan (1-3)(1-6)-b-D-glucan 74-240 96.0 No response extract from baker’s yeast Mannan (1-6)-a-D-Mannose No response No response Dextran (1-3)(1-6)-a-D-glucose No response No response Pullulan (1-3)(1-4)-a-D-glucose No response No response * The prepared (expected) concentrations of the glucan standards for the LAL assay

were: 3.125 pg/ml, 12.50 pg/ml, 50 pg/ml and 100 pg/ml. The prepared (expected)

concentrations of the glucan standards for the EIA assay were 250 ng/ml, 1000ng/ml,

2500ng/ml, and 5000 ng/ml. The glucan measured concentrations were calculated based

on a standard curve of Pachyman for the LAL assay (2 runs) and Laminarin for the EIA

assay (1 run). The measured values were then expressed as % of the expected

concentrations.

158

Table 3. Characteristics of fungal species – spore size, spore surface area and spore volume (average of n=30 spores for each fungal species).

Fungal species Spore Size Spore surface Spore volume

(µm) area (µm2) (µm3)

C. herbarum 4.76* 59.18 36.00

E. nigrum 28.10 2463 11494

P. brevicompactum 3.05 14.39 5.13

A. pullulans 11.83* 628.45 786.1

C. cladosporioides 5.64* 83.42 59.72

A. chavalieri 4.50 63.62 47.71

W. sebi 3.95 49.02 32.27

A. flavus 4.92 76.05 62.36

A. versicolor 2.96 27.53 13.58

A. penicillioides 4.45 62.21 46.14

A. unguis 3.11 30.39 15.75

S. chartarum 7.45* 148.48 172.15

* Geometric mean of width and length for ellipsoidal spores

159

Table 4. Average (1-3)-ß-D-glucan contents of twelve common indoor fungal species as measured by the LAL and EIA assay (picograms per spore, picograms per spore surface area, and picograms per spore volume).

Fungal species LAL –measured (1-3)-ß-D-glucan EIA – measured (1-3)-ß-D-glucan

pg/spore pg/µm2x103 pg/µm3 x103 pg/spore pg/µm2 x103 pg/µm3 x103

C. herbarum 8.66 146.33 240.54 (0.06)* (1.01)* (1.72)*

E. nigrum 241.57 98.08 21.02 379.97 153.25 32.72

P. brevicompactum 0.21 7.02 13.81 3.39 116.10 228.39

A. pullulans 3.76 5.98 4.78 (0.02)* (0.04)* (0.03)*

C. cladosporioides 0.25 3.00 4.19 7.20 86.30 120.55

A. chavalieri 0.22 3.46 4.61 0.24 3.81 5.08

W. sebi 0.12 2.40 3.72 9.68 197.00 300.01

A. flavus 0.03 0.39 0.48 0.10 1.34 1.64

A. versicolor 0.03 0.91 1.84 0.08 3.01 6.10

A. penicillioides 0.01 0.16 0.22 0.13 2.06 2.78

A. unguis 0.005 0.16 0.32 0.02 0.57 1.10

S. chartarum 0.004 0.03 0.33 (0.58)* (3.94)* (3.40)*

* The value for this fungal (1-3)-ß-D-glucan concentration was below the lower detection limit of the EIA assay. One half of the LOD of the pg/ml value was assigned for the (1-

3)-ß-D-glucan concentration of those fungal species.

160

Table 5. Contribution of predicted fungal (1-3)-ß-D-glucan (pg/mg dust) based on LAL and EIA-analyses and concentration of fungal species (total in all 297 dust samples, cells/mg dust).

Fungal species name Total Predicted fungal content of (1-3)-ß-D-glucan, pg/mg concentration dust (pg/m2 area) (cells/mg dust) Based on LAL Based on EIA in all 297 dust pg/mg pg/m2 pg/mg pg/m2 samples Aureobasidium pullulans 3x106 12 x106 18 x106 0.06 x106 0.12 x106 Cladosporium 1 x106 0.3 x106 3.4 x106 8 x106 97 x106 cladosporioides (Type I) Epicoccum nigrum 0.6 x106 140 x106 57 x106 220 x106 89 x106

TOTAL measured 28 x106 20 x106

161

A2: Use of (1-3)-b-D-glucan concentrations in dust as a surrogate method for estimating specific mold exposures

Iossifova Y1, Sucharew H1, Succop P1, Vesper S2, Reponen T.1†

1 Department of Environmental Health, University of Cincinnati;

2 US Environmental Protection Agency, Cincinnati, Ohio.

†Contact author: Tiina Reponen, Center for Health Related Aerosol Studies, Department of Environmental Health, University of Cincinnati, 3223 Eden Ave, PO Box 670056,

Cincinnati, Ohio 45267-0056, USA;

Email: [email protected], Fax: (513) 558-2263, Tel: (513) 558-0571

Manuscript word count: 2788

Short title: (1-3)-b-D-GLUCAN AND MOLD

Submitted to Indoor Air on 11/09/2006

162

ABSTRACT (word count = 200)

Indoor exposure to fungi has been associated with respiratory symptoms, often attributed to their major cell wall component, (1-3)-b-D-glucan. The ease and low cost of performing (1-3)-b-D-glucan analysis rather than cultivation or microscopic counting of mold spores, has prompted many to use (1-3)-b-D-glucan as a surrogate for mold exposure. The aim of this study was to examine which indoor mold species predict (1-3)- b-D-glucan concentration in field dust samples, and thus whether (1-3)-b-D-glucan can be used as a surrogate for mold exposure.

We used the quantitative polymerase chain reaction (QPCR) method to analyze

36 indoor fungal species in 297 indoor dust samples, which were also simultaneously analyzed for (1-3)-b-D-glucan concentration. Linear regression analysis, followed by factor analysis and structural equation modeling, were utilized in order to identify fungal species that mostly contribute to the (1-3)-b-D-glucan concentration in field dust samples.

The study revealed that Cladosporium and Aspergillus species were the main (1-

3)-b-D-glucan contributors followed by Epicoccum nigrum, Wallemia sebi and

Penicillium brevicompactum. The species that contributed most to the (1-3)-b-D-glucan concentration were also the most prevalent in indoor environments. However, Alternaria alternata, the third most common fungal species in indoor dust, did not seem to be a significant source of (1-3)-b-D-glucan.

Keywords: fungi, mold, (1-3)-b-D-glucan, QPCR, indoor, factor analysis

163

Practical implications

This study revealed that most prevalent species in indoor environments, such as

164

INTRODUCTION

Exposure to fungi in occupational and indoor environments is associated with building-related illness, allergic diseases like asthma and organic dust toxic syndrome

(Storey et al., 2004, IOM, 2004). Similar symptoms have been linked to occupational and indoor exposures to (1-3)-b-D-glucan (Rylander and Lin, 2000; Douwes, 2005). (1-3)- b-D-glucan is a biologically active poly-glucose molecule comprising up to 60% of the cell wall of mold, some soil bacteria and plants. Measurement of (1-3)-b-D-glucan concentration has been used in several studies as a surrogate to estimate the human exposure to mold (Schram-Bijkerk et al., 2005; Schram et al. 2005; Douwes et al. 2006;

Fogelmark et al., 2001).

Exposure to (1-3)-b-D-glucan reportedly caused stimulation of the reticulo- endothelial system, activation of neutrophils, macrophages, complement and possibly eosinophils (Douwes, 2005). These effects may play a role in the adverse health effects in humans and these effects are not dependent on the viability of the cells. Thus, (1-3)-b-D- glucan from dead mold cells may be relevant in causing health effects from mold exposures. Therefore, (1-3)-b-D-glucan may be a better predictor for health risk than the commonly used determination of viable fungal spores. In addition, the assay for (1-3)-b-

D-glucan is cheaper, quicker and more sensitive method than cultivation or microscopic analysis of mold (Rylander and Lin, 2000; Douwes et al., 1999, 2003).

We are currently conducting a prospective birth cohort study (The Cincinnati

Childhood Allergy and Air Pollution Study or CCAAPS) aimed at investigating the role of aeroallergens and diesel exhaust particles in the development of atopic respiratory disorders in an infant cohort (Ryan et al., 2005; Biagini et al., 2006). As part of this

165

study, a method was developed using a quantitative index of moldiness, called Relative

Moldiness Index (RMI). The RMI index was calculated from quantitative PCR (QPCR) measurements on the concentration of 36 species of molds in floor dust samples. The

RMI scale may be used to predict the occurrence of illness in homes as certain groups of mold species were found to be more dominant in the homes associated with wheezing

(Vesper et al., 2006). In this study, our goal is to examine which indoor mold species predict (1-3)-b-D-glucan concentration in dust samples, and thus whether (1-3)-b-D- glucan can be used as a surrogate for exposure to total mold or specific molds.

METHODS

On-site home visit and exposure assessment

As part of the CCAAPS study, when infants reached an average age of 8 months, families were visited at their homes and dust samples were vacuumed from the baby’s primary activity room floor in order to assess exposure to indoor aeroallergens and mold as described by Cho et al. (2006). In brief, for carpeted floor, samples were collected from area of 2 m2 at a vacuuming rate of 2 min/m2. For non-carpeted floor only one sample was collected from the entire room at a rate of 1 min/m2. The home dust sample was sieved (355 µm sieve), and the fine dust was divided into sub-samples and stored at -

20°C before analyses. Altogether 297 were selected for the study that included PCR analysis of a 5 mg sub-sample of the fine dust (Vesper et al., 2006). Another 50 mg sub- sample was analyzed for (1-3)-ß-D-glucan content.

166

MSQPCR analysis of dust

Methods have been reported previously for preparing conidial suspensions from dust samples, extracting DNA, performing MSQPCR analyses and preparing standard calibration curves (Vesper et al. 2004, Vesper et al., 2006). All primer and probe sequences, as well as known species comprising the assay groups, were published at the website: http://www.epa.gov/microbes/moldtech.htm. Primers and probes were synthesized commercially (Applied Biosystems, Foster City, CA; Integrated DNA

Technologies, Coralville IA; Sigma Genosys, Woodlands, TX).

Thirty-six fungal species were identified and quantified in this study. Mold concentration data having a minimum detection limit of 1 cell mg-1 dust were treated as left-censored data with appropriate statistical methods applied (Helsel, 2005). Non- detections were set as 1.

Analysis of (1-3)-b-D-glucan in dust

The (1-3)-b-D-glucan content in 297 dust samples was determined by the endpoint chromogenic Limulus Amebocyte lysate assay (LAL) (Associates of Cape Cod,

East Falmouth, MA), using laboratory materials (pipette tips, tubes, etc.) certified free of contaminating glucans by the manufacturer (Associates of Cape Cod). The protocol is described in detail by Iossifova et al. (2006). Results were reported as mg/g.

Linear Regression Analysis

Data was visualized as box-plots, histograms and scatter diagrams of each fungal species against the independent variable of (1-3)-b-D-glucan content (data not shown).

167

This revealed non-Gaussian, but skewed to the left distribution, which is usually observed for environmental data. Thus, we performed log-transformation for all data (fungal and

(1-3)-b-D-glucan concentrations). The new log-transformed data had a normal Gaussian distribution (data not shown).

The distributions of the concentrations of individual mold species were skewed to the left (tests for normality: Shapiro-Willks and Kolmogorov-Smirnov). Diagnostics on the data were also done for co-linearity because mold grows in clusters of different genera, so this would be expected. The co-linearity for all variables (n=36) were evaluated using the Pearson correlation. A linear regression model was developed for 36 mold species in 297 homes in Cincinnati. Statistical analyses were performed using SAS

(SAS Institute Inc., Cary, NC).

To test the first null hypothesis that one or more of the fungi will predict the (1-

3)-b-D-glucan content, we fit the best regression model for the log-transformed data.

Many fungal species grow together and thus a second hypothesis was that the 10 most common indoor fungal species together will better predict the (1-3)-b-D-glucan content.

To test this hypothesis these ten fungal species were forced in a backward elimination approach with a =0.05. Then, the same independent variables were forced in a forward selection approach, a =0.05.

Due to the simultaneous fungal growth, there is a potential for interaction effect between the independent variables. A third approach was to test all possible 40 pairs of interaction between the ten most common indoor fungi.

168

Factor Analysis and Structural Equation Modeling

Further analysis was conducted using the multivariate methods of factor analysis and structural equation modeling (SEM) to determine the relationships that may exist among the 36 mold species and to determine how these groups of related mold species predict (1-3)-b-D-glucan content. Several preliminary models were run in steps which led to the final SEM model. These preliminary steps included exploratory factor analysis

(EFA), confirmatory factor analysis (CFA), deriving a base SEM, and then the final

SEM. The log-transformed data was used in all analyses.

SAS procedure Calis was then used to perform the CFA. The initial CFA was based on the final EFA. In a stepwise fashion, mold species were added to the factors based on the top ten largest Lagrange multipliers that were statistically significant with positive association. Similarly, residual correlations were added based on the top ten largest

Lagrange multipliers that were statistically significant. This procedure continued until all estimated loadings and all residual correlations were statistically significant at the 0.05 level. The initial SEM was based on the final CFA with the additional model predicting

(1-3)-b-D-glucan with the factors determined in the final CFA. Factors that were not statistically significant at the 0.05 level were removed from the (1-3)-b-D-glucan model in a stepwise fashion until only significant factors remained in the model.

169

RESULTS

Linear Regression Analysis

Frequency and geometric mean of concentration in dust samples for the 36 fungal species studied are given in Table 1. There were no strong correlations between any mold species concentration and its corresponding (1-3)-b-D-glucan content (Table 1).

However, there were cross-correlations among specific fungal species, of which 12 had a coefficient of correlation r >0.5: Alternaria alternata-Aureobasidium pullulans, A. alternata-Cladosporium cladosporioides, Aspergillus penicillioides-Aspergillus chevalieri, A. penicillioides – Wallemia sebi, Aspergillus unguis-Aspergillus ustus, A. pullulans-Cladosporium herbarum, C. cladosporioides type 1 and 2, C. cladosporioides-

C. herbarum, C. cladosporioides-Cladosporium sphaerospermum, C. cladosporioides-

Epiccocum nigrum, C. spherospermum-W. sebi, and A. chevalieri-W. sebi.

The co-linearity was also diagnosed for the final model with two variables with

VIF (VIF=1). The VIF was not larger using the rule of thumb for VIF<10. No outliers were observed in the final model.

The following tests of linear regression were performed:

1) First, a backward elimination approach including all variables was used with a=0.05, which gave us a linear regression, which included Aspergillus flavus only. Next, a forward inclusion selection approach was used at the same significance level, which gave a model including A. flavus and C. herbarum. The above models were presented by the following equations:

Backward elimination model:

(1-3)-b-D-glucan = 1.637 + 0.115(A. flavus) (1)

170

Forward inclusion model:

(1-3)-b-D-glucan = 1.420 + 0.118(A. flavus) + 0.095(C. herbarum) (2)

2) Forcing the ten most common fungal species in a backward elimination and forward selection approaches (a=0.05) gave the same regression model, reported below

(equation 3). The only significant species in the model were C. cladosporioides and A. flavus.

(1-3)-b-D-glucan = 1.338 + 0.151(C. cladosporioides) + 0.140(A. flavus) (3)

The same above-described procedures were performed by changing the alpha levels to a =0.20, a =0.15 and a =0.10. The models did not change.

3) All possible 40 interaction terms between the ten most common indoor fungi were included in the forward and backward approach with all other 36 variables, a =0.05.

Both gave the same model, which included six interaction effects and only three main effects. Therefore, we forced all main effects that were involved in the interactions in a new forward selection and backward elimination approach.

171

the forward inclusion model with two variables: combined Cladosporium species and E. nigrum (Table 2).

Factor Analysis and Structural Equation Modeling

Six factors were obtained in the final EFA and were then used in the initial CFA.

The final 6-factor CFA fit the data well (c2 (278) = 235, p-value = 0.97). In the final

SEM, factors 1 and 3 were significant for predicting (1-3)-b-D-glucan with 0.116 loading for factor 1 and 0.114 for factor 3. The final SEM also fit the data well with c2 (313) =

272, p-value = 0.95.

The confirmatory factor analysis revealed the following model as significant:

(1-3)-b-D-glucan = 0.1155*factor1 + 0.1141*factor3 + 1.0000 e0 (4)

Std Err 0.0639 f1_0 0.0634 f3_0

t Value 1.8068 1.7986

DISCUSSION

Previous studies have shown that (1-3)-b-D-glucan content varies between mold species of different genera, as well as within same genus (Foto et al., 2004, Fogelmark and Rylander, 1997). This may lead to variance of health outcomes associated with exposure to different fungi. In this cohort, we have found inverse association between the concentration of airborne Cladosporium and allergen sensitization (p < 0.05), but positive associations between Penicillium/Aspergillus (p < 0.01), or Alternaria (p<0.01) and allergen sensitization (Osborne et al., 2006). The above finding may be due to

172

different (1-3)-b-D-glucan content between fungal species, as an increase in the (1-3)-b-

D-glucan content is associated with increased pathogenicity in certain fungal species

(Hogan et al., 1994; Rementeria et al., 2005; Rappleye et al., 2006). Therefore certain fungal species may contribute more than others for the total (1-3)-b-D-glucan content in indoor environments.

The present study utilized advanced statistical methods in order to identify the group of indoor fungi most contributing to the (1-3)-b-D-glucan load in residential dust samples. It revealed these species to be of the Cladosporium and Aspergillus genera. In addition E. nigrum, W. sebi and P. brevicompactum, can also contribute substantially to the (1-3)-ß-D-glucan content in dust samples. Another interesting finding of the study is that fungal species contributing most to the (1-3)-ß-D-glucan concentration are also the ones that are most prevalent in indoor environments.

Most of the biological functions related to pathogenicity and toxicity reside in the fungal cell wall, which, being the outermost part of the cell, mediates the host-fungus interplay (Chaffin et al., 1998). The microfibrillar polymers {(1-3) and (1-6)-ß-D-glucans and chitin} represent the structural components of the wall. They form a rigid skeleton that provides strong physical properties to the cell. From a quantitative point of view, b-

D-glucans are the main constituent, accounting for 47 to 60% by weight of the cell wall

(Young and Castranova, 2005). Therefore, it is expected that the indoor species of highest prevalence will contribute mostly to the (1-3)-b-D-glucan content. Although the species identified by this study certainly are of high frequency and concentration, it shall be emphasized that non-toxic, non-pathogenic species also tend to synthesize more cell wall glucans in favorable environmental conditions (Bowman and Free, 2006).

173

The factor analysis first combined the species based on the correlations among the species. Thus, species of coincident growth, concentration and frequency would be grouped in one factor. Therefore, species of high frequency and concentration were combined in one factor group, while Aspergillus species of low frequency and concentration were combined in another factor group. Thus the combined (1-3)-b-D- glucan contribution of the genera of Aspergillus and Cladosporium is more important than the individual contributions.

Furthermore, this study also suggests that (1-3)-b-D-glucan content varies between species, demonstrated by the different (1-3)-b-D-glucan loadings contributed by different fungal species in the above analyses. The latest was demonstrated in both serum and culture supernatants of clinical mold isolates, (Odabasi et al., 2006; Yoshida et al.,

1997; Yuasa et al., 1996), as well as environmental mold species (Foto et al., 2004;

Fogelmark and Rylander, 1997).

Another interesting finding is that the analysis failed to recognize Alternaria alternata, an allergenic fungus of high prevalence and concentration in indoor dust samples, as a contributor to the (1-3)-b-D-glucan content. Although immuno-modulating effects of (1-3)-b-D-glucan are suspected (Rylander and Holt, 1998; Ormstad et al.,

2000), the scientific data is not affirmative on the (1-3)-b-D-glucan being an allergenic factor (Douwes et al., 2005). Therefore, we can only speculate whether (1-3)-b-D-glucan content can modify the allergic effects of fungi. In addition, there is a possibility that the

(1-3)-b-D-glucan in Alternaria spores is not readily available for extraction and analysis by the LAL assay. Ultimately, as this is a field sample, other fungal species, such as

Cladosporium and Aspergillus species, could have introduced glucanases and made the

174

environment unfavorable for A. alternata. In such cases, we may speculate that although

A. alternata sporulates, its (1-3)-b-D-glucan can decrease, or it does not sporulate, but we detect the (1-3)-b-D-glucan of its mycelia only.

The advantage of this study is that we determined fungal species contributing to the (1-3)-b-D-glucan from real field samples, rather than individual spores or colony forming units isolates. The current study showed that certain species (such as

Cladosporium and Aspergillus) tend to cluster together, while other species may be underrepresented in field samples. Therefore, the synthesized cell wall (1-3)-b-D-glucan by fungal species in indoor environments can be different from the content of their spores or cultures in laboratory studies.

This study has shown that all three most common Cladosporium species (C. cladosporioides, C. herbarum and C. sphaerospermum) contribute to the (1-3)-b-D- glucan loading in dust samples. A recent study of the same cohort (Osborne et al, 2006) have found Cladosporium genus to be inversely associated with allergen sensitization. In the same infant cohort, we have also reported an inverse association between exposure to high (1-3)-b-D-glucan and allergen sensitization (Iossifova et al., 2006). Thus, we may expect that fungal species of high (1-3)-b-D-glucan content will be associated with lower allergen sensitization.

Aspergillus and Penicillium, as well as A. alternata are recognized as allergenic fungi due to their allergenic antigens. Thus, we can speculate that a combination of low

(1-3)-b-D-glucan content in these species, and their allergenic antigens may cause increase in the prevalence of allergenic sensitization among infants. This might explain the different direction of associations found between health outcomes and the

175

concentrations of Cladosporium vs. Aspergillus, Penicillium and Alternaria (Osborne et al. 2006).

As stated earlier, the same database of mold species and indoor dust samples was used in establishing a Relative Moldiness Index. The fungal species we identified as greatest contributors of (1-3)-b-D-glucan, were also identified as of significantly higher concentration in “more moldy homes” than in “less moldy homes” (shown in Table 1)

(Vesper et al. 2006). Ultimately, this data shows us that as (1-3)-b-D-glucan content is representative of most prevalent fungal species, as well as those associated with moldy homes as determined by the RMI. Therefore, total (1-3)-b-D-glucan may be used as a measure of the total relevant mold exposure.

Acknowledgements

This study was supported by the National Institute for Occupational Safety and

Health (NIOSH) Training Program of the University of Cincinnati Education and

Research Center Grant T42/CCT510420.

Notice

The U.S. Environmental Protection Agency (EPA) through its Office of Research and Development funded and collaborated in the research described here. It has been subjected to the Agency’s peer review and has been approved as an EPA publication.

Mention of trade names or commercial products does not constitute endorsement or recommendation by the EPA for use. Since QPCR technology is patented by the US

EPA, the Agency has a financial interest in its commercial use.

176

REFERENCES:

Biagini, J.M., LeMasters, G.K., Ryan, P.H., Levin, L., Reponen, T., Bernstein, D.I.,

Villareal, M., Hershey, G.K., Burkle, J., Lockey, J. (2006) “Environmental risk factors of rhinitis in early infancy”, Pediatr. Allergy Immunol., 17, 278-84.

Bowman, S.M., Free, S.J. (2006) “The structure and synthesis of the fungal cell wall”,

Bioessays, 28, 799-808.

Chaffin, W., Lopez-Ribot, J., Casanova, M., Gozalbo, D., Martinez, J. (1998) “Cell-wall and secreted proteins of Candida albicans: identifications, function, and expression”,

Microbiol. Mol. Biol. Rev., 62. 130-180.

Cho SH, Reponen T, Bernstein DI, Olds R, Levin L, Liu X et al. (2006) “The effect of home characteristics on dust antigen concentrations and loads in homes”, Sci. Total

Environ., 371, 31-43.

Douwes, J., Doekes, G., Montijn, R., Heederik, D., Brunekreef, B. (1996) “Measurement of b(1,3)-glucans in occupational and home environments with an inhibition enzyme immunoassay”, Appl. Environ. Microbiol., 62, 3176-3182.

Douwes, J., Thorne, P., Pearce, N., Heederick, D. (2003) “Bioaerosol health effects and exposure assessment: progress and prospects”, Ann. Occup. Hyg., 47, 187-200.

177

Douwes, J. (2005) “1-3-b-D-glucans and respiratory health: a review of the scientific evidence”, Indoor Air, 15, 160-169.

Douwes, J., van Strien, R., Doekes, G., Smit, J., Kerkhof, M., Gerritsen, J., Postma, D., de Jongste. J,, Travier, N., Brunekreef, B. (2006) “Does early indoor microbial exposure reduce the risk of asthma? The Prevention and Incidence of Asthma and Mite Allergy birth cohort study”, J. Allergy Clin. Immunol., 117, 1067-73.

Fogelmark, B., Rylander, R. (1997) “(1®3)-b-D-glucan in some indoor air fungi”,

Indoor Built Environ., 6, 291-294.

Fogelmark, B., Thorn, J., Rylander, R. (2001) “Inhalation of (1-->3)-beta-D-glucan causes airway eosinophilia”, Mediators Inflamm., 10, 13-19.

Foto, M., Plett, J., Berghout, J., Miller, J.D. (2004) “Modification of the Limulus amebocyte lysate assay for the analysis of glucan in indoor environments”, Anal.

Bioanal. Chem., 379, 156-62.

Institute of Medicine (IOM). (2004) Damp Indoor Spaces and Health, Washington, D.C.,

The National Academies Press.

Iossifova, Y., Reponen, T., Bernstein, D., Levin, L., Kalra, H., Campo, P. et al. (2006)

“House dust (1-3)-b-D-glucan and wheezing in infants”, Allergy (in revision)

178

Helsel, D.R. (2005) Nondetects and Data Analysis, Statistics for Censored Environmental

Data, Hoboken, NJ, Wiley and Sons Inc.

Hogan, L., Klein, B. (1994) “Altered expression of surface alpha-1,3-glucan in genetically related strains of Blastomyces dermatitidis that differ in virulence”, Infect.

Immun., 62, 3543-6.

Odabasi, Z., Paetznick, V., Rodriguez, J., Chen, E., Mcginnis, M., Ostrosky-Zeichner, L.

(2006) “Differences in beta-glucan levels in culture supernatants of a variety of fungi”,

Medical Mycology, 44, 267-272.

Ormstad, H., Groeng, E.C., Lovik, M., Hetland, G. (2000) “The fungal cell wall component beta-1,3-glucan has an adjuvant effect on the allergic response to ovalbumin in mice”, J. Toxicol. Environ. Health, 61, 55-67.

Osborne, M., Reponen, T., Adhikari, A., Cho, S.H., Grinshpun, S.A., Levin, L. et al.

(2006) “ Specific fungal exposures, allergic sensitization, and rhinitis in infants”, Pediatr.

Allergy Immunol., 17, 450-457.

Rappleye, C.A., Goldman, W.E. (2006) “Defining Virulence Genes in the Dimorphic

Fungi”, Annu. Rev. Microbiol., 60, 281-303.

179

Rementeria, A., Lopez-Molina, N., Ludwig, A., Vivanco, A.B., Bikandi, J., Ponton, J. et al. (2005) “Genes and molecules involved in Aspergillus fumigatus virulence”, Rev.

Iberoam. Micol., 22, 1-23.

Ryan, P.H., LeMasters, G., Biagini, J., Bernstein, D., Grinshpun, S.A., Shukla, R. et al.

(2005) “Is it traffic type, volume, or distance? Wheezing in infants living near truck and bus traffic”, J. Allergy Clin. Immunol., 116, 279-84.

Rylander, R., Holt, P.G. (1998) “(1-->3)-beta-D-glucan and endotoxin modulate immune response to inhaled allergen”, Mediators Inflamm., 7, 105-10.

Rylander, R., Lin, R.H. (2000) “(1-->3)-beta-D-glucan - relationship to indoor air-related symptoms, allergy and asthma”, Toxicology, 152, 47-52.

Schram-Bijkerk, D., Doekes, G., Douwes, J., Boeve, M., Riedler, J., Ublagger, E. et al.

(2005) “Bacterial and fungal agents in house dust and wheeze in children: the

PARSIFAL study”, Clin. Exp. Allergy, 35, 1272-8.

Schram, D., Doekes, G., Boeve, M., Douwes, J., Riedler, J., Ublagger, E. et al. (2005)

“Bacterial and fungal components in house dust of farm children, Rudolf Steiner school children and reference children-the PARSIFAL Study”, Allergy, 60, 611-8.

Storey, E., Dangman, K.H., Schenck, P., DeBernardo, R.L., Yang, C.S., Bracker, A., and

Hodgson, M.J. (2004) Guidance for Clinicians on the Recognition and Management of

180

Health Effects Related to Mold Exposure and Moisture Indoors, University of

Connecticut Health Center, Farmington, Connecticut. http://oehc.uchc.edu/clinser/MOLD%20GUIDE.pdf (accessed 01/31/05)

Vesper, S.J., Varma, M., Wymer, L.J., Dearborn, D.G., Sobolewski, J., Haugland, R.A.

(2004) “Quantitative polymerase chain reaction analysis of fungi in dust from homes of infants who developed idiopathic pulmonary hemorrhaging” J. Occup. Environ. Med., 46,

596-601.

Vesper, S., McKinstry, C., Haugland, R., Iossifova, Y., LeMasters, G., Levin, L. et al.

J. Expo. Anal. Environ. Epi. (in press).

Yoshida, K., Niki, Y., Ohno, M., Watanabe, S., Hashiguchi, K., Nakajima, M. et al.

(1997) “Clinical significance of (1-->3)-beta-D-glucan in pleural effusion and liquor”,

Kansenshogaku Zasshi, 71, 1210-5.

Young, S-H. and Castranova, V. (2005) Toxicology of (1-3)-b-D-glucans. Glucans as a marker for fungal exposure, CRC Press Taylor&Francis group.

Yuasa, K., Goto, H., Iguchi, M., Okamura, T., Ieki, R. (1996) “Evaluation of the diagnostic value of the measurement of (1-->3)-beta-D-glucan in patients with pulmonary aspergillosis”, Respiration, 63, 78-83.

181

Table 1. Frequency and concentration of 36 indoor fungal species analyzed with the QPCR method and correlations between the concentration of each species and the cumulative (1-3)-b-D-glucan concentration (n = 297 dust samples).

* Correlation between the (1-3)-b-D-glucan concentration and the cell concentration measured by the PCR

†Fungal species identified as being significantly in higher concentration in “more moldy homes” than in “non-moldy homes” (Vesper et al., 2006)

183

Table 2 Fungi associated with the indoor (1-3)-ß-D-glucan concentration.

Model Standardized Frequency indoors ( % coefficients of the total number of houses) No interactions Aspergillus flavus 0.12* 34.7% Cladosporium herbarum 0.10* 98.7%

Two-way interactions Cladosporium cladosporoides type II 0.44*** 93.9% Penicilium brevicompactum 0.51** 71.4% Aspergillus flavus 0.11** 34.7% Aspergillus unguis 0.67** 29% Interaction between Cladosporium 0.30** cladosporoides types I and II

Three-way interactions Combined Cladosporium species 0.42*** 97% Epicoccum nigrum 0.22 (p=0.20) 98.7% Interaction between Cladosporium -0.04* species and Epicoccum nigrum

*p<0.05, **p<0.01, *** p<0.001

The standardized coefficient for Alternaria alternata (X0) was 0, and thus is not included in the equation, despite its involvement in an interaction effect.

184

Table 3. Final 6-factor SEM model with Loadings. Only Factor 1 and 3 were significant, and thus are reported.

Factor 3 Loading Log X7 Aspergillus penicillioides 0.7949 93.3% Log X21 Aspergillus chevalieri 0.7652 98.0% Log X36 Wallemia sebi 0.6295 92.6% Log X13 Aspergillus versicolor 0.5031 23.0%

185

A3: House dues (1-3)-b-D-glucan and wheezing in infants

Yulia Y. Iossifova1, Tiina Reponen1†, David I. Bernstein2, Linda Levin1, Harpinder

Kalra2, Paloma Campo2, Manuel Villareal2, James Lockey1, Gurjit K. Khurana Hershey3,

Grace LeMasters1

1 Department of Environmental Health, University of Cincinnati,

2 Division of Allergy-Immunology, Department of Internal Medicine, University of

Cincinnati,

3 Division of Allergy and Immunology, Cincinnati Children’s Hospital Medical Center

†Contact author: Tiina Reponen, Center for Health Related Aerosol Studies, Department of Environmental Health, University of Cincinnati, 3223 Eden Ave, PO Box 670056,

Cincinnati, Ohio 45267-0056, USA;

Email: [email protected], Fax: (513) 558-2263, telephone: (513) 558-0571

Short title: (1-3)-b-D-GLUCAN AND WHEEZING

Revised manuscript # ALL-2006-00464

Resubmitted to: Allergy on 11/09/2006

186

ABSTRACT

Background: (1-3)-b-D-glucan is a fungal cell wall component, suspected to cause respiratory and general symptoms in adults. However, very little is known on the possible health effects of (1-3)-b-D-glucan during infancy. We examined the association between

(1-3)-b-D-glucan exposure and the prevalence of allergen sensitization and wheezing during the first year of life in a birth cohort of 574 infants born to atopic parents.

Endotoxin exposure was included as a possible confounder.

Methods: (1-3)-b-D-glucan and endotoxin exposures were measured in settled dust collected from infants’ primary activity rooms using Limulus amebocyte lysate assay.

The primary outcomes at approximately age one included parental reports of recurrent wheezing and allergen sensitization evaluated by skin prick testing to a panel of 15 aeroallergens as well as milk and egg white.

Results: Exposure to high (1-3)-b-D-glucan concentration (within 4th quartile) was associated with reduced likelihood of both recurrent wheezing [adjusted OR (aOR) =

0.39; 95% CI=0.16-0.93] and recurrent wheezing combined with allergen sensitization

(aOR=0.13; 95% CI=0.03-0.61). Similar trends were found between (1-3)-b-D-glucan concentrations and allergen sensitization (aOR=0.57, 95%CI 0.30-1.10). In contrast, recurrent wheezing with or without allergen sensitization was positively associated with low (1-3)-b-D-glucan exposure within the 1st quartile (aOR=3.04, 95%CI 1.25-7.38; aOR=4.89, 95%CI 1.02-23.57). There were no significant associations between endotoxin exposure and any of the studied health outcomes.

Conclusions: This is the first study to report that indoor exposure to high levels of (1-3)- b-D-glucan (concentration >60 µg/g) is associated with decreased risk for recurrent

187

wheezing among infants born to atopic parents. This effect was more pronounced in the subgroup of allergen sensitized infants.

Abstract Word count: 228

Keywords: allergen sensitization, (1-3)-b-D-glucan, endotoxin, infants, wheeze

188

INTRODUCTION

Exposure to indoor molds during infancy has been associated with respiratory symptoms, such as increased risk for persistent cough and wheeze (1,2). However, evaluation of human health effects of mold exposure has been hampered by lack of simple and reliable measurements of exposure. (1-3)-b-D-glucan is a biologically active polyglucose molecule comprising up to 60% of the cell wall of mold, and some soil bacteria and plants. (1-3)-b-D-glucan levels in samples of airborne or settled dust have been used in several studies as a surrogate measure of mold exposure (3-5).

Experimental exposures of mice, guinea pigs and adult human subjects to particulate (1-3)-b-D-glucan failed to elicit inflammatory responses in the nasal or bronchial airways (6-8). On the other hand, soluble (1-3)-b-D-glucan has been shown to enhance allergen induced airway inflammation by increasing eosinophil infiltration and specific IgE in guinea pigs and mice sensitized to ovalbumin (9,10).

A positive association between occupational exposure to (1-3)-b-D-glucan and general (tiredness, headache) and respiratory symptoms (nose and throat irritation, cough), airways inflammation and lung function has been found in the following workplaces: organic dusts in wood-work and paper industry (0.6-97.7 ng/m3) (11-13); household waste collection (9.2-52 ng/m3) (14,15), poultry (0.01-70 ng/m3) (16), composting (0.36-4.85 mg/m3) (17) and sewage treatment (4.8-40 ng/m3) (18,19). Even though indoor exposures are lower, similar symptoms have been found in adult populations exposed to indoor environments with elevated (1-3)-b-D-glucan (0.2-15.3 ng/m3) (20-23).

189

Data from adults may not be applicable for young children, as the immune system develops in the first years of life, and immune responses may be modified by exposure to microbial products (24). For example, early exposure to environmental endotoxin during infancy may have a protective effect on subsequent aeroallergen sensitization in childhood (25,26). Exposure to (1-3)-b-D-glucan may have a similar protective effect (3).

Although no previous studies have linked mold exposure to protective effects (2,27), our recent study has shown an inverse relationship between exposure to Cladosporium and allergen sensitization to any allergen (p < 0.05), as well as to aeroallergens (p < 0.05) in infants (28). There is limited data on health effects of (1-3)-b-D-glucan exposure in children. Rylander et al. (29) reported that exposure to increased airborne (1-3)-b-D- glucan (>15.3 ng/m3) positively correlated with upper airway symptoms in atopic school children (6-13 years old). Two recent European studies found increased dust-borne (1-3)- b-D-glucan concentrations to have a slight protective effect on atopic wheeze (1.2 fold decrease; p<0.10) in school children (5-13 years old) (3), and both asthma (aOR 0.70,

95% CI 0.30-1.60) and persistent wheeze (aOR 0.43, 95% CI 0.15-1.21) in children at age 1-4 (30).

As described above, existing data on health effects of (1-3)-b-D-glucan are contradictory and no studies to date have assessed (1-3)-b-D-glucan exposure on clinical outcomes in an infant birth cohort. We hypothesized that the prevalence of wheezing and allergen sensitization in a large birth cohort of high-risk infants is inversely related to exposure to high indoor concentrations of both (1-3)-b-D-glucan and endotoxin.

190

METHODS

Recruitment

15 aeroallergens. Infants with at least one parent having positive SPT [SPT(+)] were eligible for enrollment in the Cincinnati Childhood Allergy and Air Pollution study

(CCAAPS) as described previously (31-34). One purpose of CCAAPS is to examine gene-environmental interactions; thus, parental atopy was a critical criterion to obtain the most genetically at risk group. There were 1879 families with at least one parent reporting allergy symptoms. Of these, 1152 parents agreed to participate in the SPT, and

881 had at least one parent with a SPT(+). Thus, the infants of the 881 families were eligible and 758 agreed to participate. The participating children (n=758) were 20.1%

African-American, reflecting the greater Cincinnati, Ohio area distribution of 23.4%. The

University of Cincinnati Institutional Review Board approved the study.

Exposure assessment

When infants reached an average age of 8 months, families were visited at their homes to administer a detailed questionnaire to the parents on home characteristics, including observations of visible mold/water damage. The homes were classified into three groups: 1) no mold or water damage, 2) low visible mold (area<0.2m2), and 3) high visible mold (area =0.2 m2) (33). Dust samples were vacuumed from the baby’s primary activity room floor (area of at least 2 m2). The baby’s primary activity room was defined

191

as the one where the infant spends most of his/her time, and in >90% of cases, it was the living room or the family room. Thus bias due to selection of different types of rooms was minimized. For carpeted floors (>90% of homes), a sample was collected from an area of 2m2 (2 min/m2), while for smooth floors the sample was collected from the entire room as a rate of 1 min/m2 (33). Collected dust was weighed, sieved through a 355-mm sieve, and stored desiccated at -20oC until extraction.

(1-3)-b-D-glucan and Endotoxin Analysis

As the exposure to (1-3)-b-D-glucan was not included in the CCAAPS study, we have analyzed only dust samples with sufficient amount of dust (at least 25 mg), after analyses for all other allergens was completed. Thus from the initially collected 758 samples, only 574 were available for the additional analysis. The (1-3)-b-D-glucan and endotoxin activities in dust samples were determined by the endpoint chromogenic

Limulus Amebocyte lysate assay (LAL; Associates of Cape Cod, East Falmouth, MA).

Thus false-positive results were avoided. Endotoxin analysis was done as described by

Campo et al. (34). For the (1-3)-b-D-glucan, 50 mg of each dust sample was extracted in

2 ml of 0.6 M NaOH and shaken for 1 hour at -4oC. Twenty-five ml of Glucatell reagent was added to each well of serially diluted (1:100,000 and 1:1,000,000) dust extract and a control standard (1-3)-b-D-glucan (Pachyman), placed in a 96-well, flat-bottomed microplate. After 30 minutes incubation at 37oC, diazo-reagents were added to stop the

192

reaction. The optical density was recorded at 540 nm. The median coefficient of variation

(CV) was 9% for the intra-plate variability and 27% for the inter-plate variability. All samples of (1-3)-b-D-glucan were above the lower limit of detection (LOD) of the

Glucatell assay (5 pg/ml). Thirty-five of the 574 endotoxin samples were below the LOD of 0.0625 EU/ml and were recorded as LOD. LOD values were divided by the square root of two for the data analyses.

As currently it is not clear whether the concentration (mg/g) or loading (mg/m2) unit better represents the actual exposure (35), results were reported in both measures - expressed as µg/g and mg/m2 for (1-3)-ß-D-glucan and EU/mg and EU/m2 for endotoxin exposures.

Medical evaluation of infants

The medical evaluation of infants was performed during a clinic visit at the average age of 13 months (range 11-18 months, of these 95% were less than 15 months old). Sensitization to both food and aeroallergens before age one is an important risk factor for development of persistent wheeze symptoms and asthma in children born to atopic parents (36). Thus, infants were tested for allergen sensitization by SPT to a panel of food (milk, egg) and 15 common indoor and outdoor aeroallergens (7 pollen, 4 mold, cat, dog, German cockroach, house dust mite). An SPT(+) to at least one allergen was defined as a wheal =3 mm larger than the saline control after 15 minutes (32-34). We nalyzed dara on SPT positivity to at least one allergen (regardless whether food or aeroallergen), as well as SPT positivity to at least one aeroallergen. At clinic visits, the parents were interviewed regarding wheezing using questions adapted from the ISAAC

193

questionnaire for 4-5 years old (37). The following outcome variables were used: recurrent wheezing (= 2 episodes in the past 12 months), recurrent wheezing combined with SPT(+), and allergen sensitization (a positive SPT to at least one aeroallergen and/or food antigen). The reference group for recurrent wheezing consisted of infants with =1 wheezing episodes in the past 12 months regardless of the SPT status. The reference group for the recurrent wheezing with allergen sensitization consisted of infants with =1 wheezing episodes and SPT(-) status.

Data analysis

The associations between (1-3)-b-D-glucan and each health outcome were investigated for 574 infants. Histograms and quantile-quantile plots showed that (1-3)-b-

D-glucan levels were approximately log-normally distributed.

194

model, were breastfeeding duration (<1, 1-24, >25 weeks), and number of dogs and/or cats in the home.

After univariate analyses, allergen sensitization (SPT+), recurrent wheeze, and recurrent wheeze with allergen sensitization were analyzed by multiple logistic regression analyses, in which loge (1-3)-b-D-glucan, mother’s smoking (average number of cigarettes per day smoked in the last 12 months), and number of siblings in the same household, were continuously modelled. Categorically coded day-care attendance (yes, no), either parent asthma (yes, no), gender, race (African-American, non-African-

American), visible mold in home (none, low, high), lower respiratory condition (none, at least one of whooping cough, croup, viral infections, bronchitis/ bronchiolitis, flu, pneumonia), and upper respiratory condition (none, at least one of cold, ear infection, sinus infection, strep throat, tonsillitis, colored drainage) were also modelled. Covariates were chosen using a backward elimination technique.

Odds ratios and 95% confidence intervals for continuous variables were obtained to estimate the odds of each health outcome for an infant having a high (75th percentile) versus low (25th percentile) value of the variable. These percentiles represent the interquartile range of the predictor variable. Thus, the reference value was the lower endpoint of the inter-quartile range. Odds ratios and 95% confidence intervals were obtained for categorical predictor variables, where the reference category was chosen to facilitate the interpretation of results.

Graphical interpretations of wheeze prevalence (both wheeze outcomes) versus

(1-3)-b-D-glucan showed a non-linear relationship, which was modelled by dividing the range of (1-3)-b-D-glucan into 4 non-overlapping intervals, approximately equal to (1-3)-

195

Seattle, Washington, 2000).

RESULTS

Exposure and subject characteristics

The descriptive statistics for (1-3)-b-D-glucan and endotoxin levels are presented in Table 1. (1-3)-b-D-glucan levels in concentration units correlated significantly with those in loading units (Spearman correlation: r=0.69, p<0.001). The correlation between endotoxin and (1-3)-b-D-glucan was significant in loading units (r=0.51, p<0.001), and non-existent in concentration units (r=0.08, p=0.052). These values of correlation

196

between endotoxin and (1-3)-b-D-glucan are similar to what has been reported before

(30,39).

Among the infants, 114 (19.9%) had recurrent wheezing (defined as two or more episodes in the last 12 months), and 41 (7.1%) had recurrent wheezing combined with a positive SPT to at least one allergen (Table 2). There was a borderline significant difference between groups within the (1-3)-b-D-glucan and significant difference between the groups of visible mold exposures regarding the recurrent wheeze outcome.

Day-care attendance, parents without asthma, African-American race, and no siblings were associated with lower prevalence of recurrent wheeze and/or recurrent wheeze with allergen sensitization. As expected, those who experience lower and/or upper respiratory condition, also tend to wheeze more.

Infantile wheezing

We first investigated the association between the wheezing outcomes and (1-3)-b-

D-glucan exposure in concentration unit (mg/g) in quartiles (Table 3). This analysis showed strong significant inverse associations between recurrent wheezing and (1-3)-b-

D-glucan exposure. Recurrent wheezing was significantly less likely among infants with

197

very high (1-3)-b-D-glucan exposure levels (=61µg/g). In contrast, recurrent wheezing with or without allergen sensitization was positively associated with (1-3)-b-D-glucan exposure in the 1st quartile. The recurrent wheezing combined with allergen sensitization to any allergen was also significantly less likely in infants exposed to high (1-3)-b-D- glucan concentrations (=61µg/g). Similar trend was observed also for recurrent wheezing combined with sensitisation to aeroallergens only [SPT(aero)}, but was not statistically significant (data not shown).

Data were also analyzed using loading unit (mg/m2) for (1-3)-b-D-glucan exposure (Table 4). Similar to the data in concentration units, we observed an increase in wheezing outcomes when (1-3)-b-D-glucan exposure levels were low (1st and 2nd interquartile range), and a decrease when (1-3)-b-D-glucan exposure levels were high

(=19µg/m2) (3rd and 4th interquartile range) (Table 4). These associations however, were mostly non-significant. Only the inverse association between recurrent wheezing combined with allergen sensitization and exposure to high (1-3)-b-D-glucan loading

(within 4th quartile) was statistically significant.

In addition, as smoking is associated to a number of other risk factors, some of which such as diet were not evaluated in the study, a separate reporting of findings among smoking vs. non-smoking mothers was performed. After analyzing the recurrent

198

wheeze data separately for infants whose mothers smoke (n=83) and those whose mothers do not smoke (n=491), the same trends of decrease of occurrence with increase in (1-3)-b-D-glucan exposure was observed. Due to the smaller sample size, however, the significance was lost, except for the strength and significance of the inverse association for infants of smoking mothers, which became greater (within 3rd quartile aOR 0.41, 95%

CI 0.17-0.98; within 4th quartile aOR 0.01, 95% CI 0.00-0.45).

Visually, the association between continuously measured log-transformed beta- glucan and wheezing is presented in Fig.1. A curvilinear relationship was found which was approximated by fitting a restricted cubic spline (rcs) function to the wheezing data and the levels of (1-3)-b-D-glucan. Points where contiguous curves meet are called knots.

The spline functions that best fit the data had three knots. The cubic spline function was modeled allowing for 3 turning points (knots) in the curvilinear relationship between (1-

3)-b-D-glucan and log(odds of wheeze). Knots were located at the 5%, 50%, and 95% percentiles of (1-3)-b-D-glucan.

19mg/m2) compared to the minimum value. The same analysis was also done when (1-3)- b-D-glucan was equal to the maximum value (900 mg/g, 2966 m? g/m2) compared to the midpoint. As expected, the logistic regression analyses showed increased odds for recurrent wheezing at (1-3)-b-D-glucan levels below the turning point and decreased odds above the turning point (see aOR and CI in Fig 1). Similar trends were observed for recurrent wheeze with allergen sensitization (3-60 mg/g: aOR 6.05, 95% CI 0.84-43.79;

199

0.2-19 mg/m2: aOR 9.23, 95% CI 0.85-99.99; 60-900 mg/g: aOR 0.08, 95% CI 0.01-0.59;

19-2966 mg/m2: aOR 0.08, 95% CI 0.01-0.97). Therefore, the conclusions of inverse association between increase in (1-3)-b-D-glucan levels above the midpoint and wheeze outcomes (Tables 3 and 4) were confirmed.

Several studies have pointed out the potential modifying effects of co-exposure to

(1-3)-b-D-glucan and endotoxin (6,40), and how both may influence the wheeze outcome in children (2-4,25,26,29). Thus we tested for an interaction effect between (1-3)-b-D- glucan and endotoxin, and as such was not found in the wheezing outcomes, we included endotoxin as a confounder in the wheezing analyses. We performed two types of analyses

– with and without endotoxin included in the final logistic models. The results were similar, and endotoxin showed no effect neither on recurrent wheezing nor allergen sensitisation. In order to test whether it is endotoxin, rather than (1-3)-b-D-glucan responsible for the observed results, we also modelled endotoxin in quartiles, while the

As literature suggests that (1-3)-b-D-glucan may be related to visible mold (39), we also ran the models with no visible mold as a confounder, in order to check for over- adjustment. We obtained similar results as the trend of significant inverse association was preserved. However, the 95% CI became wider. As in our study we did not find any correlation between (1-3)-b-D-glucan and visible mold, and the role of (1-3)-b-D-glucan as a surrogate of mold exposure is uncertain due to various other sources of the (1-3)-b-

200

D-glucan, such as pollen and plants (42), we are reporting the full models, including visible mold as a confounder.

Allergen sensitization

Among the infants, 169 (29.4%) were sensitized to at least one allergen (food and/or aeroallergen), 25 were sensitized to food only and 79 to aeroallergens only (32).

Exposure to high (1-3)-b-D-glucan concentrations and loadings had borderline significant inverse association with allergen sensitization assessed by SPT(+) to at least one aeroallergen and/or food antigen [mg/g: within 3rd quartile, aOR=0.89, 95%CI 0.74-1.06; within 4th quartile, aOR=0.57, 95%CI 0.30-1.10 and mg/m2: within 3rd quartile, aOR=0.90, 95% CI 0.76-1.06; within 4th quartile, aOR=0.48, 95CI 0.18-1.26]. We also tested for interaction effect between to (1-3)-b-D-glucan and endotoxin, and such was found in the outcome including allergen sensitization to any allergens, but only in per square meter unit (p=0.02). Thus, we included the interaction effect in the SPT model when the exposure was per square meter unit: exposure to (1-3)-b-D-glucan within 3rd quartile, aOR=1.33, 95% CI 0.83-2.15; within 4th quartile, aOR=152.62, 95% CI 6.57-

3543.18. The latest result, however, may be due to the strong and significant correlation between (1-3)-b-D-glucan and endotoxin in this unit, and thus whether the result is

201

attributed to (1-3)-b-D-glucan or endotoxin, or both, is unclear. None of the other covariates in the multivariate analysis were significantly associated with positive SPT.

There were no significant associations between (1-3)-b-D-glucan and allergen sensitization assessed by SPT(+) to aeroallergens only [(mg/g: within 4th quartile, aOR=1.07; 95% CI=0.71-1.62) and (mg/m2: within 4th quartile, aOR=1.21; 95%CI=0.77-

1.89)].

DISCUSSION

This study demonstrated a significant inverse association between increasing exposure to high (1-3)-b-D-glucan concentrations (>60µg/g) and recurrent wheezing

(aOR=0.39) in high-risk infants. Even stronger association was found for recurrent wheezing combined with allergen sensitization (aOR=0.18). Although others have reported a weak inverse relationship with (1-3)-b-D-glucan and atopic wheeze in children aged 1-4 and 5-15 years (3,30), this is the first study to demonstrate a statistically significant relationship between high (1-3)-b-D-glucan exposure and reduced wheezing in infants.

We did not find significant differences in (1-3)-b-D-glucan concentrations or loadings between the homes in three visible mold categories although increasing trend was seen in (1-3)-b-D-glucan loadings. This finding is in line with the findings of

Gehring et al. (39) who only found difference in (1-3)-b-D-glucan loading, and not in concentration, between homes in two visible mold categories. This indicates that the variance in levels per square meter was largely determined by the amount of dust

202

sampled. Thus after transformation from concentration to loading unit, there was a decrease in the between samples variation.

Previous studies show that (1-3)-b-D-glucan concentrations do not consistently correlate with total culturable mold spore counts (39,43,44), and more likely reflect exposure from multiple environmental sources of (1-3)-b-D-glucan, including mold, pollen, plants, and their fragments (42,45). Furthermore, (1-3)-b-D-glucan content varies between mold species (45), and this may lead to variance of health outcomes by fungal genera (28). For example, in this cohort, we have also found inverse association between the concentration of airborne Cladosporium and SPT(+) to any allergen (p < 0.05), and

Cladosporium and SPT(+) to aeroallergens (p < 0.05), but positive associations between

Penicillium/Aspergillus and SPT(+) to any allergen (p < 0.01) and between Alternaria and SPT(+) to any allergen (p < 0.01) (28). Molds contain (1-3)-b-D-glucan but also number of other agents such as sugars and enzymes. Apart from the fungi themselves mold growth is often associated with growth of other microbes. Thus, no conclusions concerning causal relationships between (1-3)-b-D-glucan and mold can be made.

However, it seems (1-3)-b-D-glucan may be an independent measure of biologically active exposure. This may explain why visible mold exposure, unlike (1-3)-b-D-glucan, was a risk factor for recurrent wheezing in this cohort [as reported in detail by Cho et al.

(33)].

Our study demonstrated that low levels of (1-3)-b-D-glucan exposure are associated with increase in the prevalence of recurrent wheeze, while the opposite was observed for the exposure to high levels of (1-3)-b-D-glucan. Similar trend was also found for the health outcome including wheeze combined with allergen sensitization.

203

Children are a special group of interest as their immune system is evolving and thus may be more susceptible to environmental exposure factors (24). In fact, research has shown

(46) that the effective priming of aeroallergen-specific memory T-cells is initiated during infancy and is consolidated by the end of preschool years in relation to the Th1-Th2 balance. An interesting observation that Th2 (allergic) priming is preferentially favored by low-dose antigen exposure, whereas higher doses favor Th1 priming, was made by

Constant et al (1995) and Rogers et al (1999) (47,48). Moreover, epidemiological studies on exposure to indoor allergens have revealed a biphasic pattern in which sensitization risk increases with exposure levels until a plateau is reached, above which risk decreases with further increase in exposure (49-51). This same pattern was demonstrated in our study.

To date, knowledge on health effects of (1-3)-b-D-glucan in children is limited.

Rylander et al. (29) reported that both non-atopic and atopic wheezing in school children

(age 6-13 years) were positively correlated with airborne (1-3)-b-D-glucan concentrations, analyzed by the LAL assay. Douwes et al. (35) also showed that increased levels of dust-borne (1-3)-b-D-glucan were positively related to variability in peak expiratory flow in non-atopic and atopic children 7-11 years of age. In contrast,

Schram-Bijkerk et al. (3) and Douwes et al. (30) reported that dust-borne (1-3)-b-D- glucan concentrations were inversely associated with atopic wheeze in 5-13 year old children (OR=0.83; 95%CI=0.71-1.01) and asthma in 1-4 years old children (OR=0.61;

95%CI=0.28-1.33). All these studies, similarly to ours, showed that the health outcome was more significantly associated with exposure in the subgroup of atopic children.

Interestingly, in the latter two studies (3,30) inclusion of both (1-3)-b-D-glucan and

204

endotoxin in the models lead to loss of significance of either the effect of endotoxin (3) or

(1-3)-b-D-glucan (30). These two studies used the inhibition immuno assay (EIA) and had a large number of samples below the detection limit for (1-3)-b-D-glucan, which could have caused an exposure misclassification. This was a major limitation in these studies, as pointed out by the authors themselves. The low limit of detection and the consequent loss of samples is a major drawback of the EIA assay . So far, there are no published studies comparing the EIA and the LAL assay {used in the current study and by Rylander et al. (29)}. Thus, comparison of results obtained by these two assays is based on uncertain grounds.

In addition, these two previous studies reported a strong and significant correlation between endotoxin and (1-3)-b-D-glucan, thus it is uncertain whether the observed effects were driven by endotoxin, (1-3)-b-D-glucan or both. We did not find any association between the investigated health outcomes and endotoxin exposure. This finding agrees with the results of Schram-Bijkerk et al. (3) who showed that endotoxin loses significance after adjusting for (1-3)-b-D-glucan, and contradicts the study results reported by Douwes et al. (30). A separate analysis focused on the endotoxin exposure in this cohort showed that endotoxin was not independently associated with wheezing or allergen sensitization, but high endotoxin exposure in the presence of multiple dogs was associated with reduced wheezing (34). Furthermore, the lack of correlation between (1-

3)-b-D-glucan and endotoxin concentrations in our database strengthens the finding that the effect seen for (1-3)-b-D-glucan exposure is not confounded by the endotoxin exposure.

205

Our findings that homes with high (1-3)-b-D-glucan have less recurrent wheezing in sensitized infants, and that (1-3)-b-D-glucan plays more important role than endotoxin in early life wheezing is consistent with the study by Schram-Bijkerk et al. (3) and

Douwes et al. (30). A possible explanation for the stronger inverse association found in our study may be attributable to the younger age and selection of high-risk cohort (born to atopic parents). In this cohort, we have also found an inverse association between the concentration of airborne Cladosporium and allergen sensitization (28). Indeed this data is somewhat consistent with the hygiene hypothesis, which postulates that exposure to microbial products (such as endotoxin) early in life favors modification of Th2 directed immune responses (25). Studies on exposure to high endotoxin and increased or decreased frequency of wheezing during infancy have shown conflicting results

(25,26,52). Possible explanation for this controversy may be that these studies did not concomitantly assess indoor (1-3)-b-D-glucan levels. However, several studies have revealed a strong adjuvant activity of (1-3)-b-D-glucan on the systemic allergic immune response in animal models (9,10,53). Therefore it must be emphasized that the immunologic impact of (1-3)-b-D-glucan exposure on Th2 directed atopic disorders remains uncertain and requires further investigation.

The above findings were further supported by the inverse trend between (1-3)-b-

D-glucan exposure and allergen sensitization. This and the stronger inverse association between (1-3)-b-D-glucan and wheezing in allergen sensitized infants compared to wheezing in all infants, even stronger after stratification by sensitization, suggests that (1-

3)-b-D-glucan could modify allergic respiratory responses in infants. It must be emphasized that it is uncertain if this early observed effect will impact the risk of later

206

development of childhood asthma among this cohort of sensitized infants with recurrent wheezing as health outcomes associated with (1-3)-b-D-glucan exposure determined in this infant population may be transient.

To our knowledge, there is only one previous study reporting the associations between (1-3)-b-D-glucan concentration versus loading units with health outcomes.

Douwes et al. showed stronger association between peak flow variability in children and dust (1-3)-b-D-glucan in the loading unit than in the concentration unit (35). We found similar trends between the health outcomes and (1-3)-b-D-glucan exposures in the two units, but the associations were stronger for the concentration unit. These differences may be due to the different health outcome or different (1-3)-b-D-glucan analysis methods used: Douwes et al. analyzed (1-3)-b-D-glucan content by the EIA, while we used the

LAL.

We collected dust samples only from the baby’s primary activity room because the main living room allergen levels for dust allergens have shown to be higher than that in the bedding in the infants’ homes (53). Furthermore, despite the fact that infants spend most of their time in bed, a recent European study (30) has shown association only between living area exposure and asthma in 4 years old children, but none with mattress exposure. This was attributed to low allergen levels in infants’ mattresses as they were newly purchased (use of less than 3 months). In addition, the collection of dust by trained technicians using a standardized protocol and the same type of vacuum cleaner is strength of the present study as it decreases the collection bias. Further, home exposure assessment was conducted almost concurrently with health outcome assessment.

207

A potential limitation of the study generalizability is that the cohort consists only of infants born to atopic parents. However, the Third National Health and Nutrition

Examination Survey (NHANES III, 54), 1988-94, designed to obtain nationally representative information on the health of the population of the United States, showed that 54.3% of the U.S. population was sensitized to one or more allergens. This finding suggests that the results observed in the current study can be widely generalized to over

50% of the U.S. population. Research has shown that exposure to maternal smoking and aeroallergens (55) impact the development of allergen sensitization as early as in pregnancy. Thus, the absence of antenatal exposure data may be another study limitation.

In addition, the low number of recurrent wheezers with positive SPT also limits the power of the study. Unfortunately, as the CCAAPS study did not initially include the analysis for (1-3)-b-D-glucans, we were able to analyze only samples with sufficient dust amount left after analyses for all other allergens were completed. Although this introduces a selection bias, and may limit the generalizability of the results, we were still able to analyze 76% of the samples. In addition, the amount of dust and wheeze outcomes did not correlate (p=0.32). Therefore, we do not expect dust amount to bias the results.

In conclusion, we found that the concentration of (1-3)-b-D-glucan is a measure of biological exposure that is independent from observed visible mold. A significant inverse association was found between high (1-3)-b-D-glucan levels and recurrent wheezing.

This association was even stronger in a subgroup of allergen-sensitized infants. It seems that exposure to high (1-3)-b-D-glucan levels (=61µg/g, =19µg/m2) may be conducive to reduced wheezing in infants at high risk for developing asthma (as those born to atopic parents). Long-term follow-up of this cohort will help determine how early (1-3)-b-D-

208

glucan exposure affects the development of phenotypes such as atopy, allergic rhinitis and asthma in later childhood.

ACKNOWLEDGEMENTS

This study was supported by National Institute of Environmental Health Sciences

(NIEHS) grant ES11170 and the National Institute for Occupational Safety and Health

(NIOSH) Training Program of the University of Cincinnati Education and Research

Center Grant T42/CCT510420. We would like to thank our participating families, our recruitment team, Sherry Stanford and Stephanie Maier who performed infant assessment.

209

REFERENCES

1. Gent JF, Ren P, Belanger K, Triche E, Bracken MB, Holford TR et al. Levels of

household mold associated with respiratory symptoms in the first year of life in a

cohort at risk for asthma. Environ Health Perspect 2002;110:781-6.

2. Belanger K, Beckett W, Triche E, Bracken MB, Holford T, Ren P et al. Symptoms of

wheeze and persistent cough in the first year of life: associations with indoor

allergens, air contaminants, and maternal history of asthma. Am J Epidemiol

2003;158:195-202.

3. Schram-Bijkerk D, Doekes G, Douwes J, Boeve M, Riedler J, Ublagger E et al.

Bacterial and fungal agents in house dust and wheeze in children: the PARSIFAL

study. Clin Exp Allergy 2005;35:1272-8.

4. Schram D, Doekes G, Boeve M, Douwes J, Riedler J, Ublagger E et al. Bacterial and

fungal components in house dust of farm children, Rudolf Steiner school children

and reference children-the PARSIFAL Study. Allergy 2005;60:611-8.

5. Fogelmark B, Thorn J, Rylander R. Inhalation of (1-->3)-beta-D-glucan causes airway

eosinophilia. Mediators Inflamm 2001;10:13-9.

6. Fogelmark B, Sjostrand M, Rylander R. Pulmonary inflammation induced by repeated

inhalations of beta(1,3)-D-glucan and endotoxin. Int J Exp Pathol 1994;75:85-90.

7. Korpi A, Kasanen JP, Kosma VM, Rylander R, Pasanen AL. Slight respiratory

irritation but not inflammation in mice exposed to (1-->3)-beta-D-glucan aerosols.

Mediators Inflamm 2003;12:139-46.

8. Rylander R. Airway responsiveness and chest symptoms after inhalation of endotoxin

or 1,3, beta-d-glucan. Indoor Build Environ 1996;154:106-111

210

9. Rylander R, Holt PG. (1-->3)-beta-D-glucan and endotoxin modulate immune

response to inhaled allergen. Mediators Inflamm 1998;7:105-10

10. Ormstad H, Groeng EC, Lovik M, Hetland G. The fungal cell wall component beta-

1,3-glucan has an adjuvant effect on the allergic response to ovalbumin in mice. J

Toxicol Environ Health 2000;61:55-67.

11. Alwis K, Mandryk J, Hocking A. Exposure to biohazards in wood dust: bacteria,

fungi, endotoxins, and (1-3)-b-D-glucans. Appl Occup Environ Hyg 1999;14:598-

608.

12. Mandryk J, Alwis K, Hocking A. Effects of personal exposure on pulmonary function

and work-related symptoms among sawmill workers. Ann Occup Hyg

2000;44:281-289.

13. Rylander R, Thorn J, Attefors R. Airways inflammation among workers in a paper

industry. Eur Respir J 1999;13:1151-1157.

14. Thorn J. Seasonal variations in exposure to microbial cell wall components among

household waste collectors. Ann Occup Hyg 2001;45:153-156.

15. Heldal K, Halstensen A, Thorn J, Eduard W, Halstensen T. Airway inflammation in waste

handlers exposed to bioaerosols assessed by induced sputum. Eur Respir J

2003;21:641-5.

16. Rylander R, Carvalheiro M. Airways inflammation among workers in poultry houses.

Int Arch Occup Environ Health 2006;79:487-90.

211

17. Douwes J, Wouters I, Dubbeld H, van Zwieten L, Steerenberg P, Doekes G et al.

Upper airway inflammation assessed by nasal lavage in compost workers: A

relation with bio-aerosol exposure. Am J Ind Med 2000;37:459-68.

18. Rylander R. Health effects among workers in sewage treatment plants. Occup

Environ Med 1999;56:354-357.

19. Gladding T, Thorn J, Stott D. Organic dust exposure and work-related effects among

recycling workers. Am J Ind Med 2003;43:584-91.

20. Rylander R. Airborne (1®3)-b-glucan and airway disease in a day-care center before

and after renovation. Arch Environ Health 1997;52;281-285.

21. Wan, G. -H., Li C.S. Indoor endotoxin and glucan in association with airway

inflammation and systemic symptoms. Arch Environ Health 1999;54:172-179.

22. Wouters I, Hilhorst S, Kleppe P, Doekes G, Douwes J, Peretz C et al. Upper airway

inflammation and respiratory symptoms in domestic waste collectors. Occup

Environ Med 2002;59:106-112.

23. Thorn J, Rylander R. Airways inflammation and glucan in a row house area. Am J

Respir Crit Care Med 1998;157:1798-1803.

24. Holt P, Upham J, Sly P. Contemporaneous maturation of immunologic and

respiratory functions during early childhood: implications for development of

asthma prevention strategies. Review. J Allergy Clin Immunol 2005;116:16-24.

25. von Mutius E, Braun-Fahrlander C, Schierl R, Riedler J, Ehlermann S, Maisch S et al.

Exposure to endotoxin or other bacterial components might protect against the

development of atopy. Clin Exp Allergy 2000;30:1230-4.

212

26. Gereda JE, Leung DY, Thatayatikom A, Streib JE, Price MR, Klinnert MD et al.

Relation between house-dust endotoxin exposure, type 1 T-cell development, and

allergen sensitization in infants at high risk of asthma. Lancet 2000;355:1680-3.

27. Douwes J, Pearce N. Invited commentary: is indoor mold exposure a risk factor for

asthma? Am J Epidemiol 2003;158:203-6.

28. Osborne M, Reponen T, Adhikari A, Cho SH, Grinshpun S, Levin L et al. Specific

fungal exposures, allergic sensitisation, and rhinitis in infants. Pediatr Allergy

Immunol 2006;17:450-7.

29. Rylander R, Norrhall M, Engdahl U, Tunsater A, Holt P. Airways inflammation,

atopy, and (1--> 3)-beta-D-glucan exposures in two schools. Am J Respir Crit

Care Med 1998;158:1685-7.

30. Douwes J, van Strien R, Doekes G, Smit J, Kerkhof M, Gerritsen J et al. Does early

indoor microbial exposure reduce the risk of asthma? The Prevention and

Incidence of Asthma and Mite Allergy birth cohort study.

J Allergy Clin Immunol 2006;117:1067-73.

31. Ryan P, LeMasters G, Biagini J, Bernstein D, Grinshpun S, Shukla R et al. Is it traffic

type, volume, or distance? Wheezing in infants living near truck and bus traffic. J

Allergy Clin Immunol 2005;116:279-84.

32. LeMasters G, Wilson K, Levin L, Biagini J, Ryan P, Lockey J et al. High prevalence

of aeroallergen sensitization among infants of atopic parents. J Pediatr

2006;149:505-511.

213

33. Cho S-H, Reponen T, LeMasters G, Levin L, Huang J, Meklin T et al. Mold damage

in homes and wheezing in infants. Ann Allergy, Asthma Immunol 2006;97:539-

545.

34. Campo P, Kalra HK, Levin L, Reponen T, Olds R, Lummus Z et al. Influence of dog

ownership and high endotoxin on wheezing and atopy during infancy. J Allergy

Clin Immunol 2006 (in press).

35. Douwes J, Zuidhof A, Doekes G, van der Zee SC, Wouters I, Boezen M et al. (1--

>3)-beta-D-glucan and endotoxin in house dust and peak flow variability in

children.

Am J Respir Crit Care Med 2000;162:1348-54.

36. Rhodes H, Thomas P, Sporik R, Holgate S, Cogswell J. A birth cohort study of

subjects at risk of atopy: twenty-two-year follow-up of wheeze and atopic status.

Am J Respir Crit Care Med 2002;165:176-80.

37. Beasley R. The International Study of Asthma and Allergies in Childhood (ISAAC)

Steering Committee pp. 1225-1232 Worldwide variation in prevalence of

symptoms of asthma, allergic rhinoconjunctivitis, and atopic eczema: ISAAC.

Lancet 1998;351:1225-32.

38. Gold D, Burge H, Carey V, Milton D, Platts-Mills T, Weiss S. Predictors of repeated

wheeze in the first year of life: the relative roles of cockroach, birth weight, acute

lower respiratory illness, and maternal smoking. Am J Respir Crit Care Med

1999;160:227-36.

39. Gehring U, Douwes J, Doekes G, Koch A, Bischof W, Fahlbusch B et al. INGA

Study Group. Indoor Factors and Genetics in Asthma. Beta (1-->3)-glucan in

214

house dust of German homes: housing characteristics, occupant behavior, and

relations with endotoxins, allergens, and molds. Environ Health Perspect

2001;109:139-44.

40. Young S, Robinson V, Barger M, Zeidler P, Porter D, Frazer D et al. Modified

endotoxin responses in rats pretreated with 1-->3-beta-glucan (zymosan A).

Toxicol Appl Pharmacol 2002;178:172-9.

41. Rylander R, Lin RH. (1-->3)-beta-D-glucan - relationship to indoor air-related

symptoms, allergy and asthma. Review. Toxicology 2000;152:47-52.

42. Rylander R, Fogelmark B, McWilliam A, Currie A. (1-->3)-beta-D-glucan may

contribute to pollen sensitivity. Clin Exp Immunol 1999;115:383-4.

43. Douwes J, Doekes G, Heinrich J, Koch A, Bischof W, Brunekreef B. Endotoxin and

ß(1-3)-Glucan in house dust and the relation with home characteristics: a pilot

study in 25 German houses. Indoor Air 1998;8:255-263

44. Wan GH, Li CS. Indoor endotoxin and glucan in association with airway

inflammation and systemic symptoms. Arch Environ Health 1999;54:172-9.

45. Foto M, Plett J, Berghout J, Miller JD. Modification of the Limulus amebocyte lysate

assay for the analysis of glucan in indoor environments. Anal Bioanal Chem

2004;379:156-62.

46. Holt P, Macaubas C, Stumbles PA, Sly P. The role of allergy in the development of

asthma. Nature 1999;402:B12-7.

47. Constant S, Pfeiffer C, Woodard A, Pasqualini T, Bottomly K. Extent of T cell

receptor ligation can determine the functional differentiation of naive CD4+ T

215

cells.

J Exp Med 1995;182:1591-6.

48. Rogers P, Croft M. Peptide dose, affinity, and time of differentiation can contribute to

the Th1/Th2 cytokine balance. J Immunol 1999;163:1205-13.

49. Platts-Mills T, Vaughan JW, Blumenthal K, Woodfolk J, Sporik R. Decreased

prevalence of asthma among children with high exposure to cat allergen:

relevance of the modified Th2 response. Mediators Inflamm 2001;10:288-91.

50. Cullinan P, MacNeill S, Harris JM, Moffat S, White C, Mills P et al.. Early allergen

exposure, skin prick responses, and atopic wheeze at age 5 in English children: a

cohort study. Thorax 2004;59:855-61.

51. Holt P, Upham J, Sly P. Contemporaneous maturation of immunologic and

respiratory functions during early childhood: implications for development of

asthma prevention strategies. J Allergy Clin Immunol 2005;116:16-24.

52. Park JH, Gold DR, Spiegelman DL, Burge HA, Milton DK. House dust endotoxin

and wheeze in the first year of life. Am J Respir Crit Care Med 2001;163:322-8.

53. Leaderer B, Belanger K, Triche E, Holford T, Gold D, Kim Y, Jankun T et al. Dust

mite, cockroach, cat, and dog allergen concentrations in homes of asthmatic

children in the northeastern United States: impact of socioeconomic factors and

population density. Environ Health Perspect 2002;110:419-25.

54. Arbes S, Gergen P, Elliott L, Zeldin D. Prevalences of positive skin test responses to

10 common allergens in the US population: results from the third National Health

and Nutrition Examination Survey. J Allergy Clin Immunol 2005;116:377-83.

216

55. Upham J, Holt P. Environment and development of atopy. Curr Opin Allergy Clin

Immunol 2005;5:167-72.

217

FIGURE TITLES Figure 1. Smoothed plot of the adjusted prevalence rates of recurrent wheezing in relation to the log-transformed (1-3)-b-D-glucan concentration (solid lines). Dotted lines represent ± 1 standard error (SE).

218

aOR 3.90, aOR=0.32, aOR 1.41, aOR=0.80, 1 SE 0.30 1 SE

+ 95%CI 1.26-12.09 95%CI=0.11-0.97 95%CI 0.41-4.87 95%CI=0.21-3.08

0.25

0.20 0.20

5 0.15

0.1

0.10

0.10 0.05

0.05 0.00 Adjusted Prevalence Rates of Recurrent Wheeze Adjusted Prevalence Rates of Recurrent Wheeze 1.0 7.4 54.6 403.4 0.1 1.0 7.4 54.6 403.4 2980.9

µg/g µg/m2

(1-3)-ß-D-Glucan Concentration

Figure 1.

219

Table 1. Geometric mean (GM), geometric standard deviation (GSD) and interquartile (IQ) range of (1-3)-ß-D-glucan and endotoxin concentration (mg/g, EU/mg) and loading (mg/m2, EU/m2), measured in homes of 574 infants. n=574 GM GSD IQ*

220

Table 2. Characteristics of predictor variables and prevalence and percent of infants (n, % of total) reporting health outcome* by levels of (1-3)-b-D-glucan, endotoxin, demographic characteristics. Numbers in brackets for each predictor variable represents the number of infants that fall in each class of that predictor variable. Cell entries in the health outcome columns are number of subjects reporting outcome (% of column total).

221

222

Table 3. Adjusted odds ratios (aOR) and 95% confidence intervals (95% CI) for recurrent wheeze and recurrent wheeze with allergen sensitization in relation to upper vs. lower endpoints of continuously measured (1-3)-b-D-glucan quartiles (mg/g) (reference category is the value of the lower endpoint of each quartile). Recurrent Recurrent wheeze Recurrent wheeze with wheeze* with SPT(+) vs. no SPT(+) vs. no wheeze, wheeze, SPT(-)** SPT(+)*** aOR (95% CI) aOR (95% CI) aOR (95% CI) (1-3)-b-D-glucan quartile endpoints (mg/g) I.3-22 3.04 (1.25-7.38) 4.89 (1.02-23.57) 160.51 (4.85-5,311.00) II.23-60 1.29 (0.99-1.67) 1.23 (0.79-1.92) 2.54 (0.97-6.62) III.61–133 0.82 (0.65-1.05) 0.59 (0.38-0.92) 0.17 (0.05-0.57) IV.134-900 0.39 (0.16-0.93) 0.13 (0.03-0.61) 0.00 (0.00-0.07) Endotoxin fixed at continuous 0.99 (0.71-1.37) 1.17 (0.69-1.98) 1.60 (0.58-4.41) level from 39.19 to 171.17 (mg/g) Visible Mold (low vs. none) 1.18 (0.73-1.91) 1.29 (0.57-2.90) 2.64 (0.89-7.86) Visible Mold (high vs. none) 4.44 (1.63-12.05) 9.51 (2.34-38.63) 42.47 (4.70-384.14) Mother’s smoking (³20 vs. 0 5.16 (2.33-11.44) 10.17 (2.58-40.09) 10.17 (2.58-40.09) cigarettes/day) Parental asthma 1.87 (1.17-3.00) 2.22 (1.05-4.71) 2.09 (0.76-5.77) Race (Afro-American vs. other) 2.08 (1.15-3.73) 3.93 (1.57-9.84) 10.04 (2.45-41.14) Siblings (1vs. 0) 1.38 (0.77-2.47) 1.84 (0.66-5.09) 8.87 (1.85-42.51) Siblings (=2 vs. 0) 1.96 (1.08-3.57) 2.46 (0.87-6.93) 7.83 (1.60-38.38) Lower respiratory condition† 3.98 (2.47-6.41) 4.63 (2.05-10.46) 9.93 (3.06-32.16) Upper respiratory condition‡ 2.15 (1.26-3.67) 2.75 (1.08-7.04) 4.47 (1.24-16.07) * Recurrent wheeze = 2 or more wheezing episodes in the last 12 months (n=114). Infants that had one or no wheezing episodes in the last 12 months were used as the comparison group (n=460).

223

224

Table 4. Adjusted odds ratios (aOR) and 95% confidence intervals (95% CI) for recurrent wheeze and recurrent wheeze with allergen sensitization in relation to upper vs. lower endpoints of continuously measured (1-3)-b-D-glucan quartiles (mg/m2) (reference category is the value of the lower endpoint of each quartile). Recurrent wheeze* Recurrent wheeze Recurrent wheeze with with SPT(+) vs. no SPT(+) vs. no wheeze, wheeze, SPT(-)** SPT(+)*** aOR (95% CI) aOR (95% CI) aOR (95% CI) (1-3)-b-D-glucan quartile endpoints (mg/m2) I.0.2-6 1.62 (0.54-4.82) 7.51 (0.96-58.70) 147.81 (1.38-15814.02) II.7-18 1.08(0.86-1.36) 1.26 (0.82-1.91) 2.16 (0.98-4.77) III.19-58 0.95 (0.76-1.19) 0.64 (0.42-0.97) 0.60 (0.22-1.64) IV.59-2966 0.65 (0.21-2.06) 0.05 (0.00-0.51) 0.01 (0.00-4.75) Endotoxin fixed at continuous 1.07 (0.77-1.49) 1.45 (0.89-2.37) 1.40 (0.58-3.42) level from 38 to 165 (mg/m2) Visible Mold (low vs. none) 1.18 (0.73-1.90) 1.33 (0.60-2.98) 2.16 (0.77-6.01) Visible Mold (high vs. none) 4.25 (1.56-11.58) 12.56 (2.96-53.35) 32.16 (4.03-256.60) Mother’s smoking (³20 vs. 0 5.39 (2.43-11.92) 10.44 (2.65-41.14) 10.44 (2.65-41.14) cigarettes/day) Parental asthma 1.85 (1.16-2.95) 2.35 (1.10-5.04) 1.99 (0.74-5.37) Race (Afro-American vs. other) 1.94 (1.08-3.48) 3.96 (1.60-9.82) 4.47 (1.29-15.49) Siblings (1vs. 0) 1.46 (0.82-2.61) 1.84 (0.66-5.16) 5.74 (1.48-22.23) Siblings (=2 vs. 0) 2.11 (1.16-3.84) 2.51 (0.88-7.15) 6.55 (1.54-27.94) Lower respiratory condition† 4.10 (2.55-6.58) 5.33 (2.34-12.17) 10.75 (3.37-34.22) Upper respiratory condition‡ 2.11 (1.24-3.57) 2.42 (0.96-6.12) 3.69 (1.13-12.01) See footnotes for Table 3.

225

APPENDIX B: LIST OF OTHER PUBLICATIONS (NOT INCLUDED IN THE

PhD DISSERTATION) AUTHORED/CO-AUTHORED BY Ms. YULIA

IOSSIFOVA DURING HER GRADUATE STUDY IN THE UNIVERSITY OF

CINCINNATI

226

B1: Peer-Reviewed Publications:

1. Vesper S, McKinstry C, Haugland R, Iossifova Y, LeMasters G, Levin L,

Predictor of Childhood Respiratory Illness. J Expo Anal Environ Epi 2006 (in press)

2. Lee T-K, Grinshpun S, Kim K-Y, Iossifova Y, Adhikari A, Reponen T.

Relationship between indoor and outdoor airborne fungal spores, pollen and (1-3)-ß-D- glucan in homes without visible mold growth. Aerobiologia 22:227-236, 2006

3. Seo S-C, Cho S-H, Grinshpun S, Iossifova Y, Schmechel D, Rao C, Reponen

T. A new field-compatible method for the collection of fungal fragments. (submitted)

4. Tuncel S, Iossifova Y, Ravelo E, Daraiseh N, Salem S. Effectiveness of controlled workplace interventions in reducing lower back disorders. Theoretical Issues in Ergonomic Science 7(3):211-225, 2006

227

B2: Conference Proceedings and Abstracts:

1. Iossifova Y, Reponen T, Levin L, Zeigler H, Bernstein D, Kalra H, Khurana-

Hershey G, LeMasters G. House dust (1-3)-b-D-glucan and wheeze in infants. American

Thoracic Society International Conference, San Diego, May 2006

2. Reponen T, Seo S-C, Iossifova Y, Adhikari A, Grinshpun S. New-field compatible method for collection and analysis of (1-3)-b-D-glucan in fungal fragments.

International Aerosol Conference, St. Paul, Minnesota, September 2006

3. Iossifova Y, Reponen T, Levin L, Zeigler H, Bernstein D, Kalra H, Khurana-

Hershey G, LeMasters G. House dust (1-3)-b-D-glucan and wheeze in infants. University of Cincinnati Graduate Poster Forum, Cincinnati, March 2006

4. Iossifova Y, Reponen T, Bernstein D, Kalra H, Hershey GK, Masino A. (1-3)- b-D-glucan as a surrogate for mold exposure. American Industrial Hygiene Association

Conference, Anaheim, May 2005

5. Seo S-C, Reponen T, Cho S-H, Iossifova Y, Grinshpun S. A new field- compatible method for the collection of fungal fragments. American Industrial Hygiene

Association Conference, Anaheim May 2005 (Awarded best poster of session)

6. Iossifova Y, Reponen T, Bernstein D, Kalra H, Khurana-Hershey G, Masino A.

(1-3)-b-D-glucan as a surrogate for mold exposure. University of Cincinnati Graduate

Poster Forum, Cincinnati, March 2005

228

C. LAL PROTOCOLS FOR (1-3)-b-D-GLUCAN ANALYSIS

229

C1: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – DUST SAMPLES –

ENDPOINT ASSAY

Limulus Amebocyte Quantitative Assay for (1-3)-ß-D-Glucan – DUST samples

(ENDPOINT ASSAY)

GLUCATELL® - ASSOCIATES OF CAPE COD INC.

(1-3)-ß-D-Glucan Detection Reagent Kit- ENDPOINT ASSAY

GENERAL INFORMATION

Wear gloves at all times. All the material should be pyrogen-free (tips, reagents, diluents, etc...)

SAMPLE PROCESSING

If there is enough dust:

Use the Mettler AB 104-S top loading digital electronic balance, weigh 50 mg of dust directly into pyrogen-free 2 ml tubes (ISC BioExpress, cat # C-3229-1). Add 2 ml of 0.6

M NaOH (Sigma Cat #7732-18-5) to make a 25 mg/ml suspension.

If there is not enough dust:

Weigh 10 mg or less using the Mettler H20T, and 4”x4” weigh paper. Transfer the dust

(x mg) to a pyrogen-free 2 ml tube, and then add the amount of 0.6 M NaOH (y ml) required to make a 25 mg/ml suspension. Use the formula: x of mg dust = y ml of 0.6 N NaOH (for final concentration of 25 mg/ml)

230

25 mg/ml

Place the tubes in the end-over-end tube shaker & extract for 1 hour in the cold room

(40C) to unwind the triple-helix structure of the glucan and make it water-soluble.

Centrifuge for 1 minute at 5000xg (7000 rpm if using the IEC Centra 4 centrifuge, equipped with micro-centrifuge rotor, or position 5 if using Fisher micro-centrifuge).

Save the supernatant by pouring into a second set of 2 ml pyrogen-free tubes. Store at -

20oC if not analyzed within 1 week.

PREPARE DILUTIONS

(1-3)-ß-D-Glucan Standard (Pachyman)

Lyophilized (1-3)-ß-D-Glucan standard. Reconstitute the lyophilized standard with

Reagent Water provided in the kit, to make 100pg/ml solution. The amount of water is given on the vial. Vortex for 1 minute prior to make further dilutions.

Prepare dilutions of the standard and the samples in reagent water according to the table.

Use pyrogen free tubes or microplates for preparing dilutions. Mix well.

Glucan standard

ml standard (100 pg/ml) Final Concentration (pg/ml) ml Water 80 100 400 60 200 300 40 300 200 20 400 100 10 450 50 5 950 50 0 100 0

231

SAMPLE DILUTIONS

Make dilutions in a sterile microtiter plate (round-bottom or V-bottom wells) using Cape

Cod Reagent water as diluent.

Sample Numbers

#1 #2 #3 #4 #5 #6 #7 #8 #9 #10 #11 #12 A 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 B 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 C 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 D 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 E 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 F G H

Dilutions of Dust Extracts ml Row Final Dilution ml sample (in 0.6NnaOH) Water A 1:10 90 10 (from NaOH sample B 1:100 90 10 (from 1:10 diltn) C 1:1000 90 10 (from 1:100 diltn) D 1:10,000 90 10 (from 1:1000 diltn) E 1:100,000 90 10 (from 1:10,000 diltn)

Glucatell Reagent

Reconstitute with 2.8 ml of Pyrosol® Reconstitution Buffer, and swirl gently (do not

vortex). Use within 10 minutes or place at 2-8º C and use within 2 hours. Can also be

frozen at -20o C for 20 days.

232

Diazo Reagents (Substrate reagents for the Glucatell reaction with glucan –

ENDPOINT ASSAY ONLY)

The substrate reagents that will be added to the plate are prepared as follows:

Vial 1: Add the contents of Vial 1a (1M HCl solution) to Vial 1 (sodium nitrite powder)

Vial 2: Add 4 ml of Reagent Water to Vial 2 (ammonium sulfamate powder)

Vial 3: Add the contents of Vial 3a (N-methyl-Pyrrolidinone solution) to Vial 3 (NEDA

powder)

These solutions should be used the same day (i.e., within 12 hours).

ASSAY PROCEDURE- ENDPOINT ASSAY

Samples should be run using 1:10,000 and 1:100,000 dilutions. Standard curve should be

run in duplicate. Place the Assay plate in the Pyroblock, pre-warmed at 37ºC. Use the

plate included in the kit. Add 25 µl of diluent (reagent water); samples diluted 1:10,000

(10-4) and 1:100,000 (10-5), and standard at concentrations of 5, 10, 20, 40, 60, and 80

pg/ml to wells of the Assay plate as shown below.

Glucan Assay Plate (No. stands for Dust Sample Number)

Row 1 2 3 4 5 6 7 8 9 10 11 12 No. Blanks standa 1 2 3 4 5 rds A Diluent 5 10 20 40 60 80 10-4 10-4 10-4 10-4 10-4 pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml B Diluent 5 10 20 40 60 80 10-5 10-5 10-5 10-5 10-5 pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml No. 6 7 8 9 10 11 12 13 14 15 16 17 C 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 D 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 No. 18 19 20 21 22 23 24 25 26 27 28 29

233

E 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 F 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 No. 30 31 32 33 34 35 36 37 38 39 40 41 G 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 H 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 10-5 Add 25 µl/well of Glucatell reagent using a multi-channel pipette. Cover the plate with the lid,

and tap the plate carefully. Incubate in the Pyroblock for exactly 30 minutes.

Check the plate reader for accuracy of the Protocol and Template for reading the plate. The

protocol face page should show "statistical" method, no prompt for plate ID, show Absorbance

readings and means. The detailed protocol should show "endpoint" method, read at 540 nm,

blanking method –full, blank pattern A1 and B1. The Template map should show set numbers

and names for the wells containing blanks and standards in duplicate (1 = blanks, 2 = 5 pg/ml

standards, etc.). The incubation period is a good time to prepare the Diazo substrate reagents as

shown above.

Stop the reaction with 25 µl/well of Substrate Vial 1. Then add in sequence 25 µl/well of

Substrate Vial 2, and then 25 µl/well of Substrate Vial 3. Read the plate at a wavelength of 540

nm. Use an excel spreadsheet to plot a linear standard curve of pg/ml glucan standard (y axis)

versus mean OD540nm (x axis). Add a trend line (linear regression line), the equation for the

line and the R2 value to the chart. Use the equation to interpolate pg/ml of glucan from OD's of

unknowns. Use only those OD's that fall on the standard regression line. Determine pg/ml of

glucan in diluted samples, correct each value for sample dilution, and find the average pg/ml

glucan in each sample. Convert pg/ml to µg/g dust by dividing by 25,000,000 (as 1ml NaOH

was added to 25 mg dust, and 1pg=10-6 µg).

234

C2: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – DUST SAMPLES –

KINETIC ASSAY

GLUCATELL Assay for (1-3)-ß-D-Glucan – in DUST samples

(KINETIC ASSAY)

GLUCATELL® - ASSOCIATES OF CAPE COD INC.

(1-3)-ß-D-Glucan Detection Reagent Kit- KINETIC ASSAY

GENERAL INFORMATION

No gloves to be used, as gloves may be contaminated with endotoxin during production.

All the material should be pyrogen-free (tips, reagents, diluents, etc.). Perform all procedures in a bio-safety hood.

SAMPLE PROCESSING

If there is enough dust:

Use the Mettler AB 104-S top loading digital electronic balance, weigh 25 mg of dust directly into pyrogen-free 2 ml tubes (ISC BioExpress, cat # C-3229-1). Add 1 ml of 0.6

M NaOH (Sigma Cat #7732-18-5) to make a 25 mg/ml suspension.

If there is not enough dust:

Weigh 10 mg or less using the Mettler H20T, and 4”x4” weigh paper. Transfer the dust

25 mg/ml

235

Place the tubes in a rack on the shaker & extract for 1 hour at room temperature to unwind the triple-helix structure of the glucan and make it water-soluble. Centrifuge for 1 minute at 7000 rpm if using the IEC Centra 4 centrifuge, equipped with micro-centrifuge rotor. Save the supernatant by pouring into a second set of 2 ml pyrogen-free tubes.

Analyze the same day.

PREPARE DILUTIONS

(1-3)-ß-D-Glucan Standard (Pachyman)

Lyophilized (1-3)-ß-D-Glucan standard. Reconstitute the lyophilized standard with

Reagent Water provided in the kit, to make 200pg/ml solution. The amount of water is half the one given on the vial. Vortex for 1 minute prior to making further dilutions.

Prepare dilutions of the standard and the samples in reagent water according to the table.

Use pyrogen-free tubes for preparing dilutions. Expel pipette tip contents slowly to ensure complete transfer. Mix well. Vortex each standard for 1 minute before transferring to the assay plate.

Glucan standard

ml standard Final Concentration (pg/ml) ml Reagent Water

200 0 300ml of 200 pg/ml

50 300 100ml of 200 pg/ml

12.5 300 100ml of 50 pg/ml

3.125 300 100ml of 12.5 pg/ml

236

SAMPLE DILUTIONS

Make dilutions in pyrogen-free plates (Associates of Cape Cod), using Cape Cod Reagent

water as diluent.

Sample Numbers

#1 #2 #3 #4 #5 #6 #7 #8 #9 #10 #11 #12 A 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 B 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 C 1:1000 1:1000 1:1000 1:1000 0 0 0 0 0 0 0 0 D 1:104 1:104 1:104 1:104 1:104 1:104 1:104 1:104 1:104 1:104 1:104 1:104

Dilutions of Dust Extracts

ml sample (1:1 in 0.6 M Row Final Dilution ml Water NaOH)

A 1:10 90 10 (from NaOH sample

B 1:100 90 10 (from 1:10 dilution)

C 1:1000 180 20 (from 1:100 dilution)

D 1:10,000 180 20 (from 1:1000 dilution)

Run spikes for at least 5 samples. For this purpose, add 25 µl of the 50 pg/ml standard to

each sample dilution of 1:1000 and 1:10,000.

237

Glucatell Reagent

Reconstitute one vial of Glucatell reagent with 2.8 ml of Reagent water and then add 2.8

ml Pyrosol® Reconstitution Buffer. Swirl the vial gently to dissolve completely (do not

vortex). Use within 10 minutes or place at 2-8º C and use within 2 hours.

ASSAY PROCEDURE- KINETIC ASSAY

Samples should be run using 1:1000 (10-3) and 1:1 0,000 (10-4) dilutions. Standard curve

should be run in duplicate. Add 25 µl of Reagent water (blanks); samples diluted 1:1000

and 1:10,000, and standard at concentrations of 3.125, 12.5, 50, and 200 pg/ml to wells of

the assay plate as shown below. When spiked samples are used, add 25 µl of the spiked

sample to the assay.

Glucan Assay Plate

Row 1 2 3 4 5 6 7 8 9 10 11 12 A Blank 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 (RW) B Blank 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 (RW) C 3.125 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 pg/ml D 3.125 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 pg/ml E 12.5 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 10-3 pg/ml F 12.5 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 10-4 pg/ml G 50 10-3 10-3 10-3 10-3 10-3 10-4 10-4 10-4 10-4 10-4 10-4 pg/ml H 50 AV** 10-4 10-5 10-6 10-7 Dust* 10-4 10-5 10-6 10-7 10-8 pg/ml

238

*Dust control and **Asp. Versicolor control

Add 50 µl/well of Glucatell reagent using a repeater pipette. Cover the plate with its lid, and place it in the plate reader and shake for 5 seconds before reading at 405 nm. Set the software to take readings every 60 seconds for 150 minutes.

Set up the software to measure Onset Time at 0.03 O.D. (Delta T at Onset O.D.).

Use the plate reader software to plot a log/log standard curve of pg/ml (1-3)-ß-D-glucan standard (x axis) versus mean Time of Onset OD 0.03 at 405nm (y axis). Add a trend line

(linear regression line), the equation for the line and the R2 value to the chart. Use the equation to interpolate pg/ml of glucan from OD's of unknowns. Use only those OD's that fall on the standard regression line. The software gives initial concentrations, and concentrations when dilutions are accounted for. Use the last as final ones. All data is in pg/ml units. Take the highest pg/ml value for each sample, only if the sample results reached a plateau with a decrease of pg/ml or

239

C3: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – AIR SAMPLES

GLUCATELL Assay for (1-3)-ß-D-Glucan – indoor/outdoor AIR SAMPLES

(AIR FILTERS)

GLUCATELL® - ASSOCIATES OF CAPE COD INC.

(1-3)-ß-D-Glucan Detection Reagent Kit- KINETIC ASSAY (onset time)

GENERAL INFORMATION

No gloves to be used. All the material should be pyrogen and glucan free (tips, reagents, diluents, etc.). Use a bio-safety hood.

SAMPLES PROCESSING

Add 1 ml of 0.6 M NaOH to each air filter. Use a 0.5 – 5 ml pipetter with and autoclaved sterile tips. The extraction process includes vortexing each tube containing air filter with

NaOH using a touch mixer (Model 231, Fisher Scientific, Pittsburgh, PA, USA) for 2 min, followed by sonication using ultrasonic bath (FS20, Fisher Scientific) agitation for

10 mins as described by Wang et al., (2001).

Place the tubes in a rack on the shaker (220 rpm) & extract for 1 hour at room temperature to unwind the triple-helix structure of the glucan and make it water-soluble.

240

PREPARE DILUTIONS

SAMPLE DILUTIONS

Make dilutions in pyrogen-glucan free plates (Assoc Cape Cod), using Cape Cod Reagent

water as diluent. Vortex each sample before diluting.

Sample Numbers

#1 #2 #3 #4 #5 #6 #7 #8 #9 #10 #11 #12 A 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 1:10 B 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 1:100 C 1:103 1:103 1:103 1:103 1:103 1:103 1:103 1:103 1:103 1:103 1:103 1:103

Dilutions of Air Filter Extracts Final Row ml Water ml sample (1:1 in 0.6 M NaOH) Dilution A 1:10 90 10 (from NaOH sample 1:100 is the highest for indoor B 1:100 90 10 (from 1:10 dilution) samples 20 (from 1:100 dilution) 1:1000 is the highest for outdoor C 1:1000 180 samples

(1-3)-ß-D-Glucan Standard (Pachyman)

Lyophilized (1-3)-ß-D-Glucan standard. Reconstitute the lyophilized standard with

Reagent Water provided in the kit, to make 200pg/ml solution. The amount of water is

half the one given on the vial. Vortex for 1 minute prior to making further dilutions.

Prepare dilutions of the standard and the samples in reagent water according to the table.

241

Use pyrogen-free tubes for preparing dilutions. Expel pipette tip contents slowly to ensure complete transfer. Mix well. Vortex each standard for 1 minute before transferring to the assay plate.

Glucan standard

Final Concentration (pg/ml) ml Reagent ml standard Water 200 0 300ml of 200 pg/ml 100 100 100ml of 200 pg/ml 50 300 100ml of 200 pg/ml 12.5 300 100ml of 50 pg/ml 3.125 300 100ml of 12.5 pg/ml

Glucatell Reagent

Reconstitute one vial of Glucatell reagent with 2.8 ml of Reagent water and then add 2.8 ml Pyrosol® Reconstitution Buffer. Swirl the vial gently to dissolve completely (do not vortex). Use within 10 minutes or place at 2-8º C and use within 2 hours.

ASSAY PROCEDURE- KINETIC ASSAY

Samples will be run using 1:100 for indoor samples and 1,000 dilutions for outdoor samples. Samples and Standard curve will be run in duplicate.

Add 25 µl of Reagent water (blanks); samples diluted 1:100 for indoor samples, and

1:1,000 for outdoor samples, and standard at concentrations of 3.125; 12.5; 50; 100 and

200 pg/ml to wells of the Assay plate as shown below.

242

Glucan Assay Plate –indoor only or outdoor only

Row 1 2 3 4 5 6 7 8 9 10 11 12 A Blank 3.125 12.5 50 100 200 Dust* 10-4 10-5 10-6 10-7 10-8 (RW) pg/ml pg/ml pg/ml pg/ml pg/ml B Blank 3.125 12.5 50 100 200 AV** 10-4 10-5 10-6 10-7 10-8 (RW) pg/ml pg/ml pg/ml pg/ml pg/ml C 100 100 100 100 100 100 100 100 100 100 100 100 (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) D 100 100 100 100 100 100 100 100 100 100 100 100 (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) E 100 100 100 100 100 100 100 100 100 100 100 100 (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) F 100 100 100 100 100 100 100 100 100 100 100 100 (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) G 100 100 100 100 100 100 100 100 100 100 100 100 (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) H 100 100 100 100 100 100 100 100 100 100 100 100 (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) (1000) *Dust control and **Asp. Versicolor control

Add 50 µl/well of Glucatell reagent using a repeater pipette. Cover the plate with its lid,

and place it in the plate reader and shake for 5 seconds before reading at 405 nm. Set the

software to take readings every 30 seconds for 150 minutes. Set up the software to

measure Onset Time at 0.03 O.D. Use the plate reader software to plot a log/log standard

curve of pg/ml (1-3)-ß-D-glucan standard (x axis) versus mean Time of Onset OD 0.03 at

405nm (y axis). Add a trend line (linear regression line), the equation for the line and the

R2 value to the chart. Use the equation to interpolate pg/ml of glucan from OD's of the

unknowns. Use only those OD's that fall on the standard regression line. The software

gives initial concentrations, and concentrations when dilutions are accounted for. Use the

last as final ones. All data is in pg/ml units. Take the highest pg/ml value for each sample,

only if the sample results reached a plateau with a decrease of pg/ml or

repeat analysis till a plateau with a decrease in concentration is achieved. Convert pg/ml

to ng/m3 air.

243

C4: LAL PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS – PURE FUNGAL

SPORES SUSPENSIONS

GLUCATELL Assay for (1-3)-ß-D-Glucan – FUNGAL SPORES

(KINETIC ASSAY)

GLUCATELL® - ASSOCIATES OF CAPE COD INC.

(1-3)-ß-D-Glucan Detection Reagent Kit- KINETIC ASSAY (onset time)

GENERAL INFORMATION

No gloves to be used. All the material should be pyrogen and glucan free (tips, reagents, diluents, etc.). Use a bio-safety hood.

SAMPLE PROCESSING

1 Add 0.5 ml of 0.6 M NaOH to each 0.5ml fungal spores solution. Use a 0.5 – 5 ml

pipetter with Glucan free tips (ACC).

2. Place the tubes in a rack on the shaker (220 rpm) & extract for 1 hour at room

temperature to unwind the triple-helix structure of the glucan and make it water-

soluble.

244

PREPARE DILUTIONS

SAMPLE DILUTIONS

Make dilutions in pyrogen-glucan free plates (Assoc Cape Cod), using Cape Cod Reagent

water as diluent. Vortex each sample before diluting.

Sample Numbers

Dilutions of Fungal Spores Extracts

Final Row ml Water ml sample (1:1 in 0.6 M NaOH) Dilution

A 1:10 180 20 (from NaOH sample

B 1:100 180 20 (from 1:10 dilution)

C 1:1000 180 20 (from 1:100 dilution)

D 1:10000 180 20 (from 1:1000 dilution)

E 1:100000 180 20 (from 1:10000 dilution)

245

(1-3)-ß-D-glucan Standard (Pachyman)

Lyophilized (1-3)-ß-D-glucan standard. Reconstitute the lyophilized standard with

Reagent Water provided in the kit, to make 200pg/ml solution. The amount of water is half the one given on the vial. Vortex for 1 minute prior to making further dilutions.

Prepare dilutions of the standard and the samples in reagent water according to the table. Use pyrogen-free tubes for preparing dilutions. Expel pipette tip contents slowly to ensure complete transfer. Mix well. Vortex each standard for 1 minute before transferring to the assay plate.

Glucan standard

Final Concentration (pg/ml) ml Reagent ml standard Water 200 0 300ml of 200 pg/ml 100 100 100ml of 200 pg/ml 50 300 100ml of 200 pg/ml 12.5 300 100ml of 50 pg/ml 3.125 300 100ml of 12.5 pg/ml

Glucatell Reagent

ASSAY PROCEDURE- KINETIC ASSAY

Add 25 µl of Reagent water (blanks); samples, and standard at concentrations of 3.125;

12.5; 50; 100 and 200 pg/ml to wells of the Assay plate as shown below.

246

Glucan Assay Plate

Row 1 2 3 4 5 6 7 8 9 10 11 12 A Blank 3.125 12.5 50 100 200 Dust* 10-4 10-5 10-6 10-7 10-8 (RW) pg/ml pg/ml pg/ml pg/m pg/ml l B Blank 3.125 12.5 50 100 200 AV** 10-4 10-5 10-6 10-7 10-8 (RW) pg/ml pg/ml pg/ml pg/m pg/ml l C 1 1:10 1:100 1:1000 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 D 1 1:10 1:100 1:1000 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 E 1 1:10 1:100 1:1000 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 F 1 1:10 1:100 1:1000 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 G 1 1:10 1:100 1:1000 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 H 1 1:10 1:100 1:1000 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 *Dust control and **Asp. Versicolor control

Add 50 µl/well of Glucatell reagent using a repeater pipette. Cover the plate with its lid,

and place it in the plate reader and shake for 5 seconds before reading at 405 nm. Set the

software to take readings every 30 seconds for 150 minutes.

Set up the software to measure Onset Time at 0.03 O.D.

Use the plate reader software to plot a log/log standard curve of pg/ml (1-3)-ß-D-glucan

standard (x axis) versus mean Time of Onset OD 0.03 at 405nm (y axis). Add a trend line

(linear regression line), the equation for the line and the R2 value to the chart. Use the

equation to interpolate pg/ml of glucan from OD's of the unknowns. The software gives

initial concentrations, and concentrations when dilutions are accounted for. Use the last

as final ones, and multiply by 2 (as 0.5 ml solution was diluted in 0.5 ml 0.6 M NaOH

247

extraction solution). All data is in pg/ml units. Take the highest pg/ml value for each sample, only if the sample results reached a plateau with a decrease of pg/ml or

Otherwise repeat analysis till a plateau with a decrease in concentration is achieved.

248

APPENDIX D: EIA PROTOCOL FOR (1-3)-b-D-GLUCAN ANALYSIS

(1-3)-ß-D-Glucan Inhibition ELISA

Solutions:

-Laminarin for coating plates – 2ug/mL in PBS

-For 96 well plate with 200 ul per well, need 1ug/ml stock – 20 mL of

PBS and 40 ul of Laminarin to coat all plates evenly with a concentration

of 2ug/ml

-For 10mg/mL stock, 4 uL of laminarin for 20 mL of PBS

-PBST – PBS + 0.05% TWEEN20 (v/v)

-PBS stock found in 10X form, so must be diluted.

-Dilute PBS with 5 ml of PBS into 50 ml of de-ionized water

-In 500 ml of PBS, 250 ul of TWEEN necessary to complete solution

-PBSTG – PBS + 0.05% TWEEN20 (v/v) + 0.5% gelatin (w/v)

-To made up PBST, add gelatin 2.5 grams

-Heat on hot plate until completely dissolved

-B-glucan standards (Dextran,Laminarin, Mannan, Pullan, Schizophyllan)

-Diluted to 100000ng, 25000ng, 10000ng, 1000ng, 250ng, 100ng, 25ng,

and 0ng

249

_Out of powdered glucans, made .1g powder + 10 mL of water =

10mg/mL stock

-Made appropriate dilutions according to stock concentration

-Stressed importance to autoclave all stock dilutions of B-glucans to

ensure purity

1’ Ab anti-B Glucan 1:4000-1:8000 (must dilute stock from tube)

-For 1:8000 – 1.25 ul Ab + 10 mL of PBSTG = 100uL per well on 96 well plate

Procedures:

1. Coat ELISA plate with 200uL of 2ug/mL Laminarin overnight at 4 degrees C

2. Wash with 300 uL PBST X3

3. Block with 300 uL PBSTG for 30 minutes at 37 degrees C shaking

4. Add B-glucan standards or samples in 100 uL of PBSTG, mix in an equal volume

of primary anti-(1-3)-ß-D-glucan antibody (Biosupplies, Australia) in PBSTG

(1:8000)

-prepare the following dilutions of laminarin: 1 mg/ml, 100 µg/ml, 25 µg/ml, 10

µg/ml, 2.5 µg/ml, 1 µg/ml, 250 ng/ml, 100 ng/ml, 25 ng/ml and 1 ng/ml

-add in equal amounts so whole 96 well plate has 200 ul of solution in each well

(100 uL of B-glucan standard + 100 uL of primary anti-(1-3)-ß-D-glucan antibody

solution)

5. Shake plate for 90 minutes at 37 degrees C

6. Wash with 300 uL PBST X3

250

7. Add 200 uL of secondary antibody (sheep anti-mouse peroxidase, Jackson

Immuno) in PBSTG for 60 minutes shaken at 37 degrees

-Needed 1:5000 dilution of secondary Ab (4uL Ab + 20 mL of PBSTG) – fills all

96 wells with 200 uL of solution

8. Wash with 300 ul of PBST X6

9. Detect with Reagent A+B

-100 ul per well, needed approximately 10 mL for whole plate (5mL A + 5mL B)

-Doubled as well because use of 2 96 well plates

10. Wait 30 minutes for detection

11. Add 50 ul of 2N H2SO4 per well (should have changed color from blue to yellow)

251