Pergamon .I Aerosol Su Vol. 28, No. 3, pp.

3X1&392, 1997
CopyrIght ( 1997 Elsewer Saence Ltd
Pnnted in Grrat Bntain All rughts reserved
PII: SOO21-8502(96)00441-7 0021-8502:97 $17.00 + 0.00

AIR SAMPLING FOR FUNGI IN INDOOR ENVIRONMENTS

Brian Flannigan

Department of Biological Sciences, Heriot-Watt University, Edinburgh EH14 4AS, U.K.

Abstract-Mould growth in buildings is a major health issue, but most investigations of the indoor
air spora still employ culture-based methods. These are inadequate for assessing exposure, since
culturable organisms comprise a small fraction of the total of potentially allergenic/toxigenic units in
air. For epidemiological studies, measurement of airborne fungal biomass over extended periods
may be more relevant than total counts. Whilst (1+3)-/6D-gkKaU has been used to assess airborne
biomass, ergosterol may be the best indicator of exposure. For case studies, patients serum has been
used to detect specific spores on sampler slides, and both highly specific and less specific antisera
could be used either via fluorescent antibody technique or enzyme-linked immunosorbent assays. In
the future, solid-phase polymerase chain reaction (PCR) may be used to detect pathogens or other
well-characterized potentially harmful species, and in at least some groups mycotoxin/secondary
metabolite/volatile profiles may be used in identification. c 1997 Elsevier Science Ltd. Ail rights
reserved

INTRODUCTION

In temperate climates, the development of buildings with minimum energy usage has led to
tight air-conditioned buildings in which exchange of air with the outside is greatly
reduced. Water vapour, which would otherwise have been vented, condenses on cool
surfaces in the same way as in poorly insulated buildings without air-conditioning and
creates conditions for microbial growth and an associated build-up of bioaerosols. Al-
though spores of fungal pathogens such as Aspevgillusfumigatus are very seldom numerous
in indoor air and pose little hazard to healthy individuals, there are special risks for
immunosuppressed patients and other severely compromised individuals. However, it is
well known that spores of species of Aspergillus, Cladosporium and Penicillium generated in
damp buildings can cause bouts of asthma and/or rhinitis among atopic occupants. In
addition, exposure to large indoor concentrations of spores of a range of individual fungi,
from Cladosporium to the dry-rot fungus Serpula lacrymans, have been the cause of rare
instances of extrinsic allergic alveolitis (Flannigan et al., 1991). As well as having a role in
such individual cases of allergic disease, fungi are now seen as having a wider role in
respiratory health. Several large-scale epidemiological investigations in North America
have noted a strong association between reported dampness/mould in homes and reported
respiratory symptoms (Brunekreef et al., 1989; Dales et al., 1991a, b; Spengler et al., 1991)
and, in Finland, Jaakkola et al. (1993) observed a twofold increase in respiratory symptoms
among pre-school children living in homes with reported dampness or mould. Brunekreef
et al. (1989) reported that the effect on children was of similar magnitude to parental
smoking. Mould growth in homes is therefore a major health issue, and there is an urgent
need to obtain objective microbiological data in order to confirm the role of moulds
indicated by these epidemiological investigations.

THE INDOOR AIR SPORA

At least in North America, the general perception of what the air spora in homes and
non-industrial workplaces should be is of a mixture of species resembling that in outdoor
air. Certainly if the buildings are air-conditioned, or if windows and doors are kept closed in
summer, the indoor counts could be expected to be somewhat lower. Field or phylloplane
fungi such as Alternaria and Epicoccum conform to this picture, but the ratio for the sum of
species of Aspergillus and, especially, Penicillium outdoors to that indoors may either be

381
382 B. Flannigan

Table 1. Mean abundance of hyphal fragments and mam types of fungal spore in samples of
indoor and outdoor air taken on 132 days between December 1991 and September 1993 in
Ontario (Li and Kendrick, 1995)

Indoor Outdoor

Category* Spores (m ) (%B

) Spores (mm) (A,)

Hyphal fragments 146 6.3 112 3.2

AItrrnu~icc 44 1.9 74 2. I
A.speryillusiPenicilli~rm 457 19.8 131 3.8
CIudo.sporium 895 38.8 1479 42.5
Coprimu 41 1.8 78 2.3
Epicoccum 0.3 20 0.6
Gtrnotlrrmn 59 2.6 Ill 3.2
Lrptosphawiu 1x2 7.9 541 15.1
Unidentified ascosporea 65 2.8 138 4.0
Unidentified basidiospores 152 6.5 310 8.9
Other unidentified spores 206 8.9 301 8.7

* 28 other taxa. spores of which were recorded occasionally, are not included

much lower or reversed (Flannigan et al., 1991). The relative abundance of different species
of Penicillium may also differ, with some species being more conspicuous in indoor air and
others in outdoor air (Fradkin, 1987). A recent Canadian investigation in which total counts
were made from exposed slides with a Samplair particle sampler (Li and Kendrick, 1995)
illustrates the relative abundance of different categories of fungal spores, with Clndnsporium
predominating in both outdoor and indoor air and outdoor concentrations of most
categories being higher than those indoors. However, the Aspergillus/Penicillium grouping,
generally considered largely to have an indoor origin, formed < 4% of the outdoor air
spora but nearly 20% indoors (Table 1). The study also illustrates what is not revealed by
conventional viable sampling, viz., the relative abundance of the spores of Ascomycetes and
Basidiomycetes, which in this case comprised around one-third of the outdoor and one-fifth
of the total indoor air spora (Table 1). The study also emphasises the perceived role of
spores infiltrating from outdoors in determining much of the indoor air spora, with peak
counts of Cladosporium and Alternaria, although lower, coinciding with those outdoors at
the height of the growing season. In contrast, three peaks of Aspergillus~Penicillium spores
occurred indoors in January, April and September, but not outdoors.
The importance of mould growth within buildings in contributing to the indoor air spora
has been highlighted by, among others, Hunter et al. (1988) and Flannigan et al. (1993). It is
not only species of Aspergillus and Penicillium that are to be found growing in buildings,
boosting the airborne spore burden. Cludosporium spp. commonly grow on damp indoor
surfaces, as do more hydrophilic Phoma and Ulocllldium spp., and in very damp conditions
Stachyhotrys atrn (syn. S. churtarum) may be prominent (Grant et al., 1989). In an investiga-
tion of 41 homes in the American Midwest, DeKoster and Thorne (1995) noted that high
indoor viable counts were associated with high basement humidity. The mean ratio of
indoor: outdoor airborne viable fungi for basements of homes where occupants complained
of sick building syndrome (SBS) symptoms was 2.16, and in non-complaint homes was 0.56.
The corresponding ratios for the main floor were 0.84 versus 0.37.

ASSESSING THE AIR SPORA

As has been pointed out by Strachan et ul. (1990) failure to establish an objective
connection between the respiratory health status of occupants of mould-affected houses and
airborne microorganisms in these houses may be the result of inadequate quantification of
the air spora to which the occupants are exposed. Since most investigations have only
assessed numbers of culturable organisms, they have ignored numbers of non-viable or non-
culturable spores that may be as allergenic or toxigenic as their culturable counterparts and
have as significant an effect on health. There can be both qualitative and large quantitative
 Sampling indoor environments 383

differences between the total numbers of fungal particles (viable + non-viable) in indoor air
and those that can be collected, cultured, counted and identified on agar plates. Kozak et al.
(1979) graphically illustrated that the number of viable spores of individual fungi can be
below the limit of detection by established methods although the total is sufficient to cause
a respiratory problem. Thus, they did not detect Stachybotrys atra when using an Andersen
sampler in the home of an asthmatic child, but did detect the mould by means of a rotorod
sampler. It was estimated that only l-2% of S. atra spores, which caused asthma attacks in
the child, were viable. As Kozak et al. (1979) stated, in health related studies no one
sampling technique is adequate for assessment of indoor fungi; a culture-based system is
needed for identifying species which may be significant for health, and a total count or
a measure of biomass is required to assess exposure. The best approach (Flannigan, 1992) is
to use the one sampler, e.g. filter sampler, liquid impinger or cyclone sampler, for collection
of all microorganisms and divide the sample into portions for culture, total counting and/or
assessment of biomass (or any desired metabolite).
Collecting airborne microorganisms by drawing air, at a low flow rate, through a poly-
carbonate membrane in an aerosol monitor cassette for several hours (the CAMNEA
method; Palmgren et al., 1986a, b), Strom et al. (1990) found that only a small fraction of
fungal spores in the total (washed from the membrane, stained with acridine orange and
counted by a direct epifluorescence technique) were culturable. Some differences between
culturable fraction and total count in a restaurant environment are illustrated in Table 2. As
might be expected, the counts of culturable organisms obtained by the CAMNEA method
over a period of 4 h can differ greatly from short-duration grab samples taken with
a six-stage Andersen cascade impactor.
The effect of human activity on counts of viable microorganisms in indoor air is clearly
shown by the converted Andersen sampler counts for the dining area of the restaurant, with
the largest counts occurring during the busy lunchtime period. In a more recent lo-month
investigation of a group of Scottish houses (Flannigan et al., 1996), the median count of
viable airborne fungi indoors, as determined by the CAMNEA method, was 260 colony
forming units (CFU) me3 air and that of bacteria 339 rne3. The median values for total
fungal and bacterial counts were, respectively, 13,940 spores and 88,260 cells rn- 3. On
average, the viable counts for fungi and bacteria were approximately 0.57 and 0.28% of the
corresponding total counts. At > 175 : 1 (fungi) and nearly 360 : 1 (bacteria), the ratios of
total : viable counts are large, but not as large as in some healthy Swedish houses (Strom
et al., 1990), where the corresponding ratios determined by the two CAMNEA techniques
were 500: 1 and 2000: 1.
It is generally held that bacteria in indoor air are predominantly Gram-positive species
shed from the human body and that numbers of Gram-negative bacteria are relatively

Table 2. Concentration of airborne microorganisms obtained using different sampling methods in a restaurant
(after Flannigan, 1992)

Concentration (CFU m 3 air)

Andersen sampling* CAMNEA method

Plate count DEFT method

Before Early
Area open Mid-day evening a.m. p.m. a.m p.m.

Pantry
Moulds 212 141 129 3160 920 10,200 1260
Bacteria 224 152 208 138 625 34,910 5100
Dininy
Moulds 35 565 24 521 348 2010 12,360
Bacteria 106 244 88 417 1180 43,250 13,580

* Raw counts converted by the positive hole method to correct for multiple impactions of propagules at the same
sites on the agar collection plates (Andersen, 1958).
384 B. Flannigan

Table 3. Spearman correlation between Burkard personal sampler total

counts and Andersen two-stage sampler viable counts for indoor air in 41
homes (DeKoster and Thorne, 1995)

Category R values p values

Altrrnurii~ 0.43 0.02

Aspe~yillusiPeni~illilrrn 0.45 0.01
Chlosporiurn 0.65 < 0.0001
Unclassified 0.22 0.21
Total 0.80 < 0.000 I

small, unless there is some amplification site within the building, e.g. a heavily contaminated
humidifier (Flannigan, 1992). However, Gram-negative bacteria are generally more suscep-
tible than Gram-positive bacteria to desiccation, so that loss of viability (culturability) due
to continuing exposure to a stream of sampled air after deposition on a surface (particularly
if the surface is a dry membrane rather than moist agar) is likely to account for at least part
of the difference between (a) viable counts of Gram-positive and Gram-negative bacteria
and (b) total and viable counts.
Flannigan et nl. (1996) reported that the ratio between comparable viable counts and
total counts in the indoor air of a group of Scottish houses was inconsistent, probably
because the composition of the air spora differs with location and different microorganisms
have different survival rates, both in the environment and on the membrane during the
sampling period. For example, spores of Aspergilhs and Penicillium may survive for long
periods, even years, whilst the viability (or culturability) of others, e.g. S. mu. may decline
very rapidly. This can make the interpretation of results of air sampling difficult (Flannigan
and Miller, 1994). The isolation of S. utra from air samples collected on culture media
should be interpreted differently from isolation of Penidium spp. on the same plates.
However, without reporting the total concentrations of fungal spores collected, DeKoster
and Thorne (1995) stated that total concentrations, estimated using a Burkard personal
sampler, agreed well with corresponding concentrations of viable spores, obtained using an
Andersen two-stage sampler (Table 3). However, although Spearman correlation between
the counts for Cladosporium was high, that for other categories was only moderate.

PROBLEMS OF METHODOLOGY

Most microbiological investigations of indoor air still employ culture-based methods, but
sufficient attention is seldom given to four important issues: sampler performance. temporal
variability, culture media and accurate identification (Flannigan and Miller. 1994). In this
last respect, too many studies identify only to the genus level and disregard the diversity of
species, their ecology and potential significance for health, especially in important genera
such as Asperyillus and Penicillium.
Although analysis has shown that some air sampling devices can perform better than
others, the design of no currently available sampler can be described as optimal on
theoretical grounds (Nevalainen et al., 1992). All have biases and deficiencies. In a number
of laboratory investigations, particular samplers have been found to be better than others
for the intended purpose, e.g. the Andersen six-stage sampler for viable counts and the
Burkard 24 h sampler for total counts (Buttner and Stetzenbach, 1993). Efforts have been
made to determine the reliability of various air samplers suitable for fungi by means of
side-by-side comparisons in buildings. For example, Verhoeff et al. (1990a, b) reported that,
among air-samplers used in side-by-side tests in houses, the Andersen N6 sampler gave the
largest colony counts and the greatest diversity of fungi. Miller (1993) attributed this to the
longer time required to sample the same volume of air as the other samplers. Noting the
temporal variability in counts for six rooms in an office building, Stanevich and Petersen
(1990) had earlier reported that the variability of raw counts from 1 min N6 samples was six
times that of 5 min samples.
 Sampling indoor environments 385

Temporal variability is a major problem in assessing human exposure to the indoor air
spora. This is amply illustrated by Verhoeff et al. (1990b), who found that not only did the
numbers of propagules in 2-min samples taken by an Andersen N6 (one-plate) sampler
show great variability over relatively short periods but also that within-home variance
approached four times that between homes. An important factor introducing variability
into the nature and magnitude of the indoor air spora is the release of fungi from carpets
and walls or other surfaces. This depends on the type and degree of activity of occupants in
the room. All activity in buildings disturbs settled spores, but cleaning, constructional work
and any other major dust-raising activities have a particular impact (Hunter et al., 1988).
Studies in an experimental room, into which spores of Penicillium chrysogenum were
introduced and allowed to settle on a nylon carpet (Buttner and Stetzenbach, 1993)
unequivocally confirmed that foot traffic on carpets resulted in elevated counts of airborne
spores. Buttner and Stetzenbach (1993) drew attention to the possible error introduced into
the results through the activity of investigators at the sampling site and the re-entrainment
of settled spores by sampler exhaust air near the floor.
Differences in the size and sedimentation rate of spores also affect what is detected in air
samples. For example, it has been demonstrated that large Ulocladium spores released from
mould patches on walls in damp houses sediment relatively rapidly (Hunter et al., 1988) so
that, even where growth is profuse, the mould is likely to be detected in the air in quantity
only shortly after disturbance of the growth or re-entrainment of settled spores as a result of
activity. An investigation over one day in a nursery school classroom (Mouilleseaux and
Squinazi, 1991) illustrates the wide fluctuations in numbers of airborne viable spores which
can occur as a result of variation of activity levels in a room (Fig. 1).
It has often been suggested that to circumvent this temporal variability housedust should
be sampled instead, as it provides a memory of previously airborne microorganisms
which are re-entrained in the indoor air as a result of the activity of building occupants.
However, although the housedust mycobiota reflects that of the air, there are differences in
the relative abundance of some types, and a Basidiomycete common in the air within homes
in Scotland, Sistotrema brinkmannii, is only found infrequently and in small numbers in
housedust (Flannigan et al., 1993, 1996). This therefore argues against sampling of dust as
a substitute for air sampling. In addition, viable counts for settled dust are very much higher
than corresponding air sampler counts for aerosolized dust (Flannigan et al., 1994). This
suggests that many microorganisms in dust either form aggregates or are carried on dust
particles which settle very rapidly and are too large to be respirable.

Fig. 1. Variability in counts of viable airborne fungi during the course of a day in a nursery school
classroom (after Mouilleseaux and Squinazi, 1991).
386 B. Flannigan

In most investigations of fungi in indoor air, the isolation media used favour the growth
of hydrophilic species (Flannigan, 1992; Flannigan and Miller, 1994) and. by being nutri-
tionally rich, e.g. Blakeslees malt extract agar or modified Sabouraud agar, may introduce
bias in favour of rapidly growing species. An international workshop, considering the effect
of fungi in buildings on health, recommended that such rich media should not be used
(Samson et al., 1994). Further error occurs because some fungi do not compete well with
others on isolation plates, even if the medium is suitable for their growth, and are
consequently not recorded as frequently as they merit, e.g. Alternariu spp. Because some
fungi have spores distinctive enough to be recognised microscopically, two different types of
sampler operating on different principles, one for a total count of such species and one for
a viable count of those without distinctive spores, have sometimes been used in the same
health investigation (Su et ul.. 1990; Su and Spengler. 1991).
Xerophilic species may be isolated on the media mentioned above, but are only usually
seen when the propagules of faster growing species are absent or few. Since xerophilic
moulds such as Eurotium spp., Aspergillus restrictus, A. penicillioides and Wallemiu sehi are
known to be present in housedust and indoor air and are allergenic (Verhoeff et ul., 1990a, b;
Flannigan and Miller, 1994), it has been recommended that a low water activity medium,
e.g. dichloran-18% glycerol agar, should be included among the isolation media used in
indoor air studies (Samson et al.. 1994).

ALTERNATIVE APPROACHES TO ASSESSMENT

Given the problems of assessment mentioned above, there is a strong argument for using
non-cultural rapid methods to quantify the airborne microbial burden although, generally.
these do not allow identification of the microorganisms present. Such methods can be based
on chemical components, or bio-markers. common to organisms in particular groups, e.g.
chemical markers for peptidoglycan or lipopolysaccharide (LPS) in the bacterial cell
envelope (Fox et ul., 1993; Fox and Rosario, 1994), or chitin or ergosterol in. respectively,
the walls and membranes of hyphae and spores of filamentous fungi and yeast cells.
Measurements of chitin have been used as an index of fungal biomass in environments
ranging from grain (Donald and Mirocha. 1977) to wood (Swift, 1973). However, this
technique cannot be applied where insects or other arthropods are likely to be encountered.
The amount of chitin in a fragment of insect exoskeleton collected as part of an air sample is
likely to be much greater than the contribution of all fungi in the sample. There is the
additional problem of the amount of chitin differing with species, growth conditions and the
nature of the structure (hyphal fragment or spore).
As most investigations of fungal aerosols in indoor air are health-driven, it can
be considered that assessment will best be made by measuring some component or
components of the aerosol with clear biomedical effects, e.g. particular allergens. Substances
that are found in virtually all fungi of consequence in indoor air and which have
potent biomedical effects are the so-called (l-3)-/j-D-glucans in walls of hyphae and
spores. In fact, although the predominant linkage in such polyglucoses is (1+3), they
are branched mixed-linkage polymers with (1 +6) cross-linkages. In mammals, there
are receptors for fi-glucan on alveolar macrophages. neutrophils, basophils and other
cells (Czop and Kay, 1991). Exposure to the glucan causes inflammation reactions in
lymphocytes, affects interleukin-1 secretion via T-lymphocytes, stimulates bacterial
and tumour defence mechanisms, causes a decrease in numbers of pulmonary macrophages
and inhibits phagocytosis. A decrease in lymphocyte numbers in the lung wall is the
opposite of the effect of exposure to endotoxin (Fogelmark rt al.. 1994). Large concentra-
tions of airborne I-glucan have been associated with increased reporting of mucous
membrane irritation and fatigue by occupants of buildings in which there were greater than
normal numbers of complaints of building-related health effects (Rylander et al., 1992).
Disregulation of pulmonary macrophages and the associated release of inflammatory
mediators could be responsible for headache, fatigue and other neurological symptoms
(Rylander, 1995).
 Sampling indoor environments 387

Limulus amoebocyte lysate (LAL) preparations used for quantifying bacterial endotoxin
(LPS) are known to be coagulated by (l-+3)-/&glucans, although they are lOOO-fold more
sensitive to LPS than glucans (Roslansky and Novitsky, 1991). LAL can be fractionated to
produce a preparation specifically sensitive to /3-glucan (Kitagawa et al., 1991). Rylander
and his colleagues (e.g. Rylander et al., 1992) have used such a fraction for quantifying
P-glucan in indoor air. The method involves collection of an air sample on a microporous
filter similarly to the CAMNEA method for microbiological analysis (Palmgren et al.,
1986a, b). The glucan in the sample on the membrane filter is extracted by autoclaving in
a saponin solution and assayed using the glucan-specific fraction. Bacterial endotoxin in the
same sample can be extracted in saponin at room temperature before this, and assayed
using an endotoxin-specific fraction (Rylander et al., 1992). Although the technique may
appear to have more relevance to health-related studies of indoor air than traditional
methods, more research into its use is required. Because availability of the glucan-specific
LAL is restricted, studies of indoor air in which it has been applied are limited in number
and design, and doseeresponse experiments with individual species and mixtures are
lacking, fuller evaluation is clearly needed.
Like P-glucan in walls, ergosterol, the principal sterol in membranes of hyphae and
spores, provides a means of assaying for fungal biomass but again gives no information on
species present. Ergosterol measurements have previously been used as an index of fungal
biomass in housedust (Miller et al., 1988) and what appears to be a very promising new
method for determination of airborne ergosterol has now been developed (Miller and
Young, 1996) and used in investigations of homes. Since ergosterol is stable under air-dry
conditions, spores can be collected on a microporous filter as in the CAMNEA method
(Palmgren et al., 1986a, b). The sterol is extracted from the collected spores in basic aqueous
methanol, assisted by microwave heating, and then analysed by high performance liquid
chromatography (HPLC), gas chromatography (GC) or GC-mass spectroscopy (Young,
1995). Most common species in indoor air have roughly the same distribution of spore sizes,
and the ergosterol content of spores of some ten common moulds was similar after
adjustment for size, at about 3.2 fg ergosterol mg- spores (Miller and Young, 1996).
Although dependent on species and analytical method, the minimum detectable number of
spores on a filter appears to be 2&100 (Young, 1955).

IDENTIFICATION

Identification of microorganisms presents a major problem for investigations of indoor

air and, in many studies, the isolated fungi are assigned only to broad categories such as
pink yeasts or wood-rotting Basidiomycetes or to genera such as Penicillium and
Aspergillus. However, in recent cases in North America where ill-health had been attributed
by workers to their workplace, labour laws have dictated that detailed information on the
properties of fungal species contaminating indoor air be provided for the workers. With the
diversity of species in genera such as Penicillium and Aspergillus., and the differences in the
ecological, allergenic and toxigenic characteristics between species in these genera, this
clearly calls for reliable identification. Extrapolating from a survey of toxigenic fusaria, it
can however be suggested that in the literature perhaps 50% of mould identifications are
incorrect (Flannigan and Miller, 1994). Accurate identification requires skilled and experi-
enced mycologists, preferably with experience of fungi from food or soil. Even then, few
laboratories have the high level of expertise required to identify, with certainty, the species
in Penicillium and other difficult genera. The problem of dependable identification is
compounded by the world-wide decline in the teaching of taxonomy and systematics in
universities. It is becoming increasingly evident that the numbers of individuals with skills
in traditional identification methods is going to continue to decrease in the foreseeable
future. Therefore, it will be necessary to develop new methods which do not demand the
degree of training and expertise currently required. There are developments in fields such as
medical and food microbiology which indicate a range of approaches that could, at least,
partly alleviate the problem.
388 B. Flannigan

Since immunological methods are widely used in medical microbiology and most invest-
igations of indoor air are health-related, it is not unnatural to consider such methods. Here,
an interesting approach involving use of serum from patients with extrinsic allergic alveoli-
tis (EAA) symptoms to explore their home environment was adopted by Zwick er al. (1991).
Air was sampled (10 min) with a Burkard personal air sampler and spores impacted on the
glass slide were counted; total counts ranged from < lo4 spores m-3 (low) to > 10
rnp3 (high). Half of each spore sample area was then coated with dilutions of individual
patients serum, shown previously by immunodiffusion tests to contain IgG/IgM precipita-
ting antibodies to one or more of Aureobusidium pullulans, Aspergillus jumigatus, Cephalos-
porium, Penicillium species or the thermophilic actinomycetes, Faenia rectivirgula and
Thermoactinomyces oulgaris. The half of the spore sample area not flooded with patients
serum acted as a negative control. After reaction and rinsing, the whole slide was flooded
with fluorescent-labelled anti-human IgG/IgM antiserum and examined, after final rinsing,
for fluorescent spores. This technique could therefore be of value in EAA cases for
confirming both exposure and sensitization to airborne spores of particular microorgan-
isms, and also in pinpointing areas requiring air quality control. It could also be adapted by
employing antisera raised against other species or groups of microorganisms. For example,
monoclonal antibodies have been produced in other fields of study, such as plant disease
and food spoilage (Dewey et al., 1993) for detecting and quantifying particular genera, e.g.
Aspergillus, Fusarium and Penicillium, or particuar species, e.g. P. islandicum and Humicola
ianuginosa (Thermomyces lanuginosus). Those which recognise epitopes on spore walls
could be used in a similar manner to that reported by Zwick et al. (1991) for aerometric
studies, via either fluorescent antibody technique or ELISA.
Just as in other areas of environmental investigation, polymerase chain reaction (PCR)
and other molecular biology techniques are likely to find use in the detection of well-
characterised toxigenic or pathogenic microorganisms in air. Alvarez et al. (1994) have
shown the potential of solid-phase PCR (SP-PCR) for detecting specific airborne microor-
ganisms which might be unculturable because of stress caused by aerosolization, environ-
mental exposure or sampling. In laboratory experiments, they employed a strain of
Escherichiu coli which contained a plasmid with a 437-base pair insert from the silkworm
BonzhJlx mori, which was a unique marker for identification of this strain. E. coli DHl.
Aqueous suspensions of the bacterium were aerosolized and the aerosols sampled for
5 s-10 min (l-120 1) using AGI-30 all-glass liquid impingers. For SP-PCR, collection buffer
was filtered through Nytran filters, and the residue lysed before binding of DNA (with
nonspecific DNA binding being blocked). The bound DNA was then amplified in 30
denaturation/annealing/primer extension cycles. SP-PCR showed greater sensitivity than
was achieved by filtering corresponding aliquots of collection buffer through a membrane
filter and incubating that on agar medium for colony counts. Although an earlier report
that PCR would detect only viable Legionella pneumophila, Josephson rt al. (1993) showed it
could be used for detecting nonviable E. coli, Salmonella typhi and Shigella sonnei in
environmental samples. As Josephson et ul. (1993) have pointed out, care must be taken in
interpreting PCR results; positive amplification products do not mean that target organ-
isms are viable, only that target nucleic acid sequences are present.
Clearly, it cannot be expected that immunological or PCR methods will be developed for
all taxa in indoor air which have some bearing on health. However, there is a real possibility
that the identity of airborne organisms belonging to some difficult genera, such as Penicil-
lium, could be confirmed by reference to mycotoxin,/secondary metabolite and volatile
profiles. Larsen and Frisvad (1994) in whose laboratory secondary metabolite profiles have
been used successfully in chemotaxonomy of penicillia, have pointed the way here. In
a preliminary study of seven common indoor penicillia and Aspergillus cersicolor, these
authors prepared extracts from heavily sporulating cultures on agar plates, and after
clean-up subjected the extracts to analysis by HPLC. When the HPLC traces for two
penicillia isolated from the air spora in houses and grown on Sigma yeast extract-sucrose
(SYES) and wallpaper paste (WP) agar are examined, clear differences are evident (Figs
2 and 3). The isolate of P. polonicum produced the mycotoxins, penicillic acid and
 Sampling indoor environments 389

LC A 225.5 of JTJ1AlSA.D
1000y j_ of JTJlAlGA. D

600: WALL PAPER PASTE RGAR

600:

400:

-200: 2

-400: ii UV 26.710 <

. t
-600: j
. i
-800: 5
: b SIGMA YES RGRR
-1000 - . . . . . . . . . . _ - I - . . I . . -
10 20 30 40
TImr <min.)

Fig. 2. HPLC traces of cultures of Penicihm polonicum (largely conidial) on wallpaper paste agar
and yeast extract sucrose (YES) agar (Larsen and Frisvad, 1994).

LC A 225.5 x of JTJ2AlSR. II
1400 C R 225.5 of JT.TZmlGA. D
z
n
wRu_ PRPER PASTE AGAR
1200

1000

800

600

400

200

-200

-400 SIGMR YES RGAR

-600

10 -20 40
_ 30
Tfme tmln. J I
Fig. 3. HPLC traces of cultures of Penicillium expansum (largely conidial) on wallpaper paste agar
and yeast extract sucrose (YES) agar (Larsen and Frisvad, 1994).

verrucosidin (Fig. 2) whilst P. expansum produced patulin and chaetoglobosin X (Fig. 3). As
seen from the peak areas, the medium on which they were cultured markedly affected the
quantities of individual compounds present. However, the extracts of both SYES and WP
cultures of P. polonicum contained the metabolites cyclopeptin, dehydrocyclopeptin, cyclo-
penal, cyclopenin, viridicatol, 3-methoxyviridicatin, normethylverrucosidin and puberulins,
as well as penicillic acid and verrucosidin. In addition to patulin and chaetoglobosin X, P.
expansum produced citrinin and traces of chaetoglobosin C. While qualitative differences
between the mycotoxin/secondary metabolite profiles of the three strains of each individual
species examined were minor in some cases, large differences were sometimes noted in
others, e.g. P. commune and A. versicolor.
390 B. Flannigan

Table 4. Major identrfied volatile compounds produced by strains of Penrcilhn chr~soyrnum and P. commute
during growth on SYES agar (Larsen and Frisvad, 1994)

P. ch,.l.soyenutn P. coll*mul,<

IBT IBT IBT IBT IBT IBT

Compound 6041 4645 6183 6328 3468 10727

2-methyl-1-propanol ++ + + +++ +++ +t

1-heptene + + +
3-methyl-3-buten- l-01 + + + _
3-methyl-1-butanol ++ + + +
1-pentanol ++ i
1$nonadiene ++ + +
1-octen-3-01 ++ +++ +
monoterpene I _ _ _ ++ ++ ++
3-octanone ++ +7 ++ ++ ++ +++
3-octanol ++ +++ +++ ++ + +++
monoterpene 2 + +
2-methyl-isoborneol + +

Note: + + + , + + + : relatrve amounts (FID peak area) of identified compounds in sample of collected
volatiles; - : absent from sample.

Larsen and Frisvad (1994) also analysed volatile compounds emanating from agar plate
cultures of eight different species. Diffusing volatiles were collected passively on carbon
black in a tube supported on a stainless-steel net under the lid of each petri dish. The
volatiles were desorbed from the carbon black by elution with diethyl ether and the
resulting solutions analysed using a gas chromatograph with flame ionisation detector
GC-FID, and further characterized by mass spectrometry (GC-MS) or Fourier transform
infrared detector (GC-FTIRD). By way of example, the main volatile compounds identified
from isolates of two Penidlium spp. commonly associated with indoor air problems in
Danish buildings during growth on SYES are shown in Table 4. With both penicillia,
growth on SYES gave good qualitative agreement between strains of the same species.
Although not shown, the qualitative agreement between compounds from cultures on SYES
and WP was less good and the quantities of volatiles produced on SYES were generally
larger than on WP (Larsen and Frisvad, 1994). illustrating the well-known effect of
nutritional and environmental factors on production of volatile compounds by microorgan-
isms. Nevertheless, there were marked differences in the major volatiles not only between
these two species, but between all seven penicillia. Since that preliminary study, Larsen and
Frisvad (1995) have confirmed the potential value of volatile profiles from 132 isolates of 25
different terverticillate Penidium taxa, during growth on SYES, for identification. Tax-
ometric analysis showed perfect agreement between results for volatiles and previous
classification based on chemotaxonomy using biosynthetic families of non-volatile second-
ary metabolites.

CONCLUSION

Sampling methods which involve culture will continue to provide valuable information
on the types of organism in indoor air but, for health related investigations, the deficiencies
militate against their use to give quantitative estimates of exposure. At least for epi-
demiological studies, a better measure of exposure than total counts made over extended
time periods may well be fungal biomass.

AcknowledyementPI thank Dr J. David Miller, Agriculture Canada, Ottawa. for permission to quote unpublished
results.
 Sampling indoor environments 391

REFERENCES

Alvarez, A. J.. Buttner, M. P., Toranzos, G. A., Dvorsky, E. A., Toro, A., Heikes, T. B., Mertikas-Pifer, L. E. and
Stetzenbach, L. D. (1994) Use of solid-phase PCR for enhanced detection of airborne microorganisms. Appl.
Environ. Microbial. 60, 374-316.
Andersen, A. A. (1958) New sampler for the collection, sizing, and enumeration of viable airborne particles.
J. Bacterial. 76, 411484.
Brunekreef, B., Dockery, D. W., Speizer, F. E., Ware, J. H., Spengler, J. D. and Ferris, B. G. (1989) Home dampness
and respiratory morbidity in children. Amer. Rev. Respir. Dis. 140, 1363-1367.
Buttner, M. P. and Stetzenbach, L. D. (1993) Monitoring airborne fungal spores in an experimental indoor
environment to evaluate sampling methods and the effects of human activity on air sampling. Appl. Emiron.
Microhiol. 59, 2199226.
Czop, J. K. and Kay, J. (1991) Isolation and characterization of /?-glucan receptors on human mononuclear
phagocytes. J. Exp. Med. 173, 151 l-1520.
Dales, R. E., Burnett, R. and Zwanenburg, H. (1991a) Adverse health effects in adults exposed to home dampness
and molds. Am. Rev. Respir. Dis. 143, 505-509.
Dales, R. E., Zwanenburg, H., Burnett, R. and Franklin, C. A. (1991b) Respiratory health effects of home dampness
and molds among Canadian children. Amer. J. Epidemiol. 134, 196-203.
DeKoster, J. A. and Thorne, P. S. (1995) Bioaerosol concentrations in noncomplaint, complaint, and intervention
homes in the midwest. Amer. Ind. Hyg. Ass. J. 56, 573-580.
Dewey, F. M., Banham. A. H., Priestley, R. A., Martin, B., Hawes, C., Phillips, S. I. and Wareing, P. W. (1993)
Monoclonal antibodies for the detection of spoilage fungi. Int. Biodet. Biodeg. 32, 1277136.
Donald, W. W. and Mirocha, C. J. (1977) Chitin as a measure of fungal growth in stored corn and soybean seed.
Cereal Chem. 54, 466-474.
Flannigan, B. (1992) Indoor microbiological pollutants-sources, species, characterisation and evaluation. In
Chemical, Microbiological, Health and Comfort Aspects of Indoor Air QualityGGate of the Art in SBS (Edited by
Knoppel, H. and Wolkoff, P.), pp. 73398. Kluwer, Dordrecht.
Flannigan, B. and Miller, J. D. (1994) Health implications of fungi in indoor environments-an overview. In Health
Implications of Fungi in Indoor Enoironments (Edited by Samson, R. A., Flannigan, B., Flannigan, M. E.,
Verhoeff. A. P., Adan, 0. C. G. and Hoekstra, E. S.), pp. l-28. Elsevier, Amsterdam.
Flannigan, B., McCabe, E. M. and McGarry, F. (1991) Allergenic and toxigenic micro-organisms in houses.
J. Appl. Bacterial. 6lS 73s.
Flannigan, B., McCabe, E. M., Jupe, S. V. and Jeffrey, I. G. (1993) Mycological and acaralogical investigation of
complaint and non-complaint houses in Scotland. Indoor Air 93, Proc. 6th Int. Conf on Indoor Air Quality and
Climate , pp. 143 -148. Indoor Air 93, Helsinki.
Flannigan, B., Vicars, S., Pasanen, A.-L. and Pasanen, P. (1994) Bioaerosols from housedust. In Health Implications
of Fungi in Indoor Environments (Edited by Samson, R. A., Flannigan, B., Flannigan, M. E., Verhoeff, A. P.,
Adan, 0. C. G. and Hoekstra, E. S.), pp. 65574. Elsevier, Amsterdam.
Flannigan. B., McCabe, E. M. and Jupe, S. V. (1996) Quantification of air- and dust-borne deteriogenic
microorganisms in homes. In Proc. 10th Biodeterioration and Biodegradation Symp. (Edited by Sard, W.),
pp. 3777384. DECHEMA, Frankfurtam Main.
Fogelmark, B., Sjiistrand, M. and Rylander, R. (1994) Pulmonary inflammation induced by repeated inhalations of
(1+3)-p-D-glucan and endotoxin. Int. J. Exp. Pathol. 75, 85-90.
Fox, A. and Rosario, R. M. T. (1994) Quantitation of muramic acid, a marker for bacteria, in dust collected
from hospital and home air conditioning filters using gas chromatography-mass spectrometry. Indoor Air 4,
239-247.
Fox, A., Rosario, R. M. T. and Larsson, L. (1993) Monitoring of bacterial sugars and hydroxy fatty acids in dust
from air conditioners by gas chromatography-mass spectrometry. Appl. Environ. Microbial. 59, 43544360.
Fradkin, A. (1987). Sampling of Microbiological Contaminants in Indoor Air. In Sampling and Calibration for
Atmospheric Meusuremenfs, Special Technical Publication No. 957, pp. 66-77. American Society for Testing and
Materials, Philadephia, PA.
Grant, C., Hunter, C. A., Flannigan, B. and Bravery, A. F. (1989) Water activity requirements of moulds isolated
from domestic dwellings. Int. Biodet. 25, 259-284.
Hunter, C. A., Grant, C.. Flannigan, B. and Bravery, A. F. (1988) Mould in buildings: the air spora of domestic
dwellings. lnt. Biodet. 24, 81-101.
Jaakkola, J. J. K., Jaakkola, N. and Ruotsalainen, R. (1993) Home dampness and molds as determinants of
respiratory symptoms and asthma in preschool children. J. Exp. Anal. Environ. Epidemiol. 3 (Supp. l), 1299142.
Josephson, K. L., Gerba, C. P. and Pepper, I. L. (1993) Polymerase chain reaction detection of nonviable bacterial
pathogens. Appl. Environ. Microbial. 59, 35 13-35 15.
Kitagawa, K., Tsuboi, I., Kimura, S. and Sasamoto, Y. (1991) Rapid method for preparing a ,!?-glucan-specific
sensitive fraction from Limulus (Tachypleus tridentatus) amebocyte lysate. J. Chromat. 567, 267-273.
Kozak, P. P., Gallup, J., Cummins, L. H. and Gillman, S. A. (1979) Currently available methods for home mold
surveys. II. Examples of problem homes surveyed. Ann. Allergy 45, 167-176.
Larsen, T. 0. and Frisvad, J. C. (1994) Production of volatiles and presence of mycotoxins in conidia of common
Penicillia and Aspergilli. In Heulth Implications of Fungi in Indoor Environments (Edited by Samson, R. A.,
Flannigan, B.. Flannigan, M. E.. Verhoeff, A. P., Adan, 0. C. G. and Hoekstra, E. S.), pp. 251-279. Elsevier,
Amsterdam.
Larsen, T. 0. and Frisvad, J. C. (1995) Chemosystematics of Penicillium based on profiles of volatile metabolites.
Mycol. Res. 99, 116771174.
Li, D.-W. and Kendrick, B. (1995) A year-round comparison of fungal spores in indoor and outdoor air. Mycologia
87, 190-195.
Miller, J. D. (1993) Fungi and the building engineer. In Indoor Air 92: Environments for People, pp. 1477158.
ASHRAE, Atlanta.
392 B. Flanmgan

Miller, J. D., Lahamme. A.-M., Sobol. Y., Lafontaine. P. and Greenhalgh. R. (1988) Fungi and fungal products in
some Canadian houses. In?. Biodet. 24, 103 120.
Miller, J. D. and Young, J. C. (1996) The use of ergosterol to measure exposure to fungal propagules in indoor air.
J. Amer. Hyg. Assoc. (in press).
Mouilleseaux, A. and Squinazi. F. (1991) Contamination microbienne de Iair: strategie detude et exemples du
ditIerents environnements. In Proc. Sec. Fr. ilcrohiol.. 3rme Conc/r&s Notional. Institute Pasteur, Paris. (pages not
numbered).
Nevalainen, A., Pastuszka, J.. Liebhaber. F. and Willeke, K. (1992) Performance of bioaerosol samplers: collection
characteristics and sampler design considerations. Atmos. Enoiron. 26, 53 I.
Palmgren, U., Strom, G., Blomquist, G. and Malmberg, P. (1986a) Collection of airborne micro-organisms on
Nuclepore filters. estimation and analysis-mmCAMNEA method. J. Appl. Buctrriol 61, 401.
Palmgren. U., Strom, G.. Malmberg. P. and Blomquist. CT.(1986b) The Nuclepore filter method: a techmque for
enumeration of viable and nonviable airborne micro-organisms. Amer. J. Industr. Med. 10, 325.
Roslansky, P. F. and Novitsky. T. J. (1991) Sensitivity of Linlulm amebocyte lysate to LAL-reactive glucans. J.
C/in. Microhiol. 29, 2477 2483.
Rylander. R. (1995) Respiratory disease caused by bioaerosols~~exposure and diagnosis, Fungi crud Bucteriu in
Indoor Air EnDironmrnts~Hculfll &fects. Detection md Rmedicrtion (Edited by Johanning, E. and Yang. C. S.).
pp. 45 55. Eastern New York Occupational Health Program. Latham, New York.
Rylander, R., Persson, K.. Goto. H., Yuasa, K. and Tanaka, S. (1992) Airborne beta-1,3-glucan may be related to
symptoms in sick buildings. Indoor Ewiron. 1, 2633267.
Samson, R. A., Flannigan, B., Flannigan. M. E., Verhocff, A. P., Adan, 0. C. G. and Hoekstra, E. S. (1994) Heolrlt
1mpliution.s cf Funyi in Indoor Encironments. Elsevier, Amsterdam.
Spengler, J. D., Burge, H. and Su, H. J. (1991) Biological agents and their home environment. In Bu</,s, Molt/ und
Rot, Proc. Workshop on Rrsidentitrl Moisturr Problems. Hecrlth l$ects, Buildiny Damtrge. and Moisture Control.
pp. 1 l-18. National Institute of Building Sciences. Washington, D.C.
Stanevich, R. and Petersen, M. (I 990) Eflect of sampling time on airborne fungal collection. In In&or Air ~0. 5th
lnt. Conf on Indoor Air Qutrlity crnd Climute. Vol. 2, pp. 91-95. CHMC. Ottawa.
Strachan. D. P., Flannigan, B., McCabe, E. M. and McGarry, F. (1990) Quantification of airborne moulds in the
homes of children with and without wheeze Thorcrs 45, 382 387.
Strom, G., Palmgren. U., Wessen. B.. Hellstrom. B. and Kumhns, A. (1990) The sick building syndromee an effect
of microbial growth in building constructions? In It&or Air YO. 5th Int. Conf: on Indoor Air Qucrlitycuzd Clinztrtr,
Vol. I. pp. 173 178. CHMC. Ottawa.
Su, H. J. and Spengler, J. D. (1991) Association of fungal spore concentrations and childhood respiratory health. In
AAAR YI. Tenth Annucrl Mectirzq. Ahstrrrcta, Anwicm Associtrtion fir Aemsol Research, p. 54. AAAR. Bethesda,
MD.
Su, H. J.. Spengler, J. D. and Burge, H. A. (1990) Examination of microbiological concentrations and association
with childhood respiratory health. In Indoor Air YO, 5th Irrt. Corzf on Indoor Air Quality md Clirntrtr, Vol. 2. pp.
21-26. CHMC, Ottawa.
Swift, M. J. (1973) The estimation of mycelial biomass by determination of the hexoaamine content of wood tissue
decayed by fungi. Soil Biol. Biochrm. 5, 321-332.
Verhoeff, A. P., van Wijnen. J. H., Boleij. J. S. M.. Brunekreef. B.. van Reenen-Hoekstra. E. S. and Samson. R. A.
(1990a) Enumeration and identification of airborne viable mould propagules in houses, A//eryy 45, 2755284.
VerhoelT, A. P.. van Wijnen, J. H.. Fischer. P.. Boleij, J. S. M., Brunekreef, B., Boleij, J. S. M.. van Reenen-Hockstra.
E. S. and Samson. R. A. (1990b) Presence of viable mould propagules in the indoor air of houses. Toricol.
Industr. Hlth 6, 133- 145.
Young. J. C. (1955) Microwave-assisted extraction of the fungal metabolites ergosterol and total fatty acids.
J. Aqric. Fd Chrm. 43, 2904 2910.
Zwick. H., Popp. W.. Brown. 0.. Wankc, T. and Wagner. C. (1991) Personal spore sampling and indirect
immunofluorescent test for exploration of hypersensitivity pneumonitis due to mould spores. .4Ikqy 46,
277 283.