INFECTION AND IMMUNITY, Aug. 1997, p. 3080–3085 Vol. 65, No.

8
0019-9567/97/$04.0010
Copyright © 1997, American Society for Microbiology

Genetic Diversity among Clinical and Environmental

Isolates of Aspergillus fumigatus
JEAN-PAUL DEBEAUPUIS, JACQUELINE SARFATI, VALÉRIE CHAZALET,
AND JEAN-PAUL LATGÉ*

Laboratoire des Aspergillus, Institut Pasteur, 75724 Paris Cedex 15, France
Received 21 January 1997/Returned for modification 6 April 1997/Accepted 26 May 1997

To determine if cases of invasive aspergillosis (IA) were caused by strains of Aspergillus fumigatus with unique
characteristics, strains from immunosuppressed patients with IA were compared to strains obtained from
sputa of patients with cystic fibrosis and to strains from the environment. An extremely high genomic diversity
was observed among the 879 strains typed by Southern blotting with a retrotransposon-like element from A.
fumigatus (C. Neuvéglise, J. Sarfati, J. P. Latgé, and S. Paris, Nucleic Acids Res. 24:1428–1434, 1996). Analysis
of Southern blot hybridization patterns showed the absence of clustering between environmental isolates and
clinical isolates from patients with IA or cystic fibrosis. In addition, strains could not be clustered depending
on their geographical location. This study implies that practically any strain of A. fumigatus is potentially path-
ogenic and can provoke a case of IA when it encounters a favorable environment in an immunosuppressed host.

Aspergillus fumigatus is a saprobic fungus playing an essential
role in recycling carbon and nitrogen. This fungal species is a
soil inhabitant which colonizes organic debris and decaying
vegetation and releases into the atmosphere a high number of
conidia. Besides its saprobic life, A. fumigatus has also become
one of the most important human pathogens in industrialized
countries. It causes different diseases like allergic bronchopul-
monary aspergillosis, aspergilloma, and invasive aspergillosis
depending on the underlying disease as well as the immuno-
logical status of the host (5, 32). A. fumigatus is today the most
threatening aerial fungal pathogen because nosocomially ac-
quired invasive aspergillosis typically occurs in the treatment
setting for hematological malignancy. The pathogenic behavior
of this opportunistic fungus results from (i) specific biological
features of the fungus, such as the release into the air of a high
concentration of conidia of small size which are easily inhaled
and able to germinate and grow at temperatures higher than
37°C without any specific nutritional requirement, and (ii) the
increase in the use of immunosuppressive therapies which re-
sult in the lack of efficiency of the normal phagocytic host
response (1, 15).
Recent epidemiological studies using molecular techniques
have shown that, at least in some cases, infection occurs as a
consequence of exogenous acquisition of the fungus (26). Two
typing methods have been used until now. The method most
currently used is the random amplification of polymorphic
DNAs (4, 6, 16, 18, 30), although such amplified patterns are
difficult to repeat or interpret (3, 20). In contrast, a restriction
fragment length polymorphism method which uses a species-
specific inactive retrotransposon-like repeated sequence as a
marker has shown high reproducibility and reliability, allowing
computer-aided analysis of a large population of A. fumigatus
strains (11–13, 23). Previous epidemiological studies have in-
vestigated the nosocomial origin of aspergillosis cases in lim-
ited geographical areas and included fewer than 30 strains (6,
13, 16, 18). The present study includes over 800 isolates with
different hosts and geographical origins. The main purpose of
* Corresponding author. Mailing address: Laboratoire des Aspergil-
lus, Institut Pasteur, 28 rue du Docteur Roux, 75724 Paris Cedex 15,
France. Phone: 33 140 61 35 18. Fax: 33 140 61 34 19. E-mail: jplatge
@pasteur.fr.
the study was to investigate the relatedness between clinical
isolates and the known capacity to cause disease in order to
test for the presence of a secondary adaptation of the clinical
isolates to a parasitic condition.
MATERIALS AND METHODS
Isolates. The origins of the strains of A. fumigatus used in this study are given
in Table 1.
All strains from Denmark were isolated from cows with aspergillosis. Clinical
isolates of A. fumigatus were from immunosuppressed patients suffering mainly
from pulmonary or invasive aspergillosis with the exception of the strains from
the Necker and Trousseau hospitals, which were from cystic fibrosis (CF) pa-
tients.
All isolates were kept on 2% malt agar and were eventually stored as a conidial
suspension on dry silica gel (Prolabo) at room temperature (28).
DNA isolation and characterization. Protocols for DNA extraction and EcoRI
digestion were as previously described (11). Digested DNAs were loaded onto
20-cm long 0.8% agarose gels (SeaKem LE; FMC) at a ratio of 500 ng of DNA
per well (4 by 4 mm). Electrophoresis and transfer were done on the same
automated multiblotting device (Mark II; Bertin, France) at 14°C in 0.53 Tris-
borate-EDTA buffer. After 20 min of premigration at 180 V, electrophoresis was
performed for 15 h at 80 V. Transfer of DNA to Pall Biodyne B membranes
lasted 40 min at 250 V. Eight to twelve gels of 15 slots each (12 restriction digests
and 3 size markers) were run simultaneously. Membranes were treated with 0.4
M NaOH, neutralized in 0.5 M Tris-HCl (pH 7.5)–1.5 M NaCl, washed with
water, and fixed at 80°C for 30 min. Southern blot hybridization was performed
with the l3.9 probe containing the recently characterized (23) repeated sequence
Afut1 isolated from A. fumigatus. Hybridization was performed as previously
described (11) with 25 ng of probe DNA labelled with 2.5 ml of [32P]dCTP and
two exposure times of 12 and 24 h with X-Omat film (Kodak) in a cassette
without an intensifying screen.
Analysis of fingerprints. Autoradiographs were scanned with an HP Scanjet
IIcx/T (Hewlett-Packard) driven by a 466DX2/Sp personal computer (IBM). The
resulting digital images (Tagged Image Format File) were converted and nor-
malized with the GelCompar Software (Applied Maths BVBA, Kortrijk, Bel-
gium). Each track corresponding to one slot was converted to a densitometric
curve (700 points with values between 0 and 255, corresponding to the 256 gray
levels). The conversion process included a background subtraction step and
alignment on reference positions determined by size markers (corresponding to
phage l digested by XhoI, BstEII, and BssHII). In addition, an undigested phage
l giving a band of 45 kb was added to each A. fumigatus DNA sample, allowing
for the best alignment at the highest molecular weight bands. An example of a
blot and its computer treatment are shown in Fig. 1. Densitometric profiles were
clustered by the software with a matrix of similarity based on the Pearson
product-moment correlation coefficient (2). As this method compares curves as
a whole, it is independent of band definition and is not troubled by peak-
shoulders mismatches. The clustering method used was the unweighted pair
group method with arithmetic averages.
Groups of genotypic patterns were compared pairwise by principal component
analysis (PCA). Definition of the groups was made a priori and was based on the
geographical location (ranging from a localized area, e.g., one Parisian hospital

3080
VOL. 65, 1997 GENETIC DIVERSITY AMONG A. FUMIGATUS 3081

TABLE 1. Origin of A. fumigatus isolates used for this study

No. of isolates from human
No. of environmental
Country City Hospital or animal sources Provider
isolates
(no. of patients)

Denmark 26 (14a) H. Jensen

France Angers 23 (5) J. P. Bouchara
France Paris Cochin 22 10 (7) J. Dupouy-Camet
France Paris Hôtel-Dieu 85 15 (8) B. Méchin, J. Lortholary
France Paris Necker 65 37 (18) C. Hennequin
France Paris St. Antoine 21 14 (6) J. L. Poirot
France Paris St. Louis 38 53 (35) F. Traoré
France Paris Tenon 9 12 (4) P. Roux
France Paris Trousseau 338 (6) H. Vu Thien
Germany Frankfurt 4 48 (32) P. Shah
Netherlands Nijmegen 23 (6) P. Verweij
United Statesb 36 B. Lasker
Total 280 599 (141)
a
Samples from cows infected by A. fumigatus.
b
Origin not further defined.

to a more general origin, e.g., Europe) and host origin (human versus nonhuman
and clinical versus environmental) of the isolates. PCA is a nonhierarchical
method which allows densitometric data to be used directly without the appli-
cation of any similarity coefficient (29, 31). The arrays of normalized densito-
metric values were used directly as input data for PCA. The three principal axes
were used to produce a three-dimensional representation in which each single
genotype occupied a separate place, shown as a dot. The degree of discrimination
between the groups defined as indicated above was statistically evaluated with
the software. Between each pair of groups, two values were calculated: the mean
percentage relatedness (MPR) of the groups and the ratio of divergence (RD)
(2). MPR is the sum of all Pearson product-moment correlation values (rm) of
the entries of the first group (A) with the entries of the second group (B) divided
by the total number of correlation values calculated: rm,AB 5 SriA jB/nAnB, where
riA jB is the product-moment correlation between the ith and jth entries of groups
A and B, respectively, and nA and nB are the number of entries in groups A and
B, respectively. Within a group (A), this value (rm,A) is reduced to a measure of
internal homogeneity: rm,A 5 Srij/n2.
The ratio of divergence (RD) is calculated as the distance 12rm,AB between
two groups divided by the weighted internal distances 12rm,A and 12rm,B:
~nA 1 nB!~1 2 rm, A B!
RD 5
nA~1 2 rm, A! 1 nB~1 2 rm, B!
The MPR values provide information about the internal homogeneity of groups
and the overall relatedness between groups. Values of MPR higher than 90
indicate very compact groups. In the case of an intragroup analysis, such a value
would indicate that a specific hybridization profile can be associated with all of
the isolates of the group. Values between 70 and 80 indicate an even scattering
of the patterns analyzed inside a group or between groups. MPR data do not give
information about the degree of discrimination between groups. The relative
distance between groups, as revealed by the RD, gives an estimate of the degree
of discrimination between two groups, taking into account their internal homo-
geneity. RD values of 1 to 1.5 indicate that the two clusters of points from the two
groups are superimposed. RD values of greater than 2 (and up to 10 depending
on how compact the groups being compared are) indicate that the cloud of points
belongs to two statistically different groups.
Preliminary experiments on a random sample totalling 5% of the isolates
showed that the same pattern was obtained from 10 single-spore strains isolated

FIG. 1. (A) Southern blot hybridization patterns in a blot with 12 A. fumigatus isolates hybridized with the l3.9 probe. The first, eighth, and fifteenth lanes contain
molecular size markers. Lanes A and B contain identical isolates (99% homology), the isolate in lane C has 90% homology with the isolates in lanes A and B, and the
isolate in lane D has 56% homology with those in lanes A and B. (B and C) Densitometric curves obtained from digitalized images of the blot shown in panel A. Lanes
are as defined above.
3082 DEBEAUPUIS ET AL.
FIG. 2. Dendrogram (unweighted pair group method with arithmetic averages) of the 424 strains analyzed. Note that clinical isolates of A. fumigatus (marked by
vertical lines underneath dendrogram) are widespread throughout the dendrogram.
from the original slant, justifying our decision to consider every slant received as
a single strain. In addition, on .20% of the strains, repeated extraction and
hybridization of DNA samples from the same strain have shown the reproduc-
ibility of the method, with a coefficient of similarity of .95% between two
replicates. Similarly, analysis of 75 identical strains isolated over a 3-year period
from the same CF patient gave a 94.3% similarity coefficient. As a result of this
analysis, genotypic profiles presenting a similarity coefficient of $92% were
always compared individually by superimposition of densitometric curves and by
direct visual examination of the autoradiograms to validate the software analysis
and to confirm the identicalness of strains. Different exposure times were used to
match the intensity levels of the bands and the background and to make sure that
missing bands were not due to too little target DNA. For better accuracy in
pattern comparison, profiles with fuzzy bands were always discarded and DNA
extraction and hybridization were repeated. (Data are not presented in their
entirety due to the considerable number of hybridization patterns analyzed but
are available upon request.)
RESULTS
From the 879 isolates included in this study, clustering anal-
ysis of the Southern blot patterns obtained with the l3.9 probe
gave distinct and unique genotypes for 424 strains (Fig. 2).
The presence of identical genotypes among isolates was due
to our nonrandomized sampling method and corresponded to
particular epidemiological situations: (i) isolates from the
same patients (see below), (ii) environmental isolates from the
same location, and (iii) environmental and clinical isolates
obtained during nosocomial invasive aspergillosis outbreaks
(8). For PCA analysis, only one representative of the genotype
of each series of identical isolates was kept. Among the 424
strains studied, only 14 identities (14 couples) were found
between strains of different origins (for example, a sample
from a patient in Frankfurt, Germany, and a sample from a
U.S. environment). In some cases, the possibility of laboratory
contamination could not be ruled out; in other cases, contam-
ination could be eliminated when, for example, DNA samples
were prepared separately in the United States and in France.
Examination of the dendrogram in Fig. 2 shows that finger-
prints are not grouped in very distinct groups but are in a
number of small groups, indicating a gradual variability be-
tween all strains.
Comparison of clinical and environmental strains. Several
(2 to 15) isolates were obtained per patient for 56 of the 140
patients from which Aspergillus was isolated (Table 2). The
interval between the first and the last sampling dates varied
from 1 day to 6 months for patients with aspergillosis and from
12 to 32 months for patients with CF. The number of strains
isolated from CF patients was much higher than that from
aspergillosis patients. In the CF patient group, no one patient
was colonized by a single strain. Colonization of every CF
patient was due to several genotypes which were repeatedly
isolated from the same patient (data not shown). In contrast, in
INFECT. IMMUN.
the aspergillosis group, the number of genotypes isolated per
patient was close to 1. The recovery of a single genotype per
patient among 18 of the 34 patients studied indicated that in
more than 50% of the patients, infection was due to a single
strain (Table 2).
PCA analysis of clinical and environmental isolates included
328 genotypes from among three groups: (i) 136 strains iso-
lated from 115 neutropenic patients with evidence of Aspergil-
lus infection, (ii) 97 strains obtained from 26 patients with CF
and which apparently were not implicated in an infectious
process, and (iii) 155 randomly chosen environmental strains,
mainly isolated in hospitals.
PCA comparison showed no differences among the three
groups of strains. Figure 3 shows a two-dimensional represen-
tation of the three-dimensional analysis. The absence of clus-
tering was seen at all rotating angles. Pairwise comparisons
gave values of MPR and RD which were not statistically dif-
ferent (Fig. 3, inset). In conclusion, although the clinical iso-
lates were indeed the infective strains, no difference was seen
between the saprobic and clinical isolates. This result means
that every strain present in the environment can become
pathogenic, if it encounters the appropriate host.
Comparison of strains based on their geographical origin.
The 12 groups of strains (a total of 424 strains) gathered on the
basis of their common geographical origin were compared by
the different clustering methods available with the software
(dendrograms not shown) and by PCA. The PCA results
showed that the genetic variability encountered in the species
A. fumigatus was extremely high and that no discrimination
between strains was possible on the basis of their geographical
origin. Figure 4 gives an example of a comparison of strains
from (i) a Parisian hospital, (ii) the Paris area, (iii) Europe,
and (iv) the United States. Similar MPRs and RDs were found
for comparisons between 56 strains from one Parisian hospital
and either a random selection of 244 strains from the rest of
TABLE 2. A. fumigatus isolates obtained from
patients by repeated sampling
No. of No. of Strain/
No. of P1/PT
Patient status isolates per geno- patient
patients ratioa
patient types ratio
Aspergillus infection 34 2.20 46 1.35 18/34
CF 22 17.13 85 4.58 0/26
a
P1, number of patients from whom only one genotype of A. fumigatus was
isolated; PT, total number of patients.
VOL. 65, 1997 GENETIC DIVERSITY AMONG A. FUMIGATUS 3083

FIG. 3. Comparison by PCA of A. fumigatus strains isolated from patients with Aspergillus infections (PA) (■) or CF (p) or from the environment (EN) (h). The
MPR and RD values between the groups are shown in the inset.

the world (MPR, 74.9; RD, 1.3) or 140 strains from France and
160 strains from outside of France (MPR, 73.4; RD, 1.1).
DISCUSSION
This is the first large-scale study of the genetic variability
occurring in the opportunistic fungal pathogen A. fumigatus.
Although this species has been shown to display some pheno-
typic variations (17, 19, 25), A. fumigatus looks taxonomically
homogeneous on the basis of its morphology (27). In contrast,
the intraspecies variability seen at the genomic level looks very
high. As a result, comparison of different genotypes with either
a clinical or an environmental origin showed no difference
between the strains from either origin, suggesting that any

FIG. 4. PCA of A. fumigatus strains from different geographical locations. (A) A total of 62 strains from Hôtel-Dieu, Paris, France (■), versus 200 randomly chosen
strains from other Parisian hospitals (h) (MPR, 76.1; RD, 1.3). (B) A total of 150 randomly chosen strains from Paris (■) versus 124 strains from other origins in Europe
(h) (MPR, 78.9; RD, 1.0). (C) A total of 36 isolates from the United States (■) versus 200 randomly chosen strains from Europe (h) (MPR, 76.4; RD, 1.0).
3084 DEBEAUPUIS ET AL.
environmental strain of A. fumigatus is a putative infectious
strain. This conclusion has some practical implications in that
it indicates that prevention measures should be applied to any
environmental Aspergillus conidia. The absence of strains of
A. fumigatus with different pathogenicity potentials is in agree-
ment with the results of previous biochemical, molecular, and
immunological studies which were unable to identify a key
factor responsible for the pathogenicity of A. fumigatus (15).
For example, until now, no avirulent or hypervirulent mutant
has been found or constructed by genetic manipulation. How-
ever, the failure of this molecular approach may just reflect the
very few proteins purified from A. fumigatus and the low num-
ber of disrupted genes.
The ability of any strain of A. fumigatus to become patho-
genic in contact with an appropriate host can also be linked to
the absence of host specificity. No marked difference was seen
between strains isolated from cows or humans. MPR and RD
values between the two groups of strains were 78.3 and 1.2,
respectively, whereas the MPR values for the intragroup vari-
ability of the 98 strains isolated from humans and the 14 strains
isolated from cow strains were 81.7 and 78.2, respectively. This
is in contrast with true phytopathogenic fungi where patho-
types are genetically different and host-associated pathotypes
can be separated through diverse typing methods on the basis
of their host or geographical origin (21, 22).
All these studies suggest that virulence in A. fumigatus is
multifactorial and that in vivo development requires the acti-
vation of multiple genes. Questions regarding the identity of
the genes expressed and their differential levels of expression
in vivo and in vitro have never been asked seriously until now.
The potential pathogenicity of strains with very different ge-
netic backgrounds is in agreement with the absence of any
selective pressure for this fungus resulting from the accidental
and rare development of A. fumigatus in a human being. This
fungus does not need a human host to complete its biological
cycle, and encountering an immunocompetent human host is
indeed a deadly cul-de-sac for it. A. fumigatus is a true saprobic
fungus which becomes pathogenic only when the human de-
fense reactions are very weakened (for example, in the termi-
nal stage of an AIDS infection or as a result of heavy immu-
nosuppressive therapies). Virulence would seem to represent
more a lack of resistance of the human host than the expres-
sion of specific disease-related proteins. Experimental intra-
nasal infection of mice never succeeds even at the maximal
possible dose (108 conidia per mouse) unless a mouse is im-
munosuppressed (unpublished data).
The question of the origin of the variability encountered in
this species remains open. No teleomorph has been found, and
the possibility that Neosartorya fischeri represents the sexual
stage of A. fumigatus has been definitely discarded (14, 24).
Three hypotheses can be put forward. (i) Continuous genetic
exchange can occur through the parasexual cycle. A. fumigatus
remains today a continuously evolving species. Vegetative
compatible groups have been found in A. fumigatus (9), indi-
cating the possibility of a parasexual cycle in this species. How-
ever, it has never been shown that nonmeiotic parasexual re-
combination can account for recombination in a natural
population of any fungus (7). (ii) The variability could have
been fixed a long time ago through meiotic exchange at a time
when A. fumigatus had a sexual stage. Continuous exchange of
conidia through air currents all around the world, in addition
to the absence of any adaptation to a parasitic condition, would
explain the lack of recovery of identical multilocus genotypes
from geographically and temporally nonassociated hosts (6).
(iii) The third hypothesis concerns the presence of a sexual
stage in A. fumigatus which has gone undetected. Recent pop-
INFECT. IMMUN.
ulation genetic studies have suggested that such a cryptic sex-
ual stage exists in the case of another human pathogen, Coc-
cidioides imitis, which has been found to be almost completely
recombining (7). The presence of such a sexual stage would be
correlated with variations occurring through meiotic recombi-
nation. In a study with different heterokaryon compatibility
groups of Aspergillus nidulans (thus incapable of mitotic re-
combination), genetic variations were found indicating that
heterocaryon groups represent recently diverged lineages that
arose via meiotic recombination (10).
ACKNOWLEDGMENTS
Although they are not coauthors of this report, this study would not
have been possible without the help of the investigators who provided
us with strains from various origins. A special thanks is due to the main
providers: J. P. Bouchara (Laboratoire de Parasitologie-Mycologie,
Angers, France); H. Jensen (The Royal Veterinary and Agricultural
University, Department of Pharmacology and Pathobiology, Copenha-
gen, Denmark); J. Dupouy-Camet (Laboratoire de Parasitologie-My-
cologie, Hôpital Cochin, Paris, France); B. Méchin (Laboratoire de
Mycologie-Parasitologie, Hôtel-Dieu, Paris, France); C. Hennequin
(Laboratoire de Microbiologie, Hôpital Necker, Paris, France);
F. Traoré (Laboratoire de Mycologie, Hôpital Saint-Louis, Paris,
France); G. Brücker, G. Rykner, and J. Lortholary (SEHP, Assistance
Publique, Hôpitaux de Paris, Paris, France); J. L. Poirot (Laboratoire
de Parasitologie, Hôpital Saint-Antoine, Paris, France); P. Roux
(Laboratoire de Parasitologie-Mycologie, Hôpital Tenon, Paris,
France); H. Vu Thien (Laboratoire de Parasitologie-Mycologie,
Hôpital Trousseau, Paris, France); P. Shah (Zim-Infektiologie Klini-
kum der J. W. Goethe, Universität Theodore-Stern-Kai, Frankfurt,
Germany); P. Verweij (Department of Medical Microbiology, Univer-
sity Hospital, Nijmegen, The Netherlands); and B. Lasker (Centers for
Disease Control and Prevention, Division of Bacterial and Mycotic
Diseases, Atlanta, Georgia). We thank J. Taylor for helpful comments
on the manuscript and L. Vauterin (Applied Maths, Kortrijk, Belgium)
for providing data and explanations on the GelCompar software.
Research was partly funded by CNAMTS-INSERM grant 3AM051
to J.-P.L.
REFERENCES
1. Anaissie, E. J., G. P. Bodey, H. Kantarjian, J. Ro, S. E. Vartivarian, R.
Hopfer, J. Hoy, and K. Rolston. 1989. New spectrum of fungal infections in
patients with cancer. Rev. Infect. Dis. 11:369–378.
2. Applied Maths BVBA. 1996. GelCompar manual (version 4.0). Applied
Maths BVBA, Kortrijk, Belgium.
3. Arbeit, R. D., J. N. Maslow, and M. E. Mulligan. 1994. Polymerase chain
reaction-mediated genotyping in microbial epidemiology. Clin. Infect. Dis.
18:1018–1019.
4. Aufauvre-Brown, A., J. Cohen, and D. Holden. 1992. Use of randomly am-
plified polymorphic DNA markers to distinguish isolates of Aspergillus fu-
migatus. J. Clin. Microbiol. 30:2991–2993.
5. Bodey, G. P., and S. Vartivarian. 1989. Aspergillosis. Eur. J. Clin. Microbiol.
Infect. Dis. 8:413–437.
6. Buffington, J., R. Reporter, B. A. Lasker, M. M. McNeil, J. M. Lanson, L. A.
Ross, L. Mascola, and W. R. Jarvis. 1994. Investigation of an epidemic of
invasive aspergillosis: utility of molecular typing with the use of random
amplified polymorphic DNA probes. Pediatr. Infect. Dis. J. 13:386–393.
7. Burt, A., D. A. Carter, G. L. Koenig, T. J. White, and J. W. Taylor. 1996.
Molecular markers and sex in the human pathogen Coccidioides immitis
(Ascomycota). Proc. Natl. Acad. Sci. USA 93:770–773.
8. Chazalet, V., J. P. Debeaupuis, J. Sarfati, J. Lortholary, P. Ribaud, P. Shah,
F. Gluckman, G. Brücker, and J. P. Latgé. Unpublished data.
9. Chazalet, V., and J. P. Latgé. Unpublished data.
10. Geiser, D. M., M. L. Arnold, and W. E. Timberlake. 1994. Sexual origins of
British Aspergillus isolates. Proc. Natl. Acad. Sci. USA 91:1–4.
11. Girardin, H., J. P. Latgé, T. Srikantha, B. Morrow, and D. R. Soll. 1993.
Development of DNA probes for fingerprinting Aspergillus fumigatus. J. Clin.
Microbiol. 31:1547–1554.
12. Girardin, H., J. Sarfati, H. Kobayashi, J. P. Bouchara, and J. P. Latgé. 1994.
Use of DNA moderately repetitive sequence to type Aspergillus fumigatus
isolates from aspergilloma patients. J. Infect. Dis. 169:683–685.
13. Girardin, H., J. Sarfati, F. Traoré, J. Dupouy-Camet, F. Derouin, and J. P.
Latgé. 1994. Molecular epidemiology of nosocomial invasive aspergillosis.
J. Clin. Microbiol. 32:684–690.
14. Girardin, H., M. Monod, and J. P. Latgé. 1995. Molecular characterization
VOL. 65, 1997
of the food-borne fungus Neosartorya fischeri (Malloch and Cain). Appl.
Environ. Microbiol. 61:1378–1383.
15. Latgé, J. P., S. Paris, J. Sarfati, J. P. Debeaupuis, A. Beauvais, K. Jaton-
Ogay, and M. Monod. 1997. Infectivity of Aspergillus fumigatus, p. 99–110. In
H. Vanden Bossche, D. A. Stevens, and F. C. Odds (ed.), Host-fungus
interplay. National Foundation for Infectious Diseases, Bethesda, Md.
16. Leenders, A., A. Van Belkum, S. Janssen, S. De Marie, J. Kluytmans, J.
Wielenga, B. Löwenberg, and H. Verbrugh. 1996. Molecular epidemiology of
apparent outbreak of invasive aspergillosis in a hematology ward. J. Clin.
Microbiol. 34:345–351.
17. Leslie, C. E., B. Flannigan, and L. J. R. Milne. 1988. Morphological studies
on clinical isolates of Aspergillus fumigatus. J. Med. Vet. Mycol. 26:335–341.
18. Lin, D. M., P. F. Lehmann, B. H. Hamory, A. A. Padhye, E. Durry, R. W.
Pinner, and B. A. Lasker. 1995. Comparison of three typing methods for
clinical and environmental isolates of Aspergillus fumigatus. J. Clin. Micro-
biol. 33:1596–1601.
19. Loudon, K. W., J. P. Burnie, A. P. Coke, and R. C. Matthews. 1993. Appli-
cation of polymerase chain reaction to fingerprinting Aspergillus fumigatus by
random amplification of polymorphic DNA. J. Clin. Microbiol. 31:1117–
1122.
20. Loudon, K. W., A. P. Coke, and J. P. Burnie. 1995. Pseudoclusters and typing
by random amplification of polymorphic DNA of Aspergillus fumigatus.
J. Clin. Pathol. 48:183–184.
21. Mess, J. J., J. Van Doorn, E. J. A. Roebroeck, E. Van Egmond, J. Van
Aartrijk, and P. M. Boonekamp. 1994. Restriction fragment length polymor-
phisms, races and vegetative compatibility groups within a worldwide collec-
tion of Fusarium oxysporum f. sp. gladoidi. Plant Pathol. 43:362–370.
Editor: T. R. Kozel
GENETIC DIVERSITY AMONG A. FUMIGATUS 3085
22. Mouyna, I., J. L. Renard, and Y. Brygoo. 1996. DNA polymorphism among
Fusarium oxysporum f. sp. elaeidis populations from oil palm, using a re-
peated and dispersed sequence “Palm.” Curr. Genet. 30:174–180.
23. Neuvéglise, C., J. Sarfati, J. P. Latgé, and S. Paris. 1996. Afut1, a retrotrans-
poson-like element from Aspergillus fumigatus. Nucleic Acids Res. 24:1428–
1434.
24. Peterson, S. W. 1992. Neosartorya pseudofischeri sp. nov. and its relationship
to other species in Aspergillus section Fumigati. Mycol. Res. 96:547–554.
25. Rinyu, E., J. Varga, and L. Ferenczy. 1995. Phenotypic and genotypic analysis
of variability in Aspergillus fumigatus. J. Clin. Microbiol. 33:2567–2575.
26. Rogers, T. R. 1995. Epidemiology and control of nosocomial fungal infec-
tions. Curr. Opin. Infect. Dis. 8:287–290.
27. Samson, R. A., A. D. Hocking, J. I. Pitt, and A. D. King (ed.). 1992. Modern
methods in food mycology, vol. 31. Elsevier, Amsterdam, The Netherlands.
28. Smith, D., and A. H. S. Onions. 1983. A comparison of some preservation
techniques for fungi. Trans. Br. Mycol. Soc. 81:535–540.
29. Sneath, P. H. A., and R. R. Sokal. 1973. Numerical taxonomy. W. H. Free-
man & Co, San Francisco, Calif.
30. Van Belkum, A., W. G. V. Quint, B. E. De Pauw, W. J. G. Melchers, and J. F.
Meis. 1993. Typing of Aspergillus species and Aspergillus fumigatus isolates by
interrepeat polymerase chain reaction. J. Clin. Microbiol. 31:2502–2505.
31. Vauterin, L., and P. Vauterin. 1992. Computer-aided objective comparison
of electrophoresis patterns for grouping and identification of microorgan-
isms. Eur. Microbiol. 1:37–41.
32. Walsh, T. J., and D. M. Dixon. 1989. Nosocomial aspergillosis: environmen-
tal microbiology, hospital epidemiology, diagnosis and treatment. Eur. J.
Epidemiol. 5:131–142.