RESEARCH ARTICLE

Khodadadi et al., Journal of Medical Microbiology 2017;66:126–133

DOI 10.1099/jmm.0.000426

Utilization of size polymorphism in ITS1 and ITS2 regions for

identification of pathogenic yeast species
Hossein Khodadadi,1 Ladan Karimi,2 Nilufar Jalalizand,3 Hassan Adin3 and Hossein Mirhendi4,*

Abstract
Purpose. Despite the existence of a variety of available yeast-identification strategies, easier and more cost-effective
methods are required for routine use in clinical laboratories. The internal transcribed spacer (ITS) regions of fungal rRNA
genes exhibit variable sizes depending on the yeast species. In the present study, fragment size polymorphism (FSP)
analysis of the ITS1 and ITS2 regions for identification of the clinically most important yeast species was assessed.
Methodology. The ITS1 and ITS2 regions of 190 strains, including isolates of 31 standard strains and 159 clinical isolates, were
separately PCR amplified with two primer sets: ITS1–ITS2 and ITS3–ITS4. PCR products were mixed and the two-band
electrophoretic pattern of each sample was analysed according to the size of the ITS regions as predicted from the GenBank
database.
Results. Using this method and avoiding expensive tools such as sequencing or capillary electrophoresis, we were able to
differentiate nearly all pathogenic yeast species, including Candida albicans, Candida tropicalis, Candida glabrata, Candida
parapsilosis, Candida krusei, Candida guilliermondii, Candida kefyr, Candida lusitaniae, Candida rugosa, Cryptococcus neoformans
and Saccharomyces cerevisiae. The method showed limited discriminatory power to differentiate species of the Candida
parapsilosis complex. Differentiation of Candida albicans and Candida tropicalis needs already identified controls.

Conclusion. FSP method benefits from advantages such as lower cost, higher speed and wider range of species than some
commercial yeast-identification methods. We consider this method as one of the easiest molecular approaches for
identifying a wide range of human pathogenic yeast species, applicable to both diagnostic and epidemiological purposes.

INTRODUCTION PCR [6, 7], probe-based real-time PCR [8, 9] or real-time

In the last decades, the prevalence of yeast infections, particu-
larly in immunocompromised patients, has increased. Correct
yeast identification is important not only due to differences in
the susceptibility of different yeast species to the available anti-
fungal drugs [1] but also for epidemiological purposes [2].
Both phenotypical and genotypical approaches have been
used for yeast identification; however, conventional methods
based on morphological or physiological characteristics are
usually expensive, time consuming and laborious, and they
can only differentiate the more common species [1].
A number of molecular methods have been used to identify
medically important yeast species. These methods include
PCR-based techniques such as sequence analysis of different
parts of ribosomal DNA [3, 4], nested PCR [5], multiplex
PCR using species-specific primers followed by melting
temperature analysis [10, 11], RFLP [12, 13], amplified frag-
ment length polymorphism [14], randomly amplified poly-
morphic DNA [15], microarray techniques [16] and
matrix-assisted laser desorption–ionization time-of-flight
MS [17–19]. Although some of these methods have
improved yeast identification, most of them rely on expen-
sive equipment or complicated post-PCR procedures that
are not readily available in most diagnostic laboratories.
The internal transcribed spacer 1 (ITS1) and ITS2 regions,
flanking the 5.8S ribosomal DNA, are highly useful molecu-
lar targets for differentiating between common pathogenic
yeasts [20]. These regions have been used extensively for
PCR-based detection and identification of yeast pathogens
in a variety of formats. Previous studies have shown that the

Received 5 April 2016; Accepted 5 January 2017

Author affiliations: 1Department of Medical Parasitology and Mycology, School of Medicine, Shiraz University of Medical Sciences, Shiraz, Iran;
2
Dr. Beheshti Hospital, Social Security Organization, Shiraz, Iran; 3Department of Medical Parasitology and Mycology, School of Public Health,
National Institute of Health Research, Tehran University of Medical Sciences, Tehran, Iran; 4Department of Medical Parasitology and Mycology,
School of Medicine, Isfahan University of Medical Sciences, Isfahan, Iran.
*Correspondence: Hossein Mirhendi, mirhendi@tums.ac.ir
Keywords: yeast identification; internal transcribed spacer; polymorphism.
Abbreviations: FSP, fragment size polymorphism; ITS, internal transcribed spacer.

000426 ã 2017 The Authors Downloaded from www.microbiologyresearch.org by

IP: 200.27.72.243
126
On: Thu, 16 Aug 2018 17:22:02
 Khodadadi et al., Journal of Medical Microbiology 2017;66:126–133
ITS1 and ITS2 fragments of most yeast species vary in size,
exhibiting polymorphic electrophoretic patterns that are
useful for species differentiation [3, 7, 21]. However, the size
of one region, by itself, does not have sufficient discrimina-
tory power to differentiate all species; therefore, some spe-
cies remain unrecognized [3, 21].
Our aim in the present study was to apply fragment size poly-
morphism (FSP) for the analysis of the ITS1 and ITS2 regions
in order to differentiate various medically important yeasts.
We differentiated between species by considering both ITS1
and ITS2 fragments using a two-band electrophoretic pattern
as a marker. Firstly, size polymorphisms in the ITS1 and ITS2
regions were investigated by careful analysis of the sequences
in the GenBank database; then, the two-band electrophoretic
patterns of ITS PCR products on conventional agarose gels
were studied. The combined use of these unique patterns
proved helpful with regard to identifying clinical yeast isolates
with a potential to replace expensive and laborious methods
such as sequencing or capillary electrophoresis.
METHODS
For in silico sequence analysis, DNA sequences relevant to
the ITS1 and ITS2 regions of 37 medically important yeast
species were downloaded from GenBank (www.ncbi.nlm.
nih.gov/) (Table 1). These sequences were used as controls
to investigate inter- and intra-species size variation. The
exact size of both ITS regions for each downloaded query
sequence was predicted based on the commonly used fungal
universal primers ITS1 (5¢-TCCGTAGGTGAACCTGCGG-
3¢) and ITS2 (5¢-GCTGCGTTCTTCATCGATGC-3¢),
which amplify the ITS1 region, and the primers ITS3 (5¢-
GCATCGATGAAGAACGCAGC-3¢) and ITS4 (5¢-TCC
TCCGCTTATTGATATGC-3¢), which amplify the ITS2
region [22]. The arithmetic average sizes of the ITS1 and
ITS2 regions for each species and the SD were calculated to
minimize the impact of potentially unreliable data.
Thirty-three reference yeast strains from 21 different species
(Table 2) and 159 clinical isolates were used for FSP experi-
ments in this study. The reference strains came from Teikyo
University Institute of Medical Mycology (TIMM; Tokyo,
Japan), Centraalbureau voor Schimmelcultures (CBS;
Utrecht, the Netherlands) and American Type Culture Col-
lection (ATCC; Manassas, VA, USA). A total of 104 strains
were selected from a collection of yeast strains isolated from
clinical specimens, mostly from superficial infections such
as those from nail, skin or mucosa, and then submitted to
two medical mycology laboratories in Tehran, Iran. All iso-
lates had been preliminarily identified by the previously
described PCR–RFLP method [23]. As the present study
served no epidemiological purpose, some of the more
uncommon yeast species were included. Fifty-five strains
belong to a collection of yeasts isolated from Danish hospi-
talized patients with candidemia (kindly provided by Dr
Maiken Cavling Arendrup, Statens Serum Institute, Copen-
hagen, Denmark).
Downloaded from www.microbiologyresearch.org by
IP: 200.27.72.243
127
On: Thu, 16 Aug 2018 17:22:02
For DNA extraction, a suspension of fresh yeast colonies
was prepared in 200 µl of distilled water and boiled for
10 min in a water bath, centrifuged for 3 min at 2000 g and
the supernatant preserved at !20 " C until use.
The ITS1 and ITS2 regions were simultaneously PCR ampli-
fied in separate reaction tubes with the ITS1–ITS2 and
ITS3–ITS4 primer sets, respectively. Each PCR mixture con-
sisted of 12.5 µl of premix (Ampliqon), 5 pmol of each
primer, 3 µl of extracted template DNA and double distilled
water up to a final reaction volume of 25 µl. The PCR mix-
tures contained identical components for both the ITS1 and
ITS2 reactions, except for the primers. As a negative control,
in each PCR run, water instead of template DNA was used
in one of the tubes. Thermal cycling consisted of an initial
denaturation step at 95 " C for 5 min, 35 cycles of denatur-
ation at 94 " C for 30 s, annealing at 56 " C for 45 s and exten-
sion at 72 " C for 45 s. The reaction was finalized by an
extension step at 72 " C for 7 min.
Three microlitres of ITS1 and ITS2 amplicons for each yeast
sample was mixed, loaded onto 2 % agarose gels and electro-
phoresed in TBE buffer (90 mM Tris, 90 mM boric acid,
2 mM EDTA) at 10 V cm!1 for 2 h. A 100 base pair DNA
ladder (Fermentas) was used for estimating the length of the
amplicons. Gels were stained with 0.5 µg ethidium bromide
ml!1, and the two-band electrophoretic pattern was
observed visually. Yeast species were identified according to
the expected band size obtained from in silico sequence
analysis (Table 1).
A total of 121 yeast isolates identified as Candida albicans,
Candida tropicalis, Candida glabrata or Candida krusei by
the FSP method were re-identified by a chromogenic
method. Briefly, each isolate was inoculated on a CHROMa-
gar Candida (CHROMagar) plate and incubated at 35 " C for
48 h. Yeasts were identified based on the colony colour
according to the manufacturer’s instructions.
Thirty-four clinical isolates that were not identifiable with
CHROMagar Candida were subjected to PCR sequencing.
Briefly, the ITS1–5.8S–ITS2 regions of these yeasts were
amplified using the ITS1 and ITS4 primers as described
above. The PCR products were purified using a PCR purifi-
cation kit (Bioneer) and sequenced by an automated DNA
Sequencer (ABI Prism 3730 Genetic Analyzer; Applied Bio-
systems) using the ABI PRISM BigDye Terminator Cycle
Sequencing Ready Reaction kit (Applied Biosystems).
Sequences were subjected to basic local alignment search
tool-nucleotide analysis, and final species identification was
performed based on the comparison of sequences with rele-
vant sequences deposited in GenBank (www.ncbi.nlm.nih.
gov/BLAST).
RESULTS AND DISCUSSION
In this study, we used FSP across the ITS1 and ITS2 regions
for species identification of medically important yeasts.
Through this approach, yeast species were differentiated
 Khodadadi et al., Journal of Medical Microbiology 2017;66:126–133

Table 1. Size of ITS1 and ITS2 fragments according to the yeast species, calculated from sequences deposited in GenBank

ITS1 ITS2

Species Length Average SD No. of GenBank Length Average SD No. of GenBank

range (nt) length (nt) sequences included range (nt) length (nt) sequences included

Yarrowia lipolytica 139 139.00 0.00 102 239–241 239.58 0.50 96

Candida rugosa 143–146 145.00 0.77 10 273–275 273.90 0.70 11
Candida lusitaniae 146–148 147.43 0.65 58 250–257 254.95 0.71 61
Candida intermedia 149–152 150.46 1.10 13 255–258 257.30 1.10 13
Candida catenulata 152–153 152.04 0.21 21 269–271 270.08 0.41 23
Pichia ohmeri 155–157 156.10 0.39 45 270–272 271.13 0.50 44
Candida pararugosa 163–164 163.84 0.37 11 271–272 271.15 0.37 11
Pichia fermentans 164–165 164.90 0.30 11 306–307 306.45 0.50 11
Candida inconspicua 167–169 168.20 1.09 5 304–308 306.25 1.66 8
Candida krusei 180–184 182.02 0.82 75 346–348 347.07 0.43 70
Candida norvegensis 187–189 188.07 0.75 13 324–327 325.85 1.09 14
Cryptococcus 201 201.00 0.00 48 374 374.00 0.00 53
neoformans
Cryptococcus gattii 200–202 201.00 0.16 74 374–376 374.92 0.31 67
Trichosporon asahii 202–204 203.00 0.35 33 357–358 357.91 0.28 36
Candida viswanathii 212–213 212.66 0.57 3 319–320 319.40 0.54 5
Candida tropicalis 217–219 217.96 0.37 101 325–329 327.28 1.05 107
Candida albicans 218–220 218.55 0.58 242 337–339 338.00 0.39 240
Candida dubliniensis 217–218 217.96 0.19 77 341–343 342.73 0.55 88
Candida orthopsilosis 219–221 220.04 0.52 44 310–312 311.16 0.54 53
Candida utilis 219–222 220.45 0.80 11 364–365 364.18 0.40 11
Candida parapsilosis 228–231 229.02 0.30 206 309–312 311.00 0.43 230
Candida freyschussii 231 231.00 0.00 1 351 351.00 0.00 1
Rhodotorula rubra 231–233 231.95 0.40 100 403–405 403.98 0.37 95
Candida metapsilosis 235–237 236.07 0.52 40 314–315 314.90 0.30 40
Candida guilliermondii 245–249 247.76 0.86 197 377–380 378.79 0.69 172
Candida pelliculosa 261–263 262.37 0.57 69 374–375 374.91 0.28 78
Pichia veronae 265 265.00 0.00 2 371 371.00 0.00 2
Pichia fabianii 270–271 270.14 0.36 14 372 372.00 0.00 13
Candida zeylanoides 271–272 271.88 0.32 34 373–374 373.94 0.2 38
Candida famata 278 278.00 0.00 2 381 381.00 0.00 1
Candida kefyr 306–312 308.98 0.77 52 428–434 431.92 0.76 55
Candida nivariensis 360–362 361.54 0.76 25 417–418 417.91 0.28 12
Candida colliculosa 371–374 373 0.35 31 443–446 445.00 0.45 26
Candida bracarensis 385 385.00 0.00 3 440 440.00 0.00 4
Saccharomyces 440–442 440.98 0.46 81 418–422 420 0.64 89
cerevisiae
Candida glabrata 481–484 482.28 0.65 46 417–419 418.58 0.78 50
Schizosaccharomyces 500–501 500.10 0.38 7 484–487 486.51 1.13 7
pombe

using a one-step PCR without any complementary proce- combination of both fragments, two-band patterns related
dures such as sequencing, RFLP or application of probe. to the 37 yeast species were predicted, as shown in Fig. 1.
After establishment of an accurate FSP database, it was con-
A total of 1907 ITS1 and 1935 ITS2 sequences related to 37
firmed that most of the 37 yeast species included in this
different clinically important yeast species were extracted
study had specific FSP patterns (Fig. 1).
from GenBank. All sequences were analysed in order to
investigate similarities or differences in the ITS1 and ITS2 The average size of ITS1 ranged from 139 to 500 nucleotides
region sizes within and between species. Calculated arith- (nt), whereas the average size of ITS2 ranged from 240 to
metic rounded average sizes of the ITS1 and ITS2 regions 486 nt. These lowest and highest sizes correspond to Yarro-
for each species are shown in Table 1. Based on the wia lipolytica (Candida lipolytica) and Schizosaccharomyces
Downloaded from www.microbiologyresearch.org by
IP: 200.27.72.243
128
On: Thu, 16 Aug 2018 17:22:02
 Khodadadi et al., Journal of Medical Microbiology 2017;66:126–133

Table 2. Reference strains used in the FSP assays sequences (Table 1). A low SD among the downloaded ITS1
Strain Reference no.
or ITS2 sequence lengths indicates the reliability of calcu-
lated average sizes.
Candida albicans ATCC 10261, ATCC 10231, ATCC 90029,
ATCC 24432, ATCC 90028, TIMM 1768 The lengths of ITS1 fragments were smaller than those of
Candida glabrata ATCC 90030, CBS 138 the ITS2 ones in all studied yeasts except Saccharomyces cer-
Candida tropicalis ATCC 0750, TIMM 0313 evisiae, Schizosaccharomyces pombe and Candida glabrata.
Candida krusei ATCC 6258, TIMM 3404 Identical ITS1 fragment sizes were seen between Cryptococ-
Candida parapsilosis ATCC 22019, ATCC 90 018 cus neoformans and Cryptococcus gattii (mean±SD of 201
Candida guilliermondii ATCC 9058, TIMM 3400, TIMM 0257 ±0.0 and 201±0.16 nt, respectively), Candida albicans, Can-
Candida dubliniensis CBS 2747 dida dubliniensis and Candida tropicalis (218.55±0.58,
Candida viswanathii ATCC 22 981 217.96±0.19 and 217.96±0.37 nt, respectively), and Candida
Candida kefyr TIMM 0300 orthopsilosis and Candida utilis (220.04±0.52 and 220.45
Candida lusitaniae TIMM 1439 ±0.80 nt, respectively) (Table 1). On the other hand, identi-
Candida rugosa TIMM 1814 cal ITS2 fragment sizes were seen between Pichia ohmeri
Candida inconspicua CBS 180 and Candida pararugosa (271.13±0.50 and 271.15±0.37 nt,
Candida pelliculosa ATCC 8168 respectively), Pichia fermentans and Candida inconspicua
Candida intermedia ATCC 14 439 (306.45±0.50 and 306.25±1.66 nt, respectively), Cryptococ-
Candida norvegensis Ci cus neoformans, Cryptococcus gattii and Candida zeyla-
Candida famata ATCC 20 755 noides (374.00±0.00, 374.92±0.31 and 373.94±0.2 nt,
Trichosporon asahii TIMM 3140, TIMM 3411 respectively), and Candida parapsilosis and Candida orthop-
Cryptococcus neoformans ATCC 90 113
silosis (311.00±0.43 and 311.16±0.54 nt, respectively)
Cryptococcus gattii Ci
(Table 1).
Saccharomyces cerevisiae ATCC 9763 Most pathogenic yeast species differ in size across at least
Schizosaccharomyces pombe Ci one of the ITS1 and ITS2 regions; however, the size of such
Ci, Clinical yeast isolates confirmed by PCR sequencing.
regions alone was nearly identical across some species.
Therefore, considering only one region does not provide
sufficient discriminatory power to differentiate between tax-
onomically related species [7, 24]. Although usually a wide
pombe, respectively (Fig. 1, Table 1). The SD ranged from 0 range of lengths (139–500 nt) were seen in the ITS1 region
to 1.1 nt for ITS1 and from 0 to 1.66 nt for ITS2. SD lower of the studied yeasts, there was little difference between
than 1.1 was registered among average fragment sizes for all some species. The size of the ITS2 region ranged between
species except for Candida inconspicua, this with a SD of 240 and 486 nt. While Candida rugosa and Candida lusita-
1.66. A low SD could be due to the low number of available niae differ by only 2 nt across ITS1, they differ by 19 nt

600

500
482
500
445 440 441
432
418 419 486
404
374 375 379 375 371 372 374 381 373
385
400 358 364 362
347 351 420
343 338
326 319 327
Nucleotide size

311 311 315 309

306 306
300 274 270 271 271 272 278
255 257 262 265 270
240 248
229 231 232 236 ITS1
213 218 218 219 220 220
201 201 203
182 188 ITS2
200 164 165 168
152 156
139 145 147 150
Crypto. neoformans

Trichosporon asahii

Rhodotorula rubra

C. guilliermondii
C. orthopsilosis

C. metapsilosis
C. norvegensis
C. inconspicua

C. dubliniensis

Schizo. pombe
C. parapsilosis

C. zeylanoides
C. freyschussii

Pichia veronae

C. bracarensis
C. pararugosa

C. viswanathii

Pichia fabianii
P. fermentans

C. pelliculosa

C. colliculosa
C. nivariensis
C. intermedia

C. catenulata

S. cerevisiae
Crypto. gattii
C. lusitaniae

C. tropicalis

C. albicans

100
C. glabrata
Y. lipolitica

C. rugosa

C. famata
P. ohmeri

C. krusei

C. kefyr
C. utilis

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Fig. 1. Schematic two-band electrophoretic patterns of yeast ITS1 and ITS2 fragments according to in silico analysis of GenBank
sequences.

Downloaded from www.microbiologyresearch.org by

IP: 200.27.72.243
129
On: Thu, 16 Aug 2018 17:22:02
 Khodadadi et al., Journal of Medical Microbiology 2017;66:126–133

2000 bp
2000 bp
1000 bp
1000 bp

500 bp
500 bp
400 bp
300 bp 400 bp
200 bp 300 bp
100 bp
200 bp

100 bp

Fig. 2. Agarose gel electrophoresis of mixed ITS1 and ITS2 PCR amplicons from common pathogenic Candida species. Left: Candida
intermedia (lane 1), Candida norvegensis (lane 2), Candida inconspicua (lane 3), Candida viswanathii (lane 4), Cryptococcus neoformans
(lane 5), Cryptococcus gattii (lane 6), T. asahii (lane 7), Geotrichum candidum (lane 8), Saccharomyces cerevisiae (lane 9), Candida pellicu-
losa (lane 10) and Schizosaccharomyces pombe (lane 11). Right: Candida albicans (lane 1), Candida dubliniensis (lane 2), Candida tropicalis
(lane 3), Candida glabrata (lane 4), Candida parapsilosis (lane 5), Candida krusei (lane 6), Candida guilliermondii (lane 7), Candida kefyr
(lane 8), Candida lusitaniae (lane 9) and Candida rugosa (lane 10). Lanes M correspond to a 100 bp molecular size marker.

across ITS2. Likewise, Candida krusei and Candida norve-
gensis differed by 6 nt in the ITS1 region; meanwhile, there
was a 78 nt difference in the ITS2 region. The ITS1 regions
in Cryptococcus neoformans and Trichosporon asahii differ
by only 2 nt, while ITS2 regions differ by 16 nt. Unfortu-
nately, Candida albicans/Candida dubliniensis or Cryptococ-
cus neoformans/Cryptococcus gattii were not distinguishable
using this approach. Also, Candida albicans and Candida
tropicalis exhibited very similar patterns (10 nt difference
across ITS2). Therefore, differentiation of these two species
is recommended by comparing them with already identified
controls. Moreover, due to similar sizes of ITS1 and ITS2 in
members of the Candida parapsilosis complex (Candida
parapsilosis, Candida metapsilosis and Candida orthopsilo-
sis), it appears that differentiation of these closely related
species is not straightforward by normal agarose gel electro-
phoresis. Data indicate that, in pathogenic yeasts, ITS1
regions exhibit higher size diversity than ITS2 regions
(Table 1, Fig. 1).
In this study, we established a FSP database (Fig. 1) as a qual-
ity-controlled reference, comprising 37 yeast species. The
database is in agreement with the findings of previous investi-
gations [3, 7, 25]. Agarose gel electrophoresis of PCR products
of the ITS1 and ITS2 regions obtained from reference strains
(Fig. 2) showed that the patterns of the fragments are in

1000 bp

500 bp

400 bp

300 bp

200 bp

100 bp

Fig. 3. Agarose gel electrophoresis of pooled ITS1 and ITS2 PCR products of some clinical isolates. Lanes 1, 7 and 13: Candida glab-
rata; lanes 2, 8 and 14: Candida albicans; lanes 3, 9 and 15: Candida krusei; lanes 4, 10 and 16: Candida parapsilosis; lanes 5, 6, 11, 12
and 17: Candida tropicalis. Lanes M correspond to a 100 bp molecular size marker.

Downloaded from www.microbiologyresearch.org by

IP: 200.27.72.243
130
On: Thu, 16 Aug 2018 17:22:02
 Khodadadi et al., Journal of Medical Microbiology 2017;66:126–133

Table 3. Clinical yeast isolates identified by CHROMagar Candida, FSP and sequencing

FSP CHROMagar Sequencing

Species Identified Not identified Identified Not identified

Candida albicans 48 – 48 – –
Candida glabrata 34 – 34 – –
Candida guilliermondii 1 – – 1 1
Candida inconspicua – 1 – 1 1
Candida kefyr 5 – – 5 5
Candida krusei 5 – 5 – –
Candida lusitaniae 4 – – 4 2
Candida orthopsilosis – 3* – 3 3
Candida parapsilosis 7 1† – 8 8
Candida rugosa 4 – – 4 4
Candida tropicalis 36 1‡ 37 – 1
Candida magnoliae – 1§ – 1 1
Hanseniaspora uvarum – 2|| – 2 2
Pichia fabianii – 1¶ – 1 1
Pichia fermentans – 1# – 1 1
Saccharomyces cerevisiae 3 – – 3 3
Sporidiobolus salmonicolor – 1§ – 1 1
Total 147 (92.5 %) 12 (7.5 %) 124 (78 %) 35 (22 %) 34/34 (100 %)

Sequencing was performed only in cases where FSP and CHROMagar were not concordant or exhibited uncommon and unknown patterns. Although
isolates were selected randomly, uncommon and rare species were included intentionally.
*Identified as Candida parapsilosis.
†Identified as Candida orthopsilosis.
‡Identified as Candida viswanathii.
§Not identified.
||Identified as Candida nivariensis.
¶Identified as Candida famata.
#Identified as Candida inconspicua.

agreement with the predicted two-band patterns obtained Pichia fabianii, P. fermentans and Sporidiobolus salmoni-
from in silico sequence analysis (Table 1, Fig. 1). color. Details on sequence analysis of uncommon yeast
strains have been reported elsewhere [26]. The results of the
A total of 159 clinical yeast isolates were subjected to FSP
identification of clinical isolates by different methods are
analysis. Of these, 150 were identified by the FSP method. summarized in Table 3.
An example of agarose gel electrophoresis of these samples
is shown in Fig. 3. Among them, 121 isolates, including 48 Leaw et al. [3] used size variation among ITS1 and ITS2 for
from Candida albicans, 34 from Candida glabrata, 34 from yeast identification based on PCR sequencing, which is an
Candida tropicalis and 5 from Candida krusei, were sub-cul- accurate but time-consuming and expensive method. The
tured on CHROMagar Candida. According to the colour of FSP method used in the present study is not only suffi-
yeast colonies on this chromogenic medium, all the species ciently accurate for identification of the clinical isolates but
identified by this method were in agreement with those also rapid and inexpensive. Likewise, size polymorphisms of
identified by FSP. However, as CHROMagar Candida is the ITS1 and the ITS1–5.8S–ITS2 regions were previously
designed only for differentiation of the four mentioned spe- used for yeast identification [7]. Even though the approach
cies, it failed to clearly identify other species. Overall, FSP is similar to our FSP method, a large region, such as the
accurately identified 147 out of 159 (92.5 %) clinical yeast ITS1–5.8S–ITS2 region, does not provide the high resolu-
isolates (Table 3). tion provided by, for instance, ITS2. Using the short ITS2
region rather than the ITS1–5.8S–ITS2 region is preferable
Twelve clinical strains were not identified or incorrectly for molecular taxonomic studies, species barcoding and
identified by the FSP method. After sequencing of the ITS1– yeast identification [25, 27]. The mentioned studies, which
5.8S–ITS2 region of these strains, they were identified as focused on the utilization of size polymorphism of the ITS
uncommon species comprising two strains of Hansenias- regions, have shown similar limitations in terms of identify-
pora uvarum, three strains of Candida orthopsilosis and one ing taxonomically related yeast species. For example, they
strain of each of Candida inconspicua, Candida magnoliae, failed to differentiate between species belonging to the genus
Downloaded from www.microbiologyresearch.org by
IP: 200.27.72.243
131
On: Thu, 16 Aug 2018 17:22:02
 Khodadadi et al., Journal of Medical Microbiology 2017;66:126–133
Cryptococcus or between species of the Candida parapsilosis
complex [3, 7, 24, 28]. Our study should be interpreted in
the context of these limitations, plus an insufficient database
for all yeast species.
Our experimental data on reference yeast strains and the
concordance obtained between FSP and CHROMagar for
Candida species confirmed the reliability of the method.
As mentioned before, most species could be identified
based on FSP using conventional gel electrophoresis. If
uncommon yeast species are excluded, the FSP method
can potentially reach a high concordance with a gold
standard method such as PCR sequencing. By optimizing
electrophoresis conditions, such as increasing the assay
time and the agarose concentration, by using gels with
higher resolution, and applying a suitable DNA size
marker, the discriminating power of this method can be
enhanced.
In conclusion, we here presented a rapid and inexpensive
method for DNA-based identification of clinically relevant
yeasts, which uses electrophoresis of mixed amplicons of
ITS1 and ITS2. The method showed limited discrimina-
tory power to differentiate species of the Candida parapsi-
losis complex. Differentiation of Candida albicans and
Candida tropicalis relies on particular precautions such as
appropriate electrophoresis and suitable controls. Never-
theless, the method benefits from advantages such as
lower cost and higher speed than some commercial yeast-
identification methods such as VITEK yeast identification
system (bioM!erieux) and a wider range of species in com-
parison with CHROMagar. We consider this method one
of the easiest molecular approaches for accurately identi-
fying a wide range of human pathogenic yeast species,
useful for both diagnostic and epidemiological purposes.
Funding information
This work was financially supported by Tehran University of Medical
Sciences, Tehran, Iran (grant no. 86-02027-5814).
Acknowledgements
We thank the staff of Statens Serum Institute (Copenhagen, Denmark)
and Teikyo University Institute of Medical Mycology (Tokyo, Japan) for
their help and support. We also thank the Research Consultation Cen-
ter of Shiraz University of Medical Sciences for their invaluable
assistance.
Conflicts of interest
The authors declare that there are no conflicts of interest.
References
1. Pincus DH, Orenga S, Chatellier S. Yeast identification – past,
present, and future methods. Med Mycol 2007;45:97–121.
2. Silva S, Negri M, Henriques M, Oliveira R, Williams DW et al. Can-
dida glabrata, Candida parapsilosis and Candida tropicalis: biology,
epidemiology, pathogenicity and antifungal resistance. FEMS
Microbiol Rev 2012;36:288–305.
3. Leaw SN, Chang HC, Sun HF, Barton R, Bouchara JP et al. Identifi-
cation of medically important yeast species by sequence analysis
of the internal transcribed spacer regions. J Clin Microbiol 2006;
44:693–699.
4. Linton CJ, Borman AM, Cheung G, Holmes AD, Szekely A et al.
Molecular identification of unusual pathogenic yeast isolates by
Downloaded from www.microbiologyresearch.org by
IP: 200.27.72.243
132
On: Thu, 16 Aug 2018 17:22:02
large ribosomal subunit gene sequencing: 2 years of experience
at the United Kingdom mycology reference laboratory. J Clin
Microbiol 2007;45:1152–1158.
5. Taira CL, Okay TS, Delgado AF, Ceccon ME, de Almeida MT et al.
A multiplex nested PCR for the detection and identification of Can-
dida species in blood samples of critically ill paediatric patients.
BMC Infect Dis 2014;14:406.
6. Innings A, Ullberg M, Johansson A, Rubin CJ, Noreus N et al. Mul-
tiplex real-time PCR targeting the RNase P RNA gene for detection
and identification of Candida species in blood. J Clin Microbiol
2007;45:874–880.
7. Fujita SI, Senda Y, Nakaguchi S, Hashimoto T. Multiplex PCR
using internal transcribed spacer 1 and 2 regions for rapid detec-
tion and identification of yeast strains. J Clin Microbiol 2001;39:
3617–3622.
8. Foongladda S, Mongkol N, Petlum P, Chayakulkeeree M. Multi-
probe real-time PCR identification of four common Candida spe-
cies in blood culture broth. Mycopathologia 2014;177:251–261.
9. Maaroufi Y, Heymans C, de Bruyne JM, Duchateau V, Rodriguez-
Villalobos H et al. Rapid detection of Candida albicans in clinical
blood samples by using a TaqMan-based PCR assay. J Clin
Microbiol 2003;41:3293–3298.
10. Nemcova E, Cernochova M, Ruzicka F, Malisova B, Freiberger T
et al. Rapid identification of medically important Candida isolates
using high resolution melting analysis. PLoS One 2015;10:
e0116940.
11. Decat E, van Mechelen E, Saerens B, Vermeulen SJ, Boekhout T
et al. Rapid and accurate identification of isolates of Candida spe-
cies by melting peak and melting curve analysis of the internally
transcribed spacer region 2 fragment (ITS2-MCA). Res Microbiol
2013;164:110–117.
12. Trost A, Graf B, Eucker J, Sezer O, Possinger K et al. Identification
of clinically relevant yeasts by PCR/RFLP. J Microbiol Methods
2004;56:201–211.
13. Mirhendi H, Makimura K, Khoramizadeh M, Yamaguchi H. A
one-enzyme PCR-RFLP assay for identification of six medically
important Candida species. Nihon Ishinkin Gakkai Zasshi 2006;
47:225–229.
14. Borst A, Theelen B, Reinders E, Boekhout T, Fluit AC et al. Use of
amplified fragment length polymorphism analysis to identify med-
ically important Candida spp., including C. dubliniensis. J Clin
Microbiol 2003;41:1357–1362.
15. Bautista-Muñoz C, Boldo XM, Villa-Tanaca L, Hern!andez-
Rodríguez C. Identification of Candida spp. by randomly amplified
polymorphic DNA analysis and differentiation between Candida
albicans and Candida dubliniensis by direct PCR methods. J Clin
Microbiol 2003;41:414–420.
16. Huang A, Li JW, Shen ZQ, Wang XW, Jin M. High-throughput iden-
tification of clinical pathogenic fungi by hybridization to an oligo-
nucleotide microarray. J Clin Microbiol 2006;44:3299–3305.
17. Clark AE, Kaleta EJ, Arora A, Wolk DM. Matrix-assisted laser
desorption ionization-time of flight mass spectrometry: a funda-
mental shift in the routine practice of clinical microbiology. Clin
Microbiol Rev 2013;26:547–603.
18. Pinto A, Halliday C, Zahra M, van Hal S, Olma T et al. Matrix-
assisted laser desorption ionization-time of flight mass spectrom-
etry identification of yeasts is contingent on robust reference
spectra. PLoS One 2011;6:e25712.
19. Ghosh AK, Paul S, Sood P, Rudramurthy SM, Rajbanshi A et al.
Matrix-assisted laser desorption ionization time-of-flight mass
spectrometry for the rapid identification of yeasts causing blood-
stream infections. Clin Microbiol Infect 2015;21:372–378.
20. Schoch CL, Seifert KA, Huhndorf S, Robert V, Spouge JL et al.
Nuclear ribosomal internal transcribed spacer (ITS) region as a
universal DNA barcode marker for fungi. Proc Natl Acad Sci USA
2012;109:6241–6246.
21. Chen YC, Eisner JD, Kattar MM, Rassoulian-Barrett SL, Lafe K
et al. Identification of medically important yeasts using PCR-based
 Khodadadi et al., Journal of Medical Microbiology 2017;66:126–133

detection of DNA sequence polymorphisms in the internal tran-
scribed spacer 2 region of the rRNA genes. J Clin Microbiol 2000;
38:2302–2310.
22. White TJ, Bruns T, Lee S, Taylor JW. Amplification and direct
sequencing of fungal ribosomal RNA genes for phylogenetics. In:
Innis MA, Gelfand DH, Sninsky JJ and White TJ (editors). PCR
Protocols: A Guide to Methods and Applications. New York: Aca-
demic Press, Inc; 1990. pp. 315–322.
23. Mohammadi R, Mirhendi H, Rezaei-Matehkolaei A, Ghahri M,
Shidfar MR et al. Molecular identification and distribution profile
of Candida species isolated from Iranian patients. Med Mycol 2013;
51:657–663.
24. Kiss L. Limits of nuclear ribosomal DNA internal transcribed
spacer (ITS) sequences as species barcodes for Fungi. Proc Natl
Acad Sci USA 2012;109:E1811.
25. Irinyi L, Serena C, Garcia-Hermoso D, Arabatzis M, Desnos-
Ollivier M et al. International Society of Human and Animal
Mycology (ISHAM)-ITS reference DNA barcoding database—
the quality controlled standard tool for routine identification
of human and animal pathogenic fungi. Med Mycol 2015;53:
313–337.
26. Karimi L, Mirhendi H, Khodadadi H, Mohammadi R. Molecular
identification of uncommon clinical yeast species in Iran. Curr Med
Mycol 2015;1:1–6.
27. Schultz J, Wolf M. ITS2 sequence-structure analysis in phyloge-
netics: a how-to manual for molecular systematics. Mol
Phylogenet Evol 2009;52:520–523.
28. Katsu M, Kidd S, Ando A, Moretti-Branchini ML, Mikami Y et al.
The internal transcribed spacers and 5.8S rRNA gene show exten-
sive diversity among isolates of the Cryptococcus neoformans spe-
cies complex. FEMS Yeast Res 2004;4:377–388.

Five reasons to publish your next article with a Microbiology Society journal
1. The Microbiology Society is a not-for-profit organization.
2. We offer fast and rigorous peer review – average time to first decision is 4–6 weeks.
3. Our journals have a global readership with subscriptions held in research institutions around
the world.
4. 80% of our authors rate our submission process as ‘excellent’ or ‘very good’.
5. Your article will be published on an interactive journal platform with advanced metrics.

Find out more and submit your article at microbiologyresearch.org.

Downloaded from www.microbiologyresearch.org by

IP: 200.27.72.243
133
On: Thu, 16 Aug 2018 17:22:02