Eur J Plant Pathol (2019) 155:1047–1059

https://doi.org/10.1007/s10658-019-01831-x

Multiplex real-time PCR for detection and quantification

of Colletotrichum abscissum and C. gloeosporioides
on Citrus leaves
W. V. Pereira & E. Bertolini & M. Cambra &
N. S. Massola Junior

Accepted: 27 August 2019 / Published online: 30 August 2019

# Koninklijke Nederlandse Planteziektenkundige Vereniging 2019

Abstract Colletotrichum abscissum and sensitivity (98–99.3%) and specificity (94.3–97.1%) for
C. gloeosporioides are the causal agents of citrus the detection and quantification of both pathogens. This
Postbloom Fruit Drop (PFD), a major disease in several technique was 1000 and 10,000 times more sensitive
countries of the American continent. These pathogens than Nested-PCR and PCR, respectively for the detec-
infect only floral structures; however, in the absence of tion of C. abscissum. Similarly, the multiplex qPCR was
floral tissue, they survive asymptomatically on the leaf 100 and 1000 times more sensitive than Nested-PCR
tissue. Thus, detecting and quantifying pathogens on and PCR, respectivel y for the detection of
leaves are critical for epidemiological studies. C. gloeosporioides. The diagnostic parameters used to
Colletotrichum species are detected through isolation, validate the multiplex qPCR showed a high positive
serological methods and conventional PCR, but these (17–35) and low negative (0.006–0.95) likelihood ra-
techniques are not effective for accurate detection of tios, low percentage of false positives (2.8–5.6%) and
PFD agents. This study aimed to develop specific, more false negatives (0.6–1.9%), good agreement between
sensitive, faster and less laborious techniques than the the results generated between qPCR and multiplex
traditional ones routinely used. We developed, standard- qPCR, revealed by the Kappa index (0.91–0.98) and
ized and validated a multiplex real-time PCR with high coincidental results (95–99%). All parameters indicated
that the multiplex qPCR was effectively validated to
W. V. Pereira (*) : E. Bertolini : M. Cambra
detect and quantify the causal agents of PFD.
Center for Plant Protection and Biotechnology, Valencian Institute
for Agricultural Research (IVIA), Carretera Moncada-Náquera, Keywords Citrus sinensis . Postbloom fruit drop . PCR .
Km 5, 46113 Moncada, Valencia, Spain
e-mail: wvpereira@hotmail.com Nested PCR . qPCR

W. V. Pereira
Department of Plant Sciences and Crop Protection, Federal
University of Paraná, R. Funcionários 1540, Curitiba, PR Introduction
80.035-050, Brazil
Postbloom Fruit Drop (PFD) on citrus is caused by
E. Bertolini
Colletotrichum abscissum (Pinho et al. 2015) or
Department of Plant Protection, Faculty of Agronomy, Federal
University of Rio Grande do Sul, Avenida Bento Gonçalves 7712, Colletotrichum gloeosporioides (Lima et al. 2011;
Porto Alegre, RS 91540-000, Brazil McGovern et al. 2012). It occurs in several countries
on the American continent, leading to premature fruit
N. S. M. Junior
drop and compromising citrus production (Denham
Department of Plant Pathology and Nematology, University of
São Paulo/ESALQ, Avenida Pádua Dias 11, Piracicaba, SP 1979; Fagan 1979; Timmer et al. 1994; Silva-Junior
13418-900, Brazil et al. 2014).
1048 Eur J Plant Pathol (2019) 155:1047–1059

PFD mainly affects flower petals, however, it can example, to study pathogens dynamics in different areas
also affect the pistil (stigma and stylet) (Lin et al. to define strategies to control diseases, monitor latent
2001). Symptoms in the petals are characterized by pathogens and as well as the pathogen colonization in
brown or orange lesions, with the presence of acervuli the host tissues (Garrido et al. 2009; Gachon and
containing conidia surrounded by mucilage (Fagan Saindrenan 2004).
1979; Timmer et al. 1994). In the pistil, lesions are In this work, we developed, standardized and vali-
brown or black (Lin et al. 2001). The newly formed dated a multiplex qPCR for detection and quantification
fruits show pale-yellow coloring and drop quickly, of C. abscissum and C. gloeoporioides in asymptomatic
leaving the calyxes and floral disks attached to the citrus leaves to support epidemiological studies and
twigs, forming typical structures of the disease obtain more reliable diagnoses.
(Feichtenberger 1991). Although the leaves do not
show disease symptoms, they are considered source
of inoculum due the presence of appressoria that Material and methods
play a role in the pathogen survival between
flowering periods (Agostini and Timmer 1994). Fungal isolates and DNA purification
As the leaves are source of primary inoculum for
disease development, detection and quantification of Monosporic strains of C. abscissum LFN-1 (Ca142 /
PFD causal agents on leaves surfaces becomes essential FDC142) and C. gloeosporioides LFN-22 (FDC83) and
for epidemiological studies. Traditionally, diagnosis other fungal species often associated with diseases of
techniques have been used to detect and quantify these citrus plants were used in the experiments (Table 1). The
pathogens, such as pathogen isolation in semi-selective strains of C. abscissum and C. gloeosporioides used in
medium, conventional PCR using specific primers and the experiments were previously identified by Silva
serological methods with specific antibodies in the et al. (2017) and Waculicz-Andrade et al. (2017). All
immune-enzymatic test (ELISA) (Mills et al. 1992; strains were preserved in colonized filter paper kept at
Sreenivasaprasad et al. 1996; Ward et al. 2004; −80 °C and deposited in the collection of the Laboratory
Garrido et al. 2009). However, these techniques are of Plant Pathogenic Fungi of the University of São
lengthy and results can take up to 12 days. In addition, Paulo. When necessary, fragments of these paper filters
they are not capable of quantifying the pathogen accu- were transferred to the Potato-Dextrose-Agar (PDA)
rately (Garrido et al. 2009). medium (Difco, Detroit, MI, USA) and incubated at
Pathogen detection by real-time PCR (qPCR) has 25 °C under a photoperiod of 12 h for 7 days for fungal
some advantages over traditional detection techniques cultivation.
that use serology, isolation, and conventional PCR. The The genomic DNA of the strains was obtained by the
qPCR requires no gel electrophoresis after the DNA modified CTAB method (Doyle and Doyle 1990).
amplification, it is more sensitive, specific, faster and Briefly, samples were placed in 1.5 mL microtubes with
less laborious than other techniques traditionally used 400 μL of extraction CTAB buffer (2% CTAB; 100 mM
for detection of plant pathogens (Gachon and Tris-HCl, pH 8.0; 20 mM EDTA; 1.4 M NaCl, 1% PVP,
Saindrenan 2004; Mumford et al. 2006). Due to its 2% β – mercaptoethanol). After stirring for 1 min, the
specificity, qPCR allows accurate quantification of the tubes were kept at 65 °C for 30 min and shaken manu-
target pathogen. Moreover, this technique can simulta- ally every 10 min. The tubes were centrifuged at 855 g
neously amplify multiple target sequences in a single for 5 min. Four hundred microliters of supernatant were
reaction (multiplex real-time PCR). transferred to 2 mL microtubes and then 400 μL of
Multiplex real-time PCR provides advantages over isoamylic alcohol-chloroform (24/1) were added. The
the conventional qPCR by saving time and reagents, tubes were thoroughly agitated manually and centri-
since it combines multiple targets in the same reaction fuged at 18,625×g for 5 min. Two hundred microliters
(McCartney et al. 2003). It also provides higher reliabil- of the supernatant were transferred to 1.5 mL
ity of the results by reducing chances of contamination microtubes that already contained 120 μL of
as well as minimizing pipetting errors (Klein et al. 2000; isopropanol. The tubes were carefully inverted several
McCartney et al. 2003). With the quantitative informa- times and kept at −20 °C for 5 h and then centrifuged at
tion provided by this technique it is possible, for 18,625 g for 20 min. The supernatant was discarded and
Eur J Plant Pathol (2019) 155:1047–1059 1049

1 mL of 70% ethanol was added to the microtube, then conditions used as the second PCR reaction were the
stirred manually by inversion and centrifuged at same as described in the conventional PCR. Water
18,625×g for 10 min. The supernatant was discarded, was used instead of DNA for the control. Products
and the pellet was dried at room temperature for 5 h in of PCR reactions were analyzed by electrophoresis
the laminar flow chamber, followed by resuspension of in 1% agarose gel in 0.5x TBE buffer, stained with
pellet in 50 μL of ultrapure water. The microtubes SYBR Safe™ DNA gel Stain (Invitrogen
containing the extracted DNA were kept at −20 °C. Corporation, Carlsbad, CA, USA) and photographed
in an ultraviolet transilluminator.
Primers, probes and PCR conditions Primers ACUT-F1 and ACUT-R2 coupled with
ACUT-PB TaqMan probe and primers GLOE-F1 and
The primers CaInt2 (Brown et al. 1996) and CgInt GLOE-R1 coupled with GLOE-PB TaqMan probe were
(Mills et al. 1992), from the internal transcribed spacer used in qPCR and multiplex qPCR for detection and
1 (ITS1) region of ribosomal DNA, in conjunction with quantification of C. abscissum and C. gloeosporioides,
ITS4 primer, from the conserved region of the 25/28S respectively (Table 2). These primers and TaqMan
rDNA gene (White et al. 1990), were used in the con- probes for specific detection of pathogens target the
ventional PCR for the amplification of a fragment of the most divergent area of ITS1 region.
ITS region of rDNA (Table 2). The reaction was per- TaqMan probes were labeled at 5′ end position with
formed using 2.5 μL of 10x PCR buffer, 1.5 mM of 6-FAM (6-Carboxyfluorescein) for ACUT-PB probe or
MgCl2, 0.2 mM of a mixture of dNTPs, 0.5 μM of VIC (4, 7, 2′-trichloro-7′ phenyl 6-Carboxyfluorescein)
each primer, 0.04 units of Taq DNA polymerase and for GLOE-PB probe and TAMRA (6-carboxy-
2 μL of DNA sample in a total volume of 25 μL. tetramethyl-rhodamine) at 3′ end position. The multi-
Amplification was carried out for 1 cycle at 94 °C plex qPCR and qPCR were performed using ABI
for 30 s, 62 °C for 45 s and 72 °C for 90 s, followed StepOne™ Real-Time (Applied Biosystems, Foster
by 35 cycles at 94 °C for 30 s, 58 °C for 45 s and City, CA, USA) detection system. The reaction mixture
72 °C for 90 s. After the PCR reaction, the products in a final volume of 12 μL contained: 1x Path-ID™
were separated in 1% agarose gel containing 0.5x qPCR Master Mix (Applied Biosystems, Foster City,
Tris-Borate-EDTA (TBE) buffer stained with SYBR CA, USA), 1 μM of each primer, 100 nM of TaqMan
Safe™ DNA gel Stain (Invitrogen Corporation, probe and 3 μL of DNA sample. The qPCR and multi-
Carlsbad, CA, USA) and DNA bands were visual- plex qPCR protocol consisted of a cycle at 95 °C for
ized under in an ultraviolet transilluminator. 10 min followed by 45 cycles at 95 °C for 15 s and 60 °C
Nested PCR was performed using the ITS4 and ITS5 for 1 min.
primers in the first round for amplification of the ITS
regions 1 and 2, including the 5.8S rDNA. The CaInt2/
ITS4 or CgInt/ITS4 primers for identification of Specificity of primers and probes used in qPCR
C. abscissum or C. gloeosporioides, respectively were
used in the second round of PCR (Table 2). The first The specificity of primers and probes for qPCR was
PCR reaction was performed using 2 μL of 5x PCR assessed using the collection of strains described in
buffer, 2.5 mM of MgCl2, 0.25 μM of each primer, Table 1. All strains were grown in PDA medium
0.2 mM of a mixture of dNTPs, 5 units of Taq DNA (Difco, Detroit, MI, USA) at 25 °C under photoperiod
polymerase and 2 μL of DNA sample in a total volume of 12 h for 7 days. Approximately 100 mg of mycelium
of 25 μL. For the second PCR reaction, 2 μL of the was taken from the colonies and used for DNA extrac-
product of the first reaction were diluted 10 times in tion, as described earlier. The concentration of DNA of
ultrapure water and added with 2.5 μL of 10x PCR each sample was adjusted to 20 ng μL−1 for qPCR. The
buffer, 1.5 mM of MgCl2, 0.2 mM of dNTPs, 0.5 μM qPCR amplifications were carried out in triplicate in a
of each primer and 0.04 units of Taq DNA polymerase. StepOne™ Real-Time thermal cycler (Applied
Amplification of the first reaction was based on an initial Biosystems, Foster City, CA, USA). The experiment
denaturation at 95 °C for 90 s, followed by 35 cycles at was performed twice, with three repetitions per treat-
95 °C for 35 s, 55 °C for 1 min, and 72 °C for 1 min and ment. The mean and standard deviation of each treat-
a final extension at 72 °C for 1 min. The amplification ment was calculated.
1050 Eur J Plant Pathol (2019) 155:1047–1059

Table 1 Specificity of primers and probes for use in Real-Time PCR for detection of Colletotrichum abscissum and Colletotrichum
gloeosporioides

Isolate code Species Host Values of Ct ± SDa

Primers and probe for Primers and probe for

C. abscissum C. gloeosporioides

LFN-1 C. abscissum Citrus sinensis 27.68 ± 0.62 NDb

LFN-2 C. abscissum Citrus sinensis 28.84 ± 0.21 ND
LFN-3 C. abscissum Citrus sinensis 27.48 ± 0.55 ND
LFN-4 C. abscissum Citrus sinensis 27.41 ± 0.69 ND
LFN-5 C. abscissum Citrus sinensis 28.24 ± 0.15 ND
LFN-6 C. abscissum Citrus sinensis 27.33 ± 0.38 ND
LFN-7 C. abscissum Citrus sinensis 27.99 ± 0.88 ND
LFN-8 C. abscissum Citrus sinensis 28.25 ± 0.42 ND
LFN-9 C. abscissum Citrus sinensis 28.11 ± 0.31 ND
LFN-10 C. abscissum Citrus sinensis 27.75 ± 0.18 ND
LFN-11 C. abscissum Citrus sinensis 27.92 ± 0.02 ND
LFN-12 C. abscissum Citrus sinensis 27.84 ± 1.21 ND
LFN-13 C. abscissum Citrus sinensis 28.71 ± 0.88 ND
LFN-14 C. abscissum Citrus sinensis 27.37 ± 1,53 ND
LFN-15 C. abscissum Citrus sinensis 27.78 ± 0.02 ND
LFN-16 C. abscissum Citrus sinensis 28.64 ± 0.43 ND
LFN-17 C. gloeosporioides Citrus latifoli ND 24.45 ± 1.35
LFN-18 C. gloeosporioides Citrus latifoli ND 25.33 ± 0.45
LFN-19 C. gloeosporioides Citrus latifoli ND 24.84 ± 1.72
LFN-20 C. gloeosporioides Citrus latifoli ND 25.54 ± 0.57
LFN-21 C. gloeosporioides Citrus sinensis ND 25.71 ± 0.08
LFN-22 C. gloeosporioides Citrus sinensis ND 24.96 ± 0.93
LFN-23 C. gloeosporioides Citrus sinensis ND 25.14 ± 0.23
LFN-24 C. gloeosporioides Citrus sinensis ND 24.91 ± 0.88
LFN-25 C. gloeosporioides Citrus sinensis ND 25.67 ± 0.62
LFN-26 C. lindemuthianum Phaseolus vulgari ND ND
LFN-27 C. musae Musa paradisiaca ND ND
LFN-28 C. coccodes Capsicum annuum ND ND
LFN-29 C. truncatum Glycine max ND ND
LFN-30 C. capsici Capsicum annuum ND ND
LFN-31 Phyllosticta citricarpa Citrus sinensis ND ND
LFN-32 Alternaria alternata Citrus sinensis ND ND
LFN-33 Phytophthora citri Citrus sinensis ND ND
Controle – – ND ND
a
Values represent the mean Cycle Threshold (Ct) ± standard deviation (SD) for three repetitions of two independent experiments; b ND: not
detected DNA amplification until cycle 44

Comparison of sensitivity between PCR techniques were grown on oatmeal agar at 25 °C under continuous
fluorescent light for 7 days to induce conidial produc-
For comparison of the different PCR techniques, strains tion. The conidia produced were suspended in sterile
LFN-1 (C. abscissum) and LFN-22 (C. gloeosporioides) distilled water to obtain a serial dilution of 10-fold
Eur J Plant Pathol (2019) 155:1047–1059 1051

Table 2 Sequences of primers and probes used in the experiments

Target Primer/Probe Orientation Technique Sequence 5′- 3’ Reference

C. abscissum ACUT-F1 Direct qPCR CGG AGG AAA CCA AAC TCT ATT TAC A Garrido et al. (2009)
ACUT-R2 Reverse qPCR CCA GAA CCA AGA GAT CCG TTG
ACUT-PB Probe qPCR CGT CTC TTC TGA GTG GCA CAA GCA
AAT AAT TAA A
C. gloeosporioides GLO-F1 Direct qPCR GGC GGG TAG GGT CYC CGa Garrido et al. (2009)
GLO-R1 Reverse qPCR ACT CAG AAG AAA CGT CGT TAA ATC AG
GLO-PB Probe qPCR CTC CCG GCC TCC CGC CYCa
b
C. abscissum CaInt2 Direct PCR and GGG GAA GCC TCT CGC GG Brown et al. (1996)
Nested-PCR
C. gloeosporioides CgIntb Direct PCR and GGC CTC CCG CCT CCG GGC GG Mills et al. (1992);
Nested-PCR
ITS1–5·8S-ITS2 ITS4 Reverse PCR and TCC TCC GCT TAT TGA TAT GC White et al. (1990)
Nested-PCR
ITS5 Direct Nested-PCR GGA AGT AAA AGT CGT AAC AAG G
a b
Y = (C + T); Used together with reverse primer ITS4

containing from 1 × 107 to 1 × 101 conidia mL−1. One and specific probes for the detection of C. abscissum or
hundred microliters of each dilution were used for DNA C. gloeosporioides using the same conditions previous-
extraction, as described earlier. The DNA obtained was ly described. The experiment was conducted twice and
used in the conventional PCR, Nested PCR, qPCR and each comprised three repetitions per treatment. The data
multiplex qPCR. The primers, probes and conditions of were subjected to the Kolgomorov-Smirnov test to eval-
the reactions were described previously. The experiment uate the normality and analyzed by the Student’s t test
was carried out twice, with three repetitions of each for independent samples to compare the Ct means be-
dilution. tween samples with or without plant DNA.

Influence of plant DNA on qPCR sensitivity Comparative sensitivity between qPCR and multiplex
for detection of C. abscissum and C. gloeosporioides qPCR for the detection of C. abscissum
and C. gloeosporioides
DNA from healthy leaves of Citrus sinensis and from
strains of both species of Colletotrichum were used to Strains LFN-1 and LFN-22 were cultured in PDA
perform the test. Strains LFN-1 and LFN-22 were (Difco, Detroit, MI, USA) at 25 °C in the dark for 7 d.
grown in PDA medium (Difco, Detroit, MI, USA) at Approximately 100 mg of the mycelium was used for
25 °C in the dark for 7 days. Approximately 100 mg of DNA extraction, as described earlier. The DNA was
the mycelium was used for DNA extraction, as de- quantified in NanoDrop spectrophotometer (Thermo
scribed earlier. The DNA was quantified in NanoDrop Fisher Scientific, Waltham, MA, USA) and adjusted to
spectrophotometer (Thermo Fisher Scientific, Waltham, 20 ng μL−1. Three combinations were used: i) 50 μL of
MA, USA) and adjusted to 25 ng μL−1. A serial dilution DNA of C. abscissum was mixed with 50 μL of DNA of
of 10-fold from 25 ng μL−1 to 250 fg μL−1 of DNA was C. gloeosporioides; ii) 50 μL of DNA of C. abscissum
prepared. Two leaves of a healthy ‘Valencia’ orange tree was mixed with 50 μL of milli-Q water; iii) 50 μL of
were macerated in 10 mL phosphate buffered saline DNA of C. gloeosporioides was mixed with 50 μL of
(PBS) and 100 μL of the extract was used for DNA milli-Q water. The qPCR and multiplex qPCR were
extraction, as described earlier. Then, 50 μL of each performed with these samples, according to the condi-
DNA concentration of each pathogen was deposited tions described earlier. The test was carried out twice,
individually in microtubes the plant DNA at a concen- with three repetitions per treatment. The data were sub-
tration of 20 ng mL−1 or milli-Q water was added. For jected to the Kolgomorov-Smirnov test to evaluate nor-
each combination, qPCR was conducted using primers mality and analyzed by the Student’s t test for
1052 Eur J Plant Pathol (2019) 155:1047–1059

independent samples to compare the Ct means obtained the threshold cycle (Ct) and the logarithm of DNA
by qPCR and multiplex qPCR for each pathogen. concentrations.
In this assay, qPCR and multiplex qPCR were used
Validation of multiplex qPCR for the detection for DNA quantification of C. abscissum and
of C. abscissum and C. gloeosporioides C. gloeosporioides on the surface of citrus leaves.
Three-month-old leaves from a ‘Valencia’ orange tree
In total, 227 leaves from three months old ‘Valencia’ were washed in distilled water to remove any residue on
orange maintained in a greenhouse and 100 leaves of the surface and deposited on moistened filter papers
‘Fino’ lemon trees collected in commercial orchards in inside sterile plastic boxes. Subsequently, the leaves
Murcia, Spain, were used for determining some diag- were inoculated with 1 mL of suspension containing
nostic parameters. The leaves collected in the green- 1 × 105 to 1 × 102 conidia mL−1 of both
house were washed in sterilized distilled water and Colletotrichum species and maintained in plastic boxes
placed on sterile moistened filter paper inside plastic for 72 h in growth chambers at 25 °C under photoperiod
boxes. Subsequently, the leaves were inoculated by of 12 h. Each leaf was individually macerated in extrac-
spraying with a mixture of 500 μL of suspensions tion bags (12 × 15 cm) (Bioreba, Reinach AG,
containing 1 × 104 conidia/mL of strains LFN-1 and Switzerland) using 10 mL of PBS buffer. DNA was
LFN-22 (156 leaves) or sterile distilled water (71 extracted and analyzed by qPCR and multiplex qPCR
leaves). The plastic boxes with inoculated leaves were following the same conditions described previously
placed in growth chambers at 25 °C under a photoperiod using standard quantified dilutions. The experiment
of 12 h during 72 h. All inoculated and orchard- was carried out twice using five repetitions per treat-
collected leaves were packaged individually in plastic ment. The data were subjected to the Kolgomorov-
bags (12 × 15 cm) (Bioreba, Reinach AG, Switzerland) Smirnov test to evaluate normality and analyzed by the
and macerated with 10 mL of PBS buffer. One hundred Student’s t test for independent samples to compare the
microliters of each extract were used for DNA extrac- means of DNA quantity measured by qPCR and multi-
tion. The DNA extracted was analyzed by qPCR and plex qPCR for each pathogen.
multiplex qPCR as described previously.
Sensitivity, specificity, prevalence, false positive
rates, false negative rates, positive predictive values, Results
negative predictive values, positive likelihood ratio
and negative likelihood ratio were determined from the Real-time PCR protocol: Sensitivity and specificity
data obtained from the inoculated leaves. These
parameters were calculated according to Capote et al. The primers and probes used in qPCR were specific for
(2009) and Vidal et al. (2012). In addition, the correla- the detection of C. abscissum and C. gloeosporioides
tion between the data generated by qPCR and multiplex (Table 1). Of the 33 strains analyzed, all 16 strains of
qPCR was measured by the coincidental results and C. abscissum and nine strains of C. gloeosporioides
kappa index (κ) of Cohen (Cohen 1960) for inoculated were detected, with Ct values between 27.3 to 28.8 for
and orchard-collected samples. Cohen’s kappa index C. abscissum and 24.4 to 25.7 for C. gloeosporioides.
indicates the proportion of agreement beyond that ex- No other fungal species examined were detected until
pected by chance. cycle 44 (undetermined Ct values) (Table 1).
Under different inoculum concentration, the qPCR
Quantification of C. abscissum and C. gloeosporioides and multiplex qPCR conducted with primers and spe-
on citrus leaf surfaces cific probes for detection of C. abscissum were 1000 and
10,000 times more sensitive than nested PCR and con-
A standard curve was generated using a 10-fold serial ventional PCR, respectively (Table 3). The lowest de-
dilution (2.5 ng to 2.5 × 10−6 ng for C. abscissum and tectable concentration of C. abscissum by qPCR and
2.5 ng to 2.5 × 10−7 ng for C. gloeosporioides) of geno- multiplex qPCR was 1 × 102 conidia mL−1. Nested PCR
mic DNA obtained from strains LFN-1 and LFN-22, detected C. abscissum from 1 × 105 conidia mL−1 while
respectively. Three qPCR reactions per dilution were conventional PCR detected this species only from 1 ×
conducted to establish linear regression curves between 106 conidia mL−1 (Table 3). Primers and specific probes
Eur J Plant Pathol (2019) 155:1047–1059 1053

Table 3 Comparative sensitivity between Real-Time PCR (qPCR), Multiplex qPCR, Nested-PCR and conventional PCR for detection of
C. abscissum and C. gloeosporioides

Conidia.mL-1

1x107 1x106 1x105 1x104 1x103 1x102 1x101

Multiplex qPCR 21.21a 23.91 25.75 29.45 31.18 34.37 NDb

qPCR 20.54 23.55 25.89 30.62 32.44 34.52 ND

Nested-PCR
Cac

Conventional PCR

Multiplex qPCR 19.10 22.94 26.12 27.12 30.72 32.66 ND

qPCR 19.77 22.45 25.11 28.94 30.23 33.01 ND

Cgd Nested-PCR

Conventional PCR

a b
Values represent the mean Cycle Threshold (Ct) for three repetitions of two independent experiments; ND: not detected DNA
amplification until cycle 44; c Colletotrichum abscissum; d Colletotrichum gloeosporioides

for detection of C. gloeosporioides through qPCR and different concentrations tested, in the presence and ab-
multiplex qPCR were 100 and 1000 times more sensi- sence of DNA from citrus (Table 4). qPCR efficiency for
tive than Nested PCR and conventional PCR, respec- detection of C. abscissum or C. gloeosporioides at dif-
tively. The lowest concentration of C. gloeosporioides ferent DNA concentrations of these pathogens was 97.6
detected by qPCR and multiplex qPCR was 1 × 102 and 98.8%, respectively. In the presence of plant DNA,
conidia mL−1. The tests conducted with Nested PCR qPCR efficiency at different DNA concentrations of
showed detection from 1 × 104 conidia mL−1, while pathogens was 98.8% for C. abscissum and 97.2% for
conventional PCR only detected C. gloeosporioides C. gloeosporioides (Table 4).
from 1 × 105 conidia mL−1 (Table 3). The multiplex qPCR showed specificity and sensi-
The presence of DNA of the host did not interfere bility similar to those of species-specific qPCR for de-
with qPCR efficiency (Table 4). There was no DNA tection of both fungal species (Table 5). In multiplex
amplification of C. abscissum and C. gloeosporioides at qPCR, samples containing only DNA of C. abscissum
1054 Eur J Plant Pathol (2019) 155:1047–1059

Table 4 Real-Time PCR sensitivity for detection of C. abscissum and C. gloeosporioides on presence or absence of DNA from Citrus
sinensis

Values of Ct a

C. abscissum C. gloeosporioides

DNA concentration Without plant DNA With plant DNA P valueb Without plant DNA With plant DNA P value

2.5 ng 17.49 17.83 0.17 15.24 15.36 0.71

250 pg 20.35 20.78 0.30 18.45 18.26 0.66
25.0 pg 25.85 25.61 0.69 21.63 21.84 0.69
2.5 pg 27.03 27.63 0.33 25.63 25.22 0.56
250 fg 31.05 31.20 0.57 28.44 28.83 0.63
Control NDc ND – ND ND –
Efficiency (%) 97.6d 98.8 – 98.8 97.2 –
a b
Values represent the mean Cycle Threshold (Ct) for three repetitions of two independent experiments; Values of P greater than 0.05 in
Student’s t test indicate no significant differences between samples with and without the plant DNA; c ND: not detected DNA amplification
until cycle 44; d Efficiency = 10 (1- slope) -1

or C. gloeosporioides showed only one amplification 0.687 for both species. The sensibility was 0.980 and
curve for detection of these species, with Ct means 28.30 0.993 and specificity 0.943 and 0.971 for C. abscissum
and 25.28, respectively. In samples with DNA mixture and C. gloeosporioides, respectively. The ability of
of both pathogens, two amplification curves were ob- multiplex qPCR to generate false positives and false
tained, one for C. gloeosporioides, with Ct = 25.57 and negatives in the detection of C. abscissum was 0.056
another for C. abscissum with Ct = 27.89. There were no and 0.019, respectively. In the detection of
significant differences between the average Ct generated C. gloeosporioides, the false positives and false
by qPCR and multiplex qPCR within each species ex- negatives values were 0.028 and 0.006, respectively.
amined (Table 5). The predictive positive and negative values for the
detection of both pathogens were above 0.95 and
Validation of multiplex qPCR for the detection 0.98, respectively. The likelihood ratios determined
of Colletotrichum species from positive detection of C. abscissum and
C. gloeosporioides was 17.41 and 35.27, respective-
The calculated values of diagnostic parameters of mul- ly. The negative likelihood ratio determined from
tiplex qPCR for C. abscissum and C. gloeosporioides detection of C. abscissum and C. gloeosporioides
detection are shown in Table 6. The prevalence was was 0.95 and 0.006, respectively (Table 6).

Table 5 Values of Cycle Threshold (Ct) obtained from tests conducted with qPCR or multiplex qPCR for detection of C. abscissum and
C. gloeosporioides

C. abscissum C. gloeosporioides

Sample qPCR qPCR multiplex P valueb qPCR qPCR multiplex P value

LFN22 (C. gloeosporiodes) NDc ND – 25.19a 25.28 0.79

LFN1 (C. abscissum) 28.11 28.30 0.64 ND ND –
LFN1 + LFN22 28.82 27.89 0.10 25.94 25.57 0.49
Control ND ND – ND ND –
a
Values represent the mean Cycle Threshold (Ct) for three repetitions of two independent experiments; b Values of P greater than 0.05 in
Student’s t-test indicate no significant differences between average Ct of the samples amplified by qPCR and multiplex qPCR; c ND: not
detected DNA amplification until cycle 44
Eur J Plant Pathol (2019) 155:1047–1059 1055

Table 6 Parameters assessed for

multiplex qPCR validation from Parameter Leaves with C. abscissum Leaves with C. gloeosporioides
leaves inoculated with
C. abscissum and Prevalence ± SEa 0.687 ± 0.030 0.687 ± 0.030
C. gloeosporioides Sensibility ± SE 0.980 ± 0.009 0.993 ± 0.005
Specificity ± SE 0.943 ± 0.015 0.971 ± 0.011
False positives rate ± SE 0.056 ± 0.015 0.028 ± 0.011
False negatives rate ± SE 0.019 ± 0.009 0.006 ± 0.005
Positive predictive value 0.974 ± 0.010 0.987 ± 0.007
Negative predictive value 0.957 ± 0.013 0.985 ± 0.007
Positive likelihood ratio 17.41 35.27
a
Negative likelihood ratio 0.95 0.006
Standard error

The results obtained from detection of C. abscissum limit of 2.5 fg of DNA. qPCR protocol amplification
and C. gloeosporioides in samples analyzed by qPCR efficiency reached 99% for C. abscissum and 98% for
compared to those generated by multiplex qPCR show C. gloeosporioides.
coincidental results up to 98% for inoculated samples The quantification of DNA of C. abscissum and
and over 96% for orchard collected samples (Table 7). C. gloeosporioides in inoculated citrus leaves by
Similarly, the Kappa index was above 0.96 and 0.91 for qPCR and multiplex qPCR was similar in each conidia
inoculated and orchard collected samples, respectively. concentration inoculated (Table 8).

Quantification of Colletotrichum species on citrus

leaves
Discussion
The standard curve for DNA quantification of
C. abscissum and C. gloeosporioides was obtained from Sensitive and specific techniques are required to detect
three repetitions of each serial dilution of DNA extract- and quantify pathogens in plant tissue for a better un-
ed from isolates LFN-1 (C. abscissum) and LFN-22 derstanding of the population dynamics and asymptom-
(C. gloeosporioides), respectively (Fig. 1). The correla- atic presence of C. abscissum and C. gloeosporioides in
tion between Ct values and values in logarithm of dif- citrus leaves. In this study a multiplex qPCR was devel-
ferent DNA concentrations were linear for both oped for specific and sensitive quantification and detec-
C. abscissum (y = −3.714x + 17.972, R2 = 0.999), with tion of PFD causal agents on citrus.
detection limit of 25 fg of DNA, and C. gloeosporioides Both qPCR and multiplex qPCR provided high spec-
(y = −3.3068x + 16.246, R2 = 0.999), whit detection ificity to detect only isolates of C. abscissum or
C. gloeosporioides in all the samples analyzed. These
analyses were also more sensitive than other techniques
Table 7 Coincidental results (Cr) and Kappa index (Ki) calculat- routinely used to detect these pathogens. Despite differ-
ed from the data generated by qPCR and multiplex qPCR for ences between the detection techniques, qPCR and mul-
detection of C. abscissum and C gloeosporioides
tiplex qPCR showed similar sensitivity to detect patho-
Leaves with C. abscissum Leaves with gens until the concentration of 102 conidia mL−1. Since
C. gloeosporioides we used 100 μL of each suspension of conidia for DNA
extraction, the detection threshold of qPCR was 10
Cr (%) Ki Cr (%) Ki
spores for each pathogen. The detection limit was as
Set 1a 98.65 0.968 ± 0.067c 99.55 0.989 ± 0.066 low as 25 fg of DNA of C. abscissum and 2.5 fg of DNA
Set 2b
97.93 0.955 ± 0.101 96.84 0.912 ± 0.102 of C. gloeosporioides. These detection limits are lower
a
than 25 conidia or 50 pg of DNA of C. acutatum found
Leaves inoculated with 1 × 104 conidia/mL of C. abscissum and
by Debode et al. (2009) for strawberry leaves, 6.2 pg of
C. gloeosporioides; b Leaves obtained from commercial orchards
in the region of Murcia – Spain; c Values represent Ki ± standard DNA of C. coccodes described by Cullen et al. (2002)
error for potato tubers and similar to those at 20 fg of DNA of
1056 Eur J Plant Pathol (2019) 155:1047–1059

Fig. 1 Standard curve for DNA

quantification by real-time PCR
using primers and probes specific
for the detection of C. abscissum
and C. gloeosporioides

C. acutatum reported by Samuelian et al. (2011) for Multiplex real-time qPCR is a promising technique
grapevines. for detection and quantification of plant pathogens, with
Many constituents in the leaf tissue may act as inhib- advantages over other molecular techniques. Multiplex
itors of PCR amplification (Ma and Michailides 2007; qPCR saves on time and consumables, since it combines
Schrader et al. 2012). These inhibitors usually affect the multiple targets in the same reaction, enabling higher
amplification through direct interaction with DNA or productivity, fewer pipetting steps and therefore less risk
DNA polymerase, blocking its activity. The use of a of errors. Amplification of multiple targets can signifi-
PCR control is required to determine the inhibitory cantly increase the accuracy of the comparative analysis
effect of substances in the preparation of nucleic acid. with the improvement of standardization. However,
This is especially necessary to eliminate false negatives, many factors must be taken into account for the estab-
which may lead to total PCR inhibition, even in the lishment of an effective technique and high reproduc-
presence of the target sequence (Tooley et al. 2006). ibility (Klein et al. 2000; Tooley et al. 2006; Atallah
Therefore, this study showed that the presence of plant et al. 2007; Elnifro et al. 2000; Smith and Osborn 2009).
DNA did not alter the sensitivity of qPCR for the Different primers and probes in the same reaction should
detection of the pathogens under study. Similarly, not interact with each other, and the amplicons must
Atallah et al. (2007) observed that DNA amplification share the available reagents, such as primers, DNA
of Verticillium dahliae was not altered in the presence of polymerase and nucleotides, without reducing the effi-
potato DNA. ciency of the reaction. Primer concentrations should be

Table 8 DNA quantification of C. abscissum on C. gloeosporioides on leaves of Citrus sinensis inoculated with different concentrations of
inoculum

Quantity of DNA (pg)

C. abscissum C. gloeosporioides

Concentrations (conidia mL−1) qPCR qPCR multiplex P valueb qPCR qPCR multiplex P-value
5 a
1 × 10 42.1 41.7 0.74 51.9 49.5 0.51
1 × 104 22.4 21.9 0.58 30.1 29.7 0.85
1 × 103 9.1 10.1 0.45 12.4 13.6 0.97
2
1 × 10 3.8 4.3 0.48 5.6 5.9 0.78
a b
Values represent the average quantity of DNA for five repetitions of two independent experiments; Values of P greater than 0.05 in
Student’s t test indicate no significant differences between the average DNA quantity of samples amplified by qPCR and multiplex qPCR
Eur J Plant Pathol (2019) 155:1047–1059 1057

optimized to allow amplification of multiple targets in agreement, respectively (Maclure and Willett 1987).
the same reaction. It also requires a careful choice of The Kappa index values obtained by comparison of data
fluorescent dyes and quenchers to prevent spectral over- generated by qPCR and multiplex qPCR were above
lap between the fluorophores of the labeled probes. In 0.92 for artificially inoculated samples and those col-
addition, it is important to validate multiplex qPCR by lected in commercial orchards, highlighting the excel-
comparing its results with the individual qPCR and then lent concordance between these techniques. Similarly,
check its performance (Wittwer et al. 2001). In this the coincidental results indicated that qPCR and multi-
work, all these factors were taken into account for plex qPCR provided very similar results. There were no
the development of multiplex qPCR. The probes significant differences in DNA quantification of
were marked with fluorophore FAM or VIC, C. abscissum and C. gloeosporioides obtained through
avoiding the spectral overlap between them. qPCR and multiplex qPCR. These results indicated that
Primers, probes and reagents to conduct multiplex combined detection of pathogens in multiplex qPCR did
qPCR were adjusted for the tests. The samples ana- not influence the capacity of the technique to quantify
lyzed by qPCR as well as by multiplex qPCR PFD causal agents in isolation.
showed very similar Ct values with no significant In conclusion, multiplex qPCR proved to be effective
differences among them, showing that multiplex and was validated for detection and quantification of
qPCR was well adjusted for detection and quantifi- C. abscissum and C. gloeosporioides in orange leaves.
cation of C. abscissum and C. gloeosporioides. It is believed that the leaf surface is the source of primary
Sensitivity, specificity, prevalence, false positives, inoculum of PFD. However, these pathogens do not
false negatives, positive predictive values, negative pre- cause any symptom in the leaf, they use this site for their
dictive values, positive likelihood ratio, negative likeli- survival in the period between flowering events and thus
hood ratio, coincidental results and the Kappa index are serve as primary inoculum for disease outbreaks in the
important parameters to know the effectiveness of tech- flowering season. The multiplex qPCR developed in this
niques for detection of plant pathogens (Capote et al. study can be an important tool to monitor population
2009; Vidal et al. 2012). Sensitivity and specificity are dynamics of pathogens in orchards and thus promote the
indicators of the operational capacity of a technique and use of appropriate strategies for disease control. In addi-
provides information on ratios of true positives and true tion, this technique can contribute to the development of
negatives. Our study found that multiplex qPCR devel- many epidemiological studies, since it is able to effec-
oped for detection of C. abscissum and tively detect and quantify the pathogens from samples
C. gloeosporioides was highly sensitive and specific. It collected in the field. It can also be used to investigate
was able to detect assertively more than 98% of the the colonization process of pathogens in the host tissue
samples analyzed. In addition, the technique provided and, because the technique is sensitive, it may be able to
the frequency of false positives and negatives below 5 identify pathogens accurately (Garrido et al. 2009;
and 2%, respectively. The predictive positive and nega- Gachon and Saindrenan 2004).
tive values generated by the multiplex qPCR were
above 97 and 95%, respectively. These results indicated Acknowledgments This research was supported by a grant from
a high likelihood of multiplex qPCR for an accurate São Paulo Research Foundation (FAPESP 2011/11629-1), Nation-
al Council for Scientific and Technological Development (CNPq-
diagnosis. The high positive likelihood ratio indicated 140227/2013-0) and Coordination for the Improvement of Higher
the high probability of obtaining positive results in Education Personnel (CAPES-9810-11-0). We thank Dr. Antonio
samples containing pathogens. On the other hand, the Vicent Civera for allowing the use of the mycology laboratory
low negative likelihood ratio indicated a low probability Valencian Institute for Agricultural Research to conduct some
experiments.
of obtaining a positive result in samples without the
pathogens. The Kappa index is used to compare and
relate the similarity between two techniques. This index
eliminates the influence of chance and is, therefore, a References
good parameter for the validation of a method. (Vidal
et al. 2012). The Kappa index between 0 and 0.20, 0.21– Agostini, J. P., & Timmer, L. W. (1994). Population dynamics and
0.40, 0.41–0.60, 0.61–0.80, 0.81–99 and 1 represent survival of strains of Colletotrichum gloeosporioides on cit-
poor, weak, regular, good, excellent and, perfect rus in Florida. Phytopathology, 84(4), 420–425.
1058 Eur J Plant Pathol (2019) 155:1047–1059

Atallah, Z. K., Bae, J., Jansky, S. H., Rouse, D. I., & Stevenson, W. Ma, Z., & Michailides, T. J. (2007). Approaches for eliminating
R. (2007). Multiplex real-time quantitative PCR to detect and PCR inhibitors and designing PCR primers for the detection
quantify Verticillium dahliae colonization in potato lines that of phytopathogenic fungi. Crop Protection, 26(2), 145–161.
differ in response to Verticillium wilt. Phytopathology, 97(7), Maclure, M., & Willett, W. C. (1987). Misinterpretation and
865–872. misuse of the kappa statistics. American Journal of
Brown, A. E., Sreenivasaprasad, S., & Timmer, L. W. (1996). Epidemiology, 126(2), 161–169.
Molecular characterization of slow-growing orange and key McCartney, H. A., Foster, S. J., Fraaije, B. A., & Ward, E. (2003).
lime anthracnose strains of Colletotrichum from citrus as Molecular diagnostics for fungal plant pathogens. Pest
C. acutatum. Phytopathology, 86(5), 523–527. Management Science, 59(2), 129–142.
Capote, N., Bertolini, E., Olmos, A., Vidal, E., Martínez, M. C., & McGovern, R. J., Seijo, T. E., Hendricks, K., & Roberts, P. D.
Cambra, M. (2009). Direct sample preparation methods for (2012). New report of Colletotrichum gloeosporioides caus-
the detection of Plum pox vírus by real-time RT-PCR. ing postbloom fruit drop on citrus in Bermuda. Canadian
International Microbiology, 12(1), 1–6. Journal of Plant Pathology, 34(2), 187–194.
Cohen, J. (1960). A coefficient of agreement for nominal scales. Mills, P. R., Sreenivasaprasad, S., & Brown, A. E. (1992).
Educational and Psychological Measurement, 20(1), 37–46. Detection and differentiation of Colletotrichum
Cullen, D. W., Lees, A. K., Toth, I. K., & Duncan, J. M. (2002). gloeosporioides isolates using PCR. FEMS Microbiology
Detection of Colletotrichum coccodes from soil and potato Letters, 98, 137–143.
tubers by conventional and quantitative real-time PCR. Plant Mumford, R., Boonham, N., Tomlinson, J., & Barker, I. (2006).
Pathology, 51(3), 281–292. Advances in molecular phytodiagnostics – New solutions for
Debode, J., Van-Hemelrijck, W., Baeyen, S., Creemers, P., old problems. European Journal of Plant Pathology, 116(1),
Heungens, K., & Maes, M. (2009). Quantitative detection 1–19.
and monitoring of Colletotrichum acutatum in strawberry Pinho, D. B., Lopes, U. P., Pereira, O. L., Silveira, A. L., & de
leaves using real-time PCR. Plant Pathology, 58(3), 504– Goes, A. (2015). Colletotrichum abscissum Pinho & O.L.
514. Pereira, sp. nov. Persoonia, 34, 236–237.
Denham, T. G. (1979). Citrus production and premature fruit drop Samuelian, S. K., Greer, L. A., Savocchia, S., & Steel, C. C.
disease in Belize. Pans, 25, 30–36. (2011). Detection and monitoring of Greeneria uvicola and
Doyle, J. J., & Doyle, J. L. (1990). Isolation of plant DNA from Colletotrichum acutatum development on grapevines by
fresh tissue. Focus, 12(1), 13–15. real-time PCR. Plant Disease, 95(3), 298–303.
Elnifro, E. M., Ashshi, A. M., Cooper, R. J., & Klapper, P. E. Schrader, C., Schielke, A., Ellerbroek, L., & Johne, R. (2012).
(2000). Multiplex PCR: Optimization and application in PCR inhibitors – Occurrence, properties and removal.
diagnostic virology. Clinical Microbiology Reviews, 13(4), Journal of Applied Microbiology, 113(5), 1014–1026.
559–570. Silva, A. O., Savi, D. C., Gomes, F. B., Gos, F. M. W. R., Silva, G.
Fagan, H. J. (1979). Postbloom fruit drop, a new disease of citrus J., Jr., & Glienke, C. (2017). Identification of Colletotrichum
associated with a form of Colletotrichum gloeosporioides. species associated with postbloom fruit drop in Brazil
Annals of Applied Biology, 91(1), 13–20. through GAPDH sequencing analysis and multiplex PCR.
Feichtenberger, E. (1991). Queda dos frutos jovens de citros, European Journal of Plant Pathology, 147(4), 731–748.
doença induzida por uma raça virulenta do fungo Silva-Junior, G. J., Spósito, M. B., Marin, D. R., Ribeiro-Junior, P.
Colletotrichum gloeosporioides. Laranja, 12(2), 513–521. J., & Amorim, L. (2014). Spatiotemporal characterization of
citrus postbloom fruit drop in Brazil and its relationship to
Gachon, C., & Saindrenan, P. (2004). Real-time PCR monitoring
pathogen dispersal. Plant Pathology, 63(3), 519–529.
of fungal development in Arabidopsis thaliana infected by
Alternaria brassicicola and Botrytis cinerea. Plant Smith, C. J., & Osborn, A. M. (2009). Advantages and limitations
Physiology and Biochemistry, 42(5), 367–371. of quantitative PCR (Q-PCR)-based approaches in microbial
ecology. FEMS Microbiology Ecology, 67(1), 6–20.
Garrido, C., Carbú, M., Fernández-Acero, F. J., Boonham, N.,
Colyer, A., Cantoral, J. M., & Budge, G. (2009). Sreenivasaprasad, S., Sharada, K., Brown, A. E., & Mills, P. R.
Development of protocols for detection of Colletotrichum (1996). PCR-based detection of Colletotrichum acutatum on
acutatum and monitoring of strawberry anthracnose using strawberry. Plant Pathology, 45(4), 650–655.
real-time PCR. Plant Pathology, 58(1), 43–51. Timmer, L. W., Agostini, J. P., Zitko, S. E., & Zulfiqar, M. (1994).
Postbloom fruit drop of citrus, an increasingly prevalent
Klein, D., Bugl, B., Günzburg, W. H., & Salmons, B. (2000).
Accurate estimation of transduction efficiency necessitates a disease of citrus in the Americas. Plant Disease, 78(4),
329–334.
multiplex real-time PCR. Gene Therapy, 7(6), 458–463.
Tooley, P. W., Martin, F. N., Carras, M. M., & Frederick, R. D.
Lima, W. G., Spósito, M. B., Amorim, L., Gonçalves, F. P., &
(2006). Real-time fluorescent polymerase chain reaction de-
M e l o - d e - F i l h o , P. A . ( 2 0 11 ) . C o l l e t o t r i c h u m
tection of Phytophthora ramorum and Phytophthora
gloeosporioides, a new causal agent of citrus post-bloom
pseudosyringae using mitochondrial gene regions.
fruit drop. European Journal of Plant Pathology, 131(1),
Phytopathology, 96(4), 336–345.
157–165.
Vidal, E., Yokomi, R. K., Moreno, A., Bertolini, E., & Cambra, M.
Lin, Y., Stover, E., Sonoda, R., & Rosskopf, E. (2001). Stigma and
(2012). Calculation of diagnostic parameters of advanced
style necrosis is associated with postbloom fruit drop disease
serological and molecular tissue-print methods for detection
in citrus following artificial inoculation. HortScience, 36(6),
of Citrus tristeza virus: A model for other plant pathogens.
1138.
Phytopathology, 102(1), 114–121.
Eur J Plant Pathol (2019) 155:1047–1059 1059

Waculicz-Andrade, C. E., Savi, D. C., Bini, A. P., Adamoski, D., White, T. J., Bruns, T., Lee, S., & Taylor, J. W. (1990).
Goulin, E. H., Silva, G. J., Jr., Massola, N. S., Jr., Teresawa, Amplification and direct sequencing of fungal ribosomal
L. G., Kava, V., & Glienke, C. (2017). Colletotrichum RNA genes for phylogenetics. In M. A. Innis, D. H.
gloeosporioides sensu stricto: An endophytic species or cit- Gelfand, J. J. Sninsky, & T. J. White (Eds.), PCR protocols:
rus pathogen in Brazil? Australasian Plant Pathology, 46(2), A guide to methods and applications (pp. 315–322). New
191–203. York: Academic Press.
Ward, E., Foster, S. J., Fraaije, B. A., & McCartney, H. A. (2004). Wittwer, C. T., Herrmann, M. G., Gundry, C. N., & Elenitoba-
Plant pathogen diagnostics: Immunological and nucleic acid- Johnson, K. S. (2001). Real-time multiplex PCR assays.
based approaches. Annals of Applied Biology, 145(1), 1–16. Methods, 25(4), 430–442.