SEARCH

SEARCH BY CITATION

Keywords:

  • obligate biotrophic pathogen;
  • quantitative PCR;
  • host–parasite interaction

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

Quantification of obligate biotrophic parasites has been a long-standing problem in plant pathology. Many attempts have been made to determine how much of a pathogen is present in infected plant tissue. Methods of quantification included scoring disease symptoms, microscopic evaluation, determination of specific compounds like Ergosterol, and lately nucleic acid-based technologies. All of these methods have their drawbacks, and even real-time PCR may not be quantitative if for example the organism of interest has specific and differing numbers of nuclei in different infection structures. We applied reverse transcription (RT) real-time PCR to quantify Uromyces fabae within its host plant Vicia faba. We used three different genes, which have been shown to be constitutively expressed. Our analyses show an exponential increase of fungal material between 4 and 9 days post inoculation and thereafter reaching a steady state of around 45% of total RNA. We also used haustorium-specific genes to determine the amount of haustoria present at each time point. These analyses parallel the development of the whole fungus with the exception of the steady-state level, which is only around 5% of the total RNA. This indicates that RT real-time PCR is a suitable method for quantification of obligate biotrophic parasites, and also for the differentiation of developmental stages.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

All higher organisms exhibit a more or less pronounced association with a plethora of symbiotic microorganisms, some of them beneficial, some of them neutral, and some of them pathogenic. While the determination of the number of mutualistic or neutral symbionts has more of an academic value, accurate quantification of pathogen abundance is a critical issue in medicine and plant pathology. There have been numerous approaches to quantify the number of pathogens present in various host–parasite interactions at any given time point of pathogenesis. Traditionally, visual inspection and scoring of disease symptoms have been used to determine disease severity (Pei et al., 2002; Bock et al., 2008). Lately, this type of rating has been complemented by digital image analysis (Bock et al., 2008). However, because of differences in the level of susceptibility of the host plant, and of different degrees of virulence of the pathogen, disease severity may not necessarily correlate with the amount of pathogen present. Histological analysis of the pathogen within diseased tissue is another way to determine pathogen abundance (Laurans & Pilate, 1999). Light microscopic methods are often used in combination with specific stains (Tisserant et al., 1993). However, light microscopical analysis is only feasible for filamentous microorganisms like fungi and oomycetes, while bacterial or viral pathogens elude such methods. Immunological techniques, such as ELISA, have been used, but they require the production of an epitope-specific antiserum (Boyle et al., 2005). Another method is the biochemical quantification of microorganism-specific compounds, like for example Ergosterol, a cell membrane sterol found only in higher fungi (Osswald et al., 1986; Gessner et al., 1991; Manter et al., 2001). However, Ergosterol cannot be used to discriminate between different fungal species – this may be relevant when plants harbor two different pathogens or a pathogen and a fungal symbiont, and there may be differences in Ergosterol content during different developmental stages of a single pathogen (Winton et al., 2003). Lately, nucleic acid-based technologies have found entry into plant pathology (Vincelli & Tisserat, 2008). Nucleic acid-based detection methods, particularly those that rely on PCR, typically are rapid, specific, and highly sensitive (Vincelli & Tisserat, 2008). Today real-time PCR detection and identification of pathogens offers reliable means for the quantification of a variety of pathogens (Boyle et al., 2005; Barnes & Szabo, 2007). However, nucleic acid-based techniques also have their drawbacks. Using genomic DNA as template for quantitative PCR for example may result in a false estimation of the percentage of microbial matter if DNA content varies as a function of growth condition or during different developmental stages.

In this paper, we describe the application of a two-step reverse transcription (RT) real-time PCR protocol for the absolute quantification of the rust Uromyces fabae during the course of infection of its host plant Vicia faba. These analyses were performed using three constitutively expressed genes. In addition, three in planta induced genes (PIGs) (Hahn & Mendgen, 1997) were used to quantify the amount of haustoria present at any given time point during this host–pathogen interaction.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

Organisms and cultivation

Uromyces fabae (Pers.) Schroet. race I2 urediospores were used in all experiments and V. faba cv ‘con amore’ was used as the host plant. Plants (four plants per pot, ∅14 cm) were grown in standard soil in a growth chamber at a 16 : 8 h light : dark regime and 22 °C. Plants were inoculated with a conventional airbrush using urediospores suspended in 0.1% milk powder (1 mg mL−1). Inoculated plants were kept in the dark at 100% humidity for 24 h and then placed in the green house (approximately 16 : 8 h light : dark regime and 22 °C). Control plants were treated with 0.1% milk powder only.

RNA isolation

RNA isolation from infected and noninfected plants as well as from germinated spores was performed according to US Patent No. 5,973,137 (Heath, 1999). Three leaves of the same whorl were homogenized for 10 min in 14 mL lysis buffer (2% SDS, 68 mM sodium citrate, 132 mM citric acid, 10 mM EDTA, pH 3.5) using a glass potter. After the addition of 5 mL protein precipitation buffer (4 M sodium chloride, 17 mM sodium citrate, 33 mM citric acid, pH 3.5) and mixing, the solution was kept on ice for 5 min. Cell debris was removed by centrifugation for 10 min at 4 °C and 20 000 g. The supernatant was transferred to a new centrifuge tube and 14 mL isopropanol were added. After mixing, the solution was incubated at room temperature for 15 min. RNA was recovered by centrifugation at 20 000 g and 4 °C for 5 min. The supernatant was removed and the pellet washed with 1.5 mL 75% ethanol. The supernatant was carefully removed after another centrifugation step under identical conditions and the pellet dried for 10 min using a speedvac concentrator. RNA pellets were resuspended in water pretreated with diethylene pyrocarbonate (H2ODEPC) and stored at −80 °C. RNA isolations for each time point were done three times with independent sets of plants. Corresponding samples were pooled for further analysis.

Urediospores (0.5 g) were washed with 200 mL 0.01% Tween20 for 20 min. Spores were collected by filtration, resuspended in 0.5 L Tween20 and vigorously stirred at room temperature in the dark for 4 h. Progress of germination was monitored microscopically. Germinated spores were collected by filtration and transferred to a mortar prechilled with liquid nitrogen. Germlings were thoroughly ground for 20 min, continuously adding liquid nitrogen. Ground material was transferred to a centrifuge tube and after warming to 4 °C 14 mL of lysis buffer were added. Further steps were carried out as detailed above.

Isolation of haustoria from infected V. faba leafs 8 days postinoculation (dpi) was performed as described by Hahn & Mendgen (1992) and RNA was prepared using peqGold RNAPure (Peqlab, Erlangen, Germany).

All samples were subjected to a Na-acetate/EtOH precipitation, resuspended in H2ODEPC, and quantified photometrically. Samples were adjusted to a concentration of 200 ng μL−1 and integrity of RNA was verified by gel electrophoresis.

Development of primers for real-time PCR

Primers for real-time PCR were designed based on sequences of genes determined in our laboratory. Genes CON1 and CON2 represent transcripts that were identified to be constitutively expressed in the initial expression analysis performed by Hahn & Mendgen (1997) [positions H2 (CON1) and L12 (CON2) in fig. 2 of Hahn & Mendgen (1997)]. TBB1 represents the β-tubulin gene of U. fabae, which also has been shown to be constitutively expressed (Wirsel et al., 2004). THI1, RTP1, and HXT1 on the other hand represent genes whose expression has been shown to be haustorium specific (Sohn et al., 2000; Voegele et al., 2001; Kemen et al., 2005). Primers were designed using the programs gene runner V3.05 (Hastings Software Inc., Westwood, NJ) and lasergene 7 (DNASTAR Inc., Madison, WI). Primer selection was based on a minimum formation of primer secondary structure (within a single primer and among primer pairs), similar annealing temperatures and an amplicon size of 100–200 bp. Primers finally chosen for this study are listed in Table 1.

Table 1.   List of primers used in this study
GenePrimerSequence (5′–3′ orientation)Reference, accession number, position, and amplicon size
  1. Primers were synthesized by Apara Bioscience GmbH (Denzlingen, Germany) and HPLC purified.

CON1con1-fwd2TCG TAC TCT TCG CAA CTC GGHahn & Mendgen (1997)
con1-rev2TCG GAA TCC TTG TCT TGT GG792–929: 138 bp
CON2con2-fwdAGT GTT GGC GAG GCT TGA CHahn & Mendgen (1997)
con2-revGCT GGA CAT TGG TTC AGG C760–874: 115 bp
TBB1tub1-fwdGAC CGC CAT CCA AGA TTT GWirsel et al. (2004)AJ311552
tub1-revTTT CGT CCA TTC CTT CTC CAG1113–1216: 104 bp
THI1THI1-RT-F1CTT GGC TGT TGT TGC TTC TGSohn et al. (2000)AJ250426
THI1-RT-R1CGG CCC AGC TTC CAC GTG577–721: 145 bp
RTP1pig7-RT-F1AGC TAT TGT CGG GAG ATGKemen et al. (2005)AJ971426
pig7-RT-R2TGC GGA AGT GTG ATT AG124–246: 123 bp
HXT1hxt358GCA ATC CTT ACT ATC GGT TTGVoegele et al. (2001)AJ310209
hxt2112CTG GGT CAT CGT ATG GAG607–730: 124 bp

RT and real-time PCR

Generation of cDNA was performed using the QuantiTect Reverse Transcription Kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. For samples to be quantified, 200 ng of total RNA were used. Incubation was for 30 min at 42 °C. In order to be able to perform an absolute quantification of fungal RNA in a mixed sample, RNA from germinated spores and isolated haustoria was also used in amounts of 10, 20, 50, 100, and 150 ng, representing 5, 10, 25, 50, and 75% of the total RNA used for samples to be quantified.

cDNA was quantified using a SmartCyclerII Real Time PCR device (Peqlab) and the QuantiTect SYBR Green PCR Kit (Qiagen). Reactions were carried out in a final volume of 25 μL. The reaction contained 1 μL cDNA (200 ng, or for standards fractions thereof), 12.5 μL 2 × QuantiTect SYBR Green PCR Master Mix, 1.25 μL of each primer, and 9 μL H2O. Amplification conditions consisted of an initial denaturation at 95 °C for 15 min followed by 45 three-step cycles of 94 °C for 15 s, 55 °C for 20 s, and 72 °C for 20 s. Cycle threshold was manually set to 20 RFU. Following the PCR, a melting curve analysis was performed by heating the samples from 60 to 90 °C at a rate of 0.2 °C s−1.

All experiments included water instead of nucleic acids as a negative control and all PCR assays were replicated at least three times.

Results and discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

We set out to quantify the amount of pathogen present at any given time point in the obligate biotrophic interaction of the rust fungus U. fabae and its host plant V. faba. Traditionally, disease severity in this host–pathogen interaction is scored on the basis of macroscopically visible symptoms (Sillero & Rubiales, 2002). Histochemical analyses may be used to complement such ratings (Sillero & Rubiales, 2002). However, this type of quantification is very labor-intensive and only semi-quantitative at best.

Initially, we set out to use the Ergosterol content as a marker for fungal development in planta. However, using adapted extraction procedures according to established protocols (Newell et al., 1988; Martin et al., 1990) and subsequent HPLC analysis using a Nucleosil 100-5 C18 column (Macherey-Nagel, Düren, Germany) revealed that U. fabae has only a negligible Ergosterol content (data not shown). Controls using the addition of defined amounts of purified Ergosterol to diseased plant material before extraction indicated a detection limit in the range of 1 μg mL−1 extract. These results reflect similar findings by Weete et al. (2010) and demonstrated that Ergosterol content cannot be used as a parameter to quantify U. fabae in planta.

As a consequence, we focused on nucleic acid-based techniques. Most previous studies have used genomic DNA for quantification (Winton et al., 2003; Barnes & Szabo, 2007; Vincelli & Tisserat, 2008). However, different spore forms of rust fungi and the structures derived thereof show some distinct variations in DNA content. Moreover, studies using for example Puccinia striiformis, U. fabae, and Uromyces appendiculatus have indicated multinucleate conditions in different differentiation stages of these rust fungi (Staples et al., 1984; Deising et al., 1991; Chong et al., 1992). From this evidence, it has to be concluded that the amount of genomic DNA cannot be used as a reliable marker for the quantitative determination of the fungus in planta.

An alternative, which was used in this study, is the use of specific RNAs for quantification. In all organisms, some genes tend to be constitutively expressed, or to be more precise tend to exhibit constant levels of transcript abundance, regardless of the physiological condition or the differentiation stage. In U. fabae, a number of genes have been shown to be more or less constitutively expressed at a relatively high level. Among those genes were Uf-TBB1 encoding β-tubulin, a major component of the cytoskeleton (Wirsel et al., 2004), and Uf-CON1 and Uf-CON2, two genes encoding hypothetical proteins of unknown function (Hahn & Mendgen, 1997). In initial experiments, we used dot-blot analysis for the quantification of the fungal fraction in mixed samples. Figure 1 shows such an analysis with Uf-CON2 as example. Defined amounts of serial dilutions of RNA preparations from infected leaves were spotted onto Hybond-N+ membranes (GE-Healthcare, Munich, Germany) using a manifold (Gibco BRL, Gaithersburg, MD) to ensure equivalent dot sizes. As a reference, serial dilutions of RNA preparations from in vitro germinated spores (fungus only) were spotted. Quantification of spot intensity was performed using a Gel Doc 1000 System (BioRad, Munich, Germany) and the quantity one software (BioRad). The fraction of fungal RNA present in mixed samples was calculated and plotted against the time course of infection. Figure 1b shows that no fungal material could be detected before 5 dpi. Between 5 and 9 dpi, an almost exponential increase could be seen, which seemed to reach a steady-state level thereafter. While this setup yielded promising results, experiments were very labor-intensive and time-consuming.

image

Figure 1.  Dot-blot analysis for fungal quantification in planta using Uf-CON2 as an example. (a) Dot blot using serial dilutions of RNA preparations from germ tubes (GT) and infected leaves at consecutive days post inoculation (dpi: 1–9). (b) Graphical depiction of the fraction of fungal material plotted against the time course.

Download figure to PowerPoint

We therefore focused on real-time PCR analysis. mRNA was reverse transcribed in a separate reaction using the QuantiTect Reverse Transcription Kit (Qiagen). RNA prepared from germinated spores was again used to generate a standard for absolute quantification. Initially, dilutions of the reverse transcription reaction were used in real-time PCR assays to generate a standard curve. However, it turned out that the different amounts of reverse transcription reaction used as template in the subsequent real-time PCR had a dramatic influence on the efficiency of that reaction (data not shown). As a result the method was adapted such that different amounts of RNA (10, 20, 50, 100, and 150 ng of the normally used 200 ng RNA) were used in the reverse transcription reaction. Subsequently identical volumes of these reactions were used as template in real-time experiments. The standard curves for the three genes used (Uf-CON1, Uf-CON2, and Uf-TBB1) are depicted in Fig. 2a. The slopes of the three standard curves are almost identical. However, the standard curve for Uf-TBB1 is markedly shifted to higher Ct values, reflecting lower levels of transcript abundance of Uf-TBB1 compared with the two other genes (Uf-CON1 and Uf-CON2). For the quantification of haustoria, three genes (Uf-HXT1, Uf-RTP1, and Uf-THI1) were used, which have been shown to be haustorium-specifically expressed (Hahn & Mendgen, 1997; Voegele et al., 2001). Again slopes of the standard curves are almost identical (Fig. 2b). The low CT numbers indicate high levels of transcript abundance. Indeed, all three genes have been shown to be among the most highly expressed genes in haustoria, representing between 0.7% and 2.8% of the total cDNA each (Hahn & Mendgen, 1997; Voegele et al., 2001).

image

Figure 2.  Standard curves generated for the absolute quantification of the fungus in the biotrophic interaction Uromyces fabae/Vicia faba. Threshold cycles (CT) are plotted against the amount of RNA (10, 20, 50, 100, and 150 ng, representing 5%, 10%, 25%, 50%, and 75% of the maximum amount used). (a) Constitutively expressed genes (▪, Uf-CON1; ▴, Uf-CON2; ▾, Uf-TBB1); (b) haustorium-specifically expressed genes (▪, Uf-HXT1; ▴, Uf-RTP1; ▾, Uf-THI1).

Download figure to PowerPoint

These standard curves were then used to perform an absolute quantification of U. fabae in planta. Figure 3a–c depicts the fraction of the constitutively expressed genes Uf-CON1 (a), Uf-CON2 (b), and Uf-TBB1(c) of the total RNA of samples from infected leaves as a function of disease progression. These results mirror those obtained with dot plot analysis. It appears that there is a lag phase in the early days after inoculation, where hardly any fungus is detectable. Between 4 and 8 dpi, there is an exponential increase of the proportion of RNA made up by the fungus. Thereafter, the fungal fraction seemed to reach a steady-state level of around 50% of the total RNA. Results from these analyses correlated so well that data for the different genes could be integrated into a single graph (Fig. 4a). The fact that the proportion of fungal RNA does not seem to increase continuously might reflect the specific need of obligate biotrophic pathogens to keep their host plants alive in order to assure propagation. Nine days post inoculation an equilibrium seems to be established enabling further pathogen development and proliferation without damaging the host plant to a point where it ceases growth. The proportion of about 50% fungal RNA is considerably higher than the amount of 20% fungal DNA reported for a compatible interaction of the poplar rust Melampsora medusae with its host (Boyle et al., 2005). This discrepancy might either be due to the problems associated with using DNA for quantification of rust fungi mentioned above, or to different levels of pathogen present in different host–parasite interactions. Jakupovic et al. (2006) working on the same system as we did, analyzing expressed sequence tags from haustoria in an array experiment, came to similar estimates in the range of 40–50% fungal RNA. It can be concluded that the use of constitutively expressed genes allows an accurate determination of the proportion of pathogen present in a specific host–parasite interaction.

image

Figure 3.  Real-time PCR data for the quantification of Uromyces fabae in planta. (a–c) Fungal quantification [(a) Uf-CON1, (b) Uf-CON2, (c) Uf-TBB1]; (d–f) haustoria quantification [(d) Uf-HXT1, (e) Uf-RTP1, (f) Uf-THI1].

Download figure to PowerPoint

image

Figure 4.  Integration of real-time PCR data for all genes used (a) fungal quantification and (b) haustorial quantification.

Download figure to PowerPoint

We further used the method for quantification of haustoria, using haustorium-specific genes. Figure 3d–f depict the fraction of transcripts of the haustorium-specific genes Uf-HXT1 (d), Uf-RTP1 (e), and Uf-THI1 (f) in the total RNA of samples from infected leaves as a function of disease progression. Again we see a lag phase of up to day 4 or 5 after inoculation where almost no pathogen is detectable. Between 5 and 9 dpi we see an exponential increase of haustorium-specific transcript abundance. After 9 dpi, an equilibrium seems to be established with haustorium-specific genes representing about 5% of the total RNA. Data for the three different genes again correlated well, so that the results could be merged into a single graph (Fig. 4b). The curves obtained for the haustorium quantification mimic those for the total fungal quantification, albeit at a lower percentage (5% vs 50%). It can be concluded that haustorial development is tightly linked to the extent of fungal colonization of the diseased plant. Control of haustoria formation seems to be important in order not to cause too much damage to the host plant. Alternatively, the steady-state level may reflect equilibrium of newly formed and dying haustoria and intercellular mycelial cells.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

From the results presented in this paper it can be concluded that RT real-time PCR constitutes a fast, elegant, and reliable methodology to quantify a parasite in planta. A compatible interaction of the rust fungus U. fabae with its host V. faba is characterized by an initial lag phase where insufficient fungal mass for quantitative detection was present. This phase is followed by almost exponential development of the fungus in diseased plant material. Finally an equilibrium seems to be established where the fungal fraction ranges between 40% and 50% of the total RNA. This equilibrium might be a characteristic of obligate biotrophic interactions reflecting the requirement of the pathogen for a living host in order to propagate. Development of haustoria seems to be tightly linked to the development of the fungus as a whole.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusion
  7. Acknowledgements
  8. References

We are grateful to Christine Giele, Militza Stefcheva, and Otmar Ficht for technical assistance. This work was made possible by grants of the Federal Ministry of Food, Agriculture and Consumer Protection (04HS008) and the German Research Foundation (VO 595/4-1) to R.T.V.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusion
  7. Acknowledgements
  8. References