Analysis of Phakopsora pachyrhizi transcript abundance in critical pathways at four time-points during infection of a susceptible soybean cultivar using deep sequencing
© Tremblay et al.; licensee BioMed Central Ltd. 2013
Received: 8 December 2012
Accepted: 31 August 2013
Published: 11 September 2013
Phakopsora pachyrhizi, the causal agent responsible for soybean rust, is among the top hundred most virulent plant pathogens and can cause soybean yield losses of up to 80% when appropriate conditions are met. We used mRNA-Seq by Illumina to analyze pathogen transcript abundance at 15 seconds (s), 7 hours (h), 48 h, and 10 days (d) after inoculation (ai) of susceptible soybean leaves with P. pachyrhizi to gain new insights into transcript abundance in soybean and the pathogen at specific time-points during the infection including the uredinial stage.
Over three million five hundred thousand sequences were obtained for each time-point. Energy, nucleotide metabolism, and protein synthesis are major priorities for the fungus during infection and development as indicated by our transcript abundance studies. At all time-points, energy production is a necessity for P. pachyrhizi, as indicated by expression of many transcripts encoding enzymes involved in oxidative phosphorylation and carbohydrate metabolism (glycolysis, glyoxylate and dicarboxylate, pentose phosphate, pyruvate). However, at 15 sai, transcripts encoding enzymes involved in ATP production were highly abundant in order to provide enough energy for the spore to germinate, as observed by the expression of many transcripts encoding proteins involved in electron transport. At this early time-point, transcripts encoding proteins involved in RNA synthesis were also highly abundant, more so than transcripts encoding genes involved in DNA and protein synthesis. At 7 hai, shortly after germination during tube elongation and penetration, transcripts encoding enzymes involved in deoxyribonucleotide and DNA synthesis were highly abundant. At 48 hai, transcripts encoding enzymes involved in amino acid metabolism were highly abundant to provide for increased protein synthesis during haustoria maturation. During sporulation at 10 dai, the fungus still required carbohydrate metabolism, but there also was increased expression of transcripts encoding enzymes involved in fatty acid metabolism.
This information provides insight into molecular events and their timing throughout the life cycle of the P. pachyrhizi, and it may be useful in the development of new methods of broadening resistance of soybean to soybean rust.
KeywordsDeep sequencing Transcript abundance Phakopsora pachyrhizi Plant-pathogen interaction Soybean Soybean rust
Phakopsora pachyrhizi, the causal agent of soybean rust (SR), is among the top hundred most virulent plant pathogens. P. pachyrhizi can infect at least 89 different plants from bean to lupine. In the United States, soybean (Glycine max) is the only crop for which a yield effect has been reported from P. pachyrhizi infection. Infection has been reported on other crops in the U.S., but these infections are of limited scope with no apparent economic impact. These crops include scarlet runner bean, lima bean, and kidney bean. Kudzu and beggarweed have also been identified in the U.S. as non-crop host plants [1–4].
In Asia, where the pathogen originated, soybean yield losses range from 20% to 80% following P. pachyrhizi infection [5, 6]. Soybean in South America is also highly affected by SR. Since 2001, when the disease was first observed in Paraguay , soybean yield losses of 10% to 50% have been quite common . The SR disease monitoring program established in the U.S. allows farmers to respond quickly to outbreaks of SR infection by applying fungicides when and where SR disease develops to avoid yield losses. Although fungicide application decreases P. pachyrhizi development on soybean crops, the spread of the disease throughout the US has still increased over the past several years. In 2009, Mississippi recorded for the first time a yield reduction associated with SR disease, of between 8% and 25% . Since 2010, the precarious state of funding of the U.S. monitoring program has provided impetus to researchers to understand the interaction between P. pachyrhizi and its soybean host and to identify new forms of resistance to this pathogen in soybean. Most of the efforts to develop new resistance in commercial soybean cultivars have been concentrated on the identification of soybean genes from six major loci (Rpp 1-Rpp 6 and Rpp?) conferring resistance to P. pachyrhizi[10–13]. The specific gene responsible for SR resistance mediated by the Rpp 4 locus was identified by Meyer et al.  as encoding a gene belonging to the CC-NBS-LRR family of disease resistance genes showing greatest similarity to the RGC2 family of disease resistance genes from lettuce.
P. pachyrhizi is an obligate biotroph which makes it hard to study as a unique entity. Fungal spores are spread by air currents and once they land on a leaf surface and optimal growth conditions are reached, they begin to germinate and form appressoria. This first stage of rust infection, named the pre-penetration stage, happens within 24 h of a spore landing on the leaf. This is the only infection stage occurring outside the host. The germination step can be reproduced in water. Posada-Buitrago and Frederick  and Stone et al.  isolated mRNA from germinating P. pachyrhizi urediniospores and appressoria, constructed cDNA libraries, and sequenced 488 and 1,029 unique expressed sequence tags (ESTs) respectively available at NCBI. Since then more P. pachyrhizi ESTs from this specific physiological stage have been added to the NCBI database as well as ESTs from germinating spores of other plant pathogenic fungi such as Puccinia striiformis f. sp. tritici, P. triticina, Ustilago maydis and Fusarium oxysporum. Following the pre-penetration stage up to 48 hours after a spore lands, infection hyphae form that allow direct penetration inside the host and subsequent formation of primary haustoria mother cells (HMCs). This second step in infection is named the penetration stage.
From the formation of HMCs to eight days after spore landing, fungal intercellular hyphae grow between host palisade and mesophyll cells, more haustoria cells are produced, and uredinia arise as a result of hyphae aggregation. During this colonization stage, haustorial cells can be isolated using a protocol established by Hahn and Mendgen . Loehrer and Schaffrath  isolated P. pachyrhizi haustoria, extracted RNA, constructed a cDNA library, and sequenced the whole haustorial transcriptome, resulting in 111,440 de-novo assembled contigs. Transcript abundance in haustorial cells has also been studied in few other rusts including P. striiformis f. sp. tritici, Uromyces appendiculatus, U. fabae, and Melampsora lini[23–26].
Fewer studies have been conducted on transcript abundance within hyphae since they are hard to isolate from plant material. However, more transcript abundance studies have been done on cDNA libraries constructed from mRNA extracted from whole infected leaves at different time-points along the infection process. There are 5,981 ESTs from soybean cultivar Williams 82 at 6 to 8 days after inoculation (dai) with urediniospores of P. pachyrhizi isolate Taiwan 72–1 (TW72-1) and 6,390 ESTs from soybean cultivar Williams 82 at 13 to 15 dai with urediniospores of P. pachyrhizi isolate TW72-1 deposited in the NCBI database (Posada-Buitrago et al., 2006; unpublished data). Although the majority of these ESTs correspond to soybean genes, some correspond to P. pachyrhizi genes.
Between nine and ten days after urediniospores land on the leaf surface, new urediniospores are produced inside the uredinium. This represents the sporulation stage. Urediniospore production can extend up to three weeks from the generation of the first uredinium, but secondary uredinia can maintain sporulation for up to 15 weeks . During the sporulation stage, the plant cuticle is ruptured and urediniospores are released into the environment. Transcript abundance analyses were performed on urediniospores of P. pachyrhizi collected on soybean infected leaves by Posada-Buitrago et al. in 2006 (unpublished), resulting in 2,122 ESTs being deposited in the NCBI database. Similar analyses were performed on spores of P. striifromis f. sp. tritici.
Although obligate biotrophic organisms are difficult to study during the infection process, laser capture microdissection (LCM) can help to circumvent this difficulty by enabling the dissection of infected tissue precisely enough to isolate the organism within the host. Our research team used LCM to isolate uredinia of P. pachyrhizi formed 10 dai on leaves of the susceptible soybean cultivar Williams 82, and we analyzed transcript abundance at this specific stage of P. pachyrhizi development . Sanger sequencing of our library allowed us to generate a limited number of ESTs. More recently, Hacquard et al.  used LCM to isolate uredinial sites corresponding to spores and sporogenous hyphae, fungal infection tissues in spongy mesophyll, and fungal infection tissues in palisade mesophyll from poplar leaves 4 and 7 dai with Melamspora larici-populina. RNA was extracted and amplified from the cells isolated by LCM and the RNA was hybridized to whole-genome exon oligoarrays of M. larici-populina. Even though array technology provides the means to study the expression of many genes, it does not allow the discovery of new genes. Based on these limitations, we deep-sequenced cDNA libraries of soybean cultivar Williams 82 inoculated with P. pachyrhizi along a time-course of infection using the Illumina platform to generate a more in-depth transcript abundance profile of the genes expressed by soybean and P. pachyrhizi simultaneously. One objective of this study was to identify P. pachyrhizi abundant transcripts present at different times during an infection process, build a model of what is happening during the infection process based on transcript abundance and, choose target genes for silencing or overexpression in order to produce a resistant plant. This information will help us to better understand the infection process of P. pachyrhizi on soybean plants and provide new ways to engineer soybean plants to fight the pathogen attack. This manuscript described transcript abundance profile of genes expressed by P. pachyrhizi and provides insights into the uredinial stage and infection process of P. pachyrhizi at the molecular level. We show that genes encoding enzymes involved in specific biochemical pathways are expressed at specific time points to help meet the requirements of P. pachyrhizi to successfully parasitize soybean and produce urediniospores.
Results and discussion
mRNA sequencing of soybean infected with Phakopsora pachyrhizi
Summary of sequenced reads throughout a soybean- P. pachyrhizi infection study
# Raw reads
# QA/QCareads after all filtering
# De-novo transcripts
Contigs of importance, shorter than 75 bp, taking into account in our analysis
Similarity search results
NADH dehydrogenase subunit f
Pectin methylesterase 1
Gamma-glutamyl phosphate reductase
Pyruvate dehydrogenase (acetyl-transferring)
2,3-bisphosphoglycerate-independent phosphoglycerate mutase
NAD-specific glutamate dehydrogenase
As previously mentioned, sequence information from P. pachyrhizi is limited and mostly derived from cDNA libraries constructed from mRNA extracted from urediniospores and germinated urediniospores as well as from whole infected soybean leaves from 6 to 15 dai. By using a deep sequencing strategy, we sequenced and generated 27,715 contigs from four time-points which represent a little more than half of what is available at NCBI for P. pachyrhizi (49,596 ESTs). By using mRNA-Seq, we also obtained sequences from additional infection stages including 15 sai, 7 hai and 48 hai which were not yet covered at NCBI public databases. From the present data set only 8,673 contigs shared similarity to P. pachyrhizi ESTs previously identified. Thus, more than 19,000 contigs, more than two-thirds of the total number of contigs, represent newly identified P. pachyrhizi transcripts. Using genomic, transcriptomic, proteomic, and metabolomic information available from other rusts and pathogenic fungi, we interpreted our deep sequencing data in relation to the needs and requirements of P. pachyrhizi during infection of susceptible soybean leaves.
Genes encoding enzymes involved in energy production and carbohydrate metabolism were abundant at all time-points
The number of DNA sequence reads building each contig was used to determine amount of expression of each gene. This was used to compare transcript abundance among all time-points. There was an increase in the number of different transcripts identified from 15 sai to 10 dai, suggesting that as the fungus grows, spreads, and matures, it expresses more genes to complete its infection. At 48 hai and 10 dai, approximately 70% and 90% of the transcripts encoding proteins sharing similarity to known proteins identified in different databases were up-regulated while at 7 hai only approximately 30% of the transcripts encoding proteins sharing similarity to known proteins identified in different databases were up-regulated.
Genes encoding enzymes of energy production are abundant in urediniospores
Of the 6,531 contigs identified at 15 sai, 1,054 contigs displayed similarities to DNA sequences in different P. pachyrhizi EST libraries. From these, only about 30% shared significant similarity to known proteins.
Even though we collected soybean leaves as early as 15 sai, the urediniospores were not in a quiescent state since they were manipulated during inoculum preparation and plant inoculation. Many fungal transcripts were present. Indeed, many transcripts sharing similarity to genes encoding enzymes involved in energy production were identified, including several encoding electron transport proteins such as the ATP synthase beta subunit (E.C. 220.127.116.11), ferredoxin, NADH dehydrogenase (E.C. 18.104.22.168) subunit 4 and subunit f, cytochrome monohaem, electron carrier oxidoreductase, and fructose-1,6-bisphosphatase (E.C. 22.214.171.124). Transcriptome analysis followed by proteome analysis on dormant Trychophyton rubrum conidia indicated that many genes and proteins belonging to glycolysis, the pyruvate dehydrogenase complex, and the oxidative phosphorylation machinery were transcribed [30, 31]. These results indicate that transcription and translation of genes involved in energy production and carbohydrate metabolism also occurs earlier than 15 sai. In addition, early studies demonstrated that spore germination was dependent on the function of the standard, cytochrome-mediated electron transport system in Botryodiplodia theobromae and Neurospora crassa, wherein all of the enzyme components in this standard pathway, such as cytochrome-c oxidase and F1F0-type ATP synthase, appeared to be assembled and preserved. These results are consistent with what we observed at 15 sai where transcripts encoding for these enzymes were abundant.
On the other hand, we also identified transcripts at 15 sai that shared similarity to genes encoding proteins involved in transcription, specifically RNA polymerases. Active RNA polymerases have been found previously in ungerminated Rhizopus stolonifer spores  as well as some other fungi  meaning that their presence at 15 sai is not surprising.
Genes encoding enzymes involved in nucleotide metabolism are abundant during urediniospore germination
Of the 4,627 contigs identified at 7 hai representing abundant transcripts during urediniospore germination, 624 contigs displayed similarities to DNA sequences in P. pachyrhizi EST libraries. Of the remaining 4,003 contigs without similarity to publicly available P. pachyrhizi ESTs, our blastx analysis found that only 15% displayed significant similarities to genes encoding proteins listed in the various databases. Figure 4B illustrates the percentage of all identified transcripts sharing similarity to genes encoding fungal proteins belonging to various functional categories.
A high number of transcripts at 7 hai shared similarity to genes encoding enzymes involved in nucleotide metabolism, such as genes encoding cytosine deaminase, phosphoribosylaminoimidazole carboxylase (E.C. 126.96.36.199), xanthine dehydrogenase (E.C. 188.8.131.52), phosphoribosylformylglycinamidine cyclo-ligase (E. C. 184.108.40.206), and ribonucleoside triphosphate reductase (E.C. 220.127.116.11). Many of these enzymes are directly involved in deoxyribonucleotide and DNA synthesis, while others are involved in aminoimidazole ribotide (AIR) synthesis, which is a precursor of thiamine. Three major classes of macromolecules synthesized rapidly once fungal germination takes place are DNA, RNA and proteins. Interestingly, a pool of transcripts ready to be translated is not always present in ungerminated fungal spores. Indeed, dormant ascospores and conidia of Neurospora crassa lack mRNA , as do ungerminated conidia of Fusarium solani, ungerminated basidiospores of Schizophyllum commune, ungerminated spores of Dictylostelium purpureum, and dormant conidia of Penicillium chrysogenum. The low number of transcripts identified at 15 sai compared to other time-points in the present data set suggests that there is no latent pool of mRNAs present in P. pachyrhizi urediniospores as well. Even though the levels of adenosine triphosphate, guanosine triphosphate, cytidine triphosphate, and uridine triphosphate in P. pachyrhizi urediniospores are unknown, in other fungi such as dormant spores of Dictyostelium discoidenum, the relative levels of these nucleotides is greater than or equal to their levels observed in germinated spores . All together, this suggests that RNA synthesis can occur rapidly after P. pachyrhizi urediniospores are deposited on the soybean leaf and even before that point which is consistent with the high level of RNA polymerase expression 15 sai exposed in the previous section. This also occurs during germination of Rhizopus stolonifer sporangiospores, where synthesis of all classes of RNA occurred within a few minutes after the sporangiospores were placed on germination medium . Interestingly, the abundance of transcripts encoding enzymes involved in RNA synthesis decreased at 7 hai whereas there was an increase in transcript abundance sharing similarity to genes encoding enzymes involved in DNA synthesis. Proteins encoded by transcripts abundant at 7 hai involved in purine metabolism are exposed in supplementary data (Additional file 1: Figure S1). Along with DNA synthesis, transcripts encoding proteins involved in protein synthesis were lower in abundance at 7 hai, but were expressed, suggesting that protein synthesis also occurred during germination.
Figure 5B is a schematic representation of events occurring into a susceptible soybean leaf at 7 hai with P. pachyrhizi. Expression of P. pachyrhizi genes encoding F-type H + −transporting ATPase and H + −transporting ATPase (complex V), involved in electron transport is down-regulated compared to their expression at 15 sai while transcript encoding proteins building complexes I and IV are up-regulated. Other metabolic pathways also involved in energy production are similarly down-regulated (data not shown). However, transcription of RNA and synthesis of DNA and protein are active at 7 hai. Transcripts encoding enzymes involved in nucleotide metabolism are abundant at the time DNA synthesis is required by the fungus.
Genes encoding enzymes involved in amino acid metabolism and protein synthesis are abundant during haustorium formation
Of the 4,273 contigs identified at 48 hai representing abundant transcripts during haustoria formation, 1,130 contigs displayed similarities to DNA sequences in SR EST libraries. As for the two previous time-points, only about 15% of the remaining sequences displayed significant similarities to genes encoding proteins listed in various databases. Figure 4C illustrates the percentage of all transcripts sharing similarity to genes encoding fungal proteins belonging to various functional categories.
Haustoria are important structures for nutrient absorption for many fungi. Evidence of the involvement of haustoria in nutrient uptake from the host cells include increased plasma membrane H+-ATPase activity in haustorial membranes as compared to that of membranes from other infection structures [43–45]; preferential expression of genes encoding secondary transporters for amino acids in haustoria [24, 46]; and preferential localization of hexose transporter at the tip of differentiating haustoria . However, transcripts sharing similarity to genes encoding membrane transporters and plasma membrane ATPases were not present in our data at 48 hai, which represents the haustoria-formation stage. However, more recent evidences showed that haustoria function not only in nutrient absorption, but also in the suppression of host defense responses, the redirection or reprogramming of the host’s metabolic flow, and in biosynthesis of organic compounds .
There is no doubt that fungal pathogens can directly or indirectly reprogram host metabolism, but there is still no proof that this is linked to the function of haustoria. Our data set at 48 hai identified many transcripts sharing similarity to genes encoding proteins involved in amino acid metabolism and protein synthesis, including 5′-methylthioadenosine nucleosidase (E.C. 18.104.22.168), glutamine synthetase, gamma-glutamyl phosphate reductase, cysteine synthase (E.C. 22.214.171.124), methylenetetrahydrofolate reductase and ribosomal proteins. This suggests that protein synthesis occurs during this fungal growth stage in addition to nutrient transport. Proteins encoded by transcripts abundant at 7 hai involved in cysteine and methionine metabolism are exposed in supplementary data (Additional file 2: Figure S2).
Our data from the haustorial stage (48 hai) and that from P. triticina haustoria  show a limited number of transcripts common to both studies. Of 416 ESTs sharing similarity to genes encoding known proteins identified by Xu et al. , only 40% shared similarity to genes encoding fungal proteins and from these, about fifteen transcripts in the present data set at 48 hai shared similarity to genes encoding the same fungal proteins. Eleven additional transcripts shared similarity to genes that were also identified by Yin et al.  from a cDNA library from P. striiformis f. sp. tritici haustoria. They identified approximately 170 ESTs sharing similarity to genes encoding known fungal proteins. Our dataset at 48 hai did not share a lot of commonality with that of Xu et al.  and Yin et al.  examining haustorial libraries from other fungal species, since they were working on haustoria extracted from plant tissue between 4 and 8 dai. We did not isolate haustoria, but harvested a mix of fungal structures including hyphae and maturing haustoria.
Hyphal walls are mostly made of polysaccharides (60-90%) and some proteins and lipids [49, 50]. In contrast, Chong et al.  reported that rust haustorial cell walls were made principally of proteins in complexes with polysaccharides and lipids. The high abundance of transcripts sharing similarity to genes encoding enzymes involved in amino acid, lipid, and carbohydrate metabolisms at 48 hai may be related to the synthesis and expansion of hyphal and haustorial walls and membranes.
Not very much is known about the spectrum of biosynthetic reactions occurring in haustoria. However, Hahn and Mendgen  found that two of the most abundant genes isolated from their haustorial cDNA library encoded enzymes directly involved in thiamine metabolism. The expression pattern of these genes showed no significant accumulation of transcripts until 18 hours after germination (hag). The transcript abundance increased at 18 and 24 hag and finally accumulated to very high concentrations in haustoria and rust-infected leaves. Thus, Uromyces fabae performs thiamine biosynthesis very actively during parasitic growth, presumably because this vitamin is growth limiting and not available from the host plant . In our data set, transcripts sharing similarity to genes encoding a nucleoside triphosphatase (E.C. 126.96.36.199) and a hydrolase from the gdsl-motif lipase hydrolase family (E.C. 3.1.3.-) involved in thiamine metabolism were found at all time-points (Additional file 3: Figure S3). However, their estimated expression was moderate at 15 sai, declined at 7 hai, then increased at 48 hai and increased further by 10 dai. Their estimated expression at 48 hai was not as high as their estimated expression at 15 sai, and at 10 dai their expression was much higher than 15 sai. Thiamine diphosphate coming from pyruvate metabolism is converted to thiamine monophosphate by a nucleoside triphosphatase in presence of water, while the hydrolase from the gdsl-motif lipase hydrolase family converted the thiamine monophosphate into thiamine. There may be another pathway of thiamine production used. Transcripts sharing similarity to genes encoding phosphoribosylformylglycinamidine cyclo-ligase and phosphoribosylaminoimidazole carboxylase/phosphoribosylaminoimidazole-succinocarboxamide synthase were present. Both enzymes are able to convert their respective substrate into aminoimidazole ribotide (AIR) during thiamine metabolism. However, these transcripts were identified only at 7 hai and 10 dai. In addition, no transcript sharing similarity to genes encoding the enzymes responsible for the conversion of AIR into thiamine monophosphate were identified at any time-point. Even though transcripts sharing similarity to genes encoding enzymes involved in pyruvate metabolism were identified, transcripts sharing similarity to genes encoding the pyruvate dehydrogenase E1 component directly responsible for the thiamine diphosphate production was not identified. Unfortunately, there is no known physiological role associated with thiamine monophosphate, but it seems to be important early in P. pachyrhizi urediniospore germination and during haustoria maturation through to the end of the infection process.
Figure 5C is a schematic representation of events occurring into a susceptible soybean leaf at 48 hai with P. pachyrhizi. There is high abundance of transcripts encoding complexes I and V involved in oxidative phosphorylation that are important for producing energy needed for protein synthesis. Also at 48 hai, transcripts encoding enzymes involved in thiamine metabolism are up-regulated.
Genes encoding enzymes involved in carbohydrate and fatty acid metabolism are abundant during sporulation
Of the 12,163 contigs identified at 10 dai representing transcripts abundant during urediniospore production, 5,865 contigs displayed similarities to DNA sequences in SR EST libraries. Of the remaining contigs, a blastx analysis found that approximately 25% displayed significant similarities to genes encoding proteins listed in various databases. Figure 4D illustrates the percentage of all transcripts sharing similarity to genes encoding fungal proteins belonging to various functional categories.
At the late stage of infection (10 dai), a high abundance of transcripts sharing similarity to genes encoding enzymes involved in carbohydrate and lipid metabolism was identified. These transcripts included genes encoding almost all enzymes involved in glycolysis and the citrate cycle except ADP-6-phosphofructokinase (E.C. 188.8.131.52). Additional file 4: Figure S4 shows the glycolysis pathway, which includes some enzymes encoded by transcripts that had not been previously identified as abundant during the biotrophic interaction between soybean and P. pachyrhizi. Cinq genes encoding enzymes involved in glycolysis are reported here for the first time as being expressed during soybean-P. pachyrhizi interaction. They are genes encoding glucose-6-phosphate isomerase, triosephosphate isomerase, phosphoglycerate kinase, pyruvate dehydrogenase E1 component and, hexokinase. Some of these genes that encode a hexokinase were found abundant only at this specific time-point.
Fungal spore formation requires energy that can be produced by metabolism of carbohydrates. Lipids as well as carbohydrates are essential for fungal spore production since they are part of the spore membranes and they are also stored in fungal spores to serve as an energy reserve and a reserve of structural material utilized by the fungus during subsequent germination. Fourteen transcripts involved in lipid metabolism were abundant at 10 dai while not present in earlier time-points. A more detailed description of transcript abundance inside the fungus at 10 dai has previously been described by Tremblay et al. .
Figure 5D is a schematic representation of events occurring in a susceptible soybean leaf at 10 dai with P. pachyrhizi. There is high abundance of transcripts encoding complexes I, II, IV, and V involved in oxidative phosphorylation to produce energy. Carbohydrate, lipid, and thiamine metabolisms are all activated as indicated by the large number of transcripts encoding enzymes involved in these metabolic pathways at this late time-point.
Combining all time points, there were 3,331 contigs representing P. pachyrhizi genes that share similarity to genes encoding known proteins that have not been reported previously in a P. pachyrhizi.
Differential gene expression was confirmed by qPCR
Confirmation of gene expression base on mRNA-Seq assay using qPCR
NADH dehydrogenase subunit f
Ribulose-1,5-bisphosphate carboxylase oxygenase large subunit
60S ribosomal protein L18
This study identified many new P. pachyrhizi abundant transcripts at various stages of fungal development. Even though no information is available describing the function of many of these genes, identification of transcripts specific to different fungal growth stages and their similarity to genes of known function gives us a better understanding of molecular processes occurring during the interaction between P. pachyrhizi and soybean plants. In summary our study showed that transcripts encoding proteins sharing similarity to known proteins identified in different databases were mostly (68.4%) down-regulated at 7 hai compared to 15 sai while there was an impressive up-regulation at 48 hai (69.7%) and 10 dai (91.4%). More specifically, energy production is highly active in urediniospores reflected by the abundance of transcripts encoding complexes I, IV, and V involved in oxidative phosphorylation and transcripts encoding proteins involved in glycolysis such as phosphoglucomutase, fructose-1,6-bisphosphatase and triosephosphate isomerase. The energy produced by these metabolic pathways can be used by RNA polymerases, which were also highly abundant, to transcribe genes necessary for the subsequent step of fungal growth. During urediniospores germination, transcripts of genes encoding enzymes involved in nucleotide metabolism was highly abundant along with genes encoding proteins involved in protein synthesis while genes encoding enzymes involved in energy production tended to be down-regulated. At 48 hai, there was an abundance of genes encoding enzymes involved in amino acid metabolism and protein synthesis supporting the recent hypothesis that haustoria, produced during that time frame, play roles in addition to nutrient uptake. There was also abundance of transcripts sharing similarity to genes encoding enzymes involved in lipid metabolism and carbohydrate metabolism that may be associated with the synthesis and expansion of hyphal and haustorial walls and membranes. Finally, at the end of the infection process, transcripts of genes encoding enzymes involved in carbohydrate and fatty acid metabolism were the most highly abundant. Some of these genes and their proteins may potentially serve as targets for developing new modes of plant resistance to fungi by impeding fungal development. Transcripts sharing similarity to genes encoding proteins involved in thiamine metabolism are good examples. These transcripts are abundant 15 seconds after the inoculation process. Their expression decreased at 7 hai but increased at 48 hai and 10 dai when P. pachyrhizi is producing haustoria and uredinia, respectively. Fungal thiamine production in haustorium-forming fungi has previously been described as important during haustoria development and may be considered as a good target for disrupting fungal growth and decrease P. pachyrhizi infectivity on soybean plants.
Pathogen isolation and plant inoculation
The P. pachyrhizi isolate MS06-1 was obtained from urediniospores harvested from field-collected kudzu leaves in Jefferson County, Mississippi, in August 2006. Its identity was confirmed by microscopy, enzyme-linked immunosorbent assay (ELISA) and polymerase chain reaction (PCR) as previously described . Urediniospores were increased on a susceptible soybean cultivar, Williams 82 in the Stoneville Research Quarantine Facility in Mississippi. The isolate was purified by picking a single uredinium using a fine needle under an Olympus SZX12 dissecting microscope and reinoculating it on leaves of Williams 82. This inoculation-isolation cycle was repeated four times. Urediniospores from this purified culture were harvested using a Cyclone Surface Sampler (Burkard Manufacturing Co. Ltd, UK) connected to a vacuum pump at 10 to 14 dai and continuously thereafter at weekly intervals.
Inoculum was prepared using freshly collected urediniospores from Williams 82. Spore suspensions were made using sterile distilled water containing 0.01% Tween-20 (vol/vol) and then filtered through a 100-μm cell strainer (BD Biosciences, Bedford, MA) to remove any debris and clumps of urediniospores. Urediniospores were quantified using a hemocytometer and diluted to a final concentration of 1.1 × 105 spores/mL. Three plants per 10 cm-pot were prepared in three replicates (pots). Primary leaves of 3-weeks-old Williams 82 seedlings were inoculated at a rate of one milliliter of spore suspension per plant using a Preval sprayer (Yonkers, NY). The same solution minus spores was used for a mock inoculation on three pots of plants as a control. After inoculation, plants were placed in a dew chamber in the dark at 22°C overnight (approximately 16 h) and then moved to a Conviron growth chamber where temperatures were maintained at 23°C during the day and 20°C at night under a 16-h photoperiod with a light intensity of 280 μEm-2s-1. Leaves of soybean cultivar Williams 82 were collected at 15 sai, 7 hai, 48 hai and10 dai with P. pachyrhizi. Experiments were repeated once.
RNA extraction and isolation
Trifoliate leaves from each time-point collected from different pots and replications were pooled in a 50 ml conical tube, fixed in Farmer’s solution and stored at 4°C for shipping from Mississippi to Maryland. One hundred milligrams of leaf pieces from each 50 ml conical tube for each time-point were ground in liquid nitrogen, and RNA was extracted using 450 μl of buffer RLC (Qiagen). RNA was isolated from the sample using an RNeasy Plant Mini Kit (Qiagen) according to the manufacturer’s instructions. RNA was treated on the column with 80 ul of DNA™-free DNAse (Ambion) for 15 min at room temperature. Five hundred nanograms of RNA were used to evaluate its quality and integrity on a 2% agarose gel.
The mRNA was purified from ten micrograms of total RNA using oligo(dT) Dynal magnetic beads (Invitrogen). Two rounds of purification were performed and the resulting mRNA was fragmented at 70°C for five min using RNA fragmentation reagents from Ambion. The fragmented mRNA was precipitated for 30 min at −80°C with 1/10th volume of 3 M NaOAc, 40 μg of glycogen and 3 volumes of 100% EtOH. The fragmented mRNA resuspend in eight microliters of water was used as template for cDNA synthesis. One hundred nanograms of random hexamer were used in concert with the first-strand synthesis reagents from the Superscript III first-strand synthesis system (Invitrogen) following the manufacturer’s instructions. Second-strand synthesis was performed using 1× second-strand buffer (Invitrogen), 0.3 mM of dNTP mix, 2 units of RNaseH and 50 units of DNA polymerase I at 16°C for 2 h and 30 min. cDNA was then purified on a Qiaquick PCR purification column (Qiagen) and eluted in 30 μl of EB buffer (Qiagen). Then, the genomic DNA Sample Preparation Kit from Illumina  was used to repair the ends of the cDNA, add a single adenine base, and ligate the adaptor to the cDNA molecules. A 200 +/− 25 bp band was excised from a 2% agarose gel after electrophoresis in 1× TAE buffer for one h at 120 V. The cDNA was purified from the gel using a Qiaquick gel extraction kit (Qiagen) and eluted in 30 μl of EB buffer (Qiagen). The resulting cDNA was PCR enriched using Illumina primers using these conditions: 98°C for 30 sec, 15 cycles of 98°C for 10 sec, 65°C for 30 sec, 72°C for 30 sec and a final step at 72°C for five min.
Sequencing and contig building
One hundred and twenty microliters of each cDNA library, at eight picomolar (pM), was used to generate clusters on four individual flow cell lanes. A fifth control lane was hybridized with a PhiX Illumina-supplied control.
All four libraries were sequenced using a single-end recipe on the Illumina Genome Analyzer IIx (GAIIx). A set of Cluster Intensity Files (CIF) were produced which were subsequently analyzed using the Illumina Offline Base-Caller (OLB) version 1.6. A break-down of total reads per time-point is presented in Table 1.
CASAVA (Consensus Assessment of Sequence and Variation) then mapped qseq data for all lanes against the twenty soybean chromosomes. Reads aligning to the genome were then mapped against homology-based annotations using TASE (Tag counting and Analysis of Solexa Experiments), producing read-frequencies per annotation .
All reads across all four time-points not mapping to the Glycine max genome were subsequently put through two rounds of read-subtraction so as to remove contaminant reads be-it other plant or human/bacterial contamination. This subtraction ensures resultant reads to have the greatest likelihood of being potentially P. pachyrhizi. For the first subtraction, all reads per time-point were mapped against all available plant genomes from Phytozome . Following, the read-set was then mapped against the JCVI Microbial Database  and the NCBI human genome (Hg19). The remaining reads were considered as potential P. pachyrhizi (Table 1).
Once both plant-specific and contaminant human/bacterial reads were subtracted from our four read-sets, putative de-novo transcripts were assembled using Velvet and Oases [59, 60]. A k-mer hash of 27 and minimum contig length of 75 bp was specified per time-point. Assembled contigs ranged in size from 75 bp to 1,991 bp, composed of 2 to 3,911 reads for any given contig.
where Cnorm represents coverage given the contig length and K = 25 (as used throughout this study).
The NCBI nucleotide collection (nr/nt) was downloaded from NCBI . Using BLASTN and the local ‘nr/nt’ database, DNA sequences of all generated contigs were iteratively queried using the MATLAB Bioinformatics Toolbox and an expect-value of 1e-20.
For contigs without similarity to DNA sequences in the NCBI database, a second similarity search strategy was adopted to obtain additional information for those contigs where open reading frames (ORFs) were found. Using the longest ORF for each contig, similarity searches were conducted using translated peptide sequences of contigs against the NCBI non-redundant proteins database, the NCBI conserved domains database (CDD), and the Consortium for the Functional Genomics of Microbial Eukaryotes (COGEME) database (which is specific for phytopathogenic fungi and oomycetes) at an expected-value ≤ 1e-05.
Confirmation of transcript abundance by quantitative RT-PCR
PCR primer pairs
NADH dehydrogenase subunit f
60S ribosomal protein L18
ribulose-1,5-bisphosphate carboxylase / oxygenase large subunit
All primer sets were designed from regions that flanked an intron to make sure that expected size product was amplified from cDNA and not from genomic DNA. An additional control tube containing no template was included for each specific reaction using different primer sets.
qPCRs using all primer sets were performed as three technical replicates. Relative levels of gene expression were determined using the Stratagene Mx3000P Real-Time PCR system (Stratagene, La Jolla, CA) as described by the manufacturer. DNA accumulation during the reaction was measured with SYBR Green. The Ct (cycle at which there is the first clearly detectable increase in fluorescence) values were calculated using software supplied with the Stratagene Mx3000P Real-Time PCR system. The SYBR green dissociation curve of the amplified products demonstrated the production of only one product per reaction. Data analysis was performed according to the sigmoidal model  to get absolute quantification as described in Tremblay et al. .
Availability of supporting data
The data sets supporting the results of this article are available in the Sequence Read Archive repository at the National Center for Biotechnology Information, SRR445529 (http://www.ncbi.nlm.nih.gov/sra/?term=SRR445529), SRR610280 (http://www.ncbi.nlm.nih.gov/sra/?term=SRR610280), SRR610284 (http://www.ncbi.nlm.nih.gov/sra/?term=SRR6102804), SRR445528 (http://www.ncbi.nlm.nih.gov/sra/?term=SRR445528).
The authors thank Alicia Beavers for her excellent technical assistance, and Richard Joost, Eric Brewer, for their careful critical review of the manuscript. The authors gratefully acknowledge support from United Soybean Board project number T0258. This research was supported in part by the Intramural Research Program of the National Institutes of Health, National Library of Medicine. Mention of trade name, proprietary product or vendor does not constitute a guarantee or warranty of the product by the U.S. Department of Agriculture or imply its approval to the exclusion of other products or vendors that also may be suitable.
- Lynch TN, Marois JJ, Wright DL, Harmon PF, Harmon CL, Miles MR, Hartman GL: First report of soybean rust caused by Phakopsora pachyrhizi on Phaseolus spp. in the United States. Plant Dis. 2006, 90: 970-View ArticleGoogle Scholar
- Ono Y, Buritica P, Hennen JF: Delimitation of Phakopsora, Physopella and Cerotelium and their species on Leguminosae. Mycol Res. 1992, 96: 825-850. 10.1016/S0953-7562(09)81029-0.View ArticleGoogle Scholar
- Rytter JL, Dowler WM, Bromfield KR: Additional alternative hosts of Phakopsora pachyrhizi, causal agent of soybean rust. Plant Dis. 1984, 68: 818-819.View ArticleGoogle Scholar
- Sconyers LE, Kemerait RC, Brock JH, Gitaitis RD, Sanders FH, Phillips DV, Jost PH: First report of Phakopsora pachyrhizi, the causal agent of Asian Soybean Rust, on Florida beggarweed in the United States. Plant Dis. 2006, 90: 972-View ArticleGoogle Scholar
- Miles MR, Levy C, Morel W, Mueller T, Steinlage T, van Rij N, Frederick RD, Hartman GL: International fungicide efficacy trials for the management of soybean rust. Plant Dis. 2007, 91: 1450-1458. 10.1094/PDIS-91-11-1450.View ArticleGoogle Scholar
- Yeh CC, Yang CY: Yield loss cause by soybean rust, Phakopsora pachyrhizi. Plant Prot Bull (Taiwan ROC). 1975, 17: 7-8.Google Scholar
- Yorinori JT, Pavia WM, Frederick RD, Costamilan LM, Bertagnolli PF, Hartman G, Godoy C, Nunes J: Epidemics of soybean rust (Phakopsora pachyrhizi) in Brazil and Paraguay from 2001 to 2003. Phytopathology. 2003, 93: S103-Google Scholar
- Sinclair JB, Hartman GL: Soybean rust. Compendium of soybean diseases. Edited by: Hartman GL, Sinclair JB, Rupe JC. 1999, St. Paul, MN: American Phytopathological SocietyGoogle Scholar
- Mississippi Agricultural and Forestry Experimental Station, Mississippi State University Extension Service.http://msucares.com/crops/soybeans/rust/index.html,
- Calvo ES, Kiihl RAS, Garcia A, Harada A, Hiromoto DM: Two major recessive soybean genes conferring soybean rust resistance. Crop Sci. 2008, 48: 1350-1354. 10.2135/cropsci2007.10.0589.View ArticleGoogle Scholar
- Meyer JDF, Silva DCG, Yang C, Pedley KF, Zhang C, van de Mortel M, Hill JH, Shoemaker RC, Abdelnoor RV, Whitham SA, Graham MA: Identification and analyses of candidate genes for Rpp4-mediated resistance to Asian soybean rust in soybean. Plant Physiol. 2009, 150: 295-307. 10.1104/pp.108.134551.PubMed CentralView ArticlePubMedGoogle Scholar
- Posada-Buitrago ML, Frederick RD: Expressed sequence tag analysis of the soybean rust pathogen Phakopsora pachyrhizi. Fungal Genet Biol. 2005, 42: 949-962. 10.1016/j.fgb.2005.06.004.View ArticlePubMedGoogle Scholar
- Li S, Smith JR, Ray JD, Frederick RD: Identification of a new soybean rust resistance gene in PI 567102B. Theor Appl Genet. 2012, 125: 133-142. 10.1007/s00122-012-1821-y.View ArticlePubMedGoogle Scholar
- Hartwig EE: Identification of a fourth major gene conferring resistance to soybean rust. Crop Sci. 1986, 26: 1135-1136. 10.2135/cropsci1986.0011183X002600060010x.View ArticleGoogle Scholar
- Monteros MJ, Ha BK, Phillips DV, Boerma HR: SNP assay to detect the ‘Hyuuga’ red-brown lesion resistance gene for Asian soybean rust. Theor Appl Genet. 2010, 121: 1023-1032. 10.1007/s00122-010-1368-8.PubMed CentralView ArticlePubMedGoogle Scholar
- Stone CL, McMahon MB, Fortis LL, Nunez A, Smythers GW, Luster DG, Frederick RD: Gene expression and proteomic analysis of the formation of Phakopsora pachyrhizi appressoria. BMC Genomics. 2012, 13: 269-289. 10.1186/1471-2164-13-269.PubMed CentralView ArticlePubMedGoogle Scholar
- Zhang Y, Qu Z, Zheng W, Liu B, Wang X, Xue X, Xu L, Huang L, Han Q, Zhao J, Kang Z: Stage-specific gene expression during urediniospore germination in Puccinia striiformis f. sp tritici. BMC Genomics. 2008, 9: 203-10.1186/1471-2164-9-203.PubMed CentralView ArticlePubMedGoogle Scholar
- Xu J, Linning R, Fellers J, Dickinson M, Zhu W, Antonov I, Joly DL, Donaldson ME, Eilam T, Anikster Y, Banks T, Munro S, Mayo M, Wynhoven B, Ali J, Moore R, McCallum B, Borodovsky M, Saville B, Bakkeren G: Gene discovery in EST sequences from the wheat leaf rust fungus Puccinia triticina sexual spores, asexual spores and haustoria, compared to other rust and corn smut fungi. BMC Genomics. 2011, 12: 161-10.1186/1471-2164-12-161.PubMed CentralView ArticlePubMedGoogle Scholar
- Zahiri AR, Babu MR, Saville BJ: Differential gene expression during teliospore germination in Ustilago maydis. Mol Genet Genomics. 2005, 273: 394-403. 10.1007/s00438-005-1142-9.View ArticlePubMedGoogle Scholar
- Deng Y, Dong H, Jin Q, Dai C, Fang Y, Liang S, Wang K, Shao J, Lou Y, Shi W, Vakalounakis DJ, Li D: Analysis of expressed sequence Tag data and gene expression profiles involved in conidial germination of Fusarium oxysporum. Appl Environ Microb. 2006, 72: 1667-1671. 10.1128/AEM.72.2.1667-1671.2006.View ArticleGoogle Scholar
- Hahn M, Mendgen K: Isolation by ConA binding of haustoria from different rust fungi and comparison of their surface qualities. Protoplasma. 1992, 170: 95-103. 10.1007/BF01378785.View ArticleGoogle Scholar
- Loehrer M, Schaffrath U: Asian soybean rust – meet a prominent challenge in soybean cultivation, soybean. Soybean - Biochemistry, Chemistry and Physiology. Edited by: Tzi-Bun N. 2011, Croatia: InTechGoogle Scholar
- Catanzariti A-M, Dodds PN, Lawrence GJ, Ayliffe MA, Ellis JG: Haustorially expressed secreted proteins from flax rust are highly enriched for avirulence elicitors. Plant Cell. 2006, 18: 243-256. 10.1105/tpc.105.035980.PubMed CentralView ArticlePubMedGoogle Scholar
- Hahn M, Mendgen K: Characterization of In Planta–Induced Rust Genes Isolated from a Haustorium-Specific cDNA Library. Mol Plant Microbe. 1997, 10: 427-437. 10.1094/MPMI.19184.108.40.2067.View ArticleGoogle Scholar
- Puthoff DP, Neelam A, Ehrenfried ML, Scheffler BE, Ballard L, Song Q, Campbell KB, Cooper B, Tucker ML: Analysis of expressed sequence tags from Uromyces appendiculatus hyphae and haustoria and their comparison to sequences from other rust fungi. Phytopathology. 2008, 98: 1126-1135. 10.1094/PHYTO-98-10-1126.View ArticlePubMedGoogle Scholar
- Yin C, Chen X, Wang X, Han Q, Kang Z, Hulbert SH: Generation and analysis of expression sequence tags from haustoria of the wheat stripe rust fungus Puccinia striiformis f. sp. Tritici. BMC Genomics. 2009, 10: 626-10.1186/1471-2164-10-626.PubMed CentralView ArticlePubMedGoogle Scholar
- Marchetti MA, Uecker FA, Bromfield KR: Uredial development of Phakopsora pachyrhizi in soybeans. Phytopathology. 1975, 65: 822-823. 10.1094/Phyto-65-822.View ArticleGoogle Scholar
- Tremblay A, Li S, Scheffler BE, Matthews BF: Laser capture microdissection and expressed sequence tag analysis of uredinia formed by Phakopsora pachyrhizi, the causal agent of Asian soybean rust. Physiol Mol Plant P. 2009, 73: 163-174.View ArticleGoogle Scholar
- Hacquard S, Delaruelle C, Legué V, Tisserant E, Kohler A, Frey P, Martin F, Duplessis S: Laser capture microdissection of uredinia formed by Melampsora larici-populina revealed a transcriptional switch between biotrophy and sporulation. Mol Plant Microbe In. 2010, 23: 1275-1286. 10.1094/MPMI-05-10-0111.View ArticleGoogle Scholar
- Leng W, Liu T, Li R, Yang J, Wei C, Zhang W, Jin Q: Proteomic profile of dormant Trichophyton Rubrum conidia. BMC Genomics. 2008, 9: 303-10.1186/1471-2164-9-303.PubMed CentralView ArticlePubMedGoogle Scholar
- Liu T, Zhang Q, Wang L, Yu L, Leng W, Yang J, Chen L, Peng J, Ma L, Dong J, Xu X, Xue Y, Zhu Y, Zhang W, Yang L, Li W, Sun L, Wan Z, Ding G, Yu F, Tu K, Qian Z, Li R, Shen Y, Li Y, Jin Q: The use of global transcriptional analysis to reveal the biological and cellular events involved in distinct development phases of Trichophyton rubrum conidial germination. BMC Genomics. 2007, 8: 100-10.1186/1471-2164-8-100.PubMed CentralView ArticlePubMedGoogle Scholar
- Wenzler H, Brambl R: Mitochondrial biogenesis during fungal spore germination. Catalytic activity, composition, and subunit biosynthesis of oligomycin-sensitive ATPase in Botryodiplodia. J Biol Chem. 1981, 256: 7166-7172.PubMedGoogle Scholar
- Stade S, Brambl R: Mitochondrial biogenesis during fungal spore germination: respiration and cytochrome c oxidase in Neurospora crassa. J Bacteriol. 1981, 147: 757-767.PubMed CentralPubMedGoogle Scholar
- Gong C-S, Van Etten JI: Changes in soluble ribonucleic acid polymerases associated with the germination of Rhizopus stolonifer spores. Biochimica and Biophysica Acta. 1972, 272: 44-52. 10.1016/0005-2787(72)90031-7.View ArticleGoogle Scholar
- Van Etten JL, Dunkle LD, Knight RH: Nucleic acids and fungal spore germination. The fungal spore: form and function. Edited by: Weber DJ, Hess WH. 1976, New York: John Wiley & Sons, 243-300.Google Scholar
- Henney HR, Storck R: Ribosomes and ribonucleic acids in three morphological states of Neurospora. Science. 1963, 142: 1675-1677. 10.1126/science.142.3600.1675.View ArticlePubMedGoogle Scholar
- Rado TA, Cochrane VW: Ribosomal Competence and Spore Germination in Fusarium solani. J Bacteriol. 1971, 106: 301-304.PubMed CentralPubMedGoogle Scholar
- Leary JV, Morris AJ, Ellingboe AE: Isolation of functional ribosomes and polysomes from lyophilized fungi. Biochimica and Biophysica Acta. 1969, 182: 113-120. 10.1016/0005-2787(69)90526-7.View ArticleGoogle Scholar
- Feit IN, Chu LK, Iverson RM: Appearance of polyribosomes during germination of spores in cellular slime mold Dictyostelium purpureum. Exp Cell Res. 1971, 65: 439-444. 10.1016/0014-4827(71)90024-3.View ArticlePubMedGoogle Scholar
- Kornfield JM: PhD thesis. Structural and physiological aspects of germination of conidia of Penicillium chrysogenum. 1961, Madison: University of WisconsinGoogle Scholar
- Hamer JE, Cotter DA: The timing of ribonucleic acid synthesis during the germination of heat-activated Dictyostelium discoideum spores. Can J Microbiol. 1983, 29: 1390-1398. 10.1139/m83-214.View ArticlePubMedGoogle Scholar
- Roheim JR, Knight RH, Van Etten JL: Synthesis of ribonucleic acids during germination of Rhizopus stolonifer sporangiospores. Dev Biol. 1974, 41: 137-145. 10.1016/0012-1606(74)90289-9.View ArticlePubMedGoogle Scholar
- Hahn M, Deising H, Struck C, Mendgen K: Fungal morphogenesis and enzyme secretion during pathogenesis. Resistance of crop plants against fungi. Edited by: Hartleb H, Heitefuss R, Hoppe H-H. 1997, Germany: Gustav Fischer Verlag, 33-57.Google Scholar
- Struck C, Hahn M, Mendgen K: Plasma membrane H + −ATPase activity in spores, germ tubes, and haustoria of the rust fungus Uromyces viciae-fabae. Fungal Genet Biol. 1996, 20: 30-35. 10.1006/fgbi.1996.0006.View ArticlePubMedGoogle Scholar
- Struck C, Siebels C, Rommel O, Wernitz M, Hahn M: The plasma membrane H + −ATPase from the biotrophic rust fungus Uromyces fabae: molecular characterization of the gene (PMA1) and functional expression of the enzyme in yeast. Mol Plant Microbe In. 1998, 11: 458-465. 10.1094/MPMI.19220.127.116.118.View ArticleGoogle Scholar
- Hahn M, Neef U, Struck C, Göttfert M, Mendgen K: A putative amino acid transporter is specifically expressed in haustoria of the rust fungus Uromyces fabae. Mol Plant Microbe In. 1997, 10: 438-445. 10.1094/MPMI.1918.104.22.1688.View ArticleGoogle Scholar
- Voegele RT, Struck C, Hahn M, Mendgen K: The role of haustoria in sugar supply during infection of broad bean by the rust fungus Uromyces fabae. Proc Natl Acad Sci USA. 2001, 98: 8133-8138. 10.1073/pnas.131186798.PubMed CentralView ArticlePubMedGoogle Scholar
- Voegele RT, Mendgen K: Rust haustoria: nutrient uptake and beyond. New Phytol. 2003, 159: 93-100. 10.1046/j.1469-8137.2003.00761.x.View ArticleGoogle Scholar
- Bartnicki-Garcia S: Chemistry of hyphal walls of Phytophthora. J Gen Microbiol. 1966, 42: 57-69. 10.1099/00221287-42-1-57.View ArticlePubMedGoogle Scholar
- Chet I, Henis Y: Chemical composition of hyphal and sclerotial walls of Sclerotium rolfsii sacc. Can J Microbiol. 1967, 13: 137-141. 10.1139/m67-019.View ArticlePubMedGoogle Scholar
- Chong J, Harder DE, Rohringer R: Ontogeny of mono- and dikaryotic rust haustoria: cytochemical and ultrastructural studies. Phytopathology. 1981, 71: 975-983. 10.1094/Phyto-71-975.View ArticleGoogle Scholar
- Tremblay A, Hosseini P, Li S, Alkharouf NW, Matthews BF: Identification of genes expressed by Phakopsora pachyrhizi, the pathogen causing soybean rust, at a late stage of infection of susceptible soybean leaves. Plant Pathol. 2011, 61: 773-786.View ArticleGoogle Scholar
- Cheung M-S, Down TA, Latorre I, Ahringer J: Systematic bias in high-throughput sequencing data and its correction by BEADS. Nucleic Acids Res. 2011, 39: e103-10.1093/nar/gkr425.PubMed CentralView ArticlePubMedGoogle Scholar
- Li S, Moore WF, Spinks BL, Wells BC, Sciumbato GL, Robinson SJ, Libous-Bailey L: Occurrence of Asian soybean rust caused by Phakopsora pachyrhizi in Mississippi. Plant Health Prog. 2007, doi:10.1094/PHP-2007-0917-02-BRGoogle Scholar
- Hosseini P, Tremblay A, Matthews BF, Alkharouf NW: An efficient annotation and gene-expression derivation tool for Illumina Solexa Datasets. BMC Res Notes. 2010, 3: 183-10.1186/1756-0500-3-183.PubMed CentralView ArticlePubMedGoogle Scholar
- Davidsen T, Beck E, Ganapathy A, Montgomery R, Zafar N, Yang Q, Madupu R, Goetz P, Galinsky K, White O, Sutton G: The comprehensive microbial resource. Nucleic Acids Res. 2010, 38 (database issue): D340-D345.PubMed CentralView ArticlePubMedGoogle Scholar
- Zerbino DR, Birney E: Velvet: Algorithms for de novo short read assembly using de Bruijn graphs. Genome Res. 2008, 18: 821-829. 10.1101/gr.074492.107.PubMed CentralView ArticlePubMedGoogle Scholar
- Schulz MH, Zerbino D, Vingron M, Birney E: Oases: robust de novo RNA-seq assembly across the dynamic range of expression levels. Nucleic Acids Res. 2012, 28: 1086-1092.Google Scholar
- Balwierz PJ, Carninci P, Daub CO, Kawai J, Hayashizaki Y, Van Belle W, Beisel C, van Nimwegen E: Methods for analyzing deep sequencing expression data: constructing the human and mouse promoteome with deepCAGE data. Genome Biol. 2009, 10: R79-10.1186/gb-2009-10-7-r79.PubMed CentralView ArticlePubMedGoogle Scholar
- Mortazavi A, Williams BA, McCue K, Schaeffer L, Wold B: Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods. 2008, 5: 621-628. 10.1038/nmeth.1226.View ArticlePubMedGoogle Scholar
- Roberts A, Trapnell C, Donaghey J, Rinn JL, Pachter L: Improving RNA-Seq expression estimates by correcting for fragment bias. Genome Biol. 2011, 12: R22-10.1186/gb-2011-12-3-r22.PubMed CentralView ArticlePubMedGoogle Scholar
- FTP root. ftp.ncbi.nih.govGoogle Scholar
- van de Mortel M, Recknor JC, Graham MA, Nettleton D, Dittman JD, Nelson RT, Godoy CV, Abdelnoor RV, Almeida AMR, Baum TJ, Whitham SA: Distinct biphasic mRNA changes in response to Asian soybean rust infection. Mol Plant Microbe In. 2007, 20: 887-899. 10.1094/MPMI-20-8-0887.View ArticleGoogle Scholar
- Rutledge RG, Stewart D: A kinetic-based sigmoidal model for the polymerase chain reaction and its application to high-capacity absolute quantitative real-time PCR. BMC Biotechnol. 2008, 8: 47-10.1186/1472-6750-8-47.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.