A proteomic analysis of powdery mildew (Blumeria graminis f.sp. hordei) conidiospores

SUMMARY

Conidiospores are the asexual propagation units of many plant‐pathogenic fungi. In this article, we report an annotated proteome map of ungerminated conidiospores of the ascomycete barley powdery mildew pathogen, Blumeria graminis f.sp. hordei. Using a combination of two‐dimensional polyacrylamide gel electrophoresis and matrix‐assisted laser desorption ionization‐time‐of‐flight mass spectrometry, we have identified the proteins in 180 spots, which probably represent at least 123 distinct fungal gene products. Most of the identified proteins have a predicted function in carbohydrate, lipid or protein metabolism, indicating that the spore is equipped for the catabolism of storage compounds as well as for protein biosynthesis and folding on germination.

INTRODUCTION

Blumeria graminis f.sp. hordei (Bgh) is the causal agent of powdery mildew in barley (Collins et al., 2002). It is one of the major diseases affecting this monocotyledonous plant species, leading to a decrease in yield and quality in barley agriculture. Bgh is one of the most intensively studied powdery mildew fungi and, like all species of the Erysiphales (powdery mildews), it is an obligate biotrophic pathogen, requiring a living plant host for growth and propagation (Both and Spanu, 2004). Initial events in the asexual life cycle of the fungus comprise spore germination, appressorium formation (Fig. 1A), host cell penetration and the intracellular establishment of its feeding organ, the so‐called haustorium (Green et al., 2002). Following successful host cell entry, the fungal mycelium proliferates on the leaf surface and, from about 5 days post‐inoculation onwards, numerous spore carriers (conidiophores) emerge to produce large quantities of asexual conidia that are spread to other host plants via the wind (Fig. 1B; Green et al., 2002).

Figure 1.

Scanning electron micrographs of Blumeria graminis f.sp. hordei (Bgh) conidiophores and germinated conidia. The micrographs depict a germinated conidiospore on its host barley (A) (scale bar, 10 µm) and mature conidiophores bearing terminal strings of conidia (B) (scale bar, 20 µm). A, appressorium; C₁, conidium; C₂, conidiophore; H, hypha; P, primary germ tube; S, secondary germ tube. Pictures are courtesy of Sarah Maria Schmidt and Elmon Schmelzer.

Despite the extensive research carried out on the barley powdery mildew pathogen, the basic molecular biology of Bgh remains poorly understood (Zhang et al., 2005). Its obligate biotrophic lifestyle and our current inability to cultivate the fungus in vitro or effectively manipulate it genetically severely limit the experimental repertoire. Consequently, thorough experimental examination of the powdery mildews at the molecular level has lagged significantly behind that of other pathogenic fungi, in spite of their economic and agricultural importance. Only recently, two studies based on transcript profiling using cDNA microarrays comprising 2027 Bgh genes allowed new insights into Bgh biology. The first study focused on the determinants of pathogenicity (Both et al., 2005b), revealing a global shift in gene expression between the pre‐ and post‐penetrative stages of the fungus. In particular, genes encoding polypeptides related to protein biosynthesis exhibited a dramatic increase in transcript abundance, suggesting that the formation of post‐penetrative infection structures and the establishment of a stable pathogenic relationship require high levels of de novo protein biosynthesis. The second analysis highlighted a striking degree of apparent co‐expression of genes encoding enzymes of particular primary metabolic pathways. Although several genes encoding components of glycolysis were found to be coordinately up‐regulated during appressorium formation, transcripts of enzymes involved in lipid catabolism were abundant in conidia and decreased at later developmental stages (Both et al., 2005a).

To strengthen our knowledge about obligate biotrophic fungi, and to corroborate or complement transcriptome analyses, it is important to determine the protein equipment of these organisms at various developmental stages. Proteomic technologies, such as protein separation by two‐dimensional gel electrophoresis and peptide analysis by mass spectrometry (MS), generally enable the efficient identification and characterization of proteins in filamentous fungi (for a review, see Kim et al., 2007). Nevertheless, obligate biotrophs, such as powdery mildews, downy mildews and rust fungi, cannot be cultured extensively in vitro, thereby generally complicating proteomic studies. A lack of comprehensive cDNA data and an absence of partial or full genome sequences further hamper proteome profiling in most cases. Consequently, few proteomic surveys of obligate biotrophs have been reported to date (Cooper et al., 2006, 2007; Rampitsch et al., 2006).

The only period of development when powdery mildew fungi are free of any biotrophic relationship with their host is the conidium stage, enabling extensive analysis of this developmental phase. Conidia of powdery mildews are asexual, non‐motile fungal spores that are dispersed by the wind. They are one‐celled, uninucleate, vacuolated and thin‐walled organisms (Braun et al., 2002) that contain glycogen and abundant lipid droplets as storage compounds to fuel germination on contact with the host (Both et al., 2005a; McKeen, 1970). Bgh conidia are produced in chains that are connected to the colony via upright spore carriers (conidiophores; Fig. 1B). The spores normally do not germinate as long as they are attached to the conidiophore, but they do so rapidly on separation. This physiological characterization suggests that conidia are not entirely passive, but that inhibitory processes suppress germination of spores attached to the mother colony (Green et al., 2002). Further evidence for pre‐germination conidial activity is provided by studies on the secretion of conidial extracellular material (ECM). This substance seems to be released from spores immediately on contact with the leaf surface. ECM exhibits hydrolytic enzyme activities and may contain preformed enzymes, such as esterases, lipases or hydrolases (Green et al., 2002).

Little is known about the complex physiological changes that occur during fungal spore germination (Osherov and May, 2001). This, in particular, applies to conidia of fungal phytopathogens, which have only recently become the subject of detailed molecular analyses. Global transcript profiling of Bgh and Fusarium graminearum have revealed first snapshots of the changes in gene expression during the early developmental stages of fungal plant parasites (Both et al., 2005a,b; Seong et al., 2008). Reminiscent of the findings in Bgh (see above), the analysis of the F. graminearum transcriptome uncovered a major involvement of genes related to peroxisome biogenesis and fatty acid metabolism in ungerminated spores, whereas germinated spores showed a shift towards the expression of genes involved in metabolism and energy functions using carbon compounds and carbohydrates. In addition to genes related to this metabolic change, F. graminearum germinated spores exhibited elevated transcript levels of genes related to transcription and translation machineries, the cell cycle, cellular transport and cellular biogenesis (Seong et al., 2008). At present, only a few proteomic studies have complemented these transcriptome analyses and have provided initial insights into the respective fungal protein equipment (Cooper et al., 2006, 2007). Notably, however, the comparative proteomic analysis of ungerminated and germinated spores of the biotrophic bean rust fungus, Uromyces appendiculatus, revealed only a few significant differences between these two developmental stages, suggesting that the proteome of ungerminated fungal spores represents a valuable approximation of the protein inventory of germinated sporelings (Cooper et al., 2006, 2007).

In this article, we report, to our knowledge, the first proteome analysis of a biotrophic powdery mildew fungus. We have focused on ungerminated asexual conidia, as these fungal structures can be readily collected and provide a homogeneous biological source material of high purity. By combining two‐dimensional gel electrophoresis with matrix‐assisted laser desorption ionization time‐of‐flight mass spectrometry (MALDI‐TOF MS) and tandem time‐of‐flight mass spectrometry (MALDI‐TOF/TOF MS/MS), we were able to determine the identity of 180 protein spots and to establish a functionally annotated reference proteome map for Bgh conidia.

RESULTS AND DISCUSSION

Establishment of a proteome map of Bgh conidia

Two batches of approximately 600 mg of fresh Bgh conidia were collected from approximately 200 barley plants that had been inoculated 7 days earlier. From each replicate, approximately 1–2 mg of total protein were extracted and purified by trichloroacetic acid precipitation and phenol extraction. Aliquots (approximately 160 µg from the first batch) of the purified protein samples were subjected to two‐dimensional gel electrophoresis as described in ‘Experimental procedures’; the second batch of samples was used as replicates. The samples were analysed in three different pI ranges (pI 3–10, 4–7 and 6–10) on colloidal Coomassie blue‐stained gels (Fig. 2). Initially, 124 spots were cut out and analysed from the pI 3–10 range gel, 136 from the pI 4–7 range gel and 56 from the pI 6–11 range gel, adding up to a total of 316 excised spots (Fig. 3). Of these, 230 (116 from pI 3–10, 70 from pI 4–7 and 44 from pI 6–10) were identified by MS analysis as described below, giving rise to an overall identification rate of 73% (Fig. 3). As judged by manual inspection and comparison of the three gel types, 133 spots were identified from a gel in a single pI range, and 47 were identified from gels in multiple pI ranges (Fig. 4). This partial redundancy resulted in a final number of 180 distinct identified spots, represented by the entirety of 230 characterized spots from the three gel types (3, 4). Owing to the two‐dimensional gel electrophoresis separation procedure employed, these are expected to comprise predominantly soluble cytoplasmic proteins.

Figure 2.

Coomassie blue‐stained two‐dimensional gels of Blumeria graminis f.sp. hordei (Bgh) protein extracts. The photographs show two‐dimensional gel electrophoresis gels of 160 µg total Bgh protein separated on a pI 3–10 (A), pI 4–7 (B) or pI 6–10 (C) gel. Please note that the following spots co‐migrate: 20 and 99, 28 and 166, 67 and 87, 79 and 157, 132 and 148, 139 and 176 (A); 20 and 99, 52 and 118 (B); 18 and 54, 19 and 159, and 35, 153 and 163 (C).

Figure 3.

Flow chart of protein identification. The diagram illustrates the course of the protein identification process. EST, expressed sequence tag; MS, mass spectrometry; NCBInr, National Center for Biotechnology Information non‐redundant database.

Figure 4.

Venn diagram showing the distribution of protein spots in the various gel types. The diagram illustrates the dispersal and overlap of the identified polypeptide spots in the various two‐dimensional gel electrophoresis gels having separate pI ranges.

Protein identification

All protein spots were processed by automated in‐gel tryptic digestion and MALDI‐TOF MS analysis to generate peptide mass fingerprints (PMFs). Sequence information for selected peptides was then obtained from the samples by MALDI‐TOF/TOF MS/MS using laser‐induced dissociation (Suckau et al., 2003). Co‐migration of multiple proteins in two‐dimensional gel electrophoresis was observed in the case of 10 of the 230 spots successfully analysed in this study. In each of these instances, two or three distinct polypeptides were identified by a combination of MALDI‐TOF and TOF/TOF analysis (Fig. 2; Table 1). Initially, the spectra of 113 protein spots were directly assigned to at least one known protein (GI accession) by screening the National Center for Biotechnology Information non‐redundant (NCBInr) database. This procedure revealed mostly hits from fungi other than Bgh, indicating that these polypeptides represent conserved fungal proteins (Table 1). The spectra of all spots were then searched against publicly available Bgh expressed sequence tags (NCBI ESTs) and the Bgh draft genome sequence (assembly from June 2007) kindly provided by P. Spanu and coworkers (Imperial College, London, UK; unpublished data) (http://www.blugen.org/). These searches revealed that more than one‐half of the previously assigned spectra (73 of 113) also matched to Bgh ESTs and, furthermore, led to the identification of an additional 44 proteins based on Bgh ESTs and 23 proteins based on open reading frames (ORFs) predicted from Bgh genomic contigs (Supplement S1, see ‘Supporting Information’). For the latter two modes of identification the corresponding translated sequences were used for blast searches against the NCBInr database (http://blast.ncbi.nlm.nih.gov/Blast.cgi) and the best hit was used to annotate the particular polypeptide (Tables 1 and S1, and Table S1 caption, see ‘Supporting Information’). In total, 117 annotated proteins matched to Bgh ESTs (73 Bgh homologues of hits against the NCBInr database plus 44 ‘EST‐only’ hits), meaning that 63 (c. 35%) are proteins that are currently not represented by Bgh ESTs of sufficient length or quality (Fig. 3). This clearly underscores the value of investigations at the proteome level as a useful complement to transcriptome studies and/or genome annotation. The final annotation of the Bgh spore proteome (comprising 180 identified spots as described above) revealed a total of 174 spots with either direct or indirect hits to polypeptides, whereas six spots resulted in matches to Bgh ESTs with no counterpart in the protein databases (Table 1, spots 174–179). Of the 174 protein hits, 11 (6%) matched directly to published Bgh proteins, 153 (88%) corresponded to homologues from other fungal species and only 10 (6%) matched homologues of non‐fungal organisms (e.g. plants, protists, bacteria and animals). The observed paucity of Bgh‐derived protein hits is not surprising as only 51 Bgh polypeptide entries are currently (June 2008) present in the NCBInr database. It is conceivable that the proteins that do not match to fungal proteins result from organisms contaminating our biological source material.

Table 1.

Blumeria graminis f.sp. hordei (Bgh) proteins identified in this study†.

Spot name‡	2D gel§	Accession/EST/ genomic Bgh sequence	Match with Bgh EST¶	Description††	Species††	pI/obs.‡‡	MM (kDa)/obs‡‡	Cov. [%]§§	Mascot score§§
Carbohydrate metabolism
1	A,B	GI:74612186	✓	Enolase	Tuber borchii	5.3	47.9	8	342
2	A	GI:74612186	✓	Enolase	Tuber borchii	5.3	47.9	13	376
3	B	GI:74612186	✓	Enolase	Tuber borchii	5.3	47.9	8	332
4	A,B	GI:115390717	✓	Pyruvate kinase	Aspergillus terreus	6.7	58.1	2.3	112
5	A,B	GI:115390717	✓	Pyruvate kinase	Aspergillus terreus	6.7	58.1	2.3	55
6	A	AW789899		Triosephosphate isomerase (GI:156040910)	Sclerotinia sclerotiorum	5.5	27	n.a.	164 *
7	B	AW789899		Triosephosphate isomerase (GI:156040910)	Sclerotinia sclerotiorum	5.5	27	n.a.	152 *
8	B	GI:52352519	✓	Phosphoglucomutase (PGM1)	Saccharomyces kudriavzevii	5.3	60.6	3.5	76
9	A,B	GI:70986482	✓	Pyruvate dehydrogenase E1 β subunit, hyp. prot.	Aspergillus fumigatus	7.6/5	40.5/35	6.6	151
10	A	GI:115390821		Pyruvate dehydrogenase E1 α subunit	Aspergillus terreus	8.5/6	45.3	4.4	81
11	A,B	GI:110763826	✓	Phosphoglycerate kinase, hyp. prot.	Apis mellifera	9	45	3.4	65
12	B	BQ284104		Phosphoglyceromutase, hyp. prot. (GI:156042436)	Sclerotinia sclerotiorum	5.1	57.2	n.a.	180
13	A,B	GI:2494638	✓	Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH)	Blumeria graminis	6	36.5	79	633
14	A	GI:2494638		Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH)	Blumeria graminis	6/5.7	36.5/40	7.1	170
15	A	GI:2494638	✓	Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH)	Blumeria graminis	6/6.5	36.5/40	76.9	680
16	A	GI:2494638	✓	Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH)	Blumeria graminis	6	36.5/40	58	173
17	B	GI:2494638	✓	Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH)	Blumeria graminis	6	36.5/40	67.2	342
18	C	GI:2494638		Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH)	Blumeria graminis	6/8	36.5/40	7.1	129
19	C	GI:2494638		Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH)	Blumeria graminis	6/7.5	36.5/40	10.7	181
20	A,B	AW790282		Fructose‐bisphosphate aldolase (GI:154314989)	Botryotinia fuckeliana	5.4	39	n.a.	327
21	A	GI:156036300		Pyruvate carboxylase, hyp. prot.	Sclerotinia sclerotiorum	6.1	132.5	14.6	86 *
22	B	Contig_9627		Glucose‐6‐phosphate isomerase, putative (GI:111057935)	Phaeosphaeria nodorum	5.9	61	22.2	89 *
23	A,B	GI:68481811	✓	Transaldolase	Candida albicans	4.8	35.7	6.5	102
24	A,B	GI:68481811	✓	Transaldolase	Candida albicans	4.8	35.7	3.7	63
25	A	GI:50423907		Transketolase, hyp. prot.	Debaryomyces hansenii	6.1	75.4	3	117
26	A	GI:119469246	✓	6‐Phosphogluconate dehydrogenase, decarboxylating	Neosartorya fischeri	5.9	55.8	7.3	145
27	B	GI:85115938	✓	6‐Phosphogluconate dehydrogenase, hyp. prot.	Neurospora crassa	6.3/5.7	57.2	2.3	75
28	A	Contig_5253		6‐Phosphogluconolactonase, hyp. prot. (GI:156051562)	Sclerotinia sclerotiorum	5.5	28.9	50.4	143
29	A	GI:118083730	✓	Glycogen branching enzyme, hyp. prot.	Gallus gallus	6	80.1	1.7	78
30	A	Contig_2726		Glycosidase, putative (GI:154314369)	Botryotinia fuckeliana	5.8	174	32.4	247 *
31	A	GI:39947727	✓	Glycosyl hydrolase, hyp. prot.	Magnaporthe grisea	5.8	110.6	1.2	61
32	A,B	GI:67537722	✓	Malate dehydrogenase, hyp. prot.	Aspergillus nidulans	5.7	33.6	3.4	96
33	A,B	GI:67537722	✓	Malate dehydrogenase, hyp. prot.	Aspergillus nidulans	5.7	33.6	3.4	84
34	B	AW788156		Malate dehydrogenase, hyp. prot. (GI:156048488)	Sclerotinia sclerotiorum	6.7/5.5	34.5	n.a.	87
35	A,B,C	GI:5123836	✓	Malate dehydrogenase	Nicotiana tabacum	9/7	43.3	5.6	114
36	B	AW788156		Malate dehydrogenase, hyp. prot. (GI:156048488)	Sclerotinia sclerotiorum	6.7/5	34.5	n.a.	73
37	C	GI:5929964	✓	Malate dehydrogenase	Glycine max	9.1	36.1	3.5	121
38	A	GI:7160184	✓	ATP citrate lyase, subunit 2	Sordaria macrospora	5.5/7	52.2	3.5	134
39	C	GI:70992211	✓	ATP citrate lyase subunit, hyp. prot.	Aspergillus nidulans	5.8/7	53	31	216
40	A	GI:119471597	✓	ATP citrate lyase, subunit 1, hyp. prot.	Neosartorya fischeri	9.3	71.9/62	6.4	316
41	A	GI:121699808	✓	ATP citrate synthase subunit 1	Magnaporthe grisea	9.2	72.4/62	26	100 *
42	C	GI:116500157	✓	Citrate synthase type I, hyp. prot.	Coprinopsis cinerea	9.5/8	16.8/45	9.3	103
43	C	Contig_11872_107 + 11872_116		Citrate synthase (GI:156063018)	Sclerotinia sclerotiorum	7.2	50	13.8	91
44	C	Contig_7973		Isocitrate dehydrogenase, hyp. prot. (GI:119183931)	Coccidioides immitis	8.9/7	41	41.8	179
45	A	GI:27803048	✓	Aconitate hydratase, hyp. prot.	Podospora anserina	6.4	85.1	1.7	71
46	A	Contig_3948		α‐Ketoglutarate decarboxylase (GI:156054172)	Sclerotinia sclerotiorum	6.3/8	118	17.2	106 *
47	C	GI:67538802		α‐Ketoglutarate dehydrogenase, subunit, hyp. prot.	Aspergillus nidulans	6.4/7	117/97	9.3	75
48	A	GI:67524917	✓	Succinate dehydrogenase flavoprotein subunit, hyp. prot.	Aspergillus nidulans	6.1	69.4	22.3	74 *
49	B	GI:71000275		Succinate dehydrogenase subunit, hyp. prot.	Aspergillus fumigatus	6.5	71.1	2.3	72
50	A	GI:14550183	✓	NADP‐dependent mannitol dehydrogenase	Cladosporium fulvum	6.1	28.6	4.1	54
51	B,A	GI:14550183	✓	NADP‐dependent mannitol dehydrogenase	Cladosporium fulvum	6.1	28.6	4.1	50
52	C,B	GI:14550183	✓	NADP‐dependent mannitol dehydrogenase	Cladosporium fulvum	6.1	28.6/50	4.1	47
53	C	GI:46107282	✓	UDP‐glucose pyrophosphorylase, hyp. prot.	Gibberella zeae	6.1	57.1	7.8	267
54	A,C	AW788419		Nucleoside‐diphosphate‐sugar epimerase, hyp. prot. (GI:111063143)	Phaeosphaeria nodorum	6.3/8	38/40	n.a.	220
55	A	GI:111070271		Serine hydroxymethyltransferase, hyp. prot.	Phaeosphaeria nodorum	9.2	51.6	6.9	170
56	A,B	Contig_14052		Dehydrogenase acyltransferase (GI:156031305)	Sclerotinia sclerotiorum	5.5	48	59.8	156 *
Protein metabolism and modification
57	A	AW792643		Serine carboypeptidase, hyp. prot. (GI:85081820)	Neurospora crassa	5.2/4.8	62/50	n.a.	448
58	A,B	AW792643		Serine carboypeptidase, hyp. prot. (GI:85081820)	Neurospora crassa	5.2/4.8	62/49	n.a.	402
59	A	GI:119178714	✓	Zn‐dependent oligopeptidase, hyp. prot.	Coccidioides immitis	5.7	82.1	1.4	75
60	A,B	BM361215		Aminopeptidase (GI:121707310)	Aspergillus clavatus	5.1	98.4	n.a.	119
61	B	AW787940		Aminopeptidase, hyp. prot. (GI:154296638)	Botryotinia fuckeliana	5.9	56.4/25	n.a.	39
62	B	GI:146452516		Secreted/periplasmic Zn‐dependent peptidase, hyp. prot.	Lodderomyces elongisporus	6.6	130.6/95	1.2	27 (84)
63	A,B,C	BQ284067		Cyclophilin, hyp. prot. (GI:154313502)	Botryotinia fuckeliana	5.2	31.6	n.a.	312
64	A	AW788820		Cyclophilin 1, hyp. prot. (GI:33357681)	Botryotinia fuckeliana	9.1	24.2	n.a.	194 *
65	A	GI:27805447	✓	Cyclophilin	Paramecium primaurelia	9.6/7	17.7/45	7.4	68
66	B	GI46110292	✓	ATP‐dependent 26S proteasome regulatory subunit, hyp. prot.	Gibberella zeae	5	28.6/60	6.5	100
67	A	GI:46138509		26S protease regulatory subunit 6B, hyp. prot.	Gibberella zeae	5.3	46.9	52	135 *
68	A	GI:119496777		DNAJ‐class molecular chaperone, hyp. prot.	Neosartorya fischeri	5.7	45.4	3.6	73
69	C	GI:119189679		DNAJ‐class molecular chaperone, hyp. prot.	Coccidioides immitis	6	45.3	3.6	44
70	A	GI:119182507	✓	Heat shock protein 60	Coccidioides immitis	5.4	62.4	33.8	92 *
71	A	GI:39944360		Heat shock protein 70, hyp. prot.	Magnaporthe grisea	5.8	72.6	3.7	207
72	B	GI:121715382		Heat shock protein 70	Aspergillus clavatus	5.7	72.5	2.3	36 (215)
73	B	GI:39944360		Heat shock protein 70, hyp. prot.	Magnaporthe grisea	5.8	72.6	15.8	38 (199)
74	B	GI:116182094		Heat shock protein 70	Chaetomium globosum	5.1	67	5.6	278
75	A	GI:70983346		Heat shock protein 70, hyp. prot.	Aspergillus fumigatus	5.2	66.9	21	312
76	A,B	GI:42477037	✓	Heat shock protein 70	Trichophyton verrucosum	4.9	70.8	38.1	147 *
77	A	GI:123666	✓	Heat shock protein 83	Trypanosoma brucei brucei	5	80.7	5.3	233
78	B	GI:9837418	✓	Heat shock protein 90	Tetrahymena pyriformis	5	80.8	3.4	111
79	A,C	GI:85078476	✓	40S ribosomal protein S3	Neurospora crassa	9.6	28.8	43.7	178 *
80	C	GI:85078476		40S ribosomal protein S3	Neurospora crassa	9.6	28.8	36.9	79 *
81	A	GI:115402405	✓	40S ribosomal protein S2	Aspergillus terreus	11.1	28.2	5.4	62
82	A	GI:46123785	✓	40S ribosomal protein S10, hyp. prot.	Gibberella zeae	10.2	19.1	11.2	278
83	A	GI:46138661	✓	50S ribosomal protein, hyp. prot.	Gibberella zeae	11.3/9.8	15.4	6.6	55
84	A,B	GI:119177270	✓	Eukaryotic translation/initiation factor 4	Coccidioides immitis	4.9/5.4	44.8	38	304
85	B	GI:71000162		Eukaryotic translation initiation factor 4, hyp. prot.	Aspergillus fumigatus	4.9	45.7	7.4	127
86	A	GI:729823	✓	Eukaryotic translation initiation factor 5A	Neurospora crassa	5.4	18.1	7.4	105
87	A	AW790050	✓	Elongation factor 1γ (GI:85101774)	Neurospora crassa	5.5	45.4	n.a.	273
88	A	GI:17027007	✓	Elongation factor‐1α	Pycnocentrodes sp.	9.5	39.4	3	63
89	A,C	GI:17027007	✓	Elongation factor‐1α	Pycnocentrodes sp.	9.5	39.4	6.1	91
90	A,C	GI:85106981		Elongation factor 2	Neurospora crassa	6.3	93.2	26.1	145
91	B,C	GI:85106981		Elongation factor 2	Neurospora crassa	6.3	93.2/82	14.2	104
92	A	GI:83765350	✓	Elongation factor 3, conserved domains	Aspergillus oryzae	5.9	117.8	4.8	214
93	C	AW791437		Acetyl‐CoA acetyltransferase, hyp. prot. (GI:156040271)	Sclerotinia sclerotiorum	6.6	41.8	n.a.	92
94	A	GI:39940094	✓	Cell division control protein (Cdc48)	Magnaporthe grisea	4.8	89.9	39.9	194
95	A	AW790865		Serine protease 2 (GI:17385556)	Pyrenopeziza brassicae	5.8	55.6	n.a.	301
96	A,B	GI:85080590	✓	Glucose‐regulated protein homologue precursor (GRP 78)	Neurospora crassa	4.9	72.3	3.2	104
97	A	BM361602		Actin depolymerization factor/cofilin‐like, hyp. prot. (GI:46123735)	Gibberella zeae	5.2/6	16.1	n.a.	94
98	C	contig_10677		UDP‐N‐acetylglucosamine pyrophosphorylase (GI:156053648)	Sclerotinia sclerotiorum	5.4/7	56	24.6	70*
Amino acid metabolism
99	A,B	AW789533		Glutamine synthetase (GI:154298092)	Sclerotinia sclerotiorum	5.5	40.1	n.a.	103
100	B	Contig_1225		Argininosuccinate synthetase (GI:156057843)	Sclerotinia sclerotiorum	5.1	45	38.2	168 *
101	A	GI:70995343	✓	S‐Adenosylmethionine synthetase	Aspergillus fumigatus	5.6	42.2	21	211
102	A,B	GI:119495861	✓	S‐Adenosylmethionine synthetase	Neosartorya fischeri	5.5	42.1	26	170
103	A,C	GI:116205243		Chorismate synthase, hyp. prot.	Chaetomium globosum	6.5	45.6	21	45, 77*
104	A,C	GI:39946672	✓	Saccharopine dehydrogenase	Magnaporthe grisea	6.1/7	49.1	3.8	73
105	B	Contig_7205		Glycine dehydrogenase, hyp. prot. (GI:156058930)	Sclerotinia sclerotiorum	6	116/95	18.3	90 *
106	A	Contig_11926		Putative dehydrogenase/cyclohydrolase (GI:156059764)	Sclerotinia sclerotiorum	6.3/8	101	18.1	150 *
107	B,C	Contig_5443		Cystathionine gamma‐lyase, hyp. prot. (GI:156063172)	Sclerotinia sclerotiorum	6.2	45	37.2	164 *
108	A,C	GI:70999940	✓	Serine hydroxymethyltransferase, hyp. prot.	Aspergillus fumigatus	8.6	51.8	7.6	198
109	C	GI:85094603	✓	Serine hydroxymethyltransferase	Neurospora crassa	7.1	52.9	3.3	132
110	C	AW790371		Aminomethyltransferase, hyp. prot. (GI:156053183)	Sclerotinia sclerotiorum	9.1/7.5	50.7	n.a.	113 *
111	A	GI:46138463		Aspartate/tyrosine/aromatic aminotransferase, hyp. prot.	Gibberella zeae	9.3	46.6	3.8	57
112	B	contig_5844		Aspartate/tyrosine/aromatic aminotransferase, hyp. prot. (GI:156030852)	Sclerotinia sclerotiorum	5.4	46	37.8	123
113	A	GI:47132400	✓	Methionine synthase	Pichia pastoris	5.8/6.5	85.9	1.3	81
114	A	GI:6969602	✓	Methionine synthase	Cladosporium fulvum	6.4	87.1	2.7	156
115	B	GI:68475194	✓	Putative cobalamin‐independent methionine synthase	Candida albicans	5.3/6.5	85.6	1.3	70
Nucleoside, nucleotide and nucleic acid metabolism
116	C	GI:67538624	✓	GTP‐binding nuclear protein (RAN/TC4)	Aspergillus nidulans	7.7/8.5	24.9/30	6.3	73
117	A	AW789762		Polyadenylate‐binding protein (GI:115390925)	Aspergillus terreus	5.7	81.8	n.a.	63
118	A,B	GI:85107784	✓	Adenosylhomocysteinase, hyp. prot.	Neurospora crassa	5.8	49	28.5	116 *
119	B	BM361451		Adenosylhomocysteinase (GI:156062152)	Sclerotinia sclerotiorum	5.6	48.9	n.a.	79 *
120	A	BM361451		Adenosylhomocysteinase (GI:156062152)	Sclerotinia sclerotiorum	5.6	48.9	n.a.	68 *
121	C	GI:115442686	✓	Inosine‐5’‐monophosphate dehydrogenase	Aspergillus terreus	6.6	58.1	2.7	68
122	A	Contig_6578		Adenosine kinase (GI:154311437)	Botryotinia fuckeliana	5.4	37	50.4	163 *
123	A	Contig_4185		Adenylate kinase (GI:156062980)	Sclerotinia sclerotiorum	8.6	28	36.9	90 *
124	B	Contig_10205		Purine biosynthesis protein, hyp. prot. (GI:85111684)	Sclerotinia sclerotiorum	6	65	40.2	79 *
125	A	GI:46125253		Vacuolar ATPase subunit A, hyp. prot.	Gibberella zeae	6.7/5.3	91.9/62	30.8	96 *
126	B	GI:111060400		Vacuolar ATPase subunit A, hyp. prot.	Phaeosphaeria nodorum	6.6/5.3	43.8/65	24.2	78 *
127	A	GI:46107508		Vacuolar ATPase subunit B	Gibberella zeae	5.3	56.6	31.5	117 *
128	A	GI:111606563		ATP synthase β chain	Coccidioides posadasii	5.1	55.9	46.8	115 *
129	B	GI:119182499		ATP synthase β chain	Coccidioides immitis	5.1	55.8	16.1	85 *
130	B	GI:46116940		ATP synthase β chain	Gibberella zeae	5.3	54.9	19.7	66*
131	C	AW787925		ATP synthase γ chain, hyp. prot. (GI:156045651)	Sclerotinia sclerotiorum	8.3/6.5	32.1	n.a.	61 *
132	A	GI:171110		F1‐ATPase β subunit	Saccharomyces cerevisiae	5.6	54.9	5.7	147
133	A,C	GI:39972915		F1‐ATPase α‐subunit	Magnaporthe grisea	9.8/8.5	59.5	39	244
134	A	GI:70990878	✓	ADP, ATP carrier protein, hyp. prot.	Aspergillus fumigatus	10.3/8.3	33.3/25	5.2	177
Lipid, fatty acid and steroid metabolism
135	A	Contig_591_57 + _591_52		NADH‐cytochrome b5 reductase (GI:154297211)	Botryotinia fuckeliana	9.1	38	49	178
136	C	Contig_4916		Acyl‐CoA synthetase (GI:154290515)	Botryotinia fuckeliana	6.9	77	24.4	86
137	C	Contig_4916		Acyl‐CoA synthetase (GI:154290515)	Botryotinia fuckeliana	6.9	77	9	72
138	A,B	BM361770		Esterase lipase, hyp. prot. (GI:85093411)	Neurospora crassa	7.5/5	121.6/60	n.a.	129
Immunity and defence
139	A	GI:18033107	✓	Superoxide dismutase	Blumeria graminis	6.4	24.2	56	354
140	A,B	BM360950		Peroxiredoxin, hyp. prot. (GI:154312864)	Botryotinia fuckeliana	5.3	16	n.a.	151
141	A	BM360950		Peroxiredoxin, hyp. prot. (GI:154312864)	Botryotinia fuckeliana	5.7	16.3	n.a.	43
142	A,C	GI:70982394	✓	Oxidoreductase related to apoptosis‐inducing factor like, hyp. prot.	Aspergillus fumigatus	8.6	59.8	2.2	70
143	C	AW792600		Oxidoreductase related to apoptosis‐inducing factor like, hyp. prot. (GI:154304889)	Botryotinia fuckeliana	6.3	58.1	n.a.	51
144	A	GI:18033118	✓	Catalase/peroxidase	Blumeria graminis	7.3	87.5	59.3	301 *
145	B	GI:18033118	✓	Catalase/peroxidase	Blumeria graminis	7.3	87.5	24.7	141 *
146	A	GI:71012754		Ascorbate peroxidase, hyp. prot.	Ustilago maydis	9/6	43.1	4.8	128
Intracellular protein traffic
147	A	GI:111060927	✓	Hyp. prot. with RAN binding site	Phaeosphaeria nodorum	4.8	28.1 /40	13.7	214
148	A	GI:135481		Tubulin β chain	Blumeria graminis	4.7	49.7	46.2	134 *
149	A,B	GI:67540744		Actin	Aspergillus nidulans	5.5	41.7	66.2	261 *
150	A,B	GI:67540744		Actin	Aspergillus nidulans	5.5	41.7	67.8	268 *
151	B	GI:67540744		Actin	Aspergillus nidulans	5.5	41.7	32.7	115 *
Protein targeting and localization
152	A,B	GI:85113087	✓	14‐3‐3 protein homologue	Neurospora crassa	4.6	29.3	33.2	83 *
153	A,C	GI:3023852	✓	Guanine nucleotide‐binding protein subunit β‐like protein	Neurospora crassa	7	35.1	5.7	160
154	C	GI:39945474		Guanine nucleotide‐binding protein subunit β‐like protein, hyp. prot.	Magnaporthe grisea	6.9	35	5.7	140
Signal transduction
155	A	GI:396431	✓	Rab guanosine diphosphate dissociation inhibitor α	Rattus norvegicus	4.9	50.5	4.3	143
156	A,B	AW791753		RHO protein GDP dissociation inhibitor, hyp. prot. (GI:156049611)	Sclerotinia sclerotiorum	5.1	22.5	n.a.	79 *
Transport
157	A	AW789658		Porin (GI:154316512)	Botryotinia fuckeliana	9	29.9	n.a.	109
158	A,B	Contig_8027		ABC transporter signature motif (GI:154305908)	Botryotinia fuckeliana	6.5	31	27.8	265
Electron transport
159	C	AW790540		NADH‐ubiquinone oxidoreductase 40 kDa subunit (GI:154289498)	Botryotinia fuckeliana	9.4/8	28.5/40	n.a.	106
160	C	AW790630		Glutathione reductase, hyp. prot. (GI:154291666)	Botryotinia fuckeliana	6.3	51	n.a.	90
161	A,C	GI:19114408	✓	Dihydrolipoamide dehydrogenase, hyp. prot.	Schizosaccharomyces pombe	9.5/7	54.7	2.2	92
Other metabolism or biological process unclassified
162	A	GI:6325326		Spermidine synthase	Saccharomyces cerevisiae	5.2	33.3	4.4	91
163	C	AW790589		NAD(P)H‐dependent d‐xylose reductase (GI:67516283)	Aspergillus terreus	5.4/7	35.6	n.a.	78
164	A	Contig_6463		Glycerol dehydrogenase, putative (GI:119480951)	Neosartorya fischeri	6.6	35	72.5	77 *
165	A	BM361761		Aldehyde dehydrogenase, hyp. prot. (GI:154320197)	Botryotinia fuckeliana	5.2	54	n.a.	128
166	A	GI:46116836		Alcohol dehydrogenase and 3‐ketoacyl‐acyl‐carrier‐protein reductase, conserved domains	Gibberella zeae	6.4	29	6.3	86
167	C	AW791581		NAD‐binding domain 4 protein (GI:145236573)	Aspergillus niger	8.1	48.2	n.a.	151
Unknown protein
168	B	GI:50294754		Conserved hyp. prot.	Candida glabrata	9.8	96.8	13.4	76 *
169	A	Contig_3909		Conserved hyp. prot. (GI:156043053)	Sclerotinia sclerotiorum	5.1	53	79.1	282
170	B	Contig_8402_133 + _8402_137		Conserved hyp. prot. (GI:156058700)	Sclerotinia sclerotiorum	6.8/5.5	39	5.2	52
171	A,B	AW790280		Conserved hyp. prot. (GI:156052649)	Sclerotinia sclerotiorum	4.9	36.2/40	n.a.	135
172	B	AW792491		Conserved hyp. prot. (GI:121712070)	Aspergillus clavatus	6.3	41.2	n.a.	154
173	A	AW792723		Conserved hyp. prot. (GI:156061853)	Sclerotinia sclerotiorum	4.8	18	n.a.	113
174	A,B,C	BM361778		Hyp. prot., no putative conserved domains detected (GI:27453871)	n.a.	n.a.	n.a.	n.a.	221
175	A,B	BM361778		Hyp. prot., no putative conserved domains detected (GI:27453871)	n.a.	n.a.	n.a.	n.a.	217
176	A	BM361778		Hyp. prot., no putative conserved domains detected (GI:27453871)	n.a.	n.a.	n.a.	n.a.	174
177	B	BM361778		Hyp. prot., no putative conserved domains detected (GI:27453871)	n.a.	n.a.	n.a.	n.a.	172
178	B	BM361778		Hyp. prot., no putative conserved domains detected (GI:27453871)	n.a.	n.a.	n.a.	n.a.	101
179	C	BM361778		Hyp. prot., no putative conserved domains detected (GI:27453871)	n.a.	n.a.	n.a.	n.a.	217
180	A	BM361783		Hyp. prot., no putative conserved domains detected (GI:156044082)	Sclerotinia sclerotiorum	7	80.9	n.a.	63

^† For an extended version of this table, see Table S1 and caption.

^‡ Please note that spots 20 and 101, 28 and 170, 69 and 89, 136 and 152, 143 and 181 (pI 3–10 gel), 20 and 101, 53 and 121 (pI 4–7 gel), as well as 36, 157 and 167 (pI 6–10 gel), co‐migrated in the mentioned gel types.

^§ Gel A, pI 3–10 (Fig. 2A); gel B, pI 4–7 (Fig. 2B); gel C, pI 6–10 (Fig. 2C).

^¶ A tick indicates that the reported identification/annotation is supported by a significant hit to a homologous Bgh expressed sequence tag (EST) in a mass spectrometry (MS) or MS/MS search.

^†† Description and species origin of the corresponding protein accession or the closest homologue of Bgh EST/genomic sequences as determined by blast search (see http://blast.ncbi.nlm.nih.gov/Blast.cgi). In the latter cases, GI numbers of blast hits are given in parentheses (hyp. prot., hypothetical protein). n.a., not applicable, as no sequence with significant homology in any other species.

^‡‡ Molecular mass (kDa) and isoelectric point (pI) of the identified protein. These values are given according to ProteinScape software (Bruker Daltonics) when spots were identified by a protein accession. When identified from EST or genomic Bgh sequences, the expected pI and molecular masses were determined using the Expasy Compute pI/Mw tool and the best blast hit to the partial sequence. When the observed (obs.) molecular mass and/or pI was different from the expected value, both are indicated. n.a., not applicable, as best match is against an EST for which no molecular mass or pI can be calculated.

^§§ Sequence coverage (percentage of the complete protein sequence; n.a., not applicable, as best match is against an EST for which no sequence coverage can be calculated by ProteinScape software) and Mascot score (see http://www.matrixscience.com) of the hit found by MS/MS or MS (indicated by

^* ) searches. Sequence coverage is calculated from the peptide mass fingerprint (PMF) identification, when present (see Table S1). In cases in which the same protein was identified by both means, the higher score is reported. Scores falling above the significance threshold for the respective database and search type are printed in bold black type. In the case of searches of the National Center for Biotechnology Information non‐redundant (NCBInr) database, scores for fungal hits falling below the 95% significance threshold for the entire database, but above the significance threshold for a search of fungal proteins, are printed in normal black type. Subsignificant scores are indicated by grey background. For significance thresholds, see Supplement S3. The identification for spot 98 was viewed as significant because it was the top scoring hit of both the PMF and MS/MS searches, although both scores were subthreshold.

Given the relatively poor number and completeness of publicly available Bgh sequences, most identifications presented in Tables 1/S1 were found on the basis of local homology by MS/MS searching. The comparatively low sequence coverage for most proteins reflects this limitation; many matches either represent homologues from other organisms or incomplete Bgh sequences (ESTs). As all sample spectra were searched against multiple public sequence resources, the best hit is reported in each case. Where identifications could only be achieved using ORFs of the Bgh draft genome, the sequences of the translated ORFs are available as a FASTA file in the Supporting Information (Supplement S1). In the light of these constraints, the current study focuses on the functional classification of the identified proteins, rather than their correlation to individual genes in an incomplete genome.

The identified proteins ranged in calculated molecular mass from 15.4 to 174 kDa, and in calculated pI from 4.6 to 11.3, thus encompassing a broad range of biophysical properties (Table 1). The predicted molecular masses and isoelectric points for the majority of the identified proteins were consistent with the experimental data, as judged from the location of the respective polypeptide spots within the two‐dimensional gels. This further corroborates the predicted identity of the recognized polypeptides. Exceptions include, for example, spots 42, 52 and 147, which exhibited a higher molecular mass than calculated. Conversely, products related to spots such as 62, 125 and 138 showed a lower molecular mass compared with the calculated value. Likewise, the location of some spots within the gels did not conform to the calculated pI (see Table 1 for details). These deviations could result from atypical migration in the two‐dimensional gels, from post‐translational modifications or from partial degradation of intact proteins during the extraction procedure. Given the indirect identification of most polypeptides via fungal species other than Bgh, interspecies variation could also explain these discrepancies. After functional classification of the identified proteins (see below), all analysed proteins were designated with unique spot numbers indicated on the corresponding gels shown in Fig. 2. Taken together, this analysis provides a first reference map of the Bgh spore proteome.

Presence of polypeptide isoforms

Identical functional annotations of distinct spots may result from either post‐translational modifications of the same gene product or from the presence of sequence‐related isoforms (paralogues) encoded by distinct genes. In the case of species for which no fully annotated genome sequence is available, these two scenarios are often difficult or even impossible to distinguish on the basis of MS data. Owing to this constraint, we cannot provide an exact number of distinct gene products identified in our study, but base our estimate on two simplifying assumptions. Firstly, two distinct spots whose best matches from the NCBInr database come from different species may represent paralogues rather than modified forms of the same protein, as the respective data sets appear to ‘select’ distinct sequences. However, as the different hits may also arise from variations in spectral quality, and because we have not performed exhaustive spectral and sequence comparisons to determine uniqueness, even true paralogues may remain unresolved in our survey. Alternatively, modified forms of the same protein could also selectively match different sequences. Secondly, horizontal strings of spots in conjunction with identical identification almost certainly indicate a single gene product with post‐translational modifications that change the pI, but not the observed molecular mass (e.g. phosphorylation). In our analysis, examples of this type comprise spots 13–17, 32–36 and 38–39. Based on these simplifying criteria, we estimate that the 180 distinct spots identified in our study represent at least 123 unique proteins.

Functional classification

Predicted functional categories were allocated to the identified proteins based on the biological process annotation routine of panther software (Mi et al., 2007; Thomas et al., 2003; http://www.pantherdb.org/). This assignment was completed manually when proteins well described in the literature were not allotted any function. In total, proteins corresponding to 161 spots were assigned to a panther biological process heading. Six annotated polypeptides were not linked to a biological process heading, and 13 spots (six of which have the same identification) correspond to hypothetical proteins without functional annotation. This indicates that a few of the proteins are novel or uncharacterized polypeptides involved in unidentified cellular processes, some of which may be specific to the biotrophic lifestyle and/or pathogenicity.

Most of the proteins identified are classified as metabolic proteins (almost 75%; Fig. 5). Metabolic polypeptides are often highly abundant soluble proteins and are thus generally well represented in proteomic studies. This also applies to the analysis of fungal proteomes, where metabolic proteins accounted, for example, for 66% of the proteins identified in the uredospores of the phytopathogenic basidiomycete U. appendiculatus (Cooper et al., 2006). Likewise, 65% of the polypeptides identified in hyphal forms of the ascomycete Candida albicans, a major systemic pathogen of humans, correspond to metabolic proteins (Hernández et al., 2004), although only 20% of the ORFs in C. albicans encode proteins devoted to metabolism (Yin et al., 2004). Taken together, it appears that dormant asexual fungal spores exhibit an abundance of metabolic proteins similar to metabolically active fungal hyphae. This suggests that, in spores, the molecular machinery required for energy production and the biogenesis of cellular compounds that are necessary for and after spore germination is preformed to rapidly initiate vegetative growth.

Figure 5.

Pie chart representing the functional classification of the identified proteins. The diagram shows the relative abundance of the identified proteins in the various functional categories of the panther classification system. To avoid a potential imbalance owing to polypeptides that may be represented by multiple spots, the calculation is based on the set of 123 spots that probably represent distinct polypeptides.

In detail, 27% of the identified polypeptides are involved in carbohydrate metabolism (Fig. 5). Several of these proteins are responsible for energy production, such as citrate synthase and lyase, α‐ketoglutarate dehydrogenase and malate dehydrogenase, which are all involved in the tricarboxylic acid cycle. Others, for example enolase, pyruvate kinase, pyruvate dehydrogenase, phosphoglycerate kinase, phosphoglyceromutase, phosphoglucomutase, glyceraldehyde‐3‐phosphate dehydrogenase and fructose‐bisphosphate aldolase, are important enzymes of glycolysis and/or gluconeogenesis. Additional identified enzymes (e.g. transaldolase, transketolase and 6‐phosphogluconate dehydrogenase) catalyse reactions of the pentose phosphate pathway, and the glycogen branching enzyme is part of glycogen metabolism. Polypeptides involved in lipid, fatty acid and steroid metabolism represent approximately 2% of the classified proteins (Fig. 5). Glycogen and lipids have been described as major storage forms of carbon and energy in Bgh conidia, and it has been suggested that glycogen and lipid catabolism represent key sources of energy allocation in germinating powdery mildew conidia (Both et al., 2005a). Our proteomic analysis supports this assumption, indicating that crucial enzymes of these catabolic pathways are preformed in conidia, enabling swift energy conversion during spore germination.

Another functional category represented by multiple members is devoted to protein metabolism and modification (23% of the identified polypeptides; Fig. 5). This class includes various proteins with distinct functions, such as different peptidases, cyclophilins (peptidyl‐prolyl cis/trans‐isomerases), heat shock proteins, ribosomal proteins and translation/elongation factors. Taken together, these polypeptides point to a key role for protein biosynthesis and folding in Bgh conidia. This scenario is reminiscent of the situation in uredospores of the biotrophic bean rust fungus U. appendiculatus, where the majority of the identified polypeptides were also found to be involved in protein assembly and catabolism (Cooper et al., 2006). It seems that both Bgh conidiospores and rust uredospores are pre‐equipped with the necessary components to initiate protein production rapidly on germination.

Components of amino acid metabolism make up approximately 11% of the identified proteins (Fig. 5). In a broader sense, proteins of this class also contribute to protein biosynthesis and turnover, thus further corroborating the above‐mentioned hypothesis. Biological processes related to nucleoside, nucleotide and nucleic acid metabolism also account for roughly 11% of the polypeptides identified in our analysis (Fig. 5). In addition, a number of annotated proteins (4%; Fig. 5) are well‐known stress response or cell defence proteins. They are mainly redox regulatory proteins (e.g. peroxiredoxin, catalase, peroxidase) involved in protection against oxidative stress. Pathogenic microbes often trigger a localized oxidative burst in their host species at sites of attempted invasion (Hückelhoven and Kogel, 2003). The identified antioxidative proteins may be required for protection against this presumptive plant defence response (Zhang et al., 2004). The remaining identified proteins were associated with intracellular protein traffic, protein targeting and localization, signal transduction, transport and electron transport (around 2% each; Fig. 5).

CONCLUSIONS

Our study has generated the first annotated proteome map of a powdery mildew fungus. It corroborates the hypothesis that asexual fungal spores, like pollen grains of plants (Holmes‐Davis et al., 2005; Noir et al., 2005), can be considered as mobile protein factories that are ‘ready to start’ on physical contact with a suitable (host) surface. This map provides a basis for the comparative proteomic analysis of further developmental phases of these obligate biotrophic phytopathogens. The proteomic features specific to the feeding organs (haustoria), in particular, promise to be a rich source of information on the biotrophic lifestyle.

EXPERIMENTAL PROCEDURES

Growth conditions of plants and fungi

Barley (Hordeum vulgare, I10, a near‐isogenic line of cultivar Ingrid containing the Mla12 resistance gene) plants were grown from seed in a controlled environment at 17 °C, with a 16‐h light/8‐h dark cycle. Plants were inoculated at the age of 7 days by dusting fresh conidia from infected plants with sporulating Blumeria graminis f.sp. hordei (race K1). Conidia for proteomic analysis were collected from sporulating colonies at 7 days post‐inoculation using a customized vacuum system (adapted from Johnson‐Brousseau and McCormick, 2004), and stored at –80 °C until protein extraction.

Protein extraction and purification

Bgh conidia (approximately 600 mg/replicate) were ground in liquid nitrogen and a small amount of sand using a mortar and pestle. The fine powder was suspended in 1 mL of extraction buffer [50 mm Tris‐Base pH 8.0; 10 mm ethylenediaminetetraacetic acid (EDTA); 0.5% 3‐[(3‐cholamidopropyl)dimethylammonio]‐1‐propanesulphonate (CHAPS); 10 mm dithiothreitol (DTT); one protease inhibitor cocktail tablet (Roche, Mannheim, Germany)]. Extraction was performed by subjecting the sample to five repetitions of vortexing (30 s each), interrupted by short pauses on ice (30 s each). After centrifugation at 16 000 g for 5 min at 4 °C, the supernatant was removed and stored on ice. From the remaining pellet, the extraction procedure was repeated with 500 µL of extraction buffer, and the supernatants were pooled. The protein concentration of the supernatant was determined by the Bradford assay with bovine serum albumin as the standard. From the supernatants, the proteins were precipitated with one volume of 10% trichloroacetic acid in acetone overnight at –20 °C. After centrifugation at 10 000 g for 15 min at 4 °C, the supernatant was discarded and the pellet was washed with chilled 90% acetone. Residual liquid was removed and, after drying on ice, the crude extract was stored at –80 °C until electrophoretic analysis. Before loading for isoelectric focusing (IEF), the crude protein pellet was solubilized in 0.7 mL of dense sodium dodecylsulphate (SDS) buffer [0.1 m pH 8.0 Tris‐HCl; 30% (w/v) sucrose; 2% (w/v) SDS; 5% (v/v) 2‐mercaptoethanol]. Then, 0.7 mL of Tris‐buffered phenol pH 8.0 (Biomol, Hamburg, Germany) was added, and the resulting mixture was vortexed for 30 s. After centrifugation, the phenol phase was collected and five volumes of chilled 0.1 m ammonium acetate in methanol were added. Samples were kept overnight at −20 °C for precipitation. After centrifugation, the pellet was washed twice with chilled 0.1 m ammonium acetate in methanol and twice with 80% (v/v) acetone.

Two‐dimensional gel electrophoresis

Two‐dimensional polyacrylamide gel electrophoresis was performed using the NuPAGE ZOOM Benchtop Proteomics system (Invitrogen, Carlsbad, CA, USA). Proteins (160 µg) were solubilized in 160 µL of sample rehydration buffer [7 m urea; 2 m thiourea; 2% CHAPS; 0.5% ZOOM Carrier Ampholytes pH 3–10 (Invitrogen); 20 mm DTT; 0.1% bromophenol blue]. Prior to IEF, ZOOM strips (length, 7.7 cm) pI 3–10 non‐linear, pI 4–7 linear and pI 6–10 linear (Invitrogen) were incubated for 16 h in the rehydration solution containing the sample; IEF was conducted using the following step gradient: 0–175 V (1 min), 175 V (15 min), 175–2000 V (45 min), 2000 V (25 min). After IEF, the strips used for gel electrophoresis were first equilibrated in 4.5 mL of lithium dodecylsulphate (LDS) sample buffer, together with 0.5 mL of 10 × sample reducing agent (Invitrogen), and subsequently in the same solution containing 125 mm iodoacetamide without reducing agent (15 min each). The samples were separated in the second dimension on NuPAGE Novex 4%–12% Bis‐Tris ZOOM gels (8 cm × 8 cm) in 2‐(N‐morpholino)ethanesulphonic acid (MES)‐SDS running buffer (Invitrogen), and the proteins were stained with colloidal Coomassie blue using PageBlue (Fermentas, Vilnius, Lithuania).

In‐gel digestion and MS

After two‐dimensional gel electrophoresis, spots of various intensities were visually selected, robotically picked (PROTEINEERsp, Bruker Daltonics, Bremen, Germany) and tryptically digested in gel (PROTEINEERdp, Bruker Daltonics). Aliquots of the digests were automatically spotted onto AnchorChip targets (Bruker Daltonics) in a thin‐layer preparation with α‐cyano‐4‐hydroxy‐cinnamic acid for subsequent MALDI‐TOF analysis, according to Gobom et al. (2001). The mass spectra of tryptic peptides were acquired with a Bruker Daltonics Ultraflex III MALDI‐TOF/TOF spectrometer (for instrument parameters, see Supplement S2 in ‘Supporting Information’). The resulting PMFs were processed in FlexAnalysis 3.0 (Bruker Daltonics) and used to identify the corresponding proteins in the ProteinScape 1.3 database system (Protagen AG, Bruker Daltonics, Bremen, Germany), which triggered Mascot (Matrix Science Ltd., London, UK) searches.

All samples were then recrystallized on the target by the addition of 1 µL of recrystallization solution (ethanol, acetone and 0.1% trifluoroacetic acid in a 6 : 3 : 1 ratio) and subjected to MALDI‐TOF/TOF analysis. LIFT MS/MS spectra (Suckau et al., 2003) were collected for selected precursors according to the following policy: when a protein was identified from the PMF (Mascot score exceeding 95% probability threshold), two assigned peptides of sufficient strength were fragmented for verification and three unassigned peptides were fragmented for further elucidation; when initial searches indicated multiple proteins, the precursor selection was altered by hand to try to verify all initial hypothetical identifications; when initial searches yielded no hits to the PMFs, five precursors were selected on the basis of signal quality alone.

Database searching

MS and MS/MS data were used to screen the NCBInr database (version 2007_12–26) (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=Protein). The mass spectral data were also searched against Bgh ESTs in all six reading frames (NCBI EST status as of 5 February 2008) and against translated ORFs generated from the current (assembly June 2007) Bgh draft genome sequence (P. Spanu and coworkers, unpublished data; http://www.blugen.org/). Scores exceeding the Mascot 95% probability limit or the peptide identity threshold for the respective database are presented in boldface black type. Scores exceeding the 95% threshold for fungal protein searches are printed in normal black type. When identified by both MS and MS/MS data, the higher Mascot score is presented in Table 1. A complete listing of all scores and matching peptide sequences is provided in Table S1, and a listing of the Mascot significance thresholds for the various databases is listed in Supplement S3 (see ‘Supporting Information’). Sequence coverage was computed from the PMF match, when present. Matches obtained only through MS/MS searches have a correspondingly low coverage. No sequence coverage was computed for EST matches as these do not correspond to full‐size proteins.

When identified from ESTs or Bgh ORFs, blast searches were performed to obtain a match to a full‐length protein; then, the predicted pI and molecular mass (MM) were determined using the Expasy Compute pI/Mw tool (http://expasy.org/tools/pi_tool.html). For protein accessions identified directly, these values were calculated by ProteinScape software (Bruker Daltonics).

Functional categories were assigned according to the panther (Protein ANalysis THrough Evolutionary Relationships) classification system (http://www.pantherdb.org/), and refined by hand when necessary.

Supporting information

Supplement S1 FASTA file of polypeptides derived from the Blumeria graminis f.sp. hordei (Bgh) genome sequence. The conceptual translation relates to the BluGen Bgh genome assembly from June 2007 (P. Spanu and coworkers, unpublished data; http://www.blugen.org/).

Supplement S2 Instrument parameters. Additional parameters for mass spectrometry (MS) and LIFT MS/MS data acquisition with a Bruker Daltonics UltraflexIII (Suckau et al., 2003).

Supplement S3 Significance thresholds. Mascot score thresholds for significant identifications for peptide mass fingerprint (PMF) and tandem mass spectrometry (MS/MS) searches of the various sequence resources used in this study.

Table S1 Detailed mass spectrometry (MS) results. Expanded representation of MS and MS/MS database search results, including peptide mass fingerprint (PMF) match scores, MS/MS match scores and matching peptide sequences.

Table S1 Explanatory remarks for Table S1.

Please note: Wiley‐Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

Supporting info item

Supporting info item

Supporting info item

Supporting info item

Supporting info item