- Top of page
- Results & discussion
- Experimental procedures
The role of insect saliva in the first contact between an insect and a plant is crucial during feeding. Some elicitors, particularly in insect regurgitants, have been identified as inducing plant defence reactions. Here, we focused on the salivary proteome of the green peach aphid, Myzus persicae. Proteins were either directly in-solution digested or were separated by 2D SDS-PAGE before trypsin digestion. Resulting peptides were then identified by mass spectrometry coupled with database investigations. A homemade database was constituted of expressed sequence tags from the pea aphid Acyrtosiphon pisum and M. persicae. The databases were used to identify proteins related to M. persicae with a nonsequenced genome. This procedure enabled us to discover glucose oxidase, glucose dehydrogenase, NADH dehydrogenase, α-glucosidase and α-amylase in M. persicae saliva. The presence of these enzymes is discussed in terms of plant–aphid interactions.
- Top of page
- Results & discussion
- Experimental procedures
Among the 4000 species of aphids that have been described, 250 are considered pest species and are responsible for direct (phloem uptake) and indirect (virus transmission) damage (Blackman & Eastop, 2000). The green peach aphid, Myzus persicae, is able to transmit more than 100 viral diseases to more than 400 host plants and the prevalence of this aphid is significant in agronomic terms (Quaglia et al., 1993). The damaged area created by a phloem-feeding insect is not equivalent to that created by a chewing insect and the plant defence responses also differ: attacks from phloem-feeding aphids elicit weaker responses in contrast with tissue-feeding Lepidoptera larvae and mesophyl-sucking insects (Voelckel et al., 2004; De Vos et al., 2005). This is because of insect feeding behaviour: whereas chewing insects, such as caterpillars and beetles, tear tissues with mandibles and other feeding apparatus, the aphid stylet penetrates the plant between the epidermal and parenchymal cells to reach the phloem sieves, thereby inflicting minimal wounding to the plant (Miles, 1999).
The signal responsible for the activation of plant defences is not only mechanical but also chemical through the action of particular molecules, commonly called elicitors. Elicitors have been discovered in the saliva of some chewing insects: glucose oxidase (GOX) from the saliva of Helicoverpa zea (Eichenseer et al., 1999), β-glucosidase from Pieris brassicae (Mattiaci et al., 1995), volicitin from Spodoptera exigua (Alborn et al., 1997) and caeliferin from Schistocerca amerricana (Alborn et al., 2007). Both β-glucosidase and volicitin have been found to induce the release of volatile organic compounds, which are attractive to the natural enemies of the attacking insects. By contrast, it has been suggested that GOX is advantageous for the insect that produces it. The enzyme increases the caterpillar's ability to survive through the suppression of nicotine, an inducible anti-herbivore defence in tobacco plants (Eichenseer et al., 1999) and decreases transcript levels of genes encoding for terpenoid biosynthesis (Bede et al., 2006). No aphid elicitor has been discovered to date but some cues show that this matrix contains plant defence modulating compounds.
The volatile response (terpene emission in particular) of Solanum tuberosum infested by Myzus persicae aphids has been found to be significantly different from that of healthy or mechanically picked plants and has also been found to influence the foraging and reproductive behaviour of Episyrphus balteatus (Harmel et al., 2007). In Harmel et al.'s experiment, differences in volatile emissions between mechanically picked and infested leaves revealed that aphid damage was not equivalent to mechanical damage. Aphids have been found to elicit plant cell wall remodelling: genes encoding cell wall-modifying enzymes have been found to be up-regulated in aphid infested plants (Voelckel et al., 2004; Divol et al., 2005; Qubbaj et al., 2005). These structural modifications are thought to deter phloem feeding herbivores by strengthening barriers to insect probing within the host tissues.
Thompson & Goggin (2006) reviewed different studies using transcription analysis of gene expression and showed that plant phloem feeders were able to change greatly the physiology of their host plants, including photosynthetic activity, source-sink relations and secondary metabolism. Plant responses to aphids are regulated by salicylate, jasmonate and ethylene signalling pathways. Furthermore, aphids ingest phloem sap without eliciting the normal occlusion response to injury in the sieve tubes. These occlusion mechanisms are calcium-triggered. Will et al. (2007) demonstrated that aphid saliva has the ability to prevent sieve tube plugging by molecular interactions between salivary proteins and calcium. The involvement of aphid saliva certainly seems to be preponderant in these phenomena.
Moreover, the electrical penetration graph (EPG) technique allows the electrical monitoring of plant penetration by aphids with piercing mouthparts and the recording of signal waveforms reflecting different insect activities (mechanical stylet work, saliva secretion and sap ingestion) (Tjallingii, 2006). Four periods of salivary secretion have been shown by EPG; one period of gelling salivation and three periods of watery salivation: (1) intercellular sheath salivation that envelops the stylet; (2) intracellular salivation into cells along the stylet path; (3) initial phloem salivation; and (4) phloem feeding salivation (Cherqui & Tjallingii, 2000).
The proteins contained in aphid saliva are of two types: structural and enzymatic. The structural proteins provide a tube-like sheath (Cherqui & Tjallingi, 2000) and correspond to major bands on gels with estimated molecular masses of 154 and 66/69 kDa (Baumann & Baumann, 1995). The secreted salivary enzymes have been found to be hydrolases (pectinases) and oxidation/reduction enzymes (phenol oxidase and peroxidases) (Miles, 1999; Cherqui & Tjallingii, 2000).
Injected saliva may play a crucial role in the prevention of the plant's wound responses but it may also play the role of elicitor of a plant's reaction. The role of these enzymes is not clear. Only enzymatic activity has been detected in aphid saliva. Proteins have never been identified. This is the reason why we investigated the M. persicae saliva proteome. We were searching for potential plant defence elicitor(s) in accordance with the approach that has been developed with mosquito saliva. This approach has enabled the identification of, among others, anticlotting, antiplatelet and vasodilator substances that modulate the host immune response and inhibit blood coagulation (Valenzuela et al., 2002; Montgomery et al., 2004). In the present study, different methodologies were used in order to study the aphid saliva proteome: in-solution digestion was performed on aphid salivary proteins, as this is the most sensitive method for protein identification. As a result of the lack of quantification of this technique, spots excised from 2D PAGE were in-gel-digested before ESI MS-MS (nano-electrospray ion trap mass spectrometer) identification.
Results & discussion
- Top of page
- Results & discussion
- Experimental procedures
A multi-approach experiment based on both in-solution and in-gel (after 2D gel electrophoresis) protein digestion associated to complementary mass spectrometry techniques (LC (liquid chromatography) and Maldi-Tof) was performed in order to investigate the saliva proteome of an aphid species. More than 20 spots were visualized on 2D gels of aphid saliva. Nine proteins were identified with a known function in other insects, while others were related to expressed sequence tag (EST) aphid sequences (Table 2). In the in-solution digestion approach, more than 200 peptides were generated; 71 were identified as a match with known sequences in existing databases (Table 1).
Table 2. In-gel digested 2D SDS-PAGE salivary peptides resulting from the annotated expressed sequence tags (ESTs) of the Myzus persicae aphid research. (A) Peptides and associated EST where Blast searches led to known function proteins with a significant E-value. (B) Peptides only associated with EST-specific aphid databases
| ||Organism||Acc no. Blast||Mw*||pI†||E-value||Spot||Peptides||Mascot score||EST GenBank accession no.||Cov‡|
|Solid saliva related to Fig. 3A|
|Ran-binding protein||Aedes aegypti||EAT38186||293536||6.25||9e-38||7||K.MIIISC.-K.CMSSAR.L||47||EE261552||2%|
| || || || || ||1||MIIISC||32|| || |
|NADH dehydrogenase||Aedes aegypti||EAT43733|| 18100||9.69||1e-37||2, 4, 5, 11||AHIIL||31||EE264957||3%|
|Retinol dehydrogenase 13||Tribolium castaneum||XP_973517|| 34004||8.65||4e-67||2, 8, 11||CAVLERTGAK||34||EE572101||4%|
|Transcriptional intermediary factor 2||Tribolium castaneum||XP_967666||164515||6.12||7e-13||6||R.NKMLASLLAK.D R.HCGWARHWCSFR.K||39||EE262675||3%|
|Protein arginine methyltransferase 5 isoform a isoform 1||Apis mellifera||XP_394141|| 71106||7.47||2e-60||7||R.VEISAIEK.N R.VTIINEDMR.L K.ILSSITSIKER.Y||47||EE264648||5%|
|AMP dependent coa ligase||Aedes aegypti||EAT33078|| 42583||6.24||6e-49||2||TCEIEG||36||EE571292||2%|
|ATM protein||Drosophila melanogaster||AAR89513||137555||8.42||2e-34||3||NFSKVL||33||DW013327||3%|
|Soluble saliva related to Fig. 3B|
|NADH dehydrogenase||Aedes aegypti||EAT43733|| 18100||9.69||1e-37||2′, 7′||AHIIL||32||EE264957||3%|
|Retinol dehydrogenase 13||Tribolium castaneum||XP_973517|| 34004||8.65||4e-67||1′, 3′||CAVLERTGAK||34||EE572101||4%|
| ||Spot||Peptides||Mascot score||EST GenBank accession no.||Cov‡|
|Solid saliva related to Fig. 3A|
|Myzus persicae, tobacco lineage, whole aphid library cDNA|| 1||R.HSLIQFK.T, K.ALNYFENK.L K.QPGLTLEITEK.K R.SAYNYYSLYNK.Q||160||DW011417||15%|
| 3||R.HSLIQFK.T R.TTFTEFK.K K.ALNYFENK.L K.QPGLTLEITEK.K R.SAYNYYSLYNK.Q||192||DW011417||18%|
| 4||R.HSLIQFK.T K.ALNYFENK.L R.TTFTEFKK.S R.SAYNYYSLYNK.Q||166||DW011417||14%|
| 8||K.SMSPTVAQPVVA.- K.ATNEFEPTINYQTSDPQK.V|| 93||DW010315||14%|
|11||R.HSLIQFK.T K.ALNYFENK.L|| 70||DW011417|| 6%|
|Myzus persicae, tobacco lineage, a phid salivary gland library cDNA|| 5||-.ILKASVFK.K -.CAVLERTGAK.T|| 40||EE572069|| 6%|
|Myzus persicae, line G006, whole aphid library cDNA|| 7||-.CAVLERTGAK.T K.CMSSAR|| 37||EC388700|| 3%|
|Myzus persicae, tobacco lineage, aphid head library cDNA|| 5||-.TTRPHSR.N R.HSLIQFK.T K.ALNYFENK.L K.NLSINECIIVLK.I R.SAYNYYSLYNK.Q||126||EC389308|| 9%|
|Soluble saliva related to Fig. 3B|
|Myzus persicae, tobacco lineage, whole aphid library cDNA|| 1′||R.SAYNYYSLYNK.Q K.TLINYDTNIPVTSLDDDXPIDRAIL.-|| 73||DW011417||15%|
| 2′||K.VDYSAVER.A K.SMSPTVAQPVVA.|| 86||DW010315|| 9%|
| 3′||K.VDYSAVER.A -.LYSLFDPLK.V K.SMSPTVAQPVVA.- K.DALDDMHENILK.S||130||EC389290||11%|
| 6′||R.HSLIQFK.T K.ALNYFENK.L K.SYNIARSMGXSSSNDVTGMLVS.-|| 39||DW011417|| 6%|
Table 1. In-solution digested salivary peptides resulting from the annotated expressed sequence tags (ESTs) of the Myzus persicae aphid research. (A) Peptides and associated EST where Blast searches led to known function proteins with a significant E-value. (B) Peptides only associated with EST-specific aphid databases
| ||Organism||Acc no. BLAST||Mw*||pI†||E-value||Peptides||Mascot score||EST GenBank accession no.||Cov‡|
|Glucose dehydrogenase||Aedes aegypti||EAT44638||61993||6.41||7e-27||K.LEDIDLDGCAK.Y K.YMVSTTSSTAGSCR.M K.DAVVDSELNVIGISNLRAVGR.S||121||EC389056||18%|
|68893||9.27||1e-31||K.AYLSPIFGR.E K.IQPDSTTGFGIEGNMK.I|| 80||EE262240|| 8%|
|Glucose oxidase||Apis mellifera||NP_001011574||67938||6.48||9e-24||K.LEDIDLDGCAK.Y K.YMVSTTSSTAGSCR.M K.DAVVDSELNVIGISNLRAVGR.S||121||EC389056||18%|
|1e-16||K.AYLSPIFGR.E K.IQPDSTTGFGIEGNMK.I|| 80||EE262240|| 8%|
|Alpha-amylase||Aedes aegypti||EAT48298||70055||5.36||2e-61||KVDQSIMSQYQDQ|| 33||EE571055|| 5%|
|Hydroxyacyl dehydrogenase||Aedes aegypti||EAT38824||32115||6.63||1e-56||IILELLNNA|| 31||DW010534|| 5%|
|Aphid EST database references||Peptides||Mascot score||EST GenBank accession no.||Cov‡|
|Myzus persicae, line G006, whole aphid library cDNA||K.VYQVYAYTR.D R.EVISHHVILK.T R.DRIPSLDTMK.S R.GYNMITSELQETR.S||445||EC387934||21%|
|K.VYEDIER.S R.EVISHHVILK. T R.GYNMITSELQETR.S||341||EC388457||15%|
|R.DNIVEDMTK.A K.AGMPDVSSTNR.G K.TGMPDVSSTNR.G K.RDIIVEAMTK.T|| 72||EE263445||15%|
|Myzus persicae, tobacco lineage, aphid salivary gland library cDNA||R.YMVLER.G K.WNFNTR.Y R.GYNMITSK.V R.CGINPNYMIK.I||312||EE572212||30%|
|K.EVPLVYSYTR.D R.GYNMITSKVQQTR.S K.TIDDLYTFDESYFK.S|
|R.CGINPNYMIK.I K.EVPLVYSYTR.D K.TIDDLYTFDESYFK.S||281||EE571947||27%|
|R.YLCQFM.- K.SMSPTVAQPVVA.- K.ATNEFEPTINYQTSDPQK.V|| 96||EE572100||11%|
|R.YMVLER.G K.WNFNTR.Y K.GLDNITIR.K R.GYNMITSK.V|| 86||EE264749||11%|
|R.HSLIQFK.T R.TTFTEFK.K -.IGFLIVSSK.- K.ALNYFENK.L|| 79||EE264598||21%|
|K.IKLPCXK.K K.TVTEDIIERL.-|| 39||EE264595|| 3%|
|K.YMAFDMMVK.G K.FCADDSEALYQK.G|| 77||EE264560||10%|
|Myzus persicae, tobacco lineage, aphid head library cDNA clone||K.IWNDAFSNPK.A|| 90||EC389283||10%|
|-.NHIITX.- R.GNPNLLPQQK.S R.AMALAQLMNMQNR.F|| 37||EC389929||17%|
|Myzus persicae, tobacco lineage, whole aphid library cDNA||K.ECVCDGPCYSCVVSAGLDK.S|| 65||DW013464|| 9%|
|K.QSLGMVGSFSDSSAR.G|| 53||DW011294|| 7%|
|K.NLQEIENNTVK.Q|| 47||EC389211|| 6%|
|R.YLGEMEKDGQK.C K.SQLVFNIFIVLLYLCTLVSLLT AAEIGETSCR.Y|| 40||DW012626||17%|
|Myzus persicae, line F001, PLRV infected, whole aphid library cDNA||K.IVPLIANK.I R.LLTIEEAIR.M R.NGYYLNSNTR.N|| 59||EE571076|| 9%|
Digested M. persicae salivary proteins were submitted to three databases: the NCBI database, the annotated EST database related to the pea aphid A. pisum and the peach aphid M. persicae database (Table 1). Because the function of many of these cDNAs remains unknown, only sequences presenting a significant Mascot score (> 30) were blasted.
Different kinds of cDNA libraries were used to generate ESTs. Among them, some were related to salivary glands or from aphid head RNAs. The majority of peptides from in-solution digestions matched with these databases (Table 1). These results confirmed that the identified peptides are found in the head and, in particular, in the salivary glands. These specific databases were therefore very useful for identifying salivary proteins collected in the aphid's artificial diet and provided more data in peptide identification than in previous proteomic works on aphids (Francis et al., 2006).
The M. persicae database revealed three in-solution digested peptides (Table 1A) matching with the sequence from an EST of M. persicae named EC389056. After a Blast procedure, this nucleotidic sequence was found to match with two insect enzymes: glucose dehydrogenase (GLD) and GOX. This match was confirmed with the A. pisum EST database: two common peptides were found to match with an EST, presenting a 95.1% similarity with EC389056 (data not shown). Another peptide was identified that matched with an EST of an M. persicae sequence and it also matched with GLD and GOX. This EST, named EE262240 (Table 1A), was quite different from EC389056. We detected GOX activity in aphid saliva samples containing 25 µg of total proteins. We used the same samples heated at 100 °C for 30 min as a negative control and Apergillus niger pure GOX as a positive one. GOX activity in aphid saliva was found to be 0.312 ± 0.001 U/mg soluble protein and was in the same range as the labial GOX activities of several caterpillar species, between 3.260 ± 0.028 U/mg soluble protein per pair of fourth instar S. exigua labial salivary glands (Merkx-Jacques & Bede, 2005) and 0.160 ± 0.070 U/mg soluble protein per pair of H. zea labial salivary glands (Eichenseer et al., 1999).
GLD (EC 184.108.40.206) is already known to have an immunological role in insects (Cox-Foster & Stehr, 1994). This was the first enzyme demonstrated in insects to be essential in killing foreign invaders during cellular immune defence or encapsulation. The enzyme has been found to be present in an inactive form in both plasma and specific blood cells. Shortly after invasion of an abiotic or a fungal invader, the enzyme is activated and localized onto the target. GOX (EC 220.127.116.11) has been identified in the hypopharyngeal glands of the worker honeybee Apis mellifera (Ohashi et al., 1999). This enzyme is needed to convert glucose to gluconic acid and hydrogen peroxide. The gluconic acid keeps the honey acidic and, together with hydrogen peroxide, has an antiseptic action. Moreover, GOX has been reported to inhibit soybean lipoxygenase activity and GOX treatment of tobacco foliage has been found to inhibit JA (jasmonate acid) production (Bi & Felton, unpublished observations). H. zea GOX has been shown to inhibit wound-inducible nicotine production in tobacco, Nicotiana tabacum (Musser et al., 2005). Similar results have been obtained with caterpillars with a cauterized spinneret or with their labial salivary gland surgically removed (Musser et al., 2006). It has been established that aphids evade the plant wounding response and activate inefficient defence gene expression, but the mechanism used is unclear (Zhu-Salzman et al., 2005). The presence of GOX in M. persicae saliva could explain the weak induction of the wounding response and the JA-regulated genes among aphid infested plants. Indeed, H2O2 is known to be a potent inactivator of lipoxygenase (Sporn & Peters-Golden, 1998). The role of this GOX needs to be confirmed by further experiments involving protein purification.
One in-solution digested peptide included in M. persicae EST EE571055 led to a match with α-glucosidase or α-amylase (Table 1A). Glucosidases are glycoside hydrolase enzymes. Alpha-glucosidases (EC 18.104.22.168) catalyse the hydrolysis of terminal 1,4-linked alpha-D-glucose residues successively from no-reducing ends of the chains with the release of beta-D-glucose. Alpha-glucosidases are widely distributed in microorganisms, plants, mammals and insects (Nishimoto et al., 2001). Alpha-glucosidase has been identified as a digestive enzyme in the A. pisum midgut (Cristofoletti et al., 2003). It is membrane-bound and catalyses in vitro transglycosylations in the presence of an excess of the substrate sucrose. Sucrose is present in large amounts in plant phloem sap and has been shown to be used as a power supply, whereas amino acids present in low amounts have been shown to be used for the production of structural compounds (Rhodes et al., 1996). Whereas α-glucosidase acts upon α bonds, β-glucosidase cleaves β linkages. Beta-glucosidase has been found in P. brassicae regurgitant as an elicitor of cabbage volatiles, which are attractive to the parasitic wasp Cotesia glomerata (Mattiaci et al., 1995). Alpha-amylase (EC 22.214.171.124) catalyses the endohydrolysis of 1,4-alpha-D-glucosidic linkages in oligosaccharides and polysaccharides. In a study by Ohashi et al. (1999), amylase of the honeybee hypopharyngeal gland was thought to be needed to convert plant starch (found in nectar) into glucose, which is then converted into gluconic acid by GOX. Alpha-amylase, as well as α-glucosidase, has been detected in mosquito salivary glands (Effio et al., 2003).
NADH dehydrogenase (EC 126.96.36.199) has been found in both in-solution (not shown in Table 1A because of a 28 Mowse score) and in-gel digestions (spots 2, 4, 5 and 11 in Fig. 2A and Table 2A; spots 2′ and 7′ in Fig. 2B and Table 2A). It has also been identified using the A. pisum EST database in both in-solution and in-gel digestions (data not shown). This enzyme belongs to complex 1 of the mitochondrial electron transfer chain and catalyses the transfer of electrons from NADH to coenzyme Q. The complex also translocates protons across the inner membrane, leading to an increase in the electrochemical potential used to produce ATP (Fernie et al., 2004). We compared the complete mitochondrial genome (15 721 bases) of Schizaphis graminum (Thao et al., 2004) with the EST of M. persicae, blasting with NADH dehydrogenase enzyme. The nucleotide alignment showed that the EST side was similar to a part of the sequence of the S. graminum mitochondrial genome (Blast result not shown).
Figure 2. (A) 2D-PAGE of Myzus persicae solid saliva proteins. Proteins were separated on a 15% acrylamide gel before being silver stained. (B) 2D-PAGE of M. persicae soluble saliva proteins. Proteins were separated on a 15% acrylamide gel before being silver stained.
Download figure to PowerPoint
In the process of identifying the salivary proteins from M. persicae, the use of specific EST aphid databases led to the lengthening of the peptide sequences obtained from mass spectrometry. From sequences of only a few amino acids, sequences of hundreds of corresponding nucleotides were obtained using the EST genomic databases after translation and adaptation to mass spectrometry data requirements. It is important to note that the identification of many aphid salivary proteins was possible only because of the use of the specific aphid EST databases. The efficiency of matching the data depended on the aphid part used to build the EST database. Indeed, many of our positive searches were related to more specific parts of aphids, either the head or the salivary glands. This was to be expected, as working on salivary proteins would logically lead to protein identification relating to these parts of the body. Only the Swiss-Prot Database (http://www.expasy.org/sprot) was not adapted to match efficiently proteins for a nonsequenced genome organism such as M. persicae.
Several peptides resulting either from the in-solution or the in-gel digestions still failed to present significant Mascot scores or Blast results. A minority of ESTs presented significant similarities with known insect proteins. It is not surprising to have obtained only a few identified proteins: the large-scale sequencing of 40 904 ESTs from A. pisum had led to 10 082 transcripts, among which 59% showed no match to any protein of known function (Sabater-Munoz et al., 2006). One reason for the large percentage of unknown proteins may be a factor involved in EST sequencing. Many of these sequences may be 3′ untranslated regions of the transcripts, which are much less conserved than translated regions.
In conclusion, this study constituted a first investigation of an aphid saliva proteome and revealed that data resulting from complementary approaches were needed to allow the identification of some interesting salivary proteins. Among these, we identified a GOX whose enzymatic activity was confirmed in M. persicae aphid saliva. In-solution digestion was better adapted than in-gel digestion for the study of aphid saliva proteins, notably because of the gelling consistency of the collected saliva, which might cause some inadequate interference during the first dimension migration in the 2D electrophoresis. It is for this reason that proteins identified after in-solution digestion were not necessarily identified after in-gel digestion. Finally, the importance of the use of an appropriate database (in this case consisting of annotated ESTs of M. persicae) was demonstrated when investigating protein identification from nonsequenced genome organisms. It is important to note that all the peptides and ESTs presented in this paper referred first to aphid sequences. Some of these interesting results will be enhanced when a complete aphid genome, that of A. pisum, becomes available. Similar studies have already been carried out with mosquito saliva: salivary transcriptome and proteome were combined in order to identify candidate proteins implied in host defence manipulation (Valenzuela et al., 2002, 2003; Ribeiro et al., 2004).
Numerous papers have dealt with aphid enzymatic activity. Some have established a hypothetical link between saliva and its effect on plants but few have been able to confirm it. We only partially understand why aphids manipulate plant defences. Thus, further experiments are required in order to study the effect of aphid saliva on plant defences.