Agilent Metabolomic Profiling of Bacterial Leaf Blight in Rice Manual
Contents
1. om Lm ul Cc mw a H T m irii Microsoft is a U S registered trademark of TOG Mock MOE NT Microsoft Corporation 10 11 06 Peo9 9 1106 Fax9T 1309 fiock S02 Application Note Export to Excel Using another GeneSpring MS software feature all data related to the immunity features that passed the fold change filter was exported in tab delimited format Manipulation of the data in a spreadsheet allowed us to separate the data into lists of both up regulated and down regulated features Tables 2 and 3 Of particular interest were the features highlighted in the tables which were absent in one or more classes These features may be on off metabolites and were flagged for further investigation Table 2 Up regulated metabolites from the TP309 Xa21 PX099 class immunity that passed the fold change filter Retention Fold time min Mass u change 32 64 771 4705 2 4 1 11 296 9389 2 8 32 76 710 4604 4 2 41 36 167 0575 6 1 43 74 401 3279 9 7 40 60 849 5386 10 7 31 24 295 2517 11 5 25 80 329 2925 11 9 27 20 453 2855 12 4 32 66 739 4514 12 9 47 38 934 5473 18 6 46 41 660 5333 20 3 50 91 565 8811 20 7 49 41 948 5989 23 2 2 38 221 0538 24 8 45 17 817 5082 25 0 37 68 608 2646 21 8 38 53 861 5044 28 0 2 07 129 0414 35 5 2 10 122 0383 36 2 37 25 624 2587 41 8 53 60 945 6066 64 5 53 31 93 0454 97 1 2 06 385 1011 98 6 52 22 157 9583 112 7 51 76 580 4290 151 1 51 35 225 9444 203 8 www agile2. shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing performance or use of this material Agilent Technologies Inc 2007 Printed in the U S A February 14 2007 5989 6234EN
3. way ANOVA with PCA made the rice line differences much clearer and also made it easier to distinguish infection status regardless of rice line Figure 8b b 1 way ANOVA and PCA Infected TP309 and TP309 Xa21 x S 1 naa bei 7 at er Uninfected TP309 Xa21 TP309 Xa21 no treatment TP309 Xa21 mock challenge TP309 Xa21 PX099 challenge TP309 Xa21 PX099 raxST challenge Figure 8 Principal component analysis without prefiltering of data a allowed differen tiation between rice lines When PCA was combined with 1 way ANOVA b it was much easier to differentiate not only rice line but infection status regardless of rice line Application Note 2 way ANOVA measures rice line and class differences simultaneously 2 way ANOVA is a powerful tool to study combinations of treatments Parameters are compared in every combination A prior knowledge regarding these features is not necessary After performing 2 way ANOVA a Venn diagram was plotted Figure 9 Results included e 360 features explained the variance between the two rice lines e 41 features alone were sufficient to separate all classes e 30 features separated all rice lines and classes First parameter Rice Lines test Second parameter Class test 360 41 Rice line features Class features Interaction between parameters Rice Lines and Class Figure 9 Analysis by 2 way ANOVA id
4. 7010S 85 0 47 38 934 5473 18 6 C53H79N2010P 100 0 46 41 660 5333 20 3 C25H62N1902 32 0 50 91 565 8811 20 7 C4HN6016F6PS 32 0 49 41 948 5989 23 2 C58H76N804 100 0 2 30 221 0538 24 8 C6H7N503 4 0 45 17 817 5082 25 0 C34H63N1903S 100 0 37 68 608 2646 27 8 C25H45N409PS 75 Harderoporphyrin 38 53 861 5044 28 0 C47H70N607P 100 erythromycin ethylsuccinate 2 07 129 0414 35 5 C5H NO3 1 6 2 10 122 0383 36 2 C7H602 2 benzoic acid 37 25 624 2587 41 8 C30H33N1202P 85 0 53 60 945 6066 64 5 C60H79N70S 100 0 53 31 93 0454 97 1 C5H5N2 2 0 2 06 385 1011 98 6 C17H18N60PS 36 0 52 22 157 9583 112 7 C3HN30PS 1 0 51 76 580 4290 151 1 C24H57N1004P 30 Vitamin K2 51 35 225 9444 203 8 C4HF6PS 4 0 Application Note Metabolomic Profiling of Bacterial Leaf Blight in Rice Table 6 Down regulated metabolites from the TP309 Xa21 PX099 class immunity with their METLIN search results Retention time min Mass u Fold change 1 09 213 9057 3 2 45 66 452 3297 30 32 52 600 4134 3 4 1 09 73 0268 4 4 36 95 281 6063 4 9 1 09 109 1268 10 0 47 21 524 3837 21 3 1 28 103 0648 21 6 46 57 652 4474 29 2 1 43 95 9816 48 3 47 20 540 3579 101 8 was selected for further investigation The sample was rerun on the Q TOF LC MS system Separation and ion source conditions used were the same as those for the original TOF analyses MS MS spectra were acquired from each of the metabolites on the list of target masses Examination of the MS MS spectrum from the selected metabolit
5. Application Note ONCHNSCOHPCNSHCOPHSNCPONHSONPHSCQO HSNPOCHNCOSHPNOHCPNOSCPHCNPOCNHO m Metabolomic Profiling of Bacterial e a s Leaf Blight in Rice Authors Abstract Steven Fischer Rice Oryza sativa and Oryza glaberrima is one of the world s most important Senior Applications Chemist staple crops providing food for more than 3 billion people Bacterial leaf blight Agilent Technologies BLB of rice caused by the Xanthomonas oryzae pv oryzae Xoo bacteria Santa Clara California U S A leads to crop losses of up to 50 Currently the use of resistant rice cultivars is the most economical and effective way to combat BLB The interaction Theodore Sana between Xoo and rice is governed by resistance genes in rice and correspond Senior Research Scientist ing pathogenic avirulence genes in Xoo Agilent Laboratories Santa Clara California U S A In order to better understand the mechanism of infection and immunity of rice to BLB a study was undertaken to identify metabolites that are related to infection and resistance Two rice lines were studied TP309 which is suscep tible to infection by the Xoo bacterial strain PX099 and TP309 Xa21 a resistant transgenic line In addition the effect of a raxST gene knock out in PX099 was evaluated for its effect on resistance in TP309 Xa21 Appropriate controls were included A two step LC MS approach was employed Rapid differential expression analysis of samples using time of flight TOF mas
6. X099 This triggers an immune response TP309 Xa21 is not resistant to PXO099 raxST Since PX099 raxST does not produce AvrXa21 TP309 Xa21 has no way to recognize the pathogenic bacteria and does not mount an immune response Figure 5 Rice leaves showing both TP309 Xa21 diseased and healthy states Application Note 1 way analysis of variance to identify class differences Analysis of variance ANOVA is a powerful tool for analyzing the variation present in an experiment Unlike a t test which can only make pair wise comparisons ANOVA can analyze any number of data sets Multiple t tests are independent so their errors are cumulative In the case of this rice experiment with seven Classes a total of 21 t tests would have had to be performed If the probability of a type I error false positive for a single t test was 0 05 the cumulative probability of error would have been greater than 1 00 A major advantage of 1 way ANOVA is that its probability of error remains the same no matter how many conditions are included The chance of a type I error using 1 way ANOVA for this experiment is much less than with cumulative t tests Figure 6 shows the results of ANOVA P lt 0 05 after applying a Tukey post hoc test for all pair wise comparisons In the blue boxes are the number of features with statistically insignificant differences In the red boxes are the number of features with statistically significant differences The featur
7. duce the AvrXa21 peptide PX099 PX099 raxST TES KE avrXa21 peptide Figure 2 The gene raxST in Xoo encodes for a sulfotransferase like protein that is necessary for production of the AvrXa21 peptide which is what elicits an immune response in the resistant rice line TP309 Xa21 Two bacterial strains were used in this experiment The wild type PX099 includes the raxST gene and produces AvrXa21 The raxST knock out PX099 raxST lacks the raxST gene and does not produce AvrXa21 To gain greater insight into the mechanisms of infection and immunity of rice to BLB a study employing LC MS metabolite profiling was undertaken to find and identify metabolites related to infection and resistance An LC MS system composed of an Agilent 1200 Series LC and Agilent 6210 Time of Flight LC MS was selected for the profiling work based on its sub 2 ppm mass accuracy outstanding reproducibility and robustness Agilent GeneSpring MS bioinformatics software was used to analyze the complex multi class data generated by the study Application Note An Agilent LC MS system incorporating an Agilent 1200 Series LC and Agilent 6510 Quadrupole Time of Flight LC MS was chosen to identify metabolites that were found to have statistically significant variations in abundance between the control and experiment samples This system was selected because of its combination of accurate mass measurements and MS MS spectra The METLIN Personal metabolite databas
8. e Figure 13 showed a base peak representing the loss of a carboxyl group formic acid CH202 from the precursor ion A second significant peak represented the subsequent loss of CO www agilent com chem metabolomics Number of METLIN search Empirical formula formulas number of hits none 0 0 C23H44N60S 17 0 C22H57N1203PS 40 Violaxanthin Neoxanthin CHNF2P 1 0 none 0 0 none 0 0 C30H48N602 24 0 none 0 0 C29H59N13PS 48 0 none 0 0 C29H47N70P 32 0 By evaluating the MS MS spectrum against the molecular structures included in the METLIN database search results Figure 14 it was determined that only two of the six possible metabolites pyroglutamic acid and pyrrolidonecarboxylic acid could logically have produced the losses seen These two compounds are enantiomers and as such indistinguish able by mass spectrometry If obtaining the precise identity of the metabolite had been critical it could have been determined through reanalysis using standards and a chiral LC column ONCHNSCOHPCNSHCO METABOLOMICS 8A 04484 MS MS spectrum of precursor ion m z 130 0532 H 0 l oN TN NH D CO D 5 CH 0 Di y n i CH 130 05320 56 05052 measured Abundance 140 m z Figure 13 MS MS spectrum of the precursor ion at m z 130 0532 shows a base peak representing the loss of formic acid CH 0 and a peak representing a subsequent loss of CO Evaluation of this information against the structures of the
9. e different rice bacteria classes were detected Based on relatively few metabolites the rice lines and state of infection were clearly distinguishable A significant amount was also learned about instrumentation and software requirements for this type of study For example the number of replicates required is determined by natural sample variability and technical instrumentation variability In this experiment the natural variations in both the rice and bacteria dictated multiple biological replicates However because of the outstanding reproducibility of the Agilent 6210 Time of Flight LC MS system used for profiling additional technical replicates were not required Data analysis plays an essential role in large scale metabolomics studies This study convincingly demonstrated that having a range of statistical analysis tools is essential to obtaining the best results Analysis of the profiling data by principal component analysis PCA alone could barely distinguish www agilent com chem metabolomics Metabolomic Profiling of Bacterial Leaf Blight in Rice between the two rice lines The combination of PCA with one or two way analysis of variance ANOVA clearly distinguished not only the rice lines but infection state regardless of rice line Further the ability to apply fold change filtering and visualize the results made it easier to compare differences in abundance and select targets for the second phase of the study metabo
10. e was used to narrow the list of possible identities during the identification process Two rice lines TP309 and transgenic TP309 Xa21 and two bacterial strains PXO99 and PXO99 raxST were included in the study along with controls As shown in Table 1 a total of seven different classes were studied Due to natural variations in both the rice and bacteria multiple biological replicates were necessary Based on previously demonstrated repro ducibility of the LC MS system technical replicates were not necessary in this study Several possible biomarkers involved in the elicitation of defense to bacterial infection of TP309 and the resistance of TP309 Xa21 were identified Based on relatively few metabo lites the two rice lines the state of infection and state of infection within each rice line were all clearly distinguishable Table 1 Conditions tested and number of biological replicates used for each condition TP309 TP309 Xa21 Condition Class Wild type Transgenic PX099 wild type 6 6 Mock treatment control 6 6 No treatment NT control 6 6 PX099 raxST NA 6 raxST knock out www agilent com chem metabolomics Metabolomic Profiling of Bacterial Leaf Blight in Rice Inoculation of rice The Xoo bacterial strains PXO99 and PXO99 raxST were grown for 72 hours at 30 C on peptone sucrose agar Tsuchiya et al 1982 Six week old rice plants TP309 and TP309 Xa21 were cut approximately 4 cm from the tip of fully ex
11. earcher in the fundamental processes of grasses such as rice and switchgrass Dr Fiehn is an Associate Professor in the Department of Molecular and Cellular Biology amp Genome Center at U C Davis He is a leader in the emerging field of metabolomics Drs Ronald and Fiehn provided the samples for the research outlined in the application note as well as invaluable background information and advice l j Agilent Technologies Application Note About Agilent Technologies Agilent Technologies is a leading supplier of life science research systems that enable scientists to understand complex biological processes determine disease mechanisms and speed drug discovery Engineered for sensitivity reproducibility and workflow productivity Agilent s life science solutions include instrumentation microfluidics software microarrays consumables and services for genomics proteomics and metabolomics applications Learn more www agilent com chem metabolomics Buy online www agilent com chem store Find an Agilent customer center in your country www agilent com chem contactus U S and Canada 1 800 227 9770 agilent_inquiries agilent com Europe info_agilent agilent com Asia Pacific adinquiry_aplsca agilent com This item is intended for Research Use Only Not for use in diagnostic procedures Information descriptions and specifications in this publication are subject to change without notice Agilent Technologies
12. entified 30 features that together separated all classes and rice lines These 30 significant features were then processed by PCA to demonstrate the separation of rice lines and classes The TP309 Xa21 classes group together except for the TP309 Xa21 that was challenged by PXO99 raxST It grouped with the infected class TP309 PXO099 challenge Figure 10 These results are consistent with the published results for the genotype and phenotype for these rice lines and bacteria www agilent com chem metabolomics Metabolomic Profiling of Bacterial Leaf Blight in Rice 2 way ANOVA and PCA Y Infected TP309 and TP309 Xa21 p fo A ae er A ected 7 E l x TP309 Uninfected 14 r f TP309 Xa21 TP309 no treatment TP309 mock challenge TP309 PX099 challenge TP309 Xa21 no treatment TP309 Xa21 mock challenge TP309 Xa21 PX099 challenge TP309 Xa21 PX099 raxST challenge Figure 10 PCA analysis of 30 significant features found by 2 way ANOVA clearly differentiates rice lines while also grouping infected classes together Fold change filtering and visualization Fold change filtering Figure 11 was another statistical tool used to determine which features metabolites were most likely to be relevant The greater the fold change relative to mock infection the more likely it is that a particular metabolite is relevant In the GeneSpring MS software fold cha
13. er Phot Mass inspector view The mass inspector view in the GeneSpring MS software allowed comparison of the relative abundance of a particular feature across all desired classes see Figure 12 This made it easy to see if a particular feature exhibited for example an on off behavior where it was present in some experimental classes but not present in others On off behavior is a valuable indicator of possible relevance The mass inspector view also provided additional information such as retention time molecular weight standard deviation and t test significance about the feature being examined No Pe H Condition individual feature metabolite Group l1 06 MT Sroup 1 O8 Prose abundance to be compared across SCLIN m APOL SUY fet cx d i 4 classes It also displays other useful Group 209 NT 3 7 Weilin Search Croup 309 Fx099 gj feature information such as retention P Wass Tolerarit a Da k gt time molecular weight standard gs P awe Search All Samples j deviation and t test significance Sample Name Mass so RT JL s 0 Moma Raw Roo No Pe Hesta Mass Details Scale D a OO Oo oO
14. es with statistically significant differences are potential biomarkers and are potentially of greater biological interest These were analyzed further Ma 1 Mock Xa21Px099 Xa21 o CEEI Xa21 Mock oi r ha Xa21 NT Xa21 Pxo099 Ja wn Xa21_Pxo99 RaxST TP303 Mock TP303 NT TP309 Pxo099 www agilent com chem metabolomics Metabolomic Profiling of Bacterial Leaf Blight in Rice In order to find metabolites that might account for differences between rice lines the effect of PX099 on infectivity was analyzed The results of three pair wise comparisons from the ANOVA analysis e TP309 Xa21 PX099 vs TP309 Xa21 Mock e TP309 PX099 vs TP309 Mock e TP309 Mock vs TP309 Xa21 Mock were combined for further study Figure 7 The three comparisons had a total of 347 unique statistically significant features Analysis found e Immunity features 42 metabolites possibly related to resistance e Infected features 22 metabolites possibly related to infection response e Bacterial features 25 metabolites possibly produced by or in response to the bacteria e Rice line features 1 70 metabolites related to differences in the rice lines Figure 6 ANOVA P lt 0 05 of MS profiling data identified features that showed statistically insignificant differences blue and statistically significant differences red in pair wise comparisons For example comparing the result
15. he searches did not produce matches for all of the metabolites was not surprising ONCHNSCOHPCNSHCQO Table 4 Minimum and maximum element settings for A wider mass window would have increased the number of METLIN database search possible metabolite identities but it would also have increased Element Minimum Number Maximum Number the number of incorrect identities And while the METLIN Carbon 1 100 database with over 15 000 entries is probably the most Hydrogen 1 400 comprehensive commercially available database in the world Nitrogen 0 20 that number is still just a small fraction of all possible Oxygen 0 20 metabolites As will be shown when a database search Phosphorous 0 1 results in matches it can be an invaluable aid to identification SUuH In the second part of the identification process one of the Fluorine 0 6 upregulated immunity features m z 129 0414 from Table 5 Table 5 Up regulated metabolites from the TP309 Xa21 PX099 class immunity with their METLIN search results Retention Number of METLIN search time min Mass u Fold change Empirical formula formulas number of hits 32 64 771 4705 2 4 C27H68N10013P 93 0 1 11 296 9389 2 8 C7HNO8F2S 13 0 32 76 710 4604 4 2 C38H60N706 71 0 41 36 167 0575 6 1 C6H7N402 3 3 43 74 401 3279 9 7 C24H41N40 9 0 40 60 849 5386 10 7 C29H72N17010P 100 0 31 24 295 2517 11 5 C18H33N02 2 0 25 80 329 2925 11 9 C19H39N03 2 0 27 20 453 2855 12 4 C9H31N1903 20 0 32 66 739 4514 12 9 C32H65N
16. lite identification Without the wide range of statistical analysis tools in the Agilent GeneSpring MS software used for data analysis it would have been impossible to obtain as much information as was obtained from this study Proceeding from the profiling phase of the study to the metabolite identification phase was facilitated by two factors The extremely good mass accuracy of the TOF profiling data and the powerful METLIN Personal metabolite database together made it possible to narrow the list of possible metabolite identities to a managable number The MS MS spectra obtained by reanalysis using an Agilent 6510 Q TOF LC MS identified the metabolite as one of two enantiomers This also highlighted the challenges of metabolomics Even with the largest metabolite database commercially available and accurate mass MS MS spectra conclusive identification of some compounds will require further analysis In this case it would have required reanalysis using standards and a chiral LC column ONCHNSCOHPCNSHCQO 1 Song et al 1995 Science 270 1804 1806 2 Sang Wong Lee et al PNAS 103 49 18395 18400 2006 3 Weckworth et al Proteomics 2004 4 18 83 The authors would like to express special thanks to Dr Pamela Ronald and Dr Oliver Fiehn of the Genome Center at the University of California Davis Dr Ronald is Chair of the Plant Genomics Program and Professor in the Department of Plant Pathology at U C Davis and a leading res
17. mode Electrospray lonization polarity Positive ionization Drying gas flow 10 L min Drying gas temperature 250 C Nebulizer pressure 40 psig Scan range MS m z 100 1000 at 250 ms spectrum MS MS m z 100 1000 at 250 ms spectrum Collision energy 5 x 10eV Isolation medium Fragmentor voltage 170 V Skimmer voltage 65 V Octopole RF voltage 750 V Capillary voltage 4000 V Reference masses m z 121 922 Reference mass flow 10 uL min Reference nebulizer pressure 15 psig www agilent com chem metabolomics Metabolomic Profiling of Bacterial Leaf Blight in Rice adducts Na and K and the protonated molecules M H and associated adduct ions were treated as a single compound Finally the algorithm identified isotopes The monoisotopic mass and retention time was reported for each feature An empirical formula was calculated for each feature using the monoisotopic mass and isotope ratios The retention time mass pairs generated by the MassHunter Workstation software were then exported for subsequent analysis in Agilent GeneSpring MS software The workflow used was as follows see Figure 4 1 Alignment and normalization of features 2 Hierarchical clustering to check data quality 3 Identify features with differential abundances across classes using 1 way and 2 way analysis of variance ANOVA 4 Perform principle component analysis PCA to show discriminating classes 5 Visualize fold changes The result
18. nge filtering is interactive so the fold change threshold was easily altered to determine what effect a change would have on the number of features passing the filter Another analysis and visualization option available in the GeneSpring MS software but not used in this analysis is the volcano plot The volcano plot combines the results of fold change filtering and t tests in a single visualization Criteria for both tests can be varied interactively to find the most relevant features 10 34 out of 42 Masses pass filter a D a z 7 Fi Fold Diference Condition 2 Group 130 Pxo Vie E Fold Difference 2 000 Feature inspection Further analysis of individual features was performed using two additional functions of the GeneSpring MS software The first was a mass inspector view that permitted the data on individual features to be examined The second was the ability to export data related to the features on a mass list in a tab delimited format that is compatible with Microsoft Excel FT 50 314 mass 785 2317 RTSO 1 3106 masss 0 0040 Figure 12 The mass inspector view in the GeneSpring MS software allowed Conditans Group li 06 i cock Figure 11 Interactive fold change filtering of the results of 1 way ANOVA helped determine which features metabolites were most likely to be relevant F Indaractive Update Condition 1 roup 10b Fxo33 rele 1 r Scati
19. ns Data analysis lonization mode Electrospray Initial processing of the accurate mass MS profiling data lonization polarity Positive ionization was done using Agilent MassHunter Software The feature Drying gas flow 10 L min extraction and correlation algorithms in the MassHunter Drying gas temperature 250 C software located the groups of co variant ions in each Nebulizer pressure 35 psi Scan range m z 50 950 Fragmentor voltage 170 V Capillary voltage 4000 V in a chromatogram instead of just locating chromatographic E E T EE chromatogram Each of these groups represented a unique compound Thus the algorithm located all the components peaks which could have concealed multiple components Reference mass flow 10 uL min After locating components background was subtracted Both positive and negative ionization data were successfully f l acquired but this note deals only with processing of the Charge state was set to 1 The algorithm identified salt aeee trom ee Application Note Instrument Conditions LC Q TOF MS MS LC Conditions Column ZORBAX SB Aq column 2 1 x 150 mm 3 5 um Mobile phase A 0 1 formic acid in water B 0 1 formic acid in acetonitrile Gradient 2 B at O min 98 B at 46 min 98 B at 54 9 min 2 B at 55 min MS stop time 54 9 min LC stop time 55 min LC post time 7 min Column temperature 20 C Flow rate 0 3 mL min Injection volume 2 uL MS Conditions lonization
20. nt com chem metabolomics Metabolomic Profiling of Bacterial Leaf Blight in Rice Targeted metabolite identification Metabolite profiling provided a list of up regulated immunity features that were selected for the next step in metabolomic investigation targeted metabolite identification Tentative identification of the metabolites was a two step process First each of the target masses identified in the profiling process was searched in the METLIN Personal metabolite database The database was searched over a narrow 10 ppm mass window This was significantly wider than the 2 ppm Table 3 Down regulated metabolites from the TP309 Xa21 PX099 class immunity that passed the fold change filter Retention Fold time min Mass u change 1 09 213 9057 3 2 45 66 452 3297 33 32 52 600 4134 3 4 1 09 73 0268 4 4 36 95 281 6063 4 9 1 09 109 1268 10 0 47 21 524 3837 21 3 1 28 103 0648 21 6 46 57 652 4474 29 2 1 43 95 9816 48 3 47 20 540 3579 101 8 mass accuracy of the 6210 TOF but it was felt that it was better to review a few extra hits than to possibly exclude the correct match by using a too narrow window The empirical formula calculations were set with a mass error of 5 ppm and 100 as the maximum number of empircal formulas Table 4 shows the settings for elements The results of the searches of the METLIN Personal metabolite database were incorporated into the spreadsheets with the original mass lists Tables 5 and 6 That t
21. nt of resistant cultivars carrying major resistance genes has been the most effective approach to combating BLB One such gene Xa21 was successfully cloned into a rice variety Taipei 309 TP309 Once cloned Xa27 can be passed on to the next generation through self fertilization Figure 1 Rice leaves infected by bacterial leaf blight left and uninfected rice leaves right The protein product of the gene Xa2 carries both a leucine rich repeat motif and a serine threonine kinase like domain suggesting a role in cell surface recognition of a pathogen ligand and subsequent activation of an intracellular defense response This receptor directly or indirectly recognizes a signal generated via a corresponding avirulence avr gene product encoded by the pathogen in this case the AvrXa21 peptide of Xoo The formation of this putative receptor ligand complex is postulated to initiate a signaling cascade culminat ing in defense responses that halt the pathogen s progress AvrXa21 must be present for the resistant rice line TP309 Xa21 to elicit an immune response and not be infected by Xoo The raxST gene in Xoo encodes for a sulfotransferase like protein that is necessary for the production of the AvrXa21 peptide Two bacterial strains were used in this experiment The pathogenic bacterial Xoo strain PXO99 includes the raxST gene and produces the AvrXa21 peptide Figure 2 The raxST knock out strain PXO99 raxST does not pro
22. of the analysis to this point was a mass list of metabolites that showed statistically significant variations in abundance between experimental classes 6 Search of the mass list against the METLIN Personal metabolite database Samples were then rerun using targeted MS MS on the Q TOF LC MS system and further data analysis was performed 7 Comparison of metabolite database search results against the acquired Q TOF spectra ONCHNSCOHPCNSHCO METABOLOMICS POOPEEE FORE E TIREI H He F Align and normalize Hierarchical clustering Analysis of variance to find features to check for reproducibility features with statistically significant of biological replicates differences between classes Create inclusion lists for Fold change filtering to select Principle component analysis database searching and the most statistically significant of significantly different features MS MS analysis features representing class differences Figure 4 Workflow for GeneSpring MS analysis of the MS profiling data Results and Discussion Morphology of the rice bacterial interactions TP309 is not resistant to either PX099 or the knock out PX099 raxST Figure 5 TP309 lacks a mechanism that would recognize either strain of Xoo and trigger an immune PX099 raxST PX099 raxST response TP309 Xa21 is resistant to PXO99 The Xa27 gene confers resistance by producing a receptor protein that recognizes and binds the AvrXa21 peptide produced by P
23. omogenization in Transfer of liquid nitrogen cooled Extraction of supernatant adapters in Retsch mill metabolites to sample vial SS Addition of 20 C Rice sample in extraction solvent Eppendorf tube F Centrifuge to separate DNA and proteins Figure 3 Sample preparation workflow for rice leaf samples An Agilent 6210 Time of Flight LC MS equipped with an electrospray ESI ion source was used to acquire profiling data The ESI source featured a separate nebulizer for the Instrument Conditions LC TOF MS LC Conditions l l Column ZORBAX SB Aq column 2 1 x 150 mm 3 5 um continuous low level introduction of reference mass com Mobile phase pound The reference mass compound facilitates compensa A AA tion for instrument drift Data was collected at a rate of 1 MS 0 04 ane na e aee E spectrum per second in both positive and negative ion modes Gradient from m z 50 to 950 2 B at 0 min 98 B at 46 min An Agilent 6510 Quadrupole Time of Flight LC MS equipped 98 B at 54 9 min with an electrospray ion source was used to acquire accurate 2 B at 55 min mass MS MS data for metabolite identification The ESI MS stop time 54 9 min LC stop time 55 min source featured a separate nebulizer for the continuous low 5 Column temperature 20 C level introduction of reference mass compound to maximize l Flow rate 0 3 mL min mass accuracy Injection volume 2 uL 3 sec flush MS Conditio
24. panded leaves with scissors dipped in a bacterial suspension at either 10 cells per ml Kauffman et al 1973 or just peptone sucrose agar mock condition After inoculation plants were maintained in a growth chamber and allowed to grow Rice sample preparation The samples were processed according to Weckwerth et al with the following changes Approximately 20 mg segments of rice leaf were cut and weighed They were placed in liquid nitrogen cooled 2 mL Eppendorf tubes each containing a 5 mm stainless steel ball bearing Figure 3 The tubes were transferred to MM301 Retsch Mill adapter racks that had been pre cooled with liquid nitrogen Samples were homogenized for 30 seconds at 25 Hz 1 mL of solvent a 2 3 3 v v v mixture of water acetonitrile isopropanol at 20 C was used to extract the metabolites from membrane and cell wall components in the homogenized samples This solvent system was chosen to minimize extraction of waxes and to enable analysis by both LC MS and GC MS LC MS analysis An Agilent 1200 LC equipped with a ZORBAX SB Ag column 2 1 x 150 mm was used to separate the rice extracts 2 uL injections were made from 1 mL sample volumes At a flow rate of 0 3 mL min a 2 to 98 linear gradient of water acetonitrile was employed over 46 minutes followed by a solvent hold until 54 9 minutes at which time data collection was stopped 0 1 formic acid was used as a mobile phase modifier ONCHNSCOHPCNSHCQO H
25. s of a mock challenge of TP309 Xa21 with a PX099 challenge of TP309 Xa21 found 113 features with statistically significant differences in expression levels TP309 Mock TP309 NT TP309 Pxo99 Psd on oo Pa D Ma so ONCHNSCOHPCNSHCQO TP309 Xa21 PX099 vs TP309 Xa21 Mock 42 Immunity features 170 Rice line features Figure 7 Combination of the ANOVA results from 3 pair wise comparisons yielded a total of 347 unique statistically significant features Of these 42 were related to immunity 22 to infection 25 to the bacteria and 170 to the TP309 PX099 vs TP309 Mock 22 Infected features TP309 Mock vs TP309 Xa21 Mock rice lines a PCA only Y pi es ss nn TP309 no treatment TP309 mock challenge TP309 PX099 challenge TP309 Xa21 line R 4 j Principle component analysis PCA PCA is a mathematical method of compressing complex data into a few variables The objective is to discover new variables principle components which account for the majority of the differences in the data When PCA was performed in GeneSpring MS with no prefiltering of data separation of the TP309 and TP309 Xa21 rice lines was observed Figure 8a However in this experiment information regarding the immunity and defense features was the goal not just differentiation between the two rice lines Combining 1
26. s spectrometry MS was followed by targeted identification of differentially expressed metabolites using quadrupole time of flight Q TOF MS MS Clear differences in the metabolite profiles of the different rice bacteria conditions were detected Based on rela tively few metabolites the rice lines and state of infection were clearly distinguishable whee Agilent Technologies ONCHNSCOHPCNSHCO METABOLOMICS Rice Oryza sativa and Oryza glaberrima is the primary food for more than 3 billion people worldwide Over 600 million people derive more than half of their calories from rice It is the third largest commercial crop behind wheat and corn In 2005 700 million metric tons were produced world wide with a market value of USS 120 billion It is estimated that 50 of the potential yield of the world rice crop Is lost to diseases caused by bacteria fungi and viruses In 2005 300 million metric tons were lost due to disease One of the most serious bacterial diseases of rice in Africa and Asia is bacterial leaf blight BLB caused by Xanthomonas oryzae pv oryzae Xoo Figure 1 BLB is one of the oldest recorded rice diseases and has been problematic for over a century Xoo spreads rapidly from rice plant to rice plant and from field to field in water droplets Infected leaves develop lesions yellow and wilt in a matter of days In severely affected fields bacterial blight can wipe out half a farmer s rice crop Breeding and deployme
27. six possible identities generated by a search of the METLIN metabolite database reduced the list of possible identities to two Figure 14 A search of the METLIN Personal metabolite database generated a list of six possible identities for a metabolite with a molecular weight of Metabolites Metabolites 1 6 of 6 Change Query approximately 129 0414 u MID Mass Name Formula CAS KEGG Structure EE SE a N 284 129 0426 1 Pyrroline 5 carboxylic acid 3 hydroxy CHNO 22573 88 2 H OH 3251 129 042 Pyroglutamic acid CHNO 98 79 3 C01879 2 69 129 0426 Pyrrolidonecarboxylic acid CHNO 6196 129 042 Pyrroline hydroxycarborylic acid 5H NO 6343 129 0426 N Acryloylelycine S 5 3 J i 3 Q A f iN E Application Note In order to better understand the mechanisms of infection and immunity between rice and bacterial leaf blight BLB a study was undertaken to identify metabolites related to infection and resistance A two step LC MS approach was employed Rapid differential expression analysis of samples using time of flight TOF mass spectrometry MS was followed by targeted identification of differentially expressed metabolites using quadrupole time of flight Q TOF MS MS In total seven different classes were compared Clear differ ences in the metabolites of th
Download Pdf Manuals
Related Search
Agilent Metabolomic Profiling of Bacterial Leaf Blight in Rice Manual