Surface-enhanced Raman Scattering for Metabolomics Roger Jarvis by tfs31371

VIEWS: 51 PAGES: 20

									      Surface-enhanced Raman Scattering for Metabolomics
      Roger Jarvis & Roy Goodacre                                              School of Chemistry &
      Contact: roy.goodacre@manchester.ac.uk                              Manchester Interdisciplinary
                                                                          Biocentre, The University of
                                                                                          Manchester
   Levels of functional genomics                                • Metabolomics
  • Metabolomics Technology
    Development                                                     – E. coli stress (BBSRC & AZ);
                                                            Metabolomics
     – Gene                    Genomics                                 recombinant mammalian
        LDI-MS (UK EPSRC/RSC); The analysis of metabolites (typically low
                                Genome
        SERS (UK BBSRC)                                                  weight molecules) in cancer
                                                            molecular cells (BBSRC); Oral a biological
                             Transcriptomics
        mRNA
  • Imaging Transcriptome                                   organism (EPSRC); Psoriasis (Stiefel of
                                                                        at a given time, with the aim
                                                            elucidatingLabs); META-PHOR (EUdefining
                                                                           gene function and
     – Protein
        MALDI imaging (Shimadzu);
                               Proteomics
                                Proteome
                                             Bioinformatics
                                                                        FP6); Biotrace IP (EU FP6);
                                                            biochemical pathways.
                                               Integration
        Raman, FT-IR imaging                                The Metabolome (BBSRC)
                                                                        Plants
                              Metabolomics
        (ORS); SIMS (UK BBSRC) “The total biochemical composition of a cell,
      Metabolite
                               Metabolome
                                                                • Systems Biology
  • Bacterial Identification   Phenomics                    tissue or organisms at any given time (Oliver et
      Phenotype
                                Phenome                             –
                                                            al., 1998).”STREPTOMICS (EU FP6);
     – Holtorf et al. (2002)
        SERS (UK HOSDB)                                                 SYSMO (EU/BBSRC)




Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                                Why study the metabolome?

                               Knowledge of (most)                 Mainly E. coli,
        Current                   fundamental metabolic            S. cerevisieae
                                  processes

                               Develop understanding to            Measure cell
                                  investigate metabolic            components with
          Need                    network regulation                MS, FTIR, GCMS

                              Develop understanding of            Grow mutant & WT cells
                                 responses to genetic or          under different conditions
                                 environmental influences
     Ultimate                  Determine gene function             Functional genomics
       Goal                    (including Bioinformatics)


       Functional Genomics aims to assign (new) functions to (uncharacterised) genes.

    “Genomics and proteomics tell you what might happen, but metabolomics
                     tells you what actually did happen.”
                    Bill Lasley, University of California, Davis.
Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                           Metabolite analysis.
                                  Metabolic Fingerprinting                Selection of
                                          SER(R)S
                                Crude metabolite mixtures for             technology is a
                                classification. (FT-IR/Raman/DIMS)        compromise
                                                                          between speed,
                                                                          selectivity and
                                                                          sensitivity.

      Metabolite target
                                                                  Metabolic profiling
           analysis                                  Four
      SER(R)S ??
      Analysis of specific                        Approaches
                                                                    Quantification of
                                                                     SER(R)S
                                                                  pre-defined targets.
         metabolites.                                              (GC-MS, LC-MS, NMR,
                                                                     HPLC, LC/MS/MS)
 Particular interest in
 low molecular weight
                                          Metabolomics
 compounds – the                     Unbiased identification of
 substrates and                      all metabolites in sample.
 products in pathways.
                                                                         (Fiehn, 2001)
Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                      SERRS Reproducibilty
      • We want to use SERRS as a metabolic
        profiling and fingerprinting tool
      • We know that there is a question mark
        over reproducibilty
      • Metabolomics requires quantitatively
        accurate data
      • Therefore we have been looking at
        strategies for assessing objectively, the
        reproducibility of our SERRS experiments

Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                      Colloidal Batch-Batch Reproducibility
                 3


                                                                                            • 3 replicate
   Extinction




                 2                                                                          absorbance
                                                                                            measurements
                                                                                            •  (absorption) max. -
                 1                                                                          larger value equates to
                                                                                            a larger particle size
                150
                                                                                            • FWHH (full width at
   FWHH




                                                                                            half height), a larger
                100
                                                                                            FWHH indicates wider
                50
                                                                                            particle size
                                                                                            distribution.
                500
                                                                                            • Extinction - lower
  max.




                                                                                            value for the extinction
                450                                                                         indicates greater
                                                                                            aggregation
                400
                       Ag        Au        EDTA   Fructose Glucose   Oleyl-   PVP   Thiol
                       citrate   citrate                             amine

                                    Colloids prepped by Emma Oleme and Arunkumar Paneerrselvam
Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                               SERRS spectra of Cresyl Violet




                                                                                                             1277
                                                                        985
                                                                                                             *




                                                                                                      1230
                               EDTA                                     *
                                                  877
                                            846




                                                                                                                    Mean SERRS
                               778




                                                                  964




                                                                                1077
                                                                                       1102
                                            *     *



                                                                               1038
                                      810




                                                                                                             *      spectra of cresyl


                                                                              1049
                                                               949




                                                                                               1186
                                                         906




                                                                                              1167
   Raman photon count (a.u.)




                                                                                                                    violet acquired
                                                                        *                                           using the four
                               PVP                                                                                  colloidal
                                            *     *                                                          *
                                                                        *                                           substrates that
                               Ag citrate                                                                           were found to be
                                            * *                                                                     SERRS active.
                                                                                                             *
                                                                                                *
                               Au citrate         *                            *


                                     800    850         900    950      1000 1050 1100 1150 1200 1250 1300
                                                                     Raman shift (cm -1)


Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
        Signal-to-noise ratios (S/N) observed in the
          median SERRS spectra of cresyl violet
                                                  Raman shift (cm-1)
     Substrate                                                                  Mean
                               877                1049            1186   1277
     Au citrate                1.24               1.20            1.58   1.88   1.47

                               846                877             985    1277
     Ag citrate                1.23               1.28            1.81   2.10   1.60
     EDTA                      1.09                1.11           1.44   1.72   1.34
     PVP                       1.29               1.28            1.93   2.03   1.63




Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
       MANOVA on the S/N ratios calculated from the SERRS
        bands identified in spectra of cresyl violet, from four
                          active substrates


                    Ag citrate                         Au citrate   EDTA    PVP
      Raw SERRS spectra
      Wilks' L[a]      0..429      0.061                            0.122   0.300
      ~ F[b]           1.187       6.890                            4.187   1.856
      P[c]              NS         0.000                            0.006    NS
      Row normalised SERRS spectra
      Wilks' L[a]      0.543       0.141                            0.577   0.560
      ~ F[b]           0.803       3.749                            0.712   0.757
      P[c]              NS         0.009                             NS      NS

Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
      Quantification of Cresyl Violet using
                    SERRS
                                                                  • Bootstrapped
                                                                  correlation analysis
                                                                  for the log-log
                                                                  relationship to area
                                                                  under the cresyl violet
                                                                  SERRS band at 930
                                                                  cm-1

                                                                  • Dilution series from
                                                                  5 x 10-6 M to 5 x 10-2
                                                                  M, using the

                                                                  • PVP capped
                                                                  colloidal silver
                                                                  substrate.


Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                   Next question – we can find colloids that give
                statistically reproducible batch to batch SERS – but
                what happens when we start playing with chemistry?
                                                                                           1200



                                                                                           1000
                                                                                                                                 Potassium
                                                                                            800
                                                                                                                                 choride




                                                                                I732cm-1
               2500                                                                         600


                                                                Sodium                      400

                                                                chloride                    200
               2000
                                                                                              0
                                                                                                  5   10   15      20        40        55    60    70
                                                                                                                 % colloidal silver
                                                                                           3000

               1500
                                                                                           2500                            Potassium
    I732cm-1




                                                                                           2000                            nitrate
               1000




                                                                                I732cm-1
                                                                                           1500



                                                                                           1000



                                                                                           500
               500
                                                                                             0

                                                                                                  5   10   15     20        40        55    60    70
                                                                                                                % colloidal silver


                 0                                                                         3000



                      5     10   15     20        40       55      60      70              2500


                                      % colloidal silver                                   2000
                                                                                                                              Sodium
                                                                                                                              nitrate
                                                                                I732cm-1
                                                                                           1500



                          Optimisation of cytosine SERS                                    1000



                                                                                            500



                                                                                              0

                                                                                                  5   10   15     20        40        55     60    70
                                                                                                                % colloidal silver

Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                                                                         Cytosine
                                                             Power fit                                                             Power fit

                                               0.2                                                                   0.2
                                                                  Batch 1
             log10 S/N 599 cm -1




                                                                                   log10 S/N 599 cm -1
                                                                  Batch 2
                                              0.15                                                                  0.15

                                               0.1                                                                   0.1

                                              0.05                                                                  0.05
                                                                    R = 0.79295                                                           R = 0.79295

                                                0                                                                     0
                                                     -7         -6.5          -6                                           -7         -6.5          -6
                                                      log10 Concentration (M)                                               log10 Concentration (M)
                                                             Power fit                                                             Power fit
                                               2.5                                                                   2.5
                  log10 Area under 599 cm-1




                                                                                        log10 Area under 599 cm-1
                                                2                                                                     2


                                               1.5                                                                   1.5


                                                1                   R = 0.86377                                       1                   R = 0.86377
                                                                  Batch 1
                                                                  Batch 2
                                               0.5                                                                   0.5
                                                     -7         -6.5          -6                                           -7         -6.5          -6
                                                      log10 Concentration (M)                                               log10 Concentration (M)

Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                   Optimisation of surface-
                 enhanced Raman scattering
                    (SERS) experiments

                      Roger Jarvis, William Rowe, Nicola
                     Yaffe, Sven Evans, Joshua Knowles,
                        Ewan Blanch & Roy Goodacre


Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                                              Experimental

      Pseudo Full-Factorial Experiment
      • 3 colloidal silver preps at 25, 50 & 75% v/v
            –     hydroxylamine, citrate, borohydride
      •      6 aggregating agents at 1, 10 & 100 mM
            –     NaCl, KCl, Na2SO4, K2SO4, NaNO3, KNO3
      •      785 nm NIR Raman probe, 3 s integrations with ~
             (Goodness knows what!!) mW power a source, spectral
             range (150 - 2900 cm-1)
      •      Single analyte – L-cysteine (100 mM)
      •      Total of 162 experiments,5 replicate measurements for
             each giving 810 SERS spectra

      This allows us to determine the “optimal” experimental
                              conditions
Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                                                    Cont…

      Multiobjective optimisation
      • Questions
                  1.Can we use this experiment to determine the utility of an
                    directed search algorithm for optimising these conditions
                    more rapidly?
                  2.Could some form of interpolation be used to derive further
                    experiments that yield superior results?
      • Objective functions
                  1.Reproducibility: standard deviation of the Mahalanobis
                    distance between principal component scores recovered
                    from replicate spectra
                  2.Signal intensity: peaks areas calculated for 4 major bands
                    and meaned across replicates


Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
              Published results: GC-TOF mass
            spectrometer optimization via PESA-II
      O’Hagan,S., Dunn, W.B.,
      Brown, M., Knowles, J.D. and
      Kell, D.B. (2005) Closed-loop,
      multiobjective optimization of
      analytical instrumentation:
      gas chromatography/time-of-
      flight mass spectrometry of
      the metabolomes of human
      serum and of yeast
      fermentations. Analytical
      Chemistry 77(1): 290-303.


      PESA-II used to optimize                                    Optimized:
      the settings of a mass-                                     - Number of true peaks
      spectrometer to improve the                                 - Signal-to-noise ratio
      chromatograms.                                              - Sample analysis time - throughput

Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
      Typical SERS spectrum of L-cysteine and Raman bands for which peak
                            areas were calculated




                                                        647
                                       700

                                       600
                 Raman photon counts




                                       500


                                                               795



                                                                     911
                                       400




                                                                           1034
                                                    C-S
                                       300         Red –
                                                shifted due
                                       200     to binding at
                                                   silver
                                       100        surface
                                         400      600         800      1000        1200   1400
                                                              Raman shift (cm-1)


Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                   Summary of metrics calculated to quantify signal reproducibility and
                                      intensity of enhancement




             14                                                                120

             12                                                                100

             10
                                                                               80
 Frequency




                                                                   Frequency
             8
                                                                               60
             6
                                                                               40
             4

             2                                                                 20

             0                                                                  0
             0.2     0.4   0.6   0.8      1      1.2   1.4   1.6                 0   100   200   300   400   500   600   700   800
                             Mahalanobis distance                                                  peak area


                    Homogeneous distribution                                                Skewed Distribution



Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                                                                                                                                Experiment #45
                                                                                                                 2000


           Summary of Results
                                                                                                                 1500




                                                                                            Raman photon count
                          800                           45                                                      1000

                          700
                                                   54      36                                                  500

                          600
      Area under peaks




                                                                                                                   0
                                                                     51        23
                                                                                33                                400   600   800     1000      1200    1400
                          500                                                                                                                 -1
                                                                                                                               Raman shift (cm )
                                                             18
                                                             15
                                  Pareto                            24
                                                         9  63  155                                                          Experiment #54
                          400                                                                                    1500
                                   front                 27  6
                                                                   32      69
                          300                                  99  60
                                                         72




                                                                                            Raman photon count
                                                        90  96
                                                          81                                                     1000
                          200                                         20  48
                                                                             39
                                                                         57  87
                                                                                78
                                                       35               161 66
                                                                          312
                                                                          
                                                                          93
                          100                       26           21       30
                                                               
                                                       53 42 84 158  75                                        500
                                                17         71
                            0            117156131 3452100  102143 
                                        989249 2862 128157 122
                                       551121425914150 107126 7413312965
                                     67731031010180 77146  121222  134
                                     15340 4191106137 37114145136 61 83
                                                                  38 86 
                                            115119   149 160 159
                                                       44
                                         135 82 58 88 56 5 85 127 111120 
                                         46 10813818148137141125162 110 144
                                                                    123
                                      645011147104105140 162513968 89
                                           124
                                        3197   11611394118 70 130
                                      47         151  132 154 1929 794
                                                  
                                                   109
                                                   95        152   76  43
                            0.2     0.4     0.6       0.8        1       1.2          1.4    1.6
                                                   Mahalanobis distance                                           0
                                                                                                                   400   600   800     1000       1200   1400
                                                                                                                               Raman shift (cm-1)


                 Exp.                Colloid    Amount (% v/v) Agg. Agent Conc. (mM) Enhancement M. dist.
                  45              Hydroxylamine     75           K2SO4       100       662.0311  0.8539
                  36              Hydroxylamine     75          NaNO3        100       779.4253  0.7642
                  54              Hydroxylamine     75           KNO3        100       675.0239  0.6618
Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                             Multiobjective Pareto optimisation
                               using the PESA II algorithm
                       800

                       700
                                                   45
                                                                                                   • Find solutions which give
                                              54        36
                                                                                                     best trade-off between 2
                       600
                                                                                                     objectives
   Area under peaks




                                                                    51          23
                                                                                 33
                       500

                       400       Pareto
                                                        18
                                                        15
                                                               24
                                                    9  63  155                                  • PESA II is a region based
                                                      27  6

                       300
                                  front            
                                                              32
                                                          99  60
                                                                       69                           Pareto selection
                                                    72

                       200
                                                   90  96
                                                     81                                              algorithm
                                                                 20  4839
                                                  35                3
                                                                        78
                                                                    57  87
                                                                    161 66
                                                                     93
                                                                     12
                                                                                                      – Select a region or
                       100                     26           21       30
                                           17
                                                          
                                                  53 42 84 158  75
                                                       71                                             hypercube
                                    97117156131107126 100 16 12813312965
                                  47112142150  855274 102143 
                                67731031010180 77121222  134
                                 641198591449 2862157 122
                                15340 91106137 37114145136 61 83
                                                  44
                                                             38 86 
                                 147104105140146  2512068 89
                                      124 115119 148  5 127 111139 
                                                 8
                                    135  151 88 56 113125162 160 159
                                       41    58   7 34 154 1929 794
                                    46 821081381 13132  149 110 144
                                                               123
                                   31   11614194118 70 130
                                              92
                                                                                                      – Randomly select individual
                         0            55
                                     50       109
                                              95        152   76  43
                         0.2   0.4     0.6        0.8          1          1.2          1.4   1.6
                                              Mahalanobis distance
                                                                                                        from this subset
                       Analysis to be completed, however:
                       • Directed search optimises
                                                                                                   • Problem!! Our solution
                         experimental conditions in 60                                               space is quite sparse and
                         iterations                                                                  disperse!!
                       • Interpolation attempted but
                         hasn’t improved SERS
Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)
                                    www.biospec.net




   Group Leader: Professor Roy Goodacre
   Postdocs: Dr Will Allwood, Dr Robert Cormell, Dr Elon Correa,
   Dr Roger Jarvis, Dr Yankuba Kassama, Dr Iggi Shadi,
   Dr Catherine Winder, Dr Yun Xu.
   With Collabs: SERS (4), Metabolomics (2), ToF-SIMS (2)
   Research Technicians: Steffi Schuler, Richard O’Connor
   PhD Students: Felicity Currie, Katherine Hollywood, Nicoletta
   Nicolaou, Soyab Patel, Ketan Patel, Emma Wharfe, Nicola Wood,
   Dong Hyun Kim, Will Cheung, Robert Coe.




Laboratory for Bioanalytical Spectroscopy (http://biospec.net/)

								
To top