Abstract:
Novel Aspect:
Influence of high frequency QTOF MS data acquisition on identification of low level metabolites and structure elucidation by MS/MS.

Introduction:
Today, it is of crucial importance for the development phase of a new pharmaceutical drug to identify critical drug metabolites as early as possible to avoid later toxic side effects on humans. HPLC/MS techniques such as QTOF mass spectrometry are routinely used to acquire the data, which are then analyzed with the aid of specialized software applications. For mass spectral data, the mass resolution and the dynamic range are of special importance to improve information obtained. A large dynamic range is especially important to see and identify the low level metabolites. A high mass resolution is important to assign the correct structure in the interpretation of MS/MS spectra.

Methods:
The QTOF MS data from a metabolism experiment are acquired with an improved TOF detection system which has the capability to acquire data with a frequency of up to 4GHz. This provides a large in-scan dynamic range of up to 5 orders of magnitude and a high mass resolution of up to 15,000 in the relevant mass range of m/z 100 to 600. For the analysis of the MS and MS/MS data gained from the metabolite experiments the data are extracted by a specialized software tool from the metabolite sample file and from a parent drug control sample file. After analysis of the extracted data with various algorithms the final result will be displayed with identified metabolites.

Preliminary Results:
To demonstrate the improved quality of the data, which were acquired on a QTOF mass spectrometer equipped with a detection system capable to acquire data with a frequency of up to 4GHz, typical examples will be shown. The large dynamic range of 5 orders of magnitude was demonstrated by an experiment which shows the in-scan dynamic range by measurement of a low concentration compound in the presence of a very high concentration main component. Accurate mass measurements were made on coeluting compounds that were present at concentrations 5 orders of magnitude lower than the main components. The high mass resolving power of up to 15,000 in the range of m/z 100 to 600 was demonstrated by an example of coeluting compounds with isobaric mass in the MS domain and by a compound which fragments to isobaric mass fragments in the MS/MS domain. The correct molecular formulas will be assigned in a window with very small relative mass error of 2 to 5 ppm. Finally data from a metabolism experiment of the pharmaceutical compound Nefazodone will be shown which take advantage from the large dynamic range and the high mass resolution. In this experiment two isomeric low level metabolites with isobaric mass were detected in the presence of the high concentrated main drug compound. Under MS/MS conditions both isomeric metabolites undergo fragmentation and show a fragment of isobaric mass with different constitution. With the high mass resolving power of the 4GHz data acquisition it is possible to assign the correct empirical formula to each fragment. This unravels the structure for each isomeric metabolite. The whole process of data analysis will be imbedded in a specialized software tool.

Abstract:
Novel Aspect:
These studies represent the first application of MS-based metabolomics to Mtb and expand knowledge of central carbon metabolism in Mtb

Introduction:
Tuberculosis (TB) is the leading bacterial cause of deaths worldwide and a public health emergency. A major barrier to control of this pandemic is the lack of understanding of how its etiologic agent, Mycobacterium tuberculosis (Mtb), persists in the face of host immunity and chemotherapy. Such knowledge represents a potential roadmap to new drug targets. Current evidence indicates that Mtb adapts metabolism to accommodate the acidic, hypoxic, nitro-oxidative and nutrient poor conditions encountered in the host. Biochemical knowledge of Mtb metabolism at the level of metabolites however remains a conspicuous absence.

Methods:
Mtb's central carbon metabolic pathways were labeled using stable isotope-based feeding experiments. Sample preparation methods suitable for rapidly quenching and extracting Mtb's free intracellular metabolites were developed using a combination of thermal cooling, mechanical lysis and organic extraction. Hydrophobic and hydrophilic metabolites were resolved using a combination of C18-reverse phase and deactivated Type C silica chromatography, respectively, followed by ESI-TOF. Metabolites were identified using retention time-accurate mass pair qualifiers in conjunction with known standards where available.

Preliminary Results:
Stable isotope labeling studies using 13C-[U] dextrose and 2H[U]-acetate were performed to obtain unbiased global views of carbohydrate and fatty acid-derived metabolites in Mtb. These studies thus catalog Mtb's major intracellular metabolite pools and outline Mtb's carbohydrate and fatty acid metabolic pathways. Moreover, these studies led to the surprising discovery of aminobutyrate as a novel metabolite of Mtb's variant TCA cycle. These studies thus not only lay the methodologic groundwork for future metabolomics-based studies of Mtb but illustrate the unique potential for untargeted LC-MS-TOF-based metabolomics approaches to expand knowledge of Mtb physiology.

Abstract:
Novel Aspect:
An IC/MS/MS method with excellent sensitivity and reproducibility at low concentrations of perchlorate in matrices with high concentrations of salts

Introduction:
Perchlorate is commonly used as an oxidant in solid fuel propellants for rockets and missiles. Recently, perchlorate contamination was found in many aquifers associated with the Colorado River and in crops that use contaminated water for irrigation. A robust and reliable method is presented here for the quantitation of perchlorate using an Ion Chromatography (IC) system interfaced to a Triple Quadrupole (QQQ) Mass Spectrometer. Quantitation is based on the sum of two multiple reaction monitoring (MRM) transitions corresponding to both chlorine isotopes (35Cl and 37Cl) of perchlorate. The method was tested for linearity, sensitivity and reproducibility in a variety of matrices including ones with high salt content.

Methods:
An Ion Chromatography system was interfaced to a Triple Quadrupole (QQQ) Mass Spectrometer. The column used was an ion exchange column, Metrosep ASUPP7-250 column (4.0mm ID x 250mm length) with a Metrosep RP Guard column. The eluent was 10 mM Sodium Carbonate (Na2CO3) in 35% Acetonitrile. The mobile phase passed through a suppressor cartridge to remove salts before going to the mass spectrometer. The mass spectrometer was operated in the negative ESI mode. O18 labeled perchlorate was used as the Internal Standard. The three MRM transitions monitored for perchlorate and its internal standard were (sum of both Cl35 and Cl37 isotopes) = m/z 99 > 83 + 101 > 85 and (O18) = m/z 107 > 89.

Preliminary Results:
The analysis of perchlorate in water is carried out on several calibration standards in reagent water and at various concentrations in the presence of total dissolved solids consisting of chloride, carbonate, and sulfate and also in the presence of reagent water. A linearity coefficient of R2 > 0.999 is established for perchlorate standards ranging in concentrations of 0.5 to 25 ppb. Linearity for the QQQ mass spectrometer is R2 > 0.998 over the range of 0.01 to 10 ppb. Reproducibility among seven replicate injections of standards is also excellent for the QQQ with seven replicates at the 0.1 ppb level resulting in a peak area relative standard deviation (RSD) of 5.33%. The confirmation of perchlorate by QQQ mass spectrometry is also done by using qualifier ion ratios. The reproducibility is very good for the analysis of perchlorate in the presence of salt water. For example, at 1 ppb perchlorate in 1000 ppm salt water the peak area reproducibility among three injections is 0.63 % RSD.

Abstract:
Novel Aspect:
A novel multi residue method of 4 macrocyclic trichothecenes by using LC/TOF-MS and LC/MS/MS was developed.

Introduction:
Mycotoxins are secondary metabolites produced by fungi. These compounds can cause a wide range of acute and chronic systemic effects in humans and animals. The over 400 known mycotoxins are all complex organic compounds that are not volatile at ambient temperatures. Inhalant exposure to mycotoxins can occur by inhaling airborne particulates containing mycotoxins, including dust and fungal components. Toxigenic fungi such as Stachybotrys chartarum have been isolated from building materials and air samples in water damaged buildings. S. chartarum produces a number of mycotoxins, including satratoxins, rodinin, and vercalin. This work presents the development of LC/MS and LC/MS/MS methods for the multi-residue analysis of satratoxin G, satratoxin H, iso-satratoxin F, and roridine E.

Methods:
LC/MS analyses were performed on either a QQQ LC/MS or a TOF LC/MS with electrospray ionization in positive ion mode using a ZORBAX Eclipse Plus C18 column (150~2.1 mm, 3.5 ƒÊm) at 40 ‹C. The LC mobile phase was acetonitrile and 0.1% acetic acid (A) and 10mM ammonium acetate in water (B). The initial solvent composition was 70% A / 30 % B, going to 30% A / 70 % B in 30 min. Detection of macrocyclic trichothecenses (MTRs) were carried out using multiple reaction monitoring (MRM) mode for LC/MS/MS. For the TOF analysis, mass chromatograms (MC) were extracted over a narrow m/z range. The MTRs were extracted from S. chartarum with 100% methanol without additional purification.

Preliminary Results:
ESI mass spectra of satratoxin G, satratoxin H, iso-satratoxin F and roridine E were obtained under positive ion mode. Ammonium adduct ions (M+NH4)+ for satratoxins and protonated molecule (M+H)+ for roridin L2 were observed as the base peak ion by LC/MS/MS. On the other hand, analysis by TOF-MS produced the protonated molecule (M+H)+ as the base peak ion for all MTRs. Accurate mass of these base ions were acquired by LC/TOF-MS. The relative mass error ranged from -0.68 to 3.34ppm. Using the base peak ions as precursor ions produced typical product ions such as (M+H)+, (M+H-H2O)+, e.g. m/z=249 and 231 using product ion scan mode. The intensities and ratio of two major product ions were used for quantitative and qualitative analysis with MRM mode by LC/MS/MS. The detection limits (LODs) of each MTR ranged from 0.05 to 0.15 ng/mL by LC/MS/MS and from 0.6 to 3.0 ng/mL by LC/TOF-MS. The calibration curves of all MTRs were created by injecting standard solutions in the range from 0.5 to 1000 ng/mL for LC/MS/MS and 10 to 1000 ng/mL for LC/TOF-MS. Resulting calibration curves showed correlation coefficients greater than 0.999 for both LC/MS techniques. These optimized methods were then applied to analyse methanol extracts of fungal species. All MTRs were detected by LC/MS/MS and LC/TOF-MS. Calculated amount of each MTRs ranged from 1.0 to 2.3 ug. Relative mass error of observed accurate mass of the base peak ion for each MTR ranged from 0.24 to 2.12 ppm.

Abstract:
Novel Aspect:
Multivariate analysis of LC/MS data of beer was shown to find important markers for beer taste and quality.

Introduction:
Beer is a complex beverage that has been around since antiquity. There are many different methods of production and these differences affect taste. In addition, time and the environment after production affect the beer’s properties. The analytical challenge is which compounds in the complex mixture are important to the beer’s taste. A practical solution to this challenge is to use an untargeted analysis approach with multivariate data analysis to find the differences. The development of accurate mass capable LC-TOF instruments greatly aids this approach by allowing simultaneous detection of all compounds with mass accuracy. Our study here demonstrates the utility of LC/MS combined with multivariate correlative analysis to rapidly find compounds of interest in a differential analysis study.

Methods:
Different commercial beers were analyzed by LC/MS. The HPLC used was a rapid resolution instrument designed for 600 bar operation with low dispersion. The mass analyzer was a Q-TOF operated in TOF mode for the untargeted mass profiling study. The beer samples were filtered through an ultrafiltration membrane. Three technical replicates were collected for each sample using a reverse phase C18 column (i.d. 2.1x100mm, 1.8 micron). The data was processed through peak detection software and the resulting peak list was then imported into multivariate correlative analysis software. The data was frequency filtered and subjected to 1-way ANOVA followed by PCA. The resulting PCA mass list was then searched against a metabolite-specific database to get candidate chemical identities.

Preliminary Results:
The study analyzed 4 beer brands from 4 different lots at 5 different time points after production. The data when analyzed by PCA resulted in 4 separate clusters and each cluster correlated to one of the beer brands. The same data was processed by hierarchical tree clustering. The masses that were determined to be significantly different were determined to come from the choice of materials and yeast fungus. It was possible to categorize the brand of beer from these LC/MS data patterns. In the beer time studies, it was possible to detect compounds that were changing in abundance over time. These compounds are candidates for determining the age of a beer. Since a few days of storage at room temperature is enough time to change a beer’s taste, identifying these compounds is important for food safety and quality studies. LC/MS with multivariate analysis has been demonstrated to be a powerful analytical approach as differences in beer brand could be identified and beer aging could be detected and characterized by changes in chemical abundances.

Abstract:
Novel Aspect:
Absolute quantitation of protein phosphorylation using MRM with a low noise hexapole collision cell.

Introduction:
Peptide quantitation using multiple reaction monitoring (MRM) has emerged as an important methodology for biomarker validation. MRM on a triple quadruple (QQQ) mass spectrometer provides superior sensitivity and selectivity for targeted compounds in a complex sample. MRM also offers high precision in quantitation and fast scan speed, which makes it an ideal technology for validating biomarkers in a high-throughput fashion. One area that has not been explored and is of great interest is the quantitation of protein phosphorylation. Reversible protein phosphorylation plays an important role in cell signaling pathways. The percentage of phosphorylation is often very important to the signal transduction. In this study, we explored the quantitation of percentage of phosphorylation using MRM with AQUA peptides.

Methods:
The p44/p42 mitogen-activated protein kinase (MAPK) ERK1/2 was used as the target protein. Synthetic peptides with and without phosphorylation, as well as their aqua peptides were made to create standard curves. A software program, Peptide Selector, was used to help design MRM transitions. A microfluidic based nanoflow LC interfaced to a triple quadruple mass spectrometer was used to analyze the samples.

Preliminary Results:
The active ERK 1 protein typically has two phosphorylation sites (T202 and Y204) which reside in one tryptic fragment of the protein. The two sites can be phosphorylated to different degrees. This has made the quantitation of this phosphoprotein particularly challenging. Four synthetic peptides with variable phosphorylation (Lower case letter for phosphorylation, T202/Y204, t202/Y204, T202/y204, t202/y204) were made so we could correctly capture the different phosphorylation states. Using Peptide Selector, a list of MRM transitions were predicted including m/z 609.3 (y5+) for T202t, m/z 1016.910 (y16 ++) for T202t/Y204y. The peptides were then analyzed using Q-TOF for optimal MS/MS product ions. The predicted MRM transitions matched the experimental results. Standard curves of the phosphopeptides and unphosphorylated peptides were acquired. The active and control ERK1 will be spiked in human sera at different ratio and digested with trypsin. The percentage of the phosphorylation at each phosphorylation site will be measured.

Abstract:
Novel Aspect:
Determining how improved resolution impacts mass error measurements in complex mixtures

Introduction:
In mass spectrometry, there is no direct correlation between resolving power and mass accuracy. For example, when doing synthetic confirmation, Fourier Transform (FT) mass spectrometers with resolving power of over 200,000 typically exhibit only a two-fold improvement in mass accuracy over time of flight (TOF) instruments with resolution in the range of 10,000. Even with the same analyzer design recent 2-3 fold improvements in resolution showed no significant improvement in mass accuracy when dealing with pure compounds. This study examines the impact of improving resolution on TOF and QTOF systems while analyzing complex mixtures where there are compounds producing ions very close together (<50 mDa).

Methods:
Comparisons were made between early model TOF and QTOF systems with a digitization rate of 1 GHz and systems upgraded to operate at 1, 2, and 4 GHz rates. The upgraded systems also had faster responding detector amplification. Initial studies were done in MS only mode using mixtures which produced ions very close together in mass. The relative concentrations were varied and the mass accuracy obtained at each acquisition rate was measured as a function of percent purity and digitization rate. Later work was done with more complex mixtures spiked with known compounds. Finally studies were done in MS/MS mode to determine the impact of mass accuracy on interpreting the fragment ions to distinguish between different structures.

Preliminary Results:
Initial tests were done with methyl 5-acetyl-salicylate (MAS) and butyl paraben (BP) which differed in molecular weight by 36 mDa. Mixtures were prepared that varied in relative amounts from 1:1 to 128:1. At 1 Ghz with the original electronics, the mass error was 120 ppm for the 1:1 mixture reducing to 2.89 ppm for the 128:1 mixture. At even 3% of the BP, the error was 5 ppm. By contrast, at 4 Ghz where the resolution was three times better, the two components were well resolved and the mass error for MAS was 1.3 ppm for the 1:1 mixture and -0.2 ppm for the 128:1 mixture. The mass error for the BP “contaminant” was -1.4 ppm for the 128:1 mixture. The other test mix was the nominally isobaric amino acid pair glutamine (MW 146.06914) and lysine (mw 146.10553) differing by 36 mDa in molecular weight. This pair was chosen because of their importance in protein sequencing. Using a mixture with lysine as the major component and glutamine as the minor component, the protonated ions were measured with a mass error of 1.62 and -0.18 ppm respectively. Stanozolol was analyzed in MS/MS mode. It produces two fragment ions with formulas of C10H13N2 and C12H17 which differ by only 25 mDa. At 1 GHz, only one ion is observed in the region of m/z 161.1 while at 4 GHz two ions were observed with mass errors of less than 5 ppm in each case.

Abstract:
Novel Aspect:
New LCMS/MS method to determine the most relevant newer antiepileptic drugs in a single analytical run

Introduction:
Since the beginning of the nineties several novel antiepileptic drugs have been introduced. The pharmaceutical compounds lamotrigine, oxcarbazepine, felbamate, zonisamide, gabapentin, pregabalin, tiagabine, topiramate,levetiracetam, and rufinamide are traditionally summarized under the term newer antiepileptic drugs. For the therapy of seizures therapeutical drug monitoring (TDM) is an important instrument for the physican. Some clinical labs already use LC/MS techniques for monitoring the antiepileptic drugs in serum. The instituted analytical methods often allow only the determination of one analyte per run. Rapid Resolution LC combined with Triple Quadrupole Mass Spectrometry allows a focussing of the TDM. The present work describes a new LC/MS/MS method to indentify and quantify ten newer antiepileptic drugs in a single analytical run.

Methods:
Two different sample preparation steps were evaluated. Beside the etablished lab method, which consisted of a simple precipitation step, a liquid-liquid extraction was tested additionally. All sample analysis were performed on a LC/MS/MS system consisting Agilent1200 Rapid Resolution Liquid Chromatograph and an Agilent G6140A Triple Quadrupole Mass Spectrometer, operated with an electrospray ionization source in positive polarity. Determination of the optimal MRM transitions was carried out by flow injection analysis of standards at concentration levels around 1ng/ul. Different Zorbax columns with 1.8um particle size were explored in combination with different solvents, flow rates and column parameters to optimize the speed of the analysis, while maintaining a good chromatographic resolution

Preliminary Results:
The presented LC/MS/MS method, combining RR-LC and Triple Quadrupole Mass Spectrometry, allows the simultaneous determination of ten newer antiepileptic drugs in human serum. The method is fast, selective, sensitive and robust. Sample preparation was done by a simple precipitation step to remove proteins. Only 100 ul of serum is needed, which is important in the field of neuropediates. The serum was diluted with 500 ul methanol that contained also the internal standards. After centrifugation and reconstitution, the supernatant was directly injected into the LC/MS/MS system. Due to an optimized chromatography, remaining matrix did not interfere the determination of the analytes of interest. Therefore an extensive sample preparation which involved a liquid-liquid extraction step was not necessary. The total run time, using a 100*2.1mm RRHT column with 1.8um particle size, was kept at eight minutes. The multiple reaction monitoring capability of the QQQ allows for the highly selective MS/MS analysis of the target compounds and their corresponding internal standards using the more abundant product ions for quantification and the less abundant product ions for confirmation. The established detection limits were 50-150 times lower than the clinical concentrations of the antiepileptic drugs, which allowed a further dilution of the supernatant. Calibration curves were generated by analyzing five spiked serum samples and a blank. The calibration curves for all ten analytes showed a linear fit with no weighting and no origin treatment. Method Validation was performed based to an established quality control program in the lab. The method proved to be accurate and precise (intra-day precision below 2%, inter-day precision below 3%). Interlaboratory tests, as part of an external quality control, were passed successfully.

Abstract:
Novel Aspect:
This work was unique due to maximizing information from one injection and by approaching the ultimate goal of universal detection.

Introduction:
There is increasing demand in the pharmaceutical industry for higher throughput in order to analyze more samples per day on each LC/MS system. Method optimization time must be kept to a minimum and labs need to maximize the information per unit time in order to get the right answer on the first injection without detailed knowledge of the sample concentration, optimum ionization technique or polarity of ions formed. Additionally, because of the rapid chromatographic methods used, a detector with fast acquisition rates as well as fast POS/NEG polarity switching is very important. . This work describes the components of an LC/MS system optimized for such high speed chromatography with peaks as narrow as one second.

Methods:
The system used in this work operated at the high pressures required to get the rapid resolution from 2 and 3 mm ID columns packed with 1.8 micron particles. These columns were operated up to 2 ml/min. This system was fitted with a multimode ion source capable of handling flows up to 2 ml/min at 100% water without the need for splitting. In addition, this source could be operated simultaneously in electrospray and atmospheric pressure ionization, providing a better chance of producing ions in a single analysis. Using an HPLC that could operate at higher temperatures and pressures allowed the use of methanol, even though its viscosity was greater than acetonitrile.

Preliminary Results:
Hundreds of samples per day were analyzed with this system. The single quadrupole mass spectrometer was able to reach acquisition rates of 10,000 u/s without significant sacrifices in resolution. In addition, the MS was capable of ultra fast POS/NEG polarity switching. Good spectral quality was obtained across peaks as narrow as one second in width at baseline with a mass range adequate for most small molecule pharmaceutical applications. In addition, proper isotope ratios were obtained at the fastest acquisition rates which provided higher confidence in correct compound identifications.

Abstract:
Novel Aspect:
An LC/MSMS method is presented with excellent sensitivity and reproducibility in analysis of perfluorooctanoic acid

Introduction:
Perfluorooctanoic acid (PFOA), a synthetic (man-made) chemical, does not occur naturally in the environment. PFOA is used to make fluoropolymers, substances with special properties that have thousands of important manufacturing and industrial applications. PFOA is globally investigated because it is very persistent in the environment. It has been found at very low levels both in the environment and in the blood of the general U.S. population, and it caused developmental and other adverse effects in laboratory animals. A robust and reliable method is presented here for the quantitation of PFOA using an LC triple quadrupole mass spectromter. Excellent sensitivity, linearity and reproducibility are demonstrated.

Methods:
LC/MS/MS analysis was performed on an Triple Quadrupole (QQQ) Mass Spectrometer with electrospray ionization in negative ion mode. The column used was a ZORBAX Eclipse Plus C18 column (2.1x100mm, 1.8 ìm). The LC mobile phase was 10 mM ammonium acetate in water (A) and 10 mM ammonium acetate in 80:20 methanol:water (B). The initial solvent composition is 80% B then goes up to 90% B in 4 minutes. Two MRM transitions monitored for PFOA are m/z 413.0>368.9 as quantifier and m/z 413.0>168.9 as qualifier.

Preliminary Results:
The limit of detection (LOD) for PFOA in a clean standard is 33.9 fg on column with signal to noise ratio of 3:1 when using 413.0>368.9 MRM transition with Q1 peak width of 1.2 amu and Q2 peak width of 0.7 amu. The limit of quantification (LOQ) is 56.5 fg on column with signal to noise ratio of 5:1. PFOA showed a linearity coefficient of R2>0.996 over the range of 56.5 fg to 56.5 pg. The reproducibility for both LOD and LOQ were excellent. The peak area relative standard deviations (RSDs) are 3.5% and 3.4% for LOD and LOQ, respectively, using six replicates. The confirmation of PFOA by QQQ mass spectrometry is also done by using qualifier ion ratios.

Abstract:
Novel Aspect:
Systematic evaluation of increase in protein identifications in complex samples with wide dynamic range

Introduction:
It is generally acknowledged that the MudPIT approach to proteomic profiling of complex samples provides confident indication of only a subset of the actual proteins present in the sample. The number of confident protein identifications is closely related to the number of good-quality MS/MS spectra from discrete peptides. Significant factors impacting this include the peak capacity and efficiency of the separation; the MS/MS scan acquisition rate, and the sample path from injection through detection. This work examines the impact of capillary LC versus nanoflow LC on protein identification results from complex samples.

Methods:
Nano LC/MS/MS was performed on a microfluidic chip-based device coupled to a Q-TOF mass spectrometer. Capillary LC/MS/MS was done on the same mass spectrometer using a standard ESI source with a pneumatically-assisted nebulizer specifically optimized for microflow application (less than 10 ìL/min). The inner needle was designed for smaller droplet formation at low flow rates and for maintaining the fidelity of the chromatographic separation. Both systems utilized 150 mm long C-18 columns and the chromatographic conditions and sample loading were adjusted to compensate for the difference in column inner diameter. Baseline chromatographic studies were conducted using a standard protein digests. Protein identification and sequence coverage evaluations were performed using complex samples such as E. coli lysate.

Preliminary Results:
Studies using a standard BSA digest showed that the sequence coverage and number of unique peptides identified was similar for the 75 ìm microfluidic-chip and the 300 ìm capillary column using sample loads adjusted to the column i.d. However, preliminary studies with the more complex E. coli lysate showed greater differences between nanoLC and capillary LC. With this sample, the 300 ìm capillary column showed more than a 40% increase in proteins identified and 60% increase in unique peptides. Further studies will be done to examine column capacity as well as the effect of sample overloading on increasing protein identification.

Abstract:
Novel Aspect:
Within conditions studied, the degree to which deconvolution bridges the gap between GC/MS and GC/MS/MS will be defined.

Introduction:
Although the selectivity of capillary GC/quadrupole MS has proven adequate for many types of analyses, the requirement for lower detection limits in increasingly complex samples has driven GC/MS to the limits of acceptable performance. NIST AMDIS (Automated Mass Spectral Deconvolution and Identification Software) is a mathematical solution to the selectivity limitation of GC/MS, but the degree to which deconvolution extends the boundaries of performance are not well documented. GC/MS/MS, the new standard for selectivity, has obvious performance advantages that are offset by the practical constraints of cost. The study goal is a careful assessment of the degree to which deconvolution software can bridge the gap between GC/MS and GC/MS/MS for trace analysis in complex separations with overlapping matrix peaks.

Methods:
Using scan TIC as a first estimate of the amount of coelution, spiked samples will be prepared with known ratios of semi-volatile analytes of interest versus the interfering response of the matrix. Extracts of foods and biological fluids, industrial hydrocarbon samples, and mixtures of these matrices will be used to adjust the amount of chemical noise overlapping the spectral region of interest. From a base case that established the quantitative and qualitative limits of GC/MS, samples with decreasing analytes to matrix ratios are reanalyzed by GC/MS with deconvolution report software and GC/MS/MS. Quantitative accuracy and clarity of qualitative information is assess for each method.

Preliminary Results:
The ionization power of EI has the ability to generate many similar ions from overlapping compounds in a capillary GC separation. For trace analysis based upon ions extracted from high backgrounds of matrix ions, false positive and false negative results are possible. Working with up to five analytes, detection limits were determined within each matrix (analytes correctly identified and quantified within +/- 30% of the known concentration of the spiked standard). From this reference, the ratio of analyte to matrix was changed by a geometric sequence of 2, 4, 8, . . . 64 by decreasing the analyte concentration or increasing the matrix concentration. Results from deconvolution of the GC/MS data and GC/MS/MS data were compared to known, spiked concentrations for the trace level analytes. The degree to which deconvolution software was able to extend the performance of the GC/MS to lower analyte/matrix ratios was influenced by the orthogonality of the set of analyte ions from the set of matrix ions. Having a few unique analyte ions (even though not the typical quan ions) greatly improved the performance of the deconvolution process. GC/MS/MS consistently identified the analytes of interest in the increasingly complex matrices, although imprecision increased presumably due to injection and source artifacts from the coelution of a concentrated matrix.

Abstract:
Novel Aspect:
New and fully automated SPE/LC/MS/MS-method for all drugs,which are sanctioned by the german road traffic act

Introduction:
Besides cannabis, amphetamine, methamphetamine, MDMA, MDE and MDA, driving under the influence of cocaine and heroin is also sanctioned by the german road traffic act. Therefore it would be desirable to establish an analytical method which enables the simultaneous determination of cocaine and its metabolites as well as opiates along with amphetamines and cannabinoids. As preceding work has shown,a RRLC/QQQ- method was particularly suitable for a sensitive and reproducible determination of amphetamines and cannabinoids in serum. The aim of the present study was to extend this method on cocaine, benzoylecgonine, methylecgonine, 6-monoacetylmorphine, morphine and codeine. The involvement of online SPE should lead to a complete automated setup, which combines the sample preparation step with the subsequent LC/MS/MS analysis.

Methods:
Chromatographic separation and MS detection were performed using an Agilent 1200 Rapid Resolution LC and a 6410 Triple Quadrupole Mass Spectrometer. A Gerstel Multi Purpose Sampler with an automated SPE accessory was integrated into the LC/MS/MS After the extraction step, the eluate was directly injected into the LC/MS/MS system. The established method for the analysis of cannabinoids and amphetamines was modified to integrate the opiates and cocaine derivatives. The chromatographic separation was performed on a Zorbax SB-aq column with 1.8 ìm particle size and 0.1 % formic acid and methanol as eluents. The QQQ was operated with an Agilent G1978B multimode ion source. The total analysis time, including conditioning the column, was set to 15 min.

Preliminary Results:
Online SPE combined with a direct sample introduction into the LC/MS/MS system enabled an automated clean-up step immediately followed by the analysis of the drugs of abuse. The solid phase extraction was carried out in two steps. First the cannabinoids were eluted, followed by the elution of the amphetamines, opiates and cocaine with its metabolites. The total eluate was dried with the gas purge option of the SPE unit and reconstituted in the mobile phase. To improve the recovery rates of the analytes, optimization of the evaporation temperature and the purge time will be carried out in the near future. The revised LC/MS/MS method now contains all the MRM transitions of the relevant drugs of abuse, which are sanctioned by the german road traffic act. The use of a 1.8 ìm particle size RRHT column for the LC separation provided a faster analysis (cycle time 15 min) than GC/MS (cycle time 45 min). Due to the polarity differences of the different compound classes, the use of the multimode ion source allowed the detection of all compounds of interest with an optimal response for each analyte due to a change of the ionization mode on a time-based scale. The validation of the method was carried out according to the guidelines of the German Society of Toxicology and Forensic Chemistry (GTFCh). The method validation was performed by using the Microsoft Excel based validation program VALISTAT. The presented method exhibits good within-day and day-to-day reproducibility. The coefficents of variation ranged from 3 to 15 %; most of the coefficients were in the 5 - 10 % range. A new validation, which includes also the SPE step, will be presented in the final release of this publication.

Abstract:
Novel Aspect:
A novel chip, allowing efficient enrichment and separation of analytes over a broad polarity range with high sensitivity is presented.

Introduction:
Mass spectrometry has become a routine tool in a multitude of applications, as it yields qualitative, quantitative as well as structural information of the analytes. The demand for steadily increasing detection sensitivity in proteomics has pushed the miniaturization procedure of chromatography down to the nano-HPLC level. Nano-LC, however, is associated with problems and limitations, like cumbersome manageability, excessive extra-column band broadening or lacking reproducibility. The HPLCchip/ MS approach significantly overcomes those problems by integrating all connectors, fittings and transfer capillaries on one polymeric chip. As identification of drugs and drug metabolites in the early drug discovery phase is usually constrained by low sample volume and concentration, HPLC-chip/MS is introduced for the separation of pharmaceutical compounds in a highly complex biological matrix.

Methods:
The novel ultra high capacity (UHC) chip, designed for the analysis of pharmaceuticals (atenolol, atropine, metoprolol, propranolol, imipramine, caffeine, quinine as model analytes) in a broad range of polarity (log P values ranging from 0 to 4.0), contains an enrichment column with considerably larger volume (500 nl compared to 40 or 160 nl for the standard chip design) and a separation column (150x0.075 mm), directly coupled to a 10 &mum ESI spray tip (both columns packed with ZORBAX 8 nm SBC18, 5 &mum). The HPLC-chip was operated with a nano-LC, consisting of a thermostatted micro-wellplate autosampler, microdegasser, capillary-pump for enrichment, nano-pump and a HPLC-Chip/MS interface. A Triple Quadrupole-MS was used for sample identification and chip evaluation.

Preliminary Results:
The methodology was applied to a variety of pharmaceutically relevant compounds. The chip exhibits a distinctively improved trapping behavior with respect to hydrophilic analytes (90 % and 70 % peak area increase for caffeine and quinine, respectively) and with respect to separation efficiency, which can be ascribed to a re-focusing effect of hydrophilic compounds on the enrichment column in back-flush mode. The UHC-chip meets all important chromatographic standards – run-to-run reproducibility has been found to be < 0.2 % RSD, whereas column-to-column reproducibility typically was < 1.1 % RSD (0.74 % on average) and peak area precision was ranging between 3.1 and 8.1 % RSD (5.1 % on average), using the Triple Quadrupole as mass detector in MRM mode.

Compared to a narrow bore column (2.1x150 mm), operated with a standard ESI interface, the UHC-chip resulted in a 100 fold increase in signal response. To simulate typical drug metabolism and pharmacokinetics (DMPK) workflows in pharmaceutical industry, atropine and imipramine were spiked into blood serum, followed by protein precipitation. The limit of detection (LOD) was determined to be < 10 fg absolute on column.

Abstract:
Novel Aspect:
Analysis and comparison of a set of forensic toxicology samples on all the common types of LC/MS analyzers.

Introduction:
The information required for forensic and clinical toxicology ranges from initial screening through high-confidence confirmation and quantification, and all of these can be obtained using API LC/MS. This study analyzed the same set of blood extracts from actual medical examiner and DUID cases over a period of three days on five different LC/MS systems: single quadrupole, triple quadrupole, 3-D ion trap, accurate-mass TOF and accurate-mass Q-TOF, using identical LC conditions on all the systems. Examination of the resulting data shows the advantages and disadvantages of each type of mass analyzer for the analysis of low levels of drugs and metabolites in a complex biological matrix, and compares their abilities for both qualitative and quantitative analysis.

Methods:
Blood samples (replicate 1mL aliquots with deuterated IStds) from ME and DUID cases were extracted using SPE appropriate to the three analyte classes being sought: cocaine and metabolites, methadone, and benzodiazepines. The final extracts were evaporated to dryness in screwcap tubes and shipped frozen overnight to the Agilent applications laboratory. Extracts were reconstituted with the initial mobile phase composition in a volume that would allow for analysis of each sample on five systems. LC/MS analysis using fast chromatography with Zorbax RR-HT small-particle columns was carried out with the same mobile phase and gradient on a set of identical columns on multiple instruments. In addition, blood samples of the RTI set were screened directly and as extracts using DART-TOF at RTI.

Preliminary Results:
The following modes of acquisition were utilized: single quadrupole (SQ), SIM; triple quadrupole (QQQ), MRM; ion trap, auto- MS/MS; TOF, accurate-mass scan; Q-TOF, accurate-mass MS and targeted MS/MS. Quantification was carried out using the instruments’ data analysis capabilities on the SQ, QQQ, TOF and Q-TOF. Screening and identification were carried out using library search of MS/MS and MS3 spectra on the trap, and accurate-mass database search on the TOF and Q-TOF using Mass Hunter software. Comparable quantification results were obtained on the SQ, QQQ and TOF instruments for the samples for which the analytes were within the calibration range. With the sample size and preparation used in this study, accurate quantification was possible down to 5ng/mL for all analytes on the SQ and TOF instruments, and down to 1 ng/mL for most analytes on the QQQ. In addition, the accurate-mass scan data from the TOF and Q-TOF instruments allowed a retroactive search for other suspected drugs in the samples without knowing a priori that they might be present. This is a particular advantage to these analyzers in the forensic arena in which polypharmacy is often encountered but not suspected until after the initial targeted analysis. No re-injection is required for the TOF/Q-TOF data which is “all-scan all the time”, in contrast to the SIM and MRM methods which are targeted for a particular analyte list and which require some MS parameter optimization for each analyte. The results for the DART-TOF analysis of the ME samples were mixed, in that the higher-level drugs from overdose cases could be detected using the directintroduction technique, but at more typical levels of < 100ng/mL, most of the drugs found by the LC/MS systems could not be reliably detected.

Abstract:
Novel Aspect:
Use of accurate mass, retention time and molecular formula isotope ratios to identify urine metabolites from a database

Introduction:
Untargeted metabolomics workflows typically include sample preparation, data collection, feature finding, statistical analysis and compound identification. A significant challenge for compound identification is the absence of a comprehensive LC/MS based database of retention time matched compounds. In collaboration with the Scripps Research Institute we have added Retention time (RT) and Molecular Formula Generation (MFG) capabilities to increase identification confidence of compounds using content from over 15,000 mass entries. To demonstrate the utility of this customizable personal metabolite database, a library of urine metabolites was developed from a limited set of 78 standards, representing some of the compounds normally found in urine. Human urine samples were then analyzed and screened for mass and RT matches against this database

Methods:
A mixture of metabolite standards in urine was provided and ultrafiltrated. The urine was dried down and analyzed by LC/MS on a 2.1mm reverse phase column. LC/ESI- time-of-flight (TOF) mass spectrometry (MS) in positive ion mode was used to acquire mass spectra, and generated a list of masses and associated RTs. The masses were then compared against the database of urine standards. An automated Molecular Formula Generation (MFG) capability that uses isotopic pattern matching to enable higher confidence identification of compounds from accurate mass data was also incorporated and used to confirm the results of the database matches. Finally, a list of masses with a specific RT, a metabolite compound match and a putative formula(s) was generated

Preliminary Results:
Identified metabolites were annotated with a chemical formula, structure and other metadata, including CAS and KEGG identifiers. From a total of four data files collected from the urine LC/MS analysis, 1070 compounds were found to be present. From these 436 had a match to the database and only 4 did not correspond to the expected molecule formulae. This is a very high success rate for this approach of metabolite identification. For masses that do not have a database match the molecular formula generated can still be used to guide further experiments for compound identification and in this case 563 compounds were not matched but each did generate a suggested molecular formula. If identification was to be made on these compounds then there is capability to include this new information in a custom database