First-trimester metabolomic prediction of stillbirth.

Background: Stillbirth remains a major problem in both developing and developed countries. Omics evaluation of stillbirth has been highlighted as a top research priority. Objective: To identify new putative first-trimester biomarkers in maternal serum for stillbirth prediction using metabolomics-based approach. Methods: Targeted, nuclear magnetic resonance (NMR) and mass spectrometry (MS), and untargeted liquid chromatography-MS (LC-MS) metabolomic analyses were performed on first-trimester maternal serum obtained from 60 cases that subsequently had a stillbirth and 120 matched controls. Metabolites by themselves or in combination with clinical factors were used to develop logistic regression models for stillbirth prediction. Prediction of stillbirths overall, early (<28 weeks and <32 weeks), those related to growth restriction/placental disorder, and unexplained stillbirths were evaluated. Results: Targeted metabolites including glycine, acetic acid, L-carnitine, creatine, lysoPCaC18:1, PCaeC34:3, and PCaeC44:4 predicted stillbirth overall with an area under the curve [AUC, 95% confidence interval (CI)] = 0.707 (0.628-0.785). When combined with clinical predictors the AUC value increased to 0.740 (0.667-0.812). First-trimester targeted metabolites also significantly predicted early, unexplained, and placental-related stillbirths. Untargeted LC-MS features combined with other clinical predictors achieved an AUC (95%CI) = 0.860 (0.793-0.927) for the prediction of stillbirths overall. We found novel preliminary evidence that, verruculotoxin, a toxin produced by common household molds, might be linked to stillbirth. Conclusions: We have identified novel biomarkers for stillbirth using metabolomics and demonstrated the feasibility of first-trimester prediction.

IsoMS: automated processing of LC-MS data generated by a chemical isotope labeling metabolomics platform.

A chemical isotope labeling or isotope coded derivatization (ICD) metabolomics platform uses a chemical derivatization method to introduce a mass tag to all of the metabolites having a common functional group (e.g., amine), followed by LC-MS analysis of the labeled metabolites. To apply this platform to metabolomics studies involving quantitative analysis of different groups of samples, automated data processing is required. Herein, we report a data processing method based on the use of a mass spectral feature unique to the chemical labeling approach, i.e., any differential-isotope-labeled metabolites are detected as peak pairs with a fixed mass difference in a mass spectrum. A software tool, IsoMS, has been developed to process the raw data generated from one or multiple LC-MS runs by peak picking, peak pairing, peak-pair filtering, and peak-pair intensity ratio calculation. The same peak pairs detected from multiple samples are then aligned to produce a CSV file that contains the metabolite information and peak ratios relative to a control (e.g., a pooled sample). This file can be readily exported for further data and statistical analysis, which is illustrated in an example of comparing the metabolomes of human urine samples collected before and after drinking coffee. To demonstrate that this method is reliable for data processing, five (13)C2-/(12)C2-dansyl labeled metabolite standards were analyzed by LC-MS. IsoMS was able to detect these metabolites correctly. In addition, in the analysis of a (13)C2-/(12)C2-dansyl labeled human urine, IsoMS detected 2044 peak pairs, and manual inspection of these peak pairs found 90 false peak pairs, representing a false positive rate of 4.4%. IsoMS for Windows running R is freely available for noncommercial use from www.mycompoundid.org/IsoMS.

Highly efficient classification and identification of human pathogenic bacteria by MALDI-TOF MS.

Accurate and rapid identification of pathogenic microorganisms is of critical importance in disease treatment and public health. Conventional work flows are time-consuming, and procedures are multifaceted. MS can be an alternative but is limited by low efficiency for amino acid sequencing as well as low reproducibility for spectrum fingerprinting. We systematically analyzed the feasibility of applying MS for rapid and accurate bacterial identification. Directly applying bacterial colonies without further protein extraction to MALDI-TOF MS analysis revealed rich peak contents and high reproducibility. The MS spectra derived from 57 isolates comprising six human pathogenic bacterial species were analyzed using both unsupervised hierarchical clustering and supervised model construction via the Genetic Algorithm. Hierarchical clustering analysis categorized the spectra into six groups precisely corresponding to the six bacterial species. Precise classification was also maintained in an independently prepared set of bacteria even when the numbers of m/z values were reduced to six. In parallel, classification models were constructed via Genetic Algorithm analysis. A model containing 18 m/z values accurately classified independently prepared bacteria and identified those species originally not used for model construction. Moreover bacteria fewer than 10(4) cells and different species in bacterial mixtures were identified using the classification model approach. In conclusion, the application of MALDI-TOF MS in combination with a suitable model construction provides a highly accurate method for bacterial classification and identification. The approach can identify bacteria with low abundance even in mixed flora, suggesting that a rapid and accurate bacterial identification using MS techniques even before culture can be attained in the near future.

Analysis of perfluorooctanesulfonate and related fluorochemicals in water and biological tissue samples by liquid chromatography-ion trap mass spectrometry.

A method for the determination of perfluorinated compounds (PFCs) in various water and biological tissue samples was developed and validated. The contents of selected PFCs (i.e., perfluorooctanesulfonate (PFOS), perfluorooctanoate (PFOA) and perfluorodecanoate (PFDA)) in water samples were extracted by the C(18) solid-phase extraction (SPE). The biological tissue samples (frozen-dried fish and oysters) were simply extracted by liquid-solid extraction with MTBE and adding tetrabutylammonium hydrogensulfate (TBA) as an ion-pairing reagent. The analytes were then identified and quantitated by liquid chromatography-ion trap negative electrospray mass spectrometry (LC-ESI ion-trap-MS). Limits of quantitation (LOQ) were established between 0.5 and 6 ng/l in 250 ml of water sample, while 5-50 ng/g (dry weight) for biological tissue sample. Intrabatch and interbatch precision with their accuracy at two concentration levels were also investigated. Precision for these three PFCs, as indicated by RSD, proved to be less than 11 and 17%, respectively. The total contents of PFOA, PFOS and PFDA were detected in concentrations of up to 400 ng/l in various water samples, while up to 1,100 ng/g in fish and oyster samples. PFOA and PFDA was the major PFCs detected in water samples and biological tissue samples, respectively.