Anja Freiwald

Classification and identification of bacteria by mass spectrometry and computational analysis

Background In general, the definite determination of bacterial species is a tedious process and requires extensive manual labour. Novel technologies for bacterial detection and analysis can therefore help microbiologists in minimising their efforts in developing a number of microbiological applications. Methodology We present a robust, standardized procedure for automated bacterial analysis that is based on the detection of patterns of protein masses by MALDI mass spectrometry. We particularly applied the approach for classifying and identifying strains in species of the genus Erwinia. Many species of this genus are associated with disastrous plant diseases such as fire blight. Using our experimental procedure, we created a general bacterial mass spectra database that currently contains 2800 entries of bacteria of different genera. This database will be steadily expanded. To support users with a feasible analytical method, we developed and tested comprehensive software tools that are demonstrated herein. Furthermore, to gain additional analytical accuracy and reliability in the analysis we used genotyping of single nucleotide polymorphisms by mass spectrometry to unambiguously determine closely related strains that are difficult to distinguish by only relying on protein mass pattern detection. Conclusions With the method for bacterial analysis, we could identify fire blight pathogens from a variety of biological sources. The method can be used for a number of additional bacterial genera. Moreover, the mass spectrometry approach presented allows the integration of data from different biological levels such as the genome and the proteome.

Phylogenetic classification and identification of bacteria by mass spectrometry

Bacteria are a convenient source of intrinsic marker proteins, which can be detected efficiently by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. The patterns of protein masses observed can be used for accurate classification and identification of bacteria. Key to the reliability of the method is a robust and standardized procedure for sample preparations, including bacterial culturing, chemical treatment for bacterial cell wall disruption and for protein extraction, and mass spectrometry analysis. The protocol is an excellent alternative to classical microbiological classification and identification procedures, requiring minimal sample preparation efforts and costs. Without cell culturing, the protocol takes in general <1 h.

Amorfrutins are potent antidiabetic dietary natural products

Given worldwide increases in the incidence of obesity and type 2 diabetes, new strategies for preventing and treating metabolic diseases are needed. The nuclear receptor PPARγ (peroxisome proliferator-activated receptor gamma) plays a central role in lipid and glucose metabolism; however, current PPARγ-targeting drugs are characterized by undesirable side effects. Natural products from edible biomaterial provide a structurally diverse resource to alleviate complex disorders via tailored nutritional intervention. We identified a family of natural products, the amorfrutins, from edible parts of two legumes, Glycyrrhiza foetida and Amorpha fruticosa, as structurally new and powerful antidiabetics with unprecedented effects for a dietary molecule. Amorfrutins bind to and activate PPARγ, which results in selective gene expression and physiological profiles markedly different from activation by current synthetic PPARγ drugs. In diet-induced obese and db/db mice, amorfrutin treatment strongly improves insulin resistance and other metabolic and inflammatory parameters without concomitant increase of fat storage or other unwanted side effects such as hepatoxicity. These results show that selective PPARγ-activation by diet-derived ligands may constitute a promising approach to combat metabolic disease.

Transcriptomics assisted proteomic analysis of Nicotiana occidentalis infected by Candidatus Phytoplasma mali strain AT

Phytoplasmas are pathogenic bacteria within the class of Mollicutes, which are associated with more than 1000 plant diseases. In this study, we applied quantitative mass spectrometry to analyse affected pathways of the model plant tobacco (Nicotiana occidentalis) upon Candidatus Phytoplasma mali strain AT infection. Using tissue obtained from leaf midribs, 1466 plant‐assigned proteins were identified. For 1019 of these proteins, we could reproducibly quantify the expression changes of infected versus noninfected plants, of which 157 proteins were up‐ and 173 proteins were downregulated. Differential expression took place in a number of pathways, among others strong downregulation of porphyrin and chlorophyll metabolism and upregulation of alpha‐linolenic acid metabolism, which was consistent with observed increased levels of jasmonic acid, a key signal molecule of plant defence. Our data shed light on the molecular networks that are involved in defence of plants against phytoplasma infection and provide a resource for further studies.

Antidiabetic effects of chamomile flowers extract in obese mice through transcriptional stimulation of nutrient sensors of the peroxisome proliferator-activated receptor (PPAR) family.

Given the significant increases in the incidence of metabolic diseases, efficient strategies for preventing and treating of these common disorders are urgently needed. This includes the development of phytopharmaceutical products or functional foods to prevent or cure metabolic diseases. Plant extracts from edible biomaterial provide a potential resource of structurally diverse molecules that can synergistically interfere with complex disorders. In this study we describe the safe application of ethanolic chamomile (Matricaria recutita) flowers extract (CFE) for the treatment and prevention of type 2 diabetes and associated disorders. We show in vitro that this extract activates in particular nuclear receptor peroxisome proliferator-activated receptor gamma (PPARγ) and its isotypes. In a cellular context, in human primary adipocytes CFE administration (300 µg/ml) led to specific expression of target genes of PPARγ, whereas in human hepatocytes CFE-induced we detected expression changes of genes that were regulated by PPARα. In vivo treatment of insulin-resistant high-fat diet (HFD)-fed C57BL/6 mice with CFE (200 mg/kg/d) for 6 weeks considerably reduced insulin resistance, glucose intolerance, plasma triacylglycerol, non-esterified fatty acids (NEFA) and LDL/VLDL cholesterol. Co-feeding of lean C57BL/6 mice a HFD with 200 mg/kg/d CFE for 20 weeks showed effective prevention of fatty liver formation and hepatic inflammation, indicating additionally hepatoprotective effects of the extract. Moreover, CFE treatment did not reveal side effects, which have otherwise been associated with strong synthetic PPAR-targeting molecules, such as weight gain, liver disorders, hemodilution or bone cell turnover. Taken together, modulation of PPARs and other factors by chamomile flowers extract has the potential to prevent or treat type 2 diabetes and related disorders.

Lemon balm extract causes potent antihyperglycemic and antihyperlipidemic effects in insulin‐resistant obese mice

Over the last decades polyetiological metabolic diseases such as obesity and type 2 diabetes have emerged as a global epidemic. Efficient strategies for prevention and treatment include dietary intervention and the development of validated nutraceuticals. Safe extracts of edible plants provide a resource of structurally diverse molecules that can effectively interfere with multifactorial diseases. In this study, we describe the application of ethanolic lemon balm (Melissa officinalis) leaves extract for the treatment of insulin-resistance and dyslipidemia in mice. We show that lemon balm extract (LBE) activates the peroxisome proliferator-activated receptors (PPARs), which have key roles in the regulation of whole body glucose and lipid metabolism. Application of LBE (0.6 mg/mL) to human primary adipocytes resulted in specific peroxisome proliferator-activated receptor target gene expression. LBE treatment of insulin-resistant high-fat diet-fed C57BL/6 mice (200 mg/kg/day) for 6 weeks considerably reduced hyperglycemia and insulin resistance, plasma triacylglycerol, nonesterified fatty acids and LDL/VLDL cholesterol levels. Taken together, ethanolic lemon balm extract can potentially be used to prevent or concomitantly treat type 2 diabetes and associated disorders such as dyslipidemia and hypercholesterolemia.

ATR inhibition rewires cellular signaling networks induced by replication stress

The slowing down or stalling of replication forks is commonly known as replication stress and arises from multiple causes such as DNA lesions, nucleotide depletion, RNA‐DNA hybrids, and oncogene activation. The ataxia telangiectasia and Rad3‐related kinase (ATR) plays an essential role in the cellular response to replication stress and inhibition of ATR has emerged as therapeutic strategy for the treatment of cancers that exhibit high levels of replication stress. However, the cellular signaling induced by replication stress and the substrate spectrum of ATR has not been systematically investigated. In this study, we employed quantitative MS‐based proteomics to define the cellular signaling after nucleotide depletion‐induced replication stress and replication fork collapse following ATR inhibition. We demonstrate that replication stress results in increased phosphorylation of a subset of proteins, many of which are involved in RNA splicing and transcription and have previously not been associated with the cellular replication stress response. Furthermore, our data reveal the ATR‐dependent phosphorylation following replication stress and discover novel putative ATR target sites on MCM6, TOPBP1, RAD51AP1, and PSMD4. We establish that ATR inhibition rewires cellular signaling networks induced by replication stress and leads to the activation of the ATM‐driven double‐strand break repair signaling.

Comprehensive proteomic data sets for studying adipocyte-macrophage cell-cell communication.

Cellular communication is a fundamental process in biology. The interaction of adipocytes with macrophages is a key event in the development of common diseases such as type 2 diabetes. We applied an established bilayer cell coculture system and comprehensive MS detection to analyse on a proteome‐wide scale the paracrine interaction of murine adipocytes and macrophages. Altogether, we identified 4486 proteins with at least two unique peptides, of which 2392 proteins were informative for 3T3‐L1 adipocytes and 2957 proteins for RAW 264.7 macrophages. Further, we observed over 12000 phosphorylation sites, of which we could assign 3200 informative phosphopeptides with a single phosphosite for adipocytes and 4514 for macrophages. Using protein set enrichment and phosphosite analyses, we deciphered regulatory protein pathways involved in cellular stress and inflammation, which can contribute to metabolic impairment of cells including insulin resistance and other disorders. The generated datasets provide a holistic, molecular pathway‐centric view on the interplay of adipocytes and macrophages in disease processes and a resource for further studies.

Differential Analysis of Crohn's Disease and Ulcerative Colitis by Mass Spectrometry

To the Editor: Crohn’s disease (CD) and ulcerative colitis (UC) are the major types of inflammatory bowel disease (IBD). Conventional diagnosis of IBDs is based on endoscopic, histological, and radiological data. Biomarkers are postulated to enable more precise and rapid analysis to determine complex IBD subtypes. Single nucleotide polymorphisms (SNPs) were thought to be potential biomarkers for the diagnosis of CD. A recent meta-analysis of NOD2 genetic variants, i.e., SNPs Leu1007fsinsC, Arg702Trp, and Gly908Arg, has shown that the risk for CD was 2-fold higher for simple heterozygotes and 7–9-fold higher for compound heterozygotes and homozygotes compared with noncarriers. Recently, copy number variations (CNVs) in the betadefensin gene were proposed as valuable biomarkers for CD. However, later studies in additional patient cohorts could not confirm an association of these CNVs with CD. Although DNA-based biomarkers have a potential to predict disease risk, they may be limited in providing improved differential analysis and when monitoring environmental factors such as nutrition or medication. We propose that integrated molecular analysis of complex IBDs may benefit from using new approaches such as mass spectrometry detection of serum peptidomes. Serum is a clinically common source, which contains a complex array (or fingerprint) of proteolytically derived peptides that may provide a sensitive and specific disease correlate. Subsets of serum peptides can be used effectively to distinguish CD and UC patients and healthy controls. For generation of reference spectra and an appropriate analysis model, we used 60 serum samples from IBD patients (30 CD, 30 UC), which were carefully diagnosed by experienced gastroenterologists, and 45 ageand gendermatched control samples from healthy blood donors. Blood samples were treated according to standardized protocols and fractionated using reversephase (HIC-8) and ion exchange (WCX) magnetic beads (Bruker Daltonics, Leipzig, Germany). ClinProTools software v. 2.2 (Bruker Daltonics) was applied for multiple spectra data analysis and run reproducibility was calculated using the coefficient of variation (CV) over all serum samples (C, CD, UC) including 125 peak intensities. For univariate analysis, a total of 28 peptide mass signals revealed significant differences between different groups. Among them, 10 peptides were CD-specific and 8 were UC-specific. In addition, another 10 peptide signals differentiated between control and patient samples. In order to optimize the analysis, we tested different combinations of two specific peaks to find the best combination of variant class separation criteria. However, UC and CD data overlapped partly despite exhaustive testing of bivariate combinations (Fig. 1). Therefore, we finally used multivariate classification, support vector machine (SVM) analysis, to generate classification models based on an array of mass peak data and to classify mass peaks to different categories of patient samples. After model generation, recognition capacity was obtained and crossvalidation was calculated using the leave-one-out option of ClinProTools. In summary, we observed that data generated from HIC-8-treated serum were superior in differentiation between healthy and diseased samples, whereas data produced by WCX-treatment could be used to accurately distinguish between CD and UC samples. We applied HIC-8 in a first step to perform class separation of controls and diseased samples. Subsequently, for sample classification between CD and UC we used WCX data based on the preassigned disease samples. This tandem model classification for the disease/control separation allowed us to gain a cross-validation of 91.21% and a recognition capacity of 97.73% by taking the 10 best separating peaks into account. The sensitivity found by additional validation of 60 diseased and 45 control samples was 98.33% (probability that the test has a positive result when the patient is truly diseased) and the specificity was 84.09% (probability that the assay has a negative result when the patient is truly not diseased). For discrimination between CD and UC we could improve the cross-validation to 82.99% and recognition capacity of 94.94% by manually taking the 11 best peaks into account. The sensitivity observed was 79.31% and the specificity was 86.67%. Our data provide the basis for longitudinal studies including several patient subgroups to validate the diagnostic potential of multivariate classification of IBD serum peptidomes that can be combined with additional biomarker analyses to potentially improve analytic accuracy. However, serum samples in many IBD biobanks were not standardized as is required for reliable serum peptidome analysis (details described previously). Large-scale validation of the diagnostic potential of our approach would require significant efforts in setting up standardized serum samples in a number of clinical laboratories.

Discovery and characterization of protein-modifying natural products by MALDI mass spectrometry reveal potent SIRT1 and p300 inhibitors.

Natural products are a rich resource for the development of chemical probes, functional foods (nutraceuticals), and pharmaceuticals. The remarkable structural diversity of natural product or biology-oriented compound libraries in particular argues for their use in applications such as drug screenings. Enzymes that can posttranslationally modify proteins or nucleic acids are attractive drug targets and include kinases and phosphatases, methyltransferases and acetyltransferases, and deacetylases. Recently, histone-modifying enzymes like the deacetylase sirtuin 1 (SIRT1) were suggested as drug targets for treating a variety of age-related disorders such as neuropathogenic diseases, metabolic diseases, and cancer. Compound screenings for SIRT1 modulators have revealed promising enzymatic activators such as the natural product resveratrol and the synthetic compound SRT1720. However, these findings are still highly controversial due to reported assay artifacts. The employed screening assays were based on fluorescence-labeled peptide substrates and resulted in the purported but artificial enzymatic activation of SIRT1. These findings have misled many researchers over the years. Besides the generation of artifacts as observed in SIRT1 assays, there is a second general drawback of fluorescencebased assays which is broadly underestimated: Autofluorescence of compound libraries and in particular natural product libraries interferes with widely applied optical analyses such as time-resolved fluorescence resonance energy transfer (TRFRET). By contrast, mass spectrometry (MS) represents an attractive alternative as substrates are detected directly. Peptides are ionized and detected according to their mass to charge ratios (m/z). For example, deacetylation of a substrate peptide can be directly detected as a 42 Da mass peak shift (Figure 1).