NS-kNN: a modified <Emphasis Type="Italic">k</Emphasis>-nearest neighbors approach for imputing metabolomics data

Introduction

A common problem in metabolomics data analysis is the existence of a substantial number of missing values, which can complicate, bias, or even prevent certain downstream analyses. One of the most widely-used solutions to this problem is imputation of missing values using a k-nearest neighbors (kNN) algorithm to estimate missing metabolite abundances. kNN implicitly assumes that missing values are uniformly distributed at random in the dataset, but this is typically not true in metabolomics, where many values are missing because they are below the limit of detection of the analytical instrumentation.

Objectives

Here, we explore the impact of nonuniformly distributed missing values (missing not at random, or MNAR) on imputation performance. We present a new model for generating synthetic missing data and a new algorithm, No-Skip kNN (NS-kNN), that accounts for MNAR values to provide more accurate imputations.

Methods

We compare the imputation errors of the original kNN algorithm using two distance metrics, NS-kNN, and a recently developed algorithm KNN-TN, when applied to multiple experimental datasets with different types and levels of missing data.

Results

Our results show that NS-kNN typically outperforms kNN when at least 20–30% of missing values in a dataset are MNAR. NS-kNN also has lower imputation errors than KNN-TN on realistic datasets when at least 50% of missing values are MNAR.

Conclusion

Accounting for the nonuniform distribution of missing values in metabolomics data can significantly improve the results of imputation algorithms. The NS-kNN method imputes missing metabolomics data more accurately than existing kNN-based approaches when used on realistic datasets.

     

In Vitro Studies on the Putative Function of N-acetylaspartate as an Osmoregulator

Efflux and tissue content of N-acetylaspartate (NAA) and amino acids were evaluated from cultured and acutely prepared hippocampal slices in response to changes in osmolarity. The osmoregulator taurine, but not NAA, was lost from both types of slices after moderate reductions in extracellular osmolarity (−60 mOsm) for 10–48 h. Hypoosmotic shock (−166 mOsm) for 5 min resulted in unselective efflux of several amino acids from acutely prepared slices. Notably, the efflux of taurine, but not NAA, was prominent also after the shock. Efflux of NAA was markedly enhanced by NMDA and high K⁺, in particular after the stimulation period. The high K⁺-mediated efflux was decreased by high extracellular osmolarity and a NMDA-receptor antagonist. The results indicate that NAA efflux can be induced by a sudden non-physiological decrease in extracellular osmolarity but not by prolonged more moderate changes in osmolarity. The mechanisms behind the efflux of NAA by high K⁺ are complex and may involve both swelling and activation of NMDA-receptors.     

Inflammation-associated S100 proteins: new mechanisms that regulate function

This review focuses on new aspects of extracellular roles of the calgranulins. S100A8, S100A9 and S100A12 are constitutively expressed in neutrophils and induced in several cell types. The S100A8 and S100A9 genes are regulated by pro- and anti-inflammatory mediators and their functions may depend on cell type, mediators within a particular inflammatory milieu, receptors involved in their recognition and their post-translational modification. The S100A8 gene induction in macrophages is dependent on IL-10 and potentiated by immunosuppressive agents. S100A8 and S100A9 are oxidized by peroxide, hypochlorite and nitric oxide (NO). HOCl generates intra-chain sulfinamide bonds; stronger oxidation promotes cross-linked forms that are seen in human atheroma. S100A8 is >200-fold more sensitive to oxidative cross-linking than low-density lipoprotein and may reduce oxidative damage. S100A8 and S100A9 can be S-nitrosylated. S100A8–SNO suppresses mast cell activation and inflammation in the microcirculation and may act as an NO transporter to regulate vessel tone in inflammatory lesions. S100A12 activates mast cells and is a monocyte and mast cell chemoattractant; a G-protein-coupled mechanism may be involved. Structure–function studies are discussed in relation to conservation and divergence of functions in S100A8. S100A12 induces cytokines in mast cells, but not monocytes/macrophages. It forms complexes with Zn²⁺ and, by chelating Zn²⁺, S100A12 significantly inhibits MMPs. Zn²⁺ in S100A12 complexes co-localize with MMP-9 in foam cells in atheroma. In summary, S100A12 has pro-inflammatory properties that are likely to be stable in an oxidative environment, because it lacks Cys and Met residues. Conversely, S100A8 and S100A9 oxidation and S-nitrosylation may have important protective mechanisms in inflammation.     

Human Alzheimer’s disease synaptic <Emphasis Type=Italic>O</Emphasis>-GlcNAc site mapping and iTRAQ expression proteomics with ion trap mass spectrometry

Neuronal synaptic functional deficits are linked to impaired learning and memory in Alzheimer’s disease (AD). We recently demonstrated that O-GlcNAc, a novel cytosolic and nuclear carbohydrate post-translational modification, is enriched at neuronal synapses and positively regulates synaptic plasticity linked to learning and memory in mice. Reduced levels of O-GlcNAc have been observed in AD, suggesting a possible link to deficits in synaptic plasticity. Using lectin enrichment and mass spectrometry, we mapped several human cortical synaptic O-GlcNAc modification sites. Overlap in patterns of O-GlcNAcation between mouse and human appears to be high, as previously mapped mouse synaptic O-GlcNAc sites in Bassoon, Piccolo, and tubulin polymerization promoting protein p25 were identified in human. Novel O-GlcNAc modification sites were identified on Mek2 and RPN13/ADRM1. Mek2 is a signaling component of the Erk 1/2 pathway involved in synaptic plasticity. RPN13 is a component of the proteasomal degradation pathway. The potential interplay of phosphorylation with mapped O-GlcNAc sites, and possible implication of those sites in synaptic plasticity in normal versus AD states is discussed. iTRAQ is a powerful differential isotopic quantitative approach in proteomics. Pulsed Q dissociation (PQD) is a recently introduced fragmentation strategy that enables detection of low mass iTRAQ reporter ions in ion trap mass spectrometry. We optimized LTQ ion trap settings for PQD-based iTRAQ quantitation and demonstrated its utility in O-GlcNAc site mapping. Using iTRAQ, abnormal synaptic expression levels of several proteins previously implicated in AD pathology were observed in addition to novel changes in synaptic specific protein expression including Synapsin II.     

Phase-resetting curve determines how BK currents affect neuronal firing

BK channels are large conductance potassium channels gated by calcium and voltage. Paradoxically, blocking these channels has been shown experimentally to increase or decrease the firing rate of neurons, depending on the neural subtype and brain region. The mechanism for how this current can alter the firing rates of different neurons remains poorly understood. Using phase-resetting curve (PRC) theory, we determine when BK channels increase or decrease the firing rates in neural models. The addition of BK currents always decreases the firing rate when the PRC has only a positive region. When the PRC has a negative region (type II), BK currents can increase the firing rate. The influence of BK channels on firing rate in the presence of other conductances, such as I _m and I _h , as well as with different amplitudes of depolarizing input, were also investigated. These results provide a formal explanation for the apparently contradictory effects of BK channel antagonists on firing rates.     

Vanadyl bisacetylacetonate protects β cells from palmitate-induced cell death through the unfolded protein response pathway

Endoplasmic reticulum (ER) stress induced by free fatty acids (FFA) is important to β-cell loss during the development of type 2 diabetes. To test whether vanadium compounds could influence ER stress and the responses in their mechanism of antidiabetic effects, we investigated the effects and the mechanism of vanadyl bisacetylacetonate [VO(acac)₂] on β cells upon treatment with palmitate, a typical saturated FFA. The experimental results showed that VO(acac)₂ could enhance FFA-induced signaling pathways of unfolded protein responses by upregulating the prosurvival chaperone immunoglobulin heavy-chain binding protein/78-kDa glucose-regulated protein and downregulating the expression of apoptotic C/EBP homologous protein, and consequently the reduction of insulin synthesis. VO(acac)₂ also ameliorated FFA-disturbed Ca²⁺ homeostasis in β cells. Overall, VO(acac)₂ enhanced stress adaption, thus protecting β cells from palmitate-induced apoptosis. This study provides some new insights into the mechanisms of antidiabetic vanadium compounds.     

A novel zinc bis(thiosemicarbazone) complex for live cell imaging

Fluorescent zinc complexes have recently attracted a lot of interest owing to their vast applications in cellular imaging. We report the synthesis as well as physical, chemical and biological studies of a novel zinc glyoxalbis(4-methyl-4-phenyl-3-thiosemicarbazone), [Zn(GTSC)]₃, complex. As compared with the well-studied zinc biacetylbis(4-methyl-3-thiosemicarbazone), Zn(ATSM), complex, which was used as a reference, [Zn(GTSC)]₃ had 2.5-fold higher fluorescence. When cellular fluorescence was measured using flow cytometry, we observed that [Zn(GTSC)]₃ had 3.4-fold to 12-fold higher fluorescence than Zn(ATSM) in various cell lines (n = 9) of different tissue origin. Confocal fluorescence microscopy results showed that [Zn(GTSC)]₃ appeared to have a nuclear localization within 30 min of addition to MCF7 cells. Moreover, [Zn(GTSC)]₃ showed minimal cytotoxicity compared with Zn(ATSM), suggesting that [Zn(GTSC)]₃ may be less deleterious to cells when used as an imaging agent. Our data suggest that the novel [Zn(GTSC)]₃ complex can potentially serve as a biocompatible fluorescent imaging agent for live cells.     

Mass-spectrometric characterization of cisplatin binding sites on native and denatured ubiquitin

Because interactions between cisplatin and plasma proteins contribute to drug efficacy and side effects, it is important to understand both the binding sites of cisplatin on the proteins and the formation of protein–cisplatin adducts. Previous results suggest that cisplatin preferentially binds to residues on the protein surface. The present work employed electrospray ionization mass spectrometry (MS) to identify such sites on both native and denatured ubiquitin (Ub). Fourier transform (FT) MS and tandem MS (MS/MS and MS³) enable analysis of Ub–cisplatin adduct digests to locate specific cisplatin binding sites. Results indicate that there are three such binding sites, i.e., M1, T12 and T14, and D32, on native Ub. The intensity of the relevant peaks in the FT-MS spectrum of the native Ub adduct digest demonstrates that residues T12 and T14 comprise the primary cisplatin binding site under the native conditions rather than residue M1 as reported in previous research studies. It is found in the present work, however, that M1 is the primary binding site on denatured Ub. Comparison of cisplatin binding sites on native and denatured Ub in this research demonstrates that the conformation of a protein significantly influences the preference of cisplatin for specific binding sites.