Using machine learning for chemical-free histological tissue staining.

Hematoxylin and eosin staining can be hazardous, expensive, and prone to error and variability. To circumvent these issues, artificial intelligence/machine learning models such as generative adversarial networks (GANs), are being used to 'virtually' stain unstained tissue images indistinguishable from chemically stained tissue. Frameworks such as deep convolutional GANs (DCGAN) and conditional GANs (CGANs) have successfully generated highly reproducible 'stained' images. However, their utility may be limited by requiring registered, paired images which can be difficult to obtain. To avoid these dataset requirements, we attempted to use an unsupervised CycleGAN pix2pix model(5,6) to turn unpaired, unstained bright-field images into pathologist-approved digitally 'stained' images. Using formalin-fixed-paraffin-embedded liver samples, 5µm section images (20x) were obtained before and after staining to create "stained" an "unstained" datasets. Model implementation was conducted using Ubuntu 20.04.4 LTS, 32 GB RAM, Intel Core i7-9750 CPU @2.6 GHz, Nvidia GeForce RTX 2070 Mobile, Python 3.7.11 and Tensorflow 2.9.1. The CycleGAN framework utilized a u-net-based generator and discriminator from pix2pix, a CGAN. The CycleGAN used a modified loss function, cycle consistent loss that assumed unpaired images, so loss was measured twice. To our knowledge, this is the first documented application of this architecture using unpaired bright-field images. Results and suggested improvements are discussed.

Discovery of treatment for nerve agents targeting a new metabolic pathway.

The inhibition of acetylcholinesterase is regarded as the primary toxic mechanism of action for chemical warfare agents. Recently, there have been numerous reports suggesting that metabolic processes could significantly contribute to toxicity. As such, we applied a multi-omics pipeline to generate a detailed cascade of molecular events temporally occurring in guinea pigs exposed to VX. Proteomic and metabolomic profiling resulted in the identification of several enzymes and metabolic precursors involved in glycolysis and the TCA cycle. All lines of experimental evidence indicated that there was a blockade of the TCA cycle at isocitrate dehydrogenase 2, which converts isocitrate to α-ketoglutarate. Using a primary beating cardiomyocyte cell model, we were able to determine that the supplementation of α-ketoglutarate subsequently rescued cells from the acute effects of VX poisoning. This study highlights the broad impacts that VX has and how understanding these mechanisms could result in new therapeutics such as α-ketoglutarate.

Immunomagnetic separation and quantification of butyrylcholinesterase nerve agent adducts in human serum.

A novel method for extracting butyrylcholinesterase (BuChE) from serum as a means of identifying and measuring nerve agent adducts to human BuChE is presented here. Antibutyrylcholinesterase monoclonal antibodies were conjugated to protein-G ferromagnetic particles and mixed with 500 microL serum samples. The particle-antibody-BuChE product was rinsed and directly digested with pepsin. Native and isotopically enriched nonapeptides corresponding to the pepsin digest products for uninhibited BuChE, and sarin, cyclohexylsarin, VX, and Russian VX nerve agent-inhibited BuChE were synthesized for use as calibrators and internal standards, respectively. Internal standards were added to the filtered digest sample, and the samples were quantified via high performance liquid chromatography-isotope dilution-tandem mass spectrometry. The ratio of adducted to total BuChE nonapeptides was calculated for each nerve agent-exposed serum sample using data collected in a single chromatogram. Nerve agent-inhibited quality control serum pools were characterized as part of method validation; the method was observed to have extremely low background noise. The measurement of both uninhibited and inhibited BuChE peptides compensated for any variations in the pepsin digestion before the internal standard peptide was added to the sample and may prove useful in individualizing patient results following a nerve agent exposure.

Validation and application of a GC-MS method for determining soman concentration in rat plasma following low-level vapor exposure.

A method for determining the chemical warfare agent soman (GD) in rat plasma has been validated and applied to low-level inhalation exposure studies currently being conducted. This method utilizes a fluoride ion-based regeneration assay with isotope dilution followed by large volume injection gas chromatography with ammonia chemical ionization mass spectrometric detection. Following sample preparation by solid phase extraction, chromatographic separation was achieved using a 14% cyanopropylphenyl/86% dimethyl polysiloxane capillary column with a total run time of 18.16 min. Soman and the deuterated isotope ((2)H(4)-soman) internal standard were detected using the selected ion monitoring mode and quantitated using the ammonia adduction ratio of m/z ions 200/204. A reproducible linear relationship was obtained for the quantitative concentration range of 10 pg on-column to 1000 pg on-column (r(2) = 0.9995) for standards in ethyl acetate with a detection limit of 5.65 pg on-column, and an average recovery of 93% in plasma. This sensitive method was successfully applied to the analysis of soman in rat plasma immediately post-exposure, resulting in the construction of dose-response plots.