Monitoring the Bacterial Response to Antibiotic and Time Growth Using Near-infrared Spectroscopy Combined with Machine Learning

Rapid screening of bacteriostatic and bactericidal antimicrobial agents against Escherichia coli by combining machine learning (artificial intelligence) and UV-VIS spectroscopy

Antibiotics are compounds that have a particular mode of action upon the microorganism they are targeting. However, discovering and developing new antibiotics is a challenging and timely process. Antibiotic development process can take up to 10-15 years and over $1billion to develop a single new therapeutic product. Rapid screening tools to understand the mode of action of the new antimicrobial agent are considered one of the main bottle necks in the antimicrobial agent development process. Classical approaches require multifarious microbiological methods and they do not capture important biochemical and organism therapeutic-interaction mechanisms. This work aims to provide a rapid antibiotic-antimicrobial biochemical diagnostic tool to reduce the timeframes of therapeutic development, while also generating new biochemical insight into an antimicrobial-therapeutic screening assay in a complex matrix. The work evaluates the effect of antimicrobial action through "traditional" microbiological analysis techniques with a high-throughput rapid analysis method using UV-VIS spectroscopy and chemometrics. Bacteriostatic activity from tetracycline and bactericidal activity from amoxicillin were evaluated on a system using non-resistant Escherichia coli O157:H7 by confocal laser scanning microscopy (CLSM), scanning electron microscopy (SEM), and UV-VIS spectroscopy (high-throughput analysis). The data were analysed using principal component analysis (PCA) and support vector machine (SVM) classification. The rapid diagnostic technique could easily identify differences between bacteriostatic and bactericidal mechanisms and was considerably quicker than the "traditional" methods tested.

Combination of near-infrared spectroscopy with Wasserstein generative adversarial networks for rapidly detecting raw material quality for formula products

Near-infrared spectroscopy (NIRS) has emerged as a key technique for rapid quality detection owing to its fast, non-destructive, and eco-friendly characteristics. However, its practical implementation within the formulation industry is challenging owing to insufficient data, which renders model fitting difficult. The complexity of acquiring spectra and spectral reference values results in limited spectral data, aggravating the problem of low generalization, which diminishes model performance. To address this problem, we introduce what we believe to be a novel approach combining NIRS with Wasserstein generative adversarial networks (WGANs). Specifically, spectral data are collected from representative samples of raw material provided by a formula enterprise. Then, the WGAN augments the database by generating synthetic data resembling the raw spectral data. Finally, we establish various prediction models using the PLSR, SVR, LightGBM, and XGBoost algorithms. Experimental results show the NIRS-WGAN method significantly improves the performance of prediction models, with R2 and RMSE of 0.949 and 1.415 for the chemical components of sugar, respectively, and 0.922 and 0.243 for nicotine. The proposed framework effectively enhances the predictive capabilities of various models, addressing the issue caused by limited training data in NIRS prediction tasks.

Root transcriptomic provides insights on molecular mechanisms involved in the tolerance to water deficit in Pisum sativum inoculated with Pseudomonas sp

MAIN CONCLUSION

Root transcriptomics and biochemical analyses in water-stressed Pisum sativum plants inoculated with Pseudomonas spp. suggested preservation of ABA-related pathway and ROS detoxification, resulting in an improved tolerance to stress. Drought already affects agriculture in large areas of the globe and, due to climate change, these areas are predicted to become increasingly unsuitable for agriculture. For several years, plant growth-promoting bacteria (PGPB) have been used to improve legume yields, but many aspects of this interaction are still unclear. To elucidate the mechanisms through which root-associated PGPB can promote plant growth in dry environments, we investigated the response of pea plants inoculated with a potentially beneficial Pseudomonas strain (PK6) and subjected to two different water regimes. Combined biometric, biochemical, and root RNA-seq analyses revealed that PK6 improved pea growth specifically under water deficit, as inoculated plants showed an increased biomass, larger leaves, and longer roots. Abscisic acid (ABA) and proline quantification, together with the transcriptome analysis, suggested that PK6-inoculated plant response to water deficit was more diversified compared to non-inoculated plants, involving alternative metabolic pathways for the detoxification of reactive oxygen species (ROS) and the preservation of the ABA stress signaling pathway. We suggest that the metabolic response of PK6-inoculated plants was more effective in their adaptation to water deprivation, leading to their improved biometric traits. Besides confirming the positive role that PGPB can have in the growth of a legume crop under adverse conditions, this study offers novel information on the mechanisms regulating plant-bacteria interaction under varying water availability. These mechanisms and the involved genes could be exploited in the future for the development of legume varieties, which can profitably grow in dry climates.

Pre-processing Applied to Instrumental Data in Analytical Chemistry: A Brief Review of the Methods and Examples

The field of analytical chemistry has been significantly advanced by the availability of state-of-the-art instrumentation, allowing for the development of novel applications in this field. However, in many cases, the direct interpretation of the recorded data is often not straightforward, hence some level of pre-processing is required (e.g., baseline correction, derivatives, normalization, smoothing). These techniques have become a critical first step for the successful analysis of the data recorded, and it is recommended to use them before the application of chemometrics (e.g., classification, calibration development). The aim of this paper is to provide with an overview of the most used pre-processing methods applied to instrumental analytical methods (e.g., spectroscopy, chromatography). Examples of their application in near infrared and UV-VIS spectroscopy as well as in gas chromatography will be also discussed. Overall, this paper provides with a comprehensive understanding of pre-processing techniques in analytical chemistry, highlighting their importance during the analysis and interpretation of data, as well as during the development of accurate and reliable chemometric models.

Tetrazolium-Based Visually Indicating Bacteria Sensor for Colorimetric Detection of Point of Contamination

Protective equipment for detecting bacterial contamination has been in high demand with increasing interest in public health and hygiene. Herein, a fiber-based visually indicating bacteria sensor (VIBS) embedded with iodonitrotetrazolium chloride is developed for the general purpose of detecting live bacteria, and its chromogenic effectiveness is investigated for Gram-negative Escherichia coli and Gram-positive Micrococcus luteus. The developed color intensity is measured by the light absorption coefficient to the scattering coefficient (K/S) based on the Kubelka-Munk equation, and the colorimetric sensitivities of different membranes are examined by calculating the limit of detection (LOD) and the limit of quantification (LOQ). The results demonstrate that the interactions between VIBS and bacteria depend on the wetting properties of membranes. A hydrophobic membrane shows excessive interactions at high concentrations of Gram-negative E. coli bacteria, whose cell membrane is lipophilic. The membrane blended with hydrophobic and hydrophilic polymers displays linear colorimetric responses for both Gram-negative and Gram-positive bacteria strains, demonstrating a reliable sensing capability in the range of the tested bacteria concentration. This study is significant in that explorative experimentations are performed to conceive a proof of concept of a fiber-based bacteria sensor, which is readily applicable in various fields where bacteria pose a threat.

Application of Fluconazole-Loaded pH-Sensitive Lipid Nanoparticles for Enhanced Antifungal Therapy

Cryptococcus neoformans is a yeast-like fungus that can cause the life-threatening disease cryptococcal meningitis. Numerous reports have shown increased resistance of this fungus against antifungal treatments, such as fluconazole (Fluc), contributing to an 80% global mortality rate. This work presents a novel approach to improve the delivery of the antifungal agent Fluc and increase the drug's targetability and availability at the infection site. Exploiting the acidic environment surrounding a C. neoformans infected site, we have developed pH-sensitive lipid nanoparticles (LNP) encapsulating Fluc to inhibit the growth of resistant C. neoformans. The LNP-Fluc delivery system consists of a neutral lipid monoolein (MO) and a novel synthetic ionizable lipid 2-morpholinoethyl oleate (O2ME). At neutral pH, because of the presence of O2ME, the nanoparticles are neutral and exhibit a liquid crystalline hexagonal nanostructure (hexosomes). At an acidic pH, they are positively charged with a cubic nanostructure (cubosomes), which facilitates the interaction with the negatively charged fungal cell wall. This interaction results in the MIC50 and MIC90 values of the LNP-Fluc being significantly lower than that of the free-Fluc control. Confocal laser scanning microscopy and scanning electron microscopy further support the MIC values, showing fungal cells exposed to LNP-Fluc at acidic pH were heavily distorted, demonstrating efflux of cytoplasmic molecules. In contrast, fungal cells exposed to Fluc alone showed cell walls mostly intact. This current study represents a significant advancement in delivering targeted antifungal therapy to combat fungal antimicrobial resistance.

Detecting antimicrobial resistance in Escherichia coli using benchtop attenuated total reflectance-Fourier transform infrared spectroscopy and machine learning

The widespread dissemination of resistance to third-generation cephalosporins in the Enterobacterales through the production of extended-spectrum β-lactamase (ESBL) is considered a critical global crisis requiring urgent attention of clinicians and scientists alike. Rapid diagnostic methods that can identify microbial resistance profiles closer to the point of care are crucial to minimize the overuse of antimicrobial agents and improve patient outcomes. Although Fourier transform infrared (FTIR) microscopy has shown promise in distinguishing between bacterial species, the high cost and technical requirements of the IR microscope may limit broad clinical use. To address the practical needs of a clinical microbiology laboratory, here, we examine the ability of a lower cost portable benchtop attenuated total reflectance (ATR)-FTIR spectrometer to achieve antimicrobial resistance detection, using a simple, clinically aligned sampling protocol. The technical reproducibility was confirmed through multi-day analysis of an Escherichia coli type strain, which serves as quality control. We generated a dataset of 100 E. coli clinical bloodstream isolates with 63 ceftriaxone resistant bla_CTX-M ESBL gene variant strains and developed a classifier for bla_CTX-M genotype detection. After assessing 35 machine learning methods using the training set (n = 71), four methods were further optimised, and the best performing method was evaluated using the held-out testing set (n = 29). A tuned support vector machine model with a polynomial kernel, using the 700-1500 cm^-1 range achieved a sensitivity of 89.2%, and specificity of 66.7% for detecting bla_CTX-M in independent testing, approaching the reported performance of FTIR microscopy. With further algorithm improvement, these data suggest the potential deployment of a portable FTIR spectrometer as a rapid antimicrobial susceptibility prediction platform to enable the efficient use of antimicrobials.

Theoretical Simulation of Near-Infrared Spectrum of Piperine: Insight into Band Origins and the Features of Regression Models

We investigated the near-infrared spectrum of piperine using quantum mechanical calculations. We evaluated two efficient approaches, DVPT2//PM6 and DVPT2//ONIOM [PM6:B3LYP/6-311++G(2df, 2pd)] that yielded a simulated spectrum with varying accuracy versus computing time factor. We performed vibrational assignments and unveiled complex nature of the near-infrared spectrum of piperine, resulting from a high level of band convolution. The most meaningful contribution to the near-infrared absorption of piperine results from binary combination bands. With the available detailed near-infrared assignment of piperine, we interpreted the properties of partial least square regression models constructed in our earlier study to describe the piperine content in black pepper samples. Two models were compared with spectral data sets obtained with a benchtop and a miniaturized spectrometer. The two spectrometers implement distinct technology which leads to a profound instrumental difference and discrepancy in the predictive performance when analyzing piperine content. We concluded that the sensitivity of the two instruments to certain types of piperine vibrations is different and that the benchtop spectrometer unveiled higher selectivity. Such difference in obtaining chemical information from a sample can be one of the reasons why the benchtop spectrometer performs better in analyzing the piperine content of black pepper. This evidenced direct correspondence between the features critical for applied near-infrared spectroscopic routine and the underlying vibrational properties of the analyzed constituent in a complex sample.