Measurement of bacterial concentration fractions in polymicrobial mixtures by Raman microspectroscopy

Spectral encoder to extract the efficient features of Raman spectra for reliable and precise quantitative analysis

Raman spectroscopy has become a powerful analytical tool highly demanded in many applications such as microorganism sample analysis, food quality control, environmental science, and pharmaceutical analysis, owing to its non-invasiveness, simplicity, rapidity and ease of use. Among them, quantitative research using Raman spectroscopy is a crucial application field of spectral analysis. However, the entire process of quantitative modeling largely relies on the extraction of effective spectral features, particularly for measurements on complex samples or in environments with poor spectral signal quality. In this paper, we propose a method of utilizing a spectral encoder to extract effective spectral features, which can significantly enhance the reliability and precision of quantitative analysis. We built a latent encoded feature regression model; in the process of utilizing the autoencoder for reconstructing the spectrometer output, the latent feature obtained from the intermediate bottleneck layer is extracted. Then, these latent features are fed into a deep regression model for component concentration prediction. Through detailed ablation and comparative experiments, our proposed model demonstrates superior performance to common methods on single-component and multi-component mixture datasets, remarkably improving regression precision while without needing user-selected parameters and eliminating the interference of irrelevant and redundant information. Furthermore, in-depth analysis reveals that latent encoded feature possesses strong nonlinear feature representation capabilities, low computational costs, wide adaptability, and robustness against noise interference. This highlights its effectiveness in spectral regression tasks and indicates its potential in other application fields. Sufficient experimental results show that our proposed method provides a novel and effective feature extraction approach for spectral analysis, which is simple, suitable for various methods, and can meet the measurement needs of different real-world scenarios.

Fast label-free identification of bacteria by synchronous fluorescence of amino acids

Fast identification of pathogenic bacteria is an essential need for patient's diagnostic in hospitals and environmental monitoring of water and air quality. Bacterial cells consist of a very high amount of biological molecules whose content changes in response to different environmental conditions. The similarity between the molecular compositions of different bacterial cells limits the possibility to find unique markers to enable differentiation among species. Although many biological molecules in the cells absorb at the UV-Vis region, only a few of them can be detected in whole cells by their intrinsic fluorescence. Among these molecules are the amino acids phenylalanine, tyrosine, and tryptophan. In this work, we develop a rapid method for bacterial identification by synchronous fluorescence. We show that we can quantify the concentration for the 3 amino acids without any significant interference from other fluorophores in the cells and that we can differentiate among 6 pathogenic bacterial species by using the concentrations of their amino acids as a bacterial fingerprint. Fluorescent amino acids exist in all living cells. Therefore, this method has the potential to be applicative for the rapid identification of cells from all kinds of organisms.

Confocal Raman microscopy to identify bacteria in oral subgingival biofilm models

The study of oral disease progression, in relation to the accumulation of subgingival biofilm in gingivitis and periodontitis is limited, due to either the ability to monitor plaque in vitro. When compared, optical spectroscopic techniques offer advantages over traditional destructive or biofilm staining approaches, making it a suitable alternative for the analysis and continued development of three-dimensional structures. In this work, we have developed a confocal Raman spectroscopy analysis approach towards in vitro subgingival plaque models. The main objective of this study was to develop a method for differentiating multiple oral subgingival bacterial species in planktonic and biofilm conditions, using confocal Raman microscopy. Five common subgingival bacteria (Fusobacterium nucleatum, Streptococcus mutans, Veillonella dispar, Actinomyces naeslundii and Prevotella nigrescens) were used and differentiated using a 2-way orthogonal Partial Least Square with Discriminant Analysis (O2PLS-DA) for the collected spectral data. In addition to planktonic growth, mono-species biofilms cultured using the 'Zürich Model' were also analyzed. The developed method was successfully used to predict planktonic and mono-species biofilm species in a cross validation setup. The results show differences in the presence and absence of chemical bands within the Raman spectra. The O2PLS-DA model was able to successfully predict 100% of all tested planktonic samples and 90% of all mono-species biofilm samples. Using this approach we have shown that Confocal Raman microscopy can analyse and predict the identity of planktonic and mono-species biofilm species, thus enabling its potential as a technique to map oral multi-species biofilm models.

Rapid infectious disease identification by next-generation DNA sequencing

Currently, there is a critical need to rapidly identify infectious organisms in clinical samples. Next-Generation Sequencing (NGS) could surmount the deficiencies of culture-based methods; however, there are no standardized, automated programs to process NGS data. To address this deficiency, we developed the Rapid Infectious Disease Identification (RIDI™) system. The system requires minimal guidance, which reduces operator errors. The system is compatible with the three major NGS platforms. It automatically interfaces with the sequencing system, detects their data format, configures the analysis type, applies appropriate quality control, and analyzes the results. Sequence information is characterized using both the NCBI database and RIDI™ specific databases. RIDI™ was designed to identify high probability sequence matches and more divergent matches that could represent different or novel species. We challenged the system using defined American Type Culture Collection (ATCC) reference standards of 27 species, both individually and in varying combinations. The system was able to rapidly detect known organisms in <12h with multi-sample throughput. The system accurately identifies 99.5% of the DNA sequence reads at the genus-level and 75.3% at the species-level in reference standards. It has a limit of detection of 146cells/ml in simulated clinical samples, and is also able to identify the components of polymicrobial samples with 16.9% discrepancy at the genus-level and 31.2% at the species-level. Thus, the system's effectiveness may exceed current methods, especially in situations where culture methods could produce false negatives or where rapid results would influence patient outcomes.

Rejection of fluorescence background in resonance and spontaneous Raman microspectroscopy

Raman spectroscopy is often plagued by a strong fluorescent background, particularly for biological samples. If a sample is excited with a train of ultrafast pulses, a system that can temporally separate spectrally overlapping signals on a picosecond timescale can isolate promptly arriving Raman scattered light from late-arriving fluorescence light. Here we discuss the construction and operation of a complex nonlinear optical system that uses all-optical switching in the form of a low-power optical Kerr gate to isolate Raman and fluorescence signals. A single 808 nm laser with 2.4 W of average power and 80 MHz repetition rate is split, with approximately 200 mW of 808 nm light being converted to < 5 mW of 404 nm light sent to the sample to excite Raman scattering. The remaining unconverted 808 nm light is then sent to a nonlinear medium where it acts as the pump for the all-optical shutter. The shutter opens and closes in 800 fs with a peak efficiency of approximately 5%. Using this system we are able to successfully separate Raman and fluorescence signals at an 80 MHz repetition rate using pulse energies and average powers that remain biologically safe. Because the system has no spare capacity in terms of optical power, we detail several design and alignment considerations that aid in maximizing the throughput of the system. We also discuss our protocol for obtaining the spatial and temporal overlap of the signal and pump beams within the Kerr medium, as well as a detailed protocol for spectral acquisition. Finally, we report a few representative results of Raman spectra obtained in the presence of strong fluorescence using our time-gating system.

Development of a method based on chemometric analysis of Raman spectra for the discrimination of heterofermentative lactobacilli

In this work, a method based on Raman spectroscopy in combination with Principal Component Analysis (PCA) and Partial Least Square-Discriminant Analysis (PLS-DA) has been developed for the rapid differentiation of heterofermentative related lactobacilli. In a first approach, Lactobacillus kefir strains were discriminated from other species of heterofermentative lactobacilli: Lb. parakefir and Lb. brevis. After this first approach, PCA allowed for a clear differentiation between Lb. parakefir and Lb.brevis. For the first level of discrimination, PCA was performed on the whole spectra and also on delimited regions, defined taking into consideration the loading values. The best regions allowing a clear differentiation between Lb. kefir and non-Lb. kefir strains were found to be: the 1700-1500 cm(-1), 1500-1185 cm(-1) and 1800-400 (whole spectrum) cm(-1) Raman ranges. In order to develop a classification rule, PLS-DA was carried out on the mentioned regions. This method permitted the discrimination and classification of the strains under study in two groups: Lb. kefir and non-Lb. kefir. The model was further validated using lactobacilli strains from different culture collections or strains isolated from kefir grains previously identified using molecular methods. The second approach based on PCA was also performed on the whole spectra and on delimited regions, being the regions 1700-1500 cm(-1), 1500-1185 cm(-1) and 1185-1020 cm(-1), i.e., those allowing the clearest discrimination between Lb. parakefir and Lb. brevis. The results obtained in this work, allowed a clear discrimination within heterofermentative lactobacilli strains, proteins being the biological structures most determinant for this discrimination.

Identification of different bacterial species in biofilms using confocal Raman microscopy

Confocal Raman microspectroscopy is used to discriminate between different species of bacteria grown in biofilms. Tests are performed using two bacterial species, Streptococcus sanguinis and Streptococcus mutans, which are major components of oral plaque and of particular interest due to their association with healthy and cariogenic plaque, respectively. Dehydrated biofilms of these species are studied as a simplified model of dental plaque. A prediction model based on principal component analysis and logistic regression is calibrated using pure biofilms of each species and validated on pure biofilms grown months later, achieving 96% accuracy in prospective classification. When biofilms of the two species are partially mixed together, Raman-based identifications are achieved within ∼2 μm of the boundaries between species with 97% accuracy. This combination of spatial resolution and predication accuracy should be suitable for forming images of species distributions within intact two-species biofilms.

Development of a time-gated system for Raman spectroscopy of biological samples

A time gating system has been constructed that is capable of recording high quality Raman spectra of highly fluorescing biological samples while operating below the photodamage threshold. Using a collinear gating geometry and careful attention to power conservation, we have achieved all-optical switching with a one picosecond gating time and 5% peak gating efficiency. The energy per pulse in this instrument is more than 3 orders of magnitude weaker than previous reports. Using this system we have performed proof-of-concept experiments on a sample composed of perylene dissolved in toluene, and the stem of a Jasminum multiflorum plant, the latter case being particularly important for the study of plants used in production of cellulosic biofuels. In both cases, a high SNR spectrum of the high-wavenumber region of the spectrum was recorded in the presence of an overwhelming fluorescence background.

Construction of an integrated Raman- and angular-scattering microscope

We report on the construction of a multimodal microscope platform capable of gathering both elastically and inelastically scattered light from a 38 mum(2) region in both epi- and transillumination geometries. Simultaneous monitoring of elastic and inelastic scattering from a microscopic region allows noninvasive characterization of the chemistry and morphology of a living sample without the need for exogenous dyes or labels, thus allowing measurements to be made longitudinally in time on the same sample as it evolves naturally. A sample is illuminated either from above or below with a focused 785 nm TEM(00) mode laser beam, with elastic and inelastic scattering collected by two separate measurement arms. The measurements may be made either simultaneously, if identical illumination geometries are used, or sequentially, if the two modalities utilize opposing illumination paths. In the inelastic arm, Stokes-shifted light is dispersed by a spectrograph onto a charge-coupled device (CCD) array. In the elastic scattering collection arm, a relay system images the microscope's back aperture onto a CCD array. Postprocessing of the inelastic scattering to remove fluorescence signals yields high quality Raman spectra that report on the sample's chemical makeup. Comparison of the elastically scattered pupil images to generalized Lorenz-Mie theory yields estimated size distributions of scatterers within the sample.

Cell (A549)-particle (Jasada Bhasma) interactions using Raman spectroscopy

Current methods for the evaluation of cell interactions with particles are nonspecific, slow, and invasive to the cells. Raman spectroscopy is a noninvasive technique, and is used in the present study to investigate particle-cell interactions. The main focus of the present study is to employ Raman spectroscopy for investigating the interaction of human lung adenocarcinoma cell line (A549) with the particulate system Jasada Bhasma, a traditional Indian medicine. Jasada Bhasma is a unique preparation of zinc and is traditionally used for the treatment of various diseases like diabetes, age-related eye diseases, and as a health promotional tonic. The Raman spectral analysis is executed by identifying the difference in intracellular DNA/RNA, and proteins and lipids concentration between particles--treated and untreated cells. Comparison between Bhasma-treated and -untreated cells indicates that vibrational peaks corresponding to the DNA/RNA molecule show a significant increase in cells treated with the Jasada Bhasma. Apart from the DNA molecule, several other vibrational peaks related to the protein molecules also show a significant increase in A549 cells after treatment with Bhasma. These results indicate that Bhasma treatment of A549 possibly delays DNA degradation and enables retention of higher amount of protein molecules in the cells.

Raman spectroscopic measurement of relative concentrations in mixtures of oral bacteria

Near-infrared Raman spectroscopy has been used for species identification of pure microbial specimens for more than a decade. More recently, this optical method has been extended to the analysis of specimens containing multiple species. In this report, we demonstrate rapid, reagent-free quantitative analysis of a simplified model of oral plaque containing three oral bacteria species, S. mutans, S. sanguis, and S. gordonii, using near-infrared Raman spectroscopy. Raman spectra were acquired from bacterial mixtures in 200 seconds. A prediction model was calibrated by the partial least squares method and validated by additional samples. On a scale from 0 to 1, relative fractions of each species could be predicted with a root mean square error of 0.07. These results suggest that near-infrared Raman spectroscopy is potentially useful in quantification of microbial mixtures in general and oral plaques in particular.

Surface-sensitive polarized Raman spectroscopy of biological tissue

In a two-layer diffusing medium, polarized light directly backscattering off the superficial layer will partially retain its sense of polarization, whereas deeper-probing light will be increasingly depolarized by diffusion. This effect has been studied in both elastic scattering and fluorescence contexts. We apply this method to Raman scattering in two two-layer models with a highly diffusing lower layer of glucose powder and an upper layer of either clear plastic or chicken skin. We employ detection of orthogonal polarization states to generate a Raman spectrum of only the superficial layer by combining the orthogonal signals.