Franck Bonnier

Resonant Mie scattering in infrared spectroscopy of biological materials - Understanding the 'dispersion artefact'

Infrared spectroscopic cytology is potentially a powerful clinical tool. However, in order for it to be successful, practitioners must be able to extract reliably a pure absorption spectrum from a measured spectrum that often contains many confounding factors. The most intractable problem to date is the, so called, dispersion artefact which most prominently manifests itself as a sharp decrease in absorbance on the high wavenumber side of the amide I band in the measured spectrum, exhibiting a derivative-like line shape. In this paper we use synchrotron radiation FTIR micro-spectroscopy to record spectra of mono-dispersed poly(methyl methacrylate) (PMMA) spheres of systematically varying size and demonstrate that the spectral distortions in the data can be understood in terms of resonant Mie scattering. A full understanding of this effect will enable us to develop strategies for deconvolving the scattering contribution and recovering the pure absorption spectrum, thus removing one of the last technological barriers to the development of clinical spectroscopic cytology.

Reflection contributions to the dispersion artefact in FTIR spectra of single biological cells

Fourier transform infrared spectra of a single cell in transflection geometry are seen to vary significantly with position on the cell, showing a distorted derivative-like lineshape in the region of the optically dense nucleus. A similar behaviour is observable in a model system of the protein albumin doped in a potassium bromide disk. It is demonstrated that the spectrum at any point is a weighted sum of the sample reflection and transmission and that the dominance of the reflection spectrum in optically dense regions can account for some of the spectral distortions previously attributed to dispersion artefacts. Rather than being an artefact, the reflection contribution is ever present in transflection spectra and it is further demonstrated that the reflection characteristics can be used for cellular mapping.

Surface enhanced Raman scattering with gold nanoparticles: effect of particle shape

The dependence of the Surface Enhanced Raman Scattering (SERS) by gold nanoparticles on their shape is examined using the organic dye, rhodamine 6G (R6G) as probe molecule. SERS has been explored extensively for applications in sensing and imaging, but the design and optimisation of efficient substrates is still challenging. In order to understand and optimise the SERS process in nanoparticles, gold nanospheres and their aggregates, nanotriangles, and nanostars of similar dimensions were synthesised and characterised according to their average size, zeta potential and UV/visible absorption. SERS from R6G was negligible for unaggregated nanospheres at 532 nm, close to the maximum of the surface plasmon resonance (SPR) at 560 nm. Upon aggregation of the nanospheres, the SPR shifts to ∼660 nm, attributable to local surface plasmon “hotspots” between the spheres, and the SERS signal of R6G is significantly increased, at 785 nm. In monodisperse gold nanotriangles, the SPR is located at ∼800 nm, and significant SERS of R6G is observed using 785 nm as source, as is the case for gold nanostars, which exhibit a double SPR with maxima at ∼600 nm and ∼785 nm, attributable to the core sphere and vertices of the structures, respectively. In suspensions of equal nanoparticle and dye concentration, the SERS effect increases as nanospheres < nanosphere aggregates < nanotriangles < nanostars, clearly indicating that control over the number of local field hotspots can optimise the SERS efficiency. Notably, it is demonstrated that the SERS intensity per nanoparticle scales with the magnitude of the SPR absorbance at the excitation wavelength (785 nm), providing a clear guide to optimisation of the process experimentally.

Cell viability assessment using the Alamar blue assay: a comparison of 2D and 3D cell culture models.

Comparisons of 2D and 3D cell culture models in literature have indicated differences in cellular morphology and metabolism, commonly attributed the better representation of in vivo conditions of the latter cell culture environment. Thus, interest in the use of 3D collagen gels for in vitro analysis has been growing. Although comparative studies to date have indicated an enhanced resistance of cells on collagen matrices against different toxicants, in the present study it is demonstrated that non-adapted protocols can lead to misinterpretation of results obtained from classical colorimetric dye-based cytotoxic assays. Using the well established Alamar blue assay, the study demonstrates how the transfer from 2D substrates to 3D collagen matrices can affect the uptake of the resazurin itself, affecting the outcome of the assay. Using flow cytometry, it is demonstrated that the cell viability is unaffected when cells are grown on collagen matrices, thus the difference seen in the fluorescence is a result of a dilution of the resazurin dye in the collagen matrix, and an increased uptake rate due to the larger cell surface exposed to the surrounding environment, facilitating more effective diffusion through the cellular membrane. The results are supported by a rate equation based simulation, verifying that differing uptake kinetics can result in apparently different cell viability. Finally, this work highlights the feasibility to apply classical dye-based assays on collagen based 3D cell culture models. However, the diffusion and bioavailability of test substances in 3D matrices used in in vitro toxicological assays must be considered and adaption of the protocols is necessary for direct comparison with the traditional 2D models. Moreover, the observations made based on the resazurin dye can be applied to drugs or nanoparticles which freely diffuse through the collagen matrices, thus affecting the effective concentration exposed to the cells.

Improved Protocols for Vibrational Spectroscopic Analysis of Body Fluids

The applications of vibrational spectroscopy to the examination of human blood serum are explored. Although FTIR spectra can be recorded in aqueous solutions at (gelatin) concentrations as low as 100 mg/L, the high-wavenumber region remains obscured by water absorption. Using Raman spectroscopy, high quality spectra of gelatine solutions as low as 10 mg/L can be achieved, also covering the high-wavenumber regions. In human serum, spectral profiles are weak and partially obscured by water features. Dried deposits are shown to be physically and chemically inhomogeneous resulting in reduced measurement reproducibility. Concentration of the serum using commercially available centrifugal filter devices results in an improvement in the spectral intensity and quality. Additionally, in Raman spectroscopy, reduced background and significantly enhanced signal collection is achievable by measurement in an inverted geometry. The improved protocols for spectroscopic measurement of human serum are applicable to a range of bodily fluids and should accelerate potential clinical applications.

Comparison of Subcellular Responses for the Evaluation and Prediction of the Chemotherapeutic Response to Cisplatin in Lung Adenocarcinoma using Raman Spectroscopy

Confocal Raman Micro-spectroscopy (CRM) is employed to examine the chemical and physiological effects of anticancer agents, using cisplatin and A549 adenocarcinoma cells as a model compound and test system respectively. Spectral responses of the membrane and cytoplasm of the cell are analysed independently and the results are compared to previously reported spectroscopic studies of the nucleus. Moreover, Raman spectra from the proteins extracted from the control and exposed samples are acquired and analysed to confirm the origin of the molecular changes of the cell membrane and cytoplasm of the A549 cells. Multivariate data analysis techniques including Principal Component Analysis (PCA) and Partial Least Squares Regression (PLSR) along with PLS-Jackknifing are used to analyse the data measured from the cell membrane and cytoplasm of the A549 cells and results are correlated with parallel measurements from the cytotoxicity assay MTT. A PLSR model is used to differentiate between the chemical effect of the chemotherapeutic agent and the physiological response of the A549 cells and to identify regions of the spectrum that are associated with these processes respectively. The PLSR model is also employed to predict, on the basis of the Raman spectra, the effective dose as well as the level of physiological response, using spectra data from the cytoplasmic and cell membrane regions. The effectiveness of the models based on spectral datasets from the cell membrane and cytoplasm is compared to similar models constructed using spectral data from the nuclear region as well as one combining spectral data from all regions. In all cases, higher prediction accuracy is found for regression against the cisplatin dose, and for both regression against the dose and the physiological response, nuclear data yield higher precision.

Raman spectroscopic analysis of human skin tissue sections ex-vivo: evaluation of the effects of tissue processing and dewaxing

Abstract. Raman spectroscopy coupled with K-means clustering analysis (KMCA) is employed to elucidate the biochemical structure of human skin tissue sections and the effects of tissue processing. Both hand and thigh sections of human cadavers were analyzed in their unprocessed and formalin-fixed, paraffin-processed (FFPP), and subsequently dewaxed forms. In unprocessed sections, KMCA reveals clear differentiation of the stratum corneum (SC), intermediate underlying epithelium, and dermal layers for sections from both anatomical sites. The SC is seen to be relatively rich in lipidic content; the spectrum of the subjacent layers is strongly influenced by the presence of melanin, while that of the dermis is dominated by the characteristics of collagen. For a given anatomical site, little difference in layer structure and biochemistry is observed between samples from different cadavers. However, the hand and thigh sections are consistently differentiated for all cadavers, largely based on lipidic profiles. In dewaxed FFPP samples, while the SC, intermediate, and dermal layers are clearly differentiated by KMCA of Raman maps of tissue sections, the lipidic contributions to the spectra are significantly reduced, with the result that respective skin layers from different anatomical sites become indistinguishable. While efficient at removing the fixing wax, the tissue processing also efficiently removes the structurally similar lipidic components of the skin layers. In studies of dermatological processes in which lipids play an important role, such as wound healing, dewaxed samples are therefore not appropriate. Removal of the lipids does however accentuate the spectral features of the cellular and protein components, which may be more appropriate for retrospective analysis of disease progression and biochemical analysis using tissue banks.

Spectral Pre and Post Processing for Infrared and Raman Spectroscopy of Biological Tissues and Cells

Vibrational spectroscopy, both infrared absorption and Raman spectroscopy, have attracted increasing attention for biomedical applications, from in vivo and ex vivo disease diagnostics and screening, to in vitro screening of therapeutics. There remain, however, many challenges related to the accuracy of analysis of physically and chemically inhomogeneous samples, across heterogeneous sample sets. Data preprocessing is required to deal with variations in instrumental responses and intrinsic spectral backgrounds and distortions in order to extract reliable spectral data. Data postprocessing is required to extract the most reliable information from the sample sets, based on often very subtle changes in spectra associated with the targeted pathology or biochemical process. This review presents the current understanding of the factors influencing the quality of spectra recorded and the pre-processing steps commonly employed to improve on spectral quality. It further explores some of the most common techniques which have emerged for classification and analysis of the spectral data for biomedical applications. The importance of sample presentation and measurement conditions to yield the highest quality spectra in the first place is emphasised, as is the potential of model simulated datasets to validate both pre- and post-processing protocols.

A comparison of Raman, FTIR and ATR-FTIR micro spectroscopy for imaging human skin tissue sections

Raman and infrared absorption spectroscopies are compared for the analysis of human hand skin tissue sections. The tissue sections have been formalin fixed and paraffin processed, and subsequently dewaxed. Fourier Transform Infrared (FTIR) spectra are preprocessed using the resonant Mie-extended multiplicative scattering algorithm to remove spectral artefacts. FTIR images of resolution 4 cm−1, analysed using K-means cluster analysis, reveal the double layer structure of the dermis and epidermis, but no further layer differentiation is achieved using the higher spatial resolution of the Attenuated Total Reflection imaging or improved spectral resolution of 2 cm−1. At comparable spectral and spatial resolutions and measurement on the same samples, Raman scattering produces spectra of significantly higher spectral detail and can differentiate the stratum corneum from the underlying epithelial layer, and, in the absence of melanin in an artificial skin model, can further differentiate the basal layer from the overlying epithelium. The differences in the performance of the techniques are therefore not instrumentational and are discussed in terms of the technological and fundamental differences between the two complementary techniques.

Raman micro-spectroscopy for rapid screening of oral squamous cell carcinoma

Raman spectroscopy can provide a molecular-level fingerprint of the biochemical composition and structure of cells with excellent spatial resolution and could be useful to monitor changes in composition for dysplasia and early, non-invasive cancer diagnosis (carcinoma in situ), both ex-vivo and in vivo. In this study, we demonstrate this potential by collecting Raman spectra of the nucleoli, nuclei and cytoplasm from oral epithelial cancer (SCC-4) and dysplastic (pre-cancerous, DOK) cell lines and from normal oral epithelial primary cell cultures, in vitro, which were then analysed by principal component analysis (PCA) as a multivariate statistical method to discriminate the spectra. Results show significant discrimination between cancer and normal cell lines. Furthermore, the dysplastic and cancer cell lines could be discriminated based on the spectral profiles of the cytoplasmic regions. The principal component loading plot, which elucidates the biochemical features responsible for the discrimination, showed significant contributions of nucleic acid and proteins for nucleolar and nuclear sites and variation in features of lipids for the cytoplasmic area. This technique may provide a rapid screening method and have potential use in the diagnosis of dysplasia and early, non-invasive oral cancer, the treatment of which involves much less extensive and complex surgery and a reduction in associated co-morbidity for the patient.