Lena Kastl

Quantitative phase imaging for cell culture quality control

The potential of quantitative phase imaging (QPI) with digital holographic microscopy (DHM) for quantification of cell culture quality was explored. Label‐free QPI of detached single cells in suspension was performed by Michelson interferometer‐based self‐interference DHM. Two pancreatic tumor cell lines were chosen as cellular model and analyzed for refractive index, volume, and dry mass under varying culture conditions. Firstly, adequate cell numbers for reliable statistics were identified. Then, to characterize the performance and reproducibility of the method, we compared results from independently repeated measurements and quantified the cellular response to osmolality changes of the cell culture medium. Finally, it was demonstrated that the evaluation of QPI images allows the extraction of absolute cell parameters which are related to cell layer confluence states. In summary, the results show that QPI enables label‐free imaging cytometry, which provides novel complementary integral biophysical data sets for sophisticated quantification of cell culture quality with minimized sample preparation.

Continuous morphology and growth monitoring of different cell types in a single culture using quantitative phase microscopy

The minimally-invasive quantitative observation of different cell types in a single culture is of particular interest for the analysis of the impact of pharmaceuticals, pathogens or toxins on different cellular phenotypes under identical measurement conditions and to analyze interactions between different cellular specimens. Quantitative phase microscopy (QPM), provides high resolution detection of optical path length changes that is suitable for quantitative tomographic imaging and stain-free minimally-invasive live cell analysis. Due to low light intensities for object illumination, QPM minimizes the interaction with the sample and thus is in particular suitable for long term time-lapse investigations on cells in which for example morphology alterations due to toxins, drugs or genetic modifications are studied. We analyzed the feasibility of QPM, for the analysis of mixed cell cultures and explored if quantitative phase images provide sufficient information to distinguish between different cell types and to extract cell specific parameters. For the experiments quantitative phase imaging with digital holographic microscopy (DHM) was utilized. Mixed cell cultures with different cell types were continuously observed with quantitative DHM phase contrast up to 35 h. The obtained series of quantitative phase images were evaluated by adapted image segmentation algorithms. From the segmented quantitative phase images the area covered by the cells, the cellular dry mass as well as the mean cell thickness and volume were determined and used as parameters to quantify the reliability of data acquisition. The obtained results demonstrate that it is possible to characterize the growth of cell types with different morphology features separately in a single cell culture. This prospects new application fields of quantitative phase imaging in drug and toxicity testing in vitro.

Multimodal label-free growth and morphology characterization of different cell types in a single culture with quantitative digital holographic phase microscopy

For the analysis of the impact of pharmaceuticals or pathogens on different cellular phenotypes under identical measurement conditions and to analyze interactions between different cellular specimens a minimally-invasive quantitative observation of different cell types in a single culture is of particular interest. Digital holographic microscopy (DHM), a var-iant of quantitative phase microscopy (QPM), provides high resolution detection of optical path length changes that is suitable for stain-free minimally-invasive live cell analysis. Due to low light intensities for object illumination, QPM minimizes the interaction with the sample and has been demonstrated in particular to be suitable for long-term time-lapse investigations, e.g., for the detection of cell morphology alterations due to drugs and toxins. Furthermore, QPM has been demonstrated to be a versatile tool for the quantification of cellular growth and motility. Thus, we studied the feasibility of QPM for the analysis of mixed cell cultures and explored if quantitative phase images provide sufficient information to distinguish between different cell types and to extract cell specific parameters. For the experiments quantitative phase imaging with DHM was utilized. Mixed cell cultures with different cell types were observed with quantitative DHM phase contrast up to 35 h. The obtained series of quantitative phase images were evaluated by adapted algorithms for image segmentation. From the segmented images the area covered by the cells, the cellular dry mass and the mean cell thickness were calculated and used in the further analysis as parameters to quantify the reliability of the measurement principle. The obtained results demonstrate that it is possible to characterize the growth of cell types with different mor-phology features separately in a single culture.

Multivariate classification of fourier transform infrared hyperspectral images of skin cancer cells

A multilevel framework for the multiclass classification of spectra extracted from Fourier transform infrared images is described. This learning structure was employed to discriminate the spectra extracted from hyperspectral images of two batches of four different skin cultured cells (two normal and two tumor), where the cells of one batch had been stained with fluorescence live cell dyes. Different options were explored in each stage of the framework, specifically in the spectral pre-processing and the employed classification algorithm. Special care was taken to optimize the learning models and to objectively estimate the generalization performance by means of cross-validation. A very high discriminative performance was obtained for all the unstained skin cell types. However, the presence of the stains introduces spectral artifacts that worsen the class separation, as has been demonstrated in several classification experiments.

Standardized cell samples for midIR technology development

The application of midIR spectroscopy towards human cell and tissue samples is impaired by the need for technical solutions and lacking sample standards for technology development. We here present the standardization of stable test samples for the continuous development and testing of novel optical system components. We have selected cell lines representing the major cellular skin constituents keratinocytes and fibroblasts (NIH-3T3, HaCaT). In addition, two skin cancer cell types (A-375 and SK-MEL-28 cells) were analyzed. Cells were seeded on CaF2 substrates and measured dried and under aqueous medium as well as fixated and unfixated. Several independent cell preparations were analyzed with an FTIR spectrometer in the wave number range from 1000 - 4000 cm-1. The obtained data demonstrate that fixed and dehydrated cells on CaF2 can be stored in pure ethanol for several weeks without significant losses in quality of the spectral properties. The established protocol of cell seeding on CaF2 substrates, chemical fixation, dehydration, storage under ethanol and air-drying is suitable for the production of reliable midIR standards. The retrieved spectra demonstrate that fixed cells on CaF2 can be prepared reproducibly; with stable midIR spectral properties over several weeks and properties mimicking reliable unfixed cells. Moreover, the fixated samples on CaF2 show clear differences in the cell type specific spectra that can be identified by principle component analysis. In summary, the standardized cell culture samples on CaF2 substrates are suitable for the development of a midIR device and the optimization of the specific midIR spectra.

Performance of mid infrared spectroscopy in skin cancer cell type identification

Marker free optical spectroscopy is a powerful tool for the rapid inspection of pathologically suspicious skin lesions and the non-invasive detection of early skin tumors. This goal can be reached by the combination of signal localization and the spectroscopical detection of chemical cell signatures. We here present the development and application of mid infrared spectroscopy (midIR) for the analysis of skin tumor cell types and three dimensional tissue phantoms towards the application of midIR spectroscopy for fast and reliable skin diagnostics. We developed standardized in vitro skin systems with increasing complexity, from single skin cell types as fibroblasts, keratinocytes and melanoma cells, to mixtures of these and finally three dimensional skin cancer phantoms. The cell systems were characterized with different systems in the midIR range up to 12 μm. The analysis of the spectra by novel data processing algorithms demonstrated the clear separation of all cell types, especially melanoma cells. Special attention and algorithm training was required for closely related mesenchymal cell types as dedifferentiated melanoma cells and fibroblasts. Proof of concept experiments with mixtures of in vivo fluorescence labelled skin cell types allowed the test of the new algorithms performance for the identification of specific cell types. The intense training of the software systems with various samples resulted in a increased sensitivity and specificity of the combined midIR and software system. These data highlight the potential of midIR spectroscopy as sensitive and specific future optical biopsy technology.

Potential of mid IR spectroscopy in the rapid label free identification of skin malignancies

The rapid inspection of suspicious skin lesions for pathological cell types is the objective of optical point of care diagnostics technologies. A marker free fast diagnosis of skin malignancies would overcome the limitations of the current gold standard surgical biopsy. The time consuming and costly biopsy procedure requires the inspection of each sample by a trained pathologist, which limits the analysis of potentially malignant lesions. Optical technologies like RAMAN or infrared spectroscopy, which provide both, localization and chemical information, can be used to differentiate malignant from healthy tissue by the analysis of multi cell structures and cell type specific spectra. We here report the application of midIR spectroscopy towards fast and reliable skin diagnostics. Within the European research project MINERVA we developed standardized in vitro skin systems with increasing complexity, from single skin cell types as fibroblasts, keratinocytes and melanoma cells, to mixtures of these and finally three dimensional human skin equivalents. The standards were characterized in the established midIR range and also with newly developed systems for fast imaging up to 12 μm. The analysis of the spectra by novel data processing algorithms demonstrated the clear separation of all cell types, especially the tumor cells. The signals from single cell layers were sufficient for cell type differentiation. We have compared different midIR systems and found all of them suitable for specific cell type identification. Our data demonstrate the potential of midIR spectroscopy for fast image acquisition and an improved data processing as sensitive and specific optical biopsy technology.

Multimodal optical phenotyping of cancer cells

There is a growing interest in label-free, optical techniques like digital holographic microscopy (DHM) and optical cell stretching, since the interaction with samples is minimized. Because optical manipulation strongly depends on the optical and physiological properties of the investigated material, we combined the usage of these methods for the characterization of pancreatic tumor cells. Our results demonstrate that cells of distinct differentiation levels, or different expression in only one protein, show differences in their deformability. Additionally, the DHM results showed only few variations in the refractive index, indicating that it does not significantly influence the results of the optical cell stretching. Thus, the combined usage of the two technologies represents a promising new approach for tumor cell characterization.

Optomechanical properties of cancer cells revealed by light-induced deformation and quantitative phase microscopy

There is a growing interest in cell biology and clinical diagnostics in label-free, optical techniques as the interaction with the sample is minimized and substances like dyes or fixatives do not affect the investigated cells. Such techniques include digital holographic microscopy (DHM) and the optical stretching by fiber optical two beam traps. DHM enables quantitative phase contrast imaging and thereby the determination of the cellular refractive index, dry mass and the volume, whereas optical cell stretching reveals the deformability of cells. Since optical stretching strongly depends on the optical properties and the shape of the investigated material we combined the usage of fiber optical stretching and DHM for the characterization of pancreatic tumor cells. The risk of tumors is their potential to metastasize, spread through the bloodstream and build distal tumors/metastases. The grade of dedifferentiation in which the cells lose their cell type specific properties is a measure for this metastatic potential. The less differentiated the cells are, the higher is their risk to metastasize. Our results demonstrate that pancreatic tumor cells, which are from the same tumor but vary in their grade of differentiation, show significant differences in their deformability. The retrieved data show that differentiated cells have a higher stiffness than less differentiated cells of the same tumor. Even cells that differ only in the expression of a single tumor suppressor gene which is responsible for cell-cell adhesions can be distinguished by their mechanical properties. Additionally, results from DHM measurements yield that the refractive index shows only few variations, indicating that it does not significantly influence optical cell stretching. The obtained results show a promising new approach for the phenotyping of different cell types, especially in tumor cell characterization and cancer diagnostics.

Hyperspectral digital holographic microscopy approach for reduction of coherence induced disturbances in quantitative phase imaging of biological specimens

Coherence induced noise and parasitic reflections in the experimental setup are main restrictions that limit the resolution and measurement accuracy in laser light-based digital holographic microscopy (DHM). We explored, if coherence properties of partial coherent light sources can be mimicked by utilizing spectrally tunable lasers. Moreover, the performance for label-free quantitative phase imaging of biological specimens is illustrated utilizing an experimental configuration including a commercial microscope and tunable super continuum laser sources with a wavelength range of up to 230 nm.