Signal, noise and resolution in linear and nonlinear structured-illumination microscopy.

Structured-illumination microscopy allows widefield fluorescence imaging with resolution beyond the classical diffraction limit. Its linear form extends resolution by a factor of two, and its nonlinear form by an in-principle infinite factor, the effective resolution in practice being determined by noise. In this paper, we analyse the noise properties and achievable resolution of linear and nonlinear 1D and 2D patterned SIM from a frequency-space perspective. We develop an analytical theory for a general case of linear or nonlinear fluorescent imaging, and verify the analytical calculations with numerical simulation for a special case where nonlinearity is produced by photoswitching of fluorescent labels. We compare the performance of two alternative implementations, using either two-dimensional (2D) illumination patterns or sequentially rotated one-dimensional (ID) patterns. We show that 1D patterns are advantageous in the linear case, and that in the nonlinear case 2D patterns provide a slight signal-to-noise advantage under idealised conditions, but perform worse than 1D patterns in the presence of nonswitchable fluorescent background. LAY DESCRIPTION: Structured-illumination microscopy (SIM) is a high-resolution light microscopy technique that allows imaging of fluorescence at a resolution about twice the classical diffraction limit. There are various ways that the illumination can be structured, but it is not obvious how the choice of illumination pattern affects the final image quality, especially in view of the noise. We present a detailed performance analysis considering two illumination techniques: sequential illumination with line-gratings that are shifted and rotated during image acquisition and two-dimensional (2D) illumination structures requiring only shift operations. Our analysis is based on analytical theory, supported by simulations of images considering noise. We also extend our analysis to a nonlinear variant of SIM, with which enhanced resolution can be achieved, limited only by noise. This includes nonlinear SIM based on the light-induced switching of the fluorescent molecules between a bright and a dark state. We find sequential illumination with line-gratings to be advantageous in ordinary (linear) SIM, whereas 2D patterns provides a slight signal-to-noise advantage under idealised conditions in nonlinear SIM if there is no nonswitching background.

Combination screening in vitro identifies synergistically acting KP372-1 and cytarabine against acute myeloid leukemia.

Cytogenetic lesions often alter kinase signaling in acute myeloid leukemia (AML) and the addition of kinase inhibitors to the treatment arsenal is of interest. We have screened a kinase inhibitor library and performed combination testing to find promising drug-combinations for synergistic killing of AML cells. Cytotoxicity of 160 compounds in the library InhibitorSelect™ 384-Well Protein Kinase Inhibitor I was measured using the fluorometric microculture cytotoxicity assay (FMCA) in three AML cell lines. The 15 most potent substances were evaluated for dose-response. The 6 most cytotoxic compounds underwent combination synergy analysis based on the FMCA readouts after either simultaneous or sequential drug addition in AML cell lines. The 4 combinations showing the highest level of synergy were evaluated in 5 primary AML samples. Synergistic calculations were performed using the combination interaction analysis package COMBIA, written in R, using the Bliss independence model. Based on obtained results, an iterative combination search was performed using the therapeutic algorithmic combinatorial screen (TACS) algorithm. Of 160 substances, cell survival was ⩽50% at <0.5µM for Cdk/Crk inhibitor, KP372-1, synthetic fascaplysin, herbimycin A, PDGF receptor tyrosine kinase inhibitor IV and reference-drug cytarabine. KP372-1, synthetic fascaplysin or herbimycin A obtained synergy when combined with cytarabine in AML cell lines MV4-11 and HL-60. KP372-1 added 24h before cytarabine gave similar results in patient cells. The iterative search gave further improved synergy between cytarabine and KP372-1. In conclusion, our in vitro studies suggest that combining KP372-1 and cytarabine is a potent and synergistic drug combination in AML.

In vitro discovery of promising anti-cancer drug combinations using iterative maximisation of a therapeutic index.

In vitro-based search for promising anti-cancer drug combinations may provide important leads to improved cancer therapies. Currently there are no integrated computational-experimental methods specifically designed to search for combinations, maximizing a predefined therapeutic index (TI) defined in terms of appropriate model systems. Here, such a pipeline is presented allowing the search for optimal combinations among an arbitrary number of drugs while also taking experimental variability into account. The TI optimized is the cytotoxicity difference (in vitro) between a target model and an adverse side effect model. Focusing on colorectal carcinoma (CRC), the pipeline provided several combinations that are effective in six different CRC models with limited cytotoxicity in normal cell models. Herein we describe the identification of the combination (Trichostatin A, Afungin, 17-AAG) and present results from subsequent characterisations, including efficacy in primary cultures of tumour cells from CRC patients. We hypothesize that its effect derives from potentiation of the proteotoxic action of 17-AAG by Trichostatin A and Afungin. The discovered drug combinations against CRC are significant findings themselves and also indicate that the proposed strategy has great potential for suggesting drug combination treatments suitable for other cancer types as well as for other complex diseases.

Interferometer-based structured-illumination microscopy utilizing complementary phase relationship through constructive and destructive image detection by two cameras.

In an interferometer-based fluorescence microscope, a beam splitter is often used to combine two emission wavefronts interferometrically. There are two perpendicular paths along which the interference fringes can propagate and normally only one is used for imaging. However, the other path also contains useful information. Here we introduced a second camera to our interferometer-based three-dimensional structured-illumination microscope (I(5)S) to capture the fringes along the normally unused path, which are out of phase by π relative to the fringes along the other path. Based on this complementary phase relationship and the well-defined phase interrelationships among the I(5)S data components, we can deduce and then computationally eliminate the path length errors within the interferometer loop using the simultaneously recorded fringes along the two imaging paths. This self-correction capability can greatly relax the requirement for eliminating the path length differences before and maintaining that status during each imaging session, which are practically challenging tasks. Experimental data is shown to support the theory.

Digital gene expression profiling of primary acute lymphoblastic leukemia cells.

We determined the genome-wide digital gene expression (DGE) profiles of primary acute lymphoblastic leukemia (ALL) cells from 21 patients taking advantage of 'second-generation' sequencing technology. Patients included in this study represent four cytogenetically distinct subtypes of B-cell precursor (BCP) ALL and T-cell lineage ALL (T-ALL). The robustness of DGE combined with supervised classification by nearest shrunken centroids (NSC) was validated experimentally and by comparison with published expression data for large sets of ALL samples. Genes that were differentially expressed between BCP ALL subtypes were enriched to distinct signaling pathways with dic(9;20) enriched to TP53 signaling, t(9;22) to interferon signaling, as well as high hyperdiploidy and t(12;21) to apoptosis signaling. We also observed antisense tags expressed from the non-coding strand of ~50% of annotated genes, many of which were expressed in a subtype-specific pattern. Antisense tags from 17 gene regions unambiguously discriminated between the BCP ALL and T-ALL subtypes, and antisense tags from 76 gene regions discriminated between the 4 BCP subtypes. We observed a significant overlap of gene regions with alternative polyadenylation and antisense transcription (P<1 × 10(-15)). Our study using DGE profiling provided new insights into the RNA expression patterns in ALL cells.

Interrogating health-related public databases from a food toxicology perspective: computational analysis of scoring data.

Over the last 15 years, an expanding number of databases with information on noxious effects of substances on mammalian organisms and the environment have been made available on the Internet. This set of databases is a key source of information for risk assessment within several areas of toxicology. Here we present features and relationships across a relatively wide set of publicly accessible databases broadly within toxicology, in part by clustering multi-score representations of such repositories, to support risk assessment within food toxicology. For this purpose 36 databases were each scrutinized, using 18 test substances from six different categories as probes. Results have been analyzed by means of various uni- and multi-variate statistical operations. The former included a special index devised to afford context-specific rating of databases across a highly heterogeneous data matrix, whereas the latter involved cluster analysis, enabling the identification of database assemblies with overall shared characteristics. One database - HSDB - was outstanding due to rich and qualified information for most test substances, but an appreciable fraction of the interrogated repositories showed good to decent scoring. Among the six chosen substance groups, Food contact materials had the most comprehensive toxicological information, followed by the Pesticides category.

In vitro drug sensitivity-gene expression correlations involve a tissue of origin dependency.

A major concern of chemogenomics is to associate drug activity with biological variables. Several reports have clustered cell line drug activity profiles as well as drug activity-gene expression correlation profiles and noted that the resulting groupings differ but still reflect mechanism of action. The present paper shows that these discrepancies can be viewed as a weighting of drug-drug distances, the weights depending on which cell lines the two drugs differ in.

Computational detection of allergenic proteins attains a new level of accuracy with in silico variable-length peptide extraction and machine learning.

The placing of novel or new-in-the-context proteins on the market, appearing in genetically modified foods, certain bio-pharmaceuticals and some household products leads to human exposure to proteins that may elicit allergic responses. Accurate methods to detect allergens are therefore necessary to ensure consumer/patient safety. We demonstrate that it is possible to reach a new level of accuracy in computational detection of allergenic proteins by presenting a novel detector, Detection based on Filtered Length-adjusted Allergen Peptides (DFLAP). The DFLAP algorithm extracts variable length allergen sequence fragments and employs modern machine learning techniques in the form of a support vector machine. In particular, this new detector shows hitherto unmatched specificity when challenged to the Swiss-Prot repository without appreciable loss of sensitivity. DFLAP is also the first reported detector that successfully discriminates between allergens and non-allergens occurring in protein families known to hold both categories. Allergenicity assessment for specific protein sequences of interest using DFLAP is possible via ulfh@slv.se.

Identification of molecular mechanisms for cellular drug resistance by combining drug activity and gene expression profiles.

Acquired drug resistance is a major problem in cancer treatment. To explore the genes involved in chemosensitivity and resistance, 10 human tumour cell lines, including parental cells and resistant subtypes selected for resistance against doxorubicin, melphalan, teniposide and vincristine, were profiled for mRNA expression of 7400 genes using cDNA microarray technology. The drug activity of 66 cancer agents was evaluated on the cell lines, and correlations between drug activity and gene expression were calculated and ranked. Hierarchical clustering of drugs based on their drug-gene correlations yielded clusters of drugs with similar mechanism of action. Genes correlated with drug sensitivity and resistance were imported into the PathwayAssist software to identify putative molecular pathways involved. A substantial number of both proapoptotic and antiapoptotic genes such as signal transducer and activator of transcription 1, mitogen-activated protein kinase 1 and focal adhesion kinase were found to be associated to drug resistance, whereas genes linked to cell cycle control and proliferation, such as cell division cycle 25A and signal transducer of activator of transcription 5A, were associated to general drug sensitivity. The results indicate that combined information from drug activity and gene expression in a resistance-based cell line panel may provide new knowledge of the genes involved in anticancer drug resistance and become a useful tool in drug development.

Phase-retrieved pupil functions in wide-field fluorescence microscopy.

Pupil functions are compact and modifiable descriptions of the three-dimensional (3D) imaging properties of wide-field optical systems. The pupil function of a microscope can be computationally estimated from the measured point spread function (PSF) using phase retrieval algorithms. The compaction of a 3D PSF into a 2D pupil function suppresses artefacts and measurement noise without resorting to rotational averaging. We show here that such 'phase-retrieved' pupil functions can reproduce features in the optical path, both near the sample and in the microscope. Unlike the PSF, the pupil function can be easily modified to include known aberrations, such as those induced by index-mismatched mounting media, simply by multiplying the pupil function by a calculated aberration function. PSFs calculated from such a modified pupil function closely match the corresponding measured PSFs collected under the aberrated imaging conditions. When used for image deconvolution of simulated objects, these phase-retrieved, calculated PSFs perform similarly to directly measured PSFs.

Statistical evaluation of local alignment features predicting allergenicity using supervised classification algorithms.

BACKGROUND: Recently, two promising alignment-based features predicting food allergenicity using the k nearest neighbor (kNN) classifier were reported. These features are the alignment score and alignment length of the best local alignment obtained in a database of known allergen sequences. METHODS: In the work reported here a much more comprehensive statistical evaluation of the potential of these features was performed, this time for the prediction of allergenicity in general. The evaluation consisted of the following four key components. (1) A new high quality database consisting of 318 carefully selected, non-redundant allergens and 1,007 sequences carefully selected to be non-allergens. (2) Three different supervised algorithms: the kNN classifier, the Bayesian linear Gaussian classifier, and the Bayesian quadratic Gaussian classifier. (3) A large set of local alignment procedures defined using the FASTA3 alignment program by means of a wide range of different parameter settings. (4) Novel performance curves, alternative to conventional receiver-operating characteristic curves, to display not only average behaviors but also statistical variations due to small data sets. RESULTS: The linear Gaussian classifier proved most useful among the tested supervised machine learning algorithms, closely followed by the quadratic Gaussian equivalent and kNN. The overall best classification results were obtained with a novel feature vector consisting of the combined alignment scores derived from local alignment procedures using different substitution matrices. CONCLUSIONS: The models reported here should be useful as a part of an integrated assessment scheme for potential protein allergenicity and for future comparisons with alternative bioinformatic approaches.

A probabilistic derivation of the partial least-squares algorithm.

Traditionally the partial least-squares (PLS) algorithm, commonly used in chemistry for ill-conditioned multivariate linear regression, has been derived (motivated) and presented in terms of data matrices. In this work the PLS algorithm is derived probabilistically in terms of stochastic variables where sample estimates calculated using data matrices are employed at the end. The derivation, which offers a probabilistic motivation to each step of the PLS algorithm, is performed for the general multiresponse case and without reference to any latent variable model of the response variable and also without any so-called "inner relation". On the basis of the derivation, some theoretical issues of the PLS algorithm are briefly considered: the complexity of the original motivation of PLS regression which involves an "inner relation"; the original motivation behind the prediction stage of the PLS algorithm; the relationship between uncorrelated and orthogonal latent variables; the limited possibilities to make natural interpretations of the latent variables extracted.

Computational adaptive optics for live three-dimensional biological imaging.

Light microscopy of thick biological samples, such as tissues, is often limited by aberrations caused by refractive index variations within the sample itself. This problem is particularly severe for live imaging, a field of great current excitement due to the development of inherently fluorescent proteins. We describe a method of removing such aberrations computationally by mapping the refractive index of the sample using differential interference contrast microscopy, modeling the aberrations by ray tracing through this index map, and using space-variant deconvolution to remove aberrations. This approach will open possibilities to study weakly labeled molecules in difficult-to-image live specimens.

Surpassing the lateral resolution limit by a factor of two using structured illumination microscopy.

Lateral resolution that exceeds the classical diffraction limit by a factor of two is achieved by using spatially structured illumination in a wide-field fluorescence microscope. The sample is illuminated with a series of excitation light patterns, which cause normally inaccessible high-resolution information to be encoded into the observed image. The recorded images are linearly processed to extract the new information and produce a reconstruction with twice the normal resolution. Unlike confocal microscopy, the resolution improvement is achieved with no need to discard any of the emission light. The method produces images of strikingly increased clarity compared to both conventional and confocal microscopes.

Extended resolution fluorescence microscopy.

Fluorescence microscopy is an essential tool of modern biology, but, like all forms of optical imaging, it is subject to physical limits on its resolving power. In recent years, several exciting techniques have been introduced to exceed these limits, including standing wave microscopy, 4Pi confocal microscopy, I5M and structured illumination microscopy. Several such techniques have been definitively demonstrated for the first time during the past year.

I5M: 3D widefield light microscopy with better than 100 nm axial resolution.

Sevenfold improved axial resolution has been achieved in three-dimensional widefield fluorescence microscopy, using a novel interferometric technique in which the sample is observed and/or illuminated from both sides simultaneously using two opposing objective lenses. Separate interference effects in the excitation light and the emitted light give access to higher resolution axial information about the sample than can be reached by conventional widefield or confocal microscopes. Here we report the experimental verification of this resolution performance on complex biological samples.

Split spectrum algorithms rely on instantaneous phase information-a geometrical approach [US NDE].

Split spectrum algorithms for suppression of interference noise in ultrasonic nondestructive evaluation are known to work well when properly tuned. However, the algorithms are sensitive to certain parameter values and it is not clear how the algorithms use the phase and amplitude information available. This is partly because most split spectrum algorithms have been suggested heuristically without any detailed signal and noise model. A first step towards a model based approach to split spectrum is initiated. Based on a fairly detailed signal model, geometrical interpretations of the split spectrum concept are presented in an attempt to trace the phase and amplitude information utilized by the algorithms. In the light of gained theoretical knowledge, the conventional algorithms, polarity thresholding, minimization and geometric mean are evaluated. The geometrical approach is related to statistical pattern recognition and neural network approaches and the signal model is verified experimentally on real ultrasonic signals.

Scanning tunneling microscopy of planar biomembranes.

We combined planar membrane monolayer techniques with scanning tunneling microscopy (STM) to measure the thickness of metal-coated purple membrane (PM) isolated from Halobacterium halobium. Although the metal coating precluded obtaining high-resolution lateral information, it facilitated obtaining high-resolution vertical information. For example, the apparent mean thickness of planar PM and variations in thickness of enzyme-treated PM could be detected and quantified at sub-nanometer resolution.

Measuring changes in membrane thickness by scanning tunneling microscopy.

We investigated the feasibility of using the scanning tunneling microscope (STM) as a morphometric tool to measure the thickness of biomembranes. Planar monolayers of oriented purple membrane (PM) were prepared, nitrogen-dried or freeze-etched, and coated with metal. PM thickness was quantified by STM and transmission electron microscopy. STM calibration and the effect of contamination-mediated surface deformation on measurements of PM thickness were evaluated. The thickness of PM attached to mica and glass and the effect of papain on PM thickness were also examined. The apparent thickness of enzymatically modified PM increased after papain treatment. The mean thickness of both nitrogen-dried PM on mica and freeze-etched PM on glass was 4.6 nm. After papain treatment PM thickness on mica increased to 4.8 nm and on glass to 5.4 nm. These results demonstrate that STM analysis of metal-coated planar membrane monolayers can be used to measure changes in average membrane thickness at sub-nanometer resolution.