Protein Crystallization Images Research Articles

Fluorescence enhancement of organic dyes by femtosecond laser-induced cavitation bubbles for crystal imaging.

Fluorescence from organic dyes can be applied in many research fields such as imaging, bio-sensing and diagnosis. One shortcoming of fluorescence imaging is the limitation in emission intensity. Amplification of fluorescence signals can be achieved by the enhancement of localized electromagnetic fields. Metallic nanoparticles are widely applied to produce plasmon resonance, but they cause thermal damage to fragile bio-materials. In this study, we propose a method for nanoparticle-free fluorescence enhancement by ultrafast laser-induced cavitation bubbles in organic dye solutions. Fluorescence enhancement without the use of nanoparticles prevents potential hazards including thermal effects and biotoxicity. In order to achieve fluorescence enhancement in neat dye solution, cavitation bubbles were induced by focusing an 800 nm ultrafast laser beam. Another 400 nm laser beam was used to pump the gain medium. Fluorescence enhancement was observed in various dye solutions. The intensity and spectra of the fluorescence emission can be controlled by changing the power and focus of the excitation laser. According to time-resolved microscopy and simulation results, the cavity formed by the laser-induced bubbles results in the enhancement of the localized electromagnetic field and induces the amplification of the fluorescence signal. The bubble-enhanced fluorescence emission was used for imaging of protein crystals without causing thermal damage to the samples. This study provides an effective method for bio-compatible fluorescence enhancement and has application prospects in fields such as bio-imaging.

Rocking-curve imaging of high-quality protein crystals by non-dispersive X-ray optics in the double-crystal configuration

The double-crystal configuration is useful for the evaluation of strain and defects in single crystals. In this study, rocking curve measurements by X-ray topography in the double-crystal configuration were demonstrated using perfect crystals of the protein glucose isomerase (GI). The setup enables precise evaluation of perfection in protein crystals with the nearly nondispersive X-ray optics. It reveals the uniform perfection of GI crystals according to the theory of X-ray diffraction. It is expected that unknown imperfections in various protein crystals of lower quality will be revealed by the nondispersive configuration using perfect protein crystals.

Depth-of-field extension in optical imaging for rapid crystal screening.

The depth of field (DoF) was extended 2.8-fold to achieve rapid crystal screening by retrofitting a custom-designed micro-retarder array (µRA) in the optical beam path of a nonlinear optical microscope. The merits of the proposed strategy for DoF enhancement were assessed in applications of second-harmonic generation imaging of protein crystals. It was found that DoF extension increased the number of crystals detected while simultaneously reducing the number of `z-slices' required for screening. Experimental measurements of the wavelength-dependence of the extended DoF were in excellent agreement with theoretical predictions. These results provide a simple and broadly applicable approach to increase the throughput of existing nonlinear optical imaging methods for protein crystal screening.

Inexpensive robotic system for standard and fluorescent imaging of protein crystals.

Protein-crystallization imaging and classification is a labor-intensive process typically performed either by humans or by instruments that currently cost well over $100 000. This cost puts the use of crystallization-trial imaging outside the reach of most academic laboratories, and also start-up biotechnology firms, where resources are scarce. An imaging system has been designed and prototyped which automatically captures images from multi-well protein-crystallization experiments using both standard and fluorescent imaging techniques at a cost 28 times lower than current market rates. The machine uses a Panowin F1 3D printer as a base and controls it using G-code commands sent from a Python script running on a desktop computer. A graphical user interface (GUI) was developed to enable users to control the machine and facilitate image capture, classification and editing. A 488 nm laser diode and a 525 nm filter were incorporated to allow in situ fluorescent imaging of proteins trace-labeled with a fluorophore, Alexa Fluor 488. The instrument was primarily designed using a 3D printer and augmented using commercially available parts, and this publication aims to serve as a guide for comparable in-laboratory robotics projects.

A low-cost method for visible fluorescence imaging.

A wide variety of crystallization solutions are screened to establish conditions that promote the growth of a diffraction-quality crystal. Screening these conditions requires the assessment of many crystallization plates for the presence of crystals. Automated systems for screening and imaging are very expensive. A simple approach to imaging trace fluorescently labeled protein crystals in crystallization plates has been devised, and can be implemented at a cost as low as $50. The proteins β-lactoglobulin B, trypsin and purified concanavalin A (ConA) were trace fluorescently labeled using three different fluorescent probes: Cascade Yellow (CY), Carboxyrhodamine 6G (CR) and Pacific Blue (PB). A crystallization screening plate was set up using β-lactoglobulin B labeled with CR, trypsin labeled with CY, ConA labeled with each probe, and a mixture consisting of 50% PB-labeled ConA and 50% CR-labeled ConA. The wells of these plates were imaged using a commercially available macro-imaging lens attachment for smart devices that have a camera. Several types of macro lens attachments were tested with smartphones and tablets. Images with the highest quality were obtained with an iPhone 6S and an AUKEY Ora 10× macro lens. Depending upon the fluorescent probe employed and its Stokes shift, a light-emitting diode or a laser diode was used for excitation. An emission filter was used for the imaging of protein crystals labeled with CR and crystals with two-color fluorescence. This approach can also be used with microscopy systems commonly used to observe crystallization plates.

Feature analysis for classification of trace fluorescent labeled protein crystallization images

BackgroundLarge number of features are extracted from protein crystallization trial images to improve the accuracy of classifiers for predicting the presence of crystals or phases of the crystallization process. The excessive number of features and computationally intensive image processing methods to extract these features make utilization of automated classification tools on stand-alone computing systems inconvenient due to the required time to complete the classification tasks. Combinations of image feature sets, feature reduction and classification techniques for crystallization images benefiting from trace fluorescence labeling are investigated.ResultsFeatures are categorized into intensity, graph, histogram, texture, shape adaptive, and region features (using binarized images generated by Otsu’s, green percentile, and morphological thresholding). The effects of normalization, feature reduction with principle components analysis (PCA), and feature selection using random forest classifier are also analyzed. The time required to extract feature categories is computed and an estimated time of extraction is provided for feature category combinations. We have conducted around 8624 experiments (different combinations of feature categories, binarization methods, feature reduction/selection, normalization, and crystal categories). The best experimental results are obtained using combinations of intensity features, region features using Otsu’s thresholding, region features using green percentile G90 thresholding, region features using green percentile G99 thresholding, graph features, and histogram features. Using this feature set combination, 96% accuracy (without misclassifying crystals as non-crystals) was achieved for the first level of classification to determine presence of crystals. Since missing a crystal is not desired, our algorithm is adjusted to achieve a high sensitivity rate. In the second level classification, 74.2% accuracy for (5-class) crystal sub-category classification. Best classification rates were achieved using random forest classifier.ContributionsThe feature extraction and classification could be completed in about 2 s per image on a stand-alone computing system, which is suitable for real time analysis. These results enable research groups to select features according to their hardware setups for real-time analysis.

Highly Sensitive Coherent Anti-Stokes Raman Scattering Imaging of Protein Crystals.

Serial crystallography at last generation X-ray synchrotron sources and free electron lasers enabled data collection with micrometer and even submicrometer size crystals, which have resulted in amazing progress in structural biology. However, imaging of small crystals, which although is highly demanded, remains a challenge, especially in the case of membrane protein crystals. Here we describe a new extremely sensitive method of the imaging of protein crystals that is based on coherent anti-Stokes Raman scattering.

Super-Thresholding: Supervised Thresholding of Protein Crystal Images

In general, a single thresholding technique is developed or enhanced to separate foreground objects from background for a domain of images. This idea may not generate satisfactory results for all images in a dataset, since different images may require different types of thresholding methods for proper binarization or segmentation. To overcome this limitation, in this study, we propose a novel approach called "super-thresholding" that utilizes a supervised classifier to decide an appropriate thresholding method for a specific image. This method provides a generic framework that allows selection of the best thresholding method among different thresholding techniques that are beneficial for the problem domain. A classifier model is built using features extracted priori from the original image only or posteriori by analyzing the outputs of thresholding methods and the original image. This model is applied to identify the thresholding method for new images of the domain. We performed our method on protein crystallization images, and then we compared our results with six thresholding techniques. Numerical results are provided using four different correctness measurements. Super-thresholding outperforms the best single thresholding method around 10 percent, and it gives the best performance for protein crystallization dataset in our experiments.

Three-dimensional reconstruction of the size and shape of protein microcrystals using Bragg coherent diffractive imaging

Three-dimensional imaging of protein crystals during x-ray diffraction experiments opens up a range of possibilities for optimizing crystal quality and gaining new insights into the fundamental processes that drive radiation damage. Obtaining this information at the appropriate length-scales however is extremely challenging. One approach that has been recently demonstrated as a promising avenue for characterizing the size and shape of protein crystals at nanometre length-scales is Bragg coherent diffractive imaging (BCDI). BCDI is a recently developed technique that is able to recover the phase of the continuous diffraction intensity signal around individual Bragg peaks. When data is collected at multiple points on a rocking curve, a reciprocal space map (RSM) can be assembled and then inverted using BCDI to obtain a three-dimensional image of the crystal. The first demonstration of two-dimensional biological BCDI was reported by Boutet et al on holoferritin, recently this work was extended to the study of radiation damage in micron-sized protein crystals. Here we present the first three-dimensional reconstructions of a Lysozyme protein crystal using BDI. The results are validated against RSM and transmission electron microscopy data and have implications for both radiation damage studies and for developing new approaches for structure retrieval from micron-sized protein crystals.

FocusALL: Focal Stacking of Microscopic Images Using Modified Harris Corner Response Measure.

Automated image analysis of microscopic images such as protein crystallization images and cellular images is one of the important research areas. If objects in a scene appear at different depths with respect to the camera's focal point, objects outside the depth of field usually appear blurred. Therefore, scientists capture a collection of images with different depths of field. Focal stacking is a technique of creating a single focused image from a stack of images collected with different depths of field. In this paper, we introduce a novel focal stacking technique, FocusALL, which is based on our modified Harris Corner Response Measure. We also propose enhanced FocusALL for application on images collected under high resolution and varying illumination. FocusALL resolves problems related to the assumption that focus regions have high contrast and high intensity. Especially, FocusALL generates sharper boundaries around protein crystal regions and good in focus images for high resolution images in reasonable time. FocusALL outperforms other methods on protein crystallization images and performs comparably well on other datasets such as retinal epithelial images and simulated datasets.

Learning Deep Convolutional Neural Networks for X-Ray Protein Crystallization Image Analysis

Obtaining a protein's 3D structure is crucial to the understanding of its functions and interactions with other proteins. It is critical to accelerate the protein crystallization process with improved accuracy for understanding cancer and designing drugs. Systematic high-throughput approaches in protein crystallization have been widely applied, generating a large number of protein crystallization-trial images. Therefore, an efficient and effective automatic analysis for these images is a top priority. In this paper, we present a novel system, CrystalNet, for automatically labeling outcomes of protein crystallization-trial images. CrystalNet is a deep convolutional neural network that automatically extracts features from X-ray protein crystallization images for classification. We show that (1) CrystalNet can provide real-time labels for crystallization images effectively, requiring approximately 2 seconds to provide labels for all 1536 images of crystallization microassay on each plate; (2) compared with the state-of-the-art classification systems in crystallization image analysis, our technique demonstrates an improvement of 8% in accuracy, and achieve 90.8% accuracy in classification. As a part of the high-throughput pipeline which generates millions of images a year, CrystalNet can lead to a substantial reduction of labor-intensive screening.

Intercalating dyes for enhanced contrast in second-harmonic generation imaging of protein crystals.

The second-harmonic generation (SHG) activity of protein crystals was found to be enhanced by up to ∼1000-fold by the intercalation of SHG phores within the crystal lattice. Unlike the intercalation of fluorophores, the SHG phores produced no significant background SHG from solvated dye or from dye intercalated into amorphous aggregates. The polarization-dependent SHG is consistent with the chromophores adopting the symmetry of the crystal lattice. In addition, the degree of enhancement for different symmetries of dyes is consistent with theoretical predictions based on the molecular nonlinear optical response. Kinetics studies indicate that intercalation arises over a timeframe of several minutes in lysozyme, with detectable enhancements within seconds. These results provide a potential means to increase the overall diversity of protein crystals and crystal sizes amenable to characterization by SHG microscopy.

Evaluation of Normalization and PCA on the Performance of Classifiers for Protein Crystallization Images.

In this paper, we investigate the performance of classification of protein crystallization images captured during protein crystal growth process. We group protein crystallization images into 3 categories: noncrystals, likely leads (conditions that may yield formation of crystals) and crystals. In this research, we only consider the subcategories of noncrystal and likely leads protein crystallization images separately. We use 5 different classifiers to solve this problem and we applied some data preprocessing methods such as principal component analysis (PCA), min-max (MM) normalization and z-score (ZS) normalization methods to our datasets in order to evaluate their effects on classifiers for the noncrystal and likely leads datasets. We performed our experiments on 1606 noncrystal and 245 likely leads images independently. We had satisfactory results for both datasets. We reached 96.8% accuracy for noncrystal dataset and 94.8% accuracy for likely leads dataset. Our target is to investigate the best classifiers with optimal preprocessing techniques on both noncrystal and likely leads datasets.

Evaluation of Semi-supervised Learning for Classification of Protein Crystallization Imagery.

In this paper, we investigate the performance of two wrapper methods for semi-supervised learning algorithms for classification of protein crystallization images with limited labeled images. Firstly, we evaluate the performance of semi-supervised approach using self-training with naïve Bayesian (NB) and sequential minimum optimization (SMO) as the base classifiers. The confidence values returned by these classifiers are used to select high confident predictions to be used for self-training. Secondly, we analyze the performance of Yet Another Two Stage Idea (YATSI) semi-supervised learning using NB, SMO, multilayer perceptron (MLP), J48 and random forest (RF) classifiers. These results are compared with the basic supervised learning using the same training sets. We perform our experiments on a dataset consisting of 2250 protein crystallization images for different proportions of training and test data. Our results indicate that NB and SMO using both self-training and YATSI semi-supervised approaches improve accuracies with respect to supervised learning. On the other hand, MLP, J48 and RF perform better using basic supervised learning. Overall, random forest classifier yields the best accuracy with supervised learning for our dataset.

Confocal microscopy on the beamline: novel three-dimensional imaging and sample positioning

Confocal microscopy, a technique that has been extensively applied in cellular biological studies, may also be applied to the visualization and three-dimensional imaging of protein crystals at high resolution on synchrotron beamlines. Protein crystal samples are examined using a commercially available confocal microscope adapted for cryogenic use. A preliminary test using a custom confocal design adapted for beamline use is also presented. The confocal optics configuration is compatible with nonlinear imaging techniques such as two-photon excited fluorescence imaging and second harmonic generation. The possibilities of this method are explored using two modes: fluorescence and reflection confocal. In fluorescence mode, small amounts of dye are introduced into the crystal through soaking or growth conditions. Under such conditions, protein crystals are easily resolved from salts and amorphous precipitates, which do not generally take up dye. Reflection mode, which does not require dye, still exhibits greater resolution and sensitivity to surface detail than conventional wide-field microscopy as a result of the confocal optics configuration. The inherent three-dimensional nature of the method means that on-axis sample views (along the direction of the X-ray beam) can be reconstructed from an off-axis configuration, simplifying the beamline setup and providing uniquely detailed views of cryogenically cooled crystals.

Three-dimensional, non-invasive, cross-sectional imaging of protein crystals using ultrahigh resolution optical coherence tomography

Micro-scale, non-invasive, three-dimensional cross-sectional imaging of protein crystals was successfully accomplished using ultra-high resolution optical coherence tomography (UHR-OCT) with low noise, Gaussian like supercontinuum. This technique facilitated visualization of protein crystals even those in medium that also contained substantial amounts of precipitates. We found the enhancement of the scattered signal from protein crystal by inclusion of agarose gel in the crystallization medium. Crystals of a protein and a salt in the same sample when visualized by UHR-OCT showed distinct physical characteristics, suggesting that protein and salt crystals may, in general, be distinguishable by UHR-OCT. UHR-OCT is a nondestructive and rapid method, which should therefore find use in automated systems designed to visualize crystals.

Imaging of Protein Crystals with Two-Photon Microscopy

Second-order nonlinear optical imaging of chiral crystals (SONICC), which portrays second-harmonic generation (SHG) by noncentrosymmetric crystals, is emerging as a powerful imaging technique for protein crystals in media opaque to visible light because of its high signal-to-noise ratio. Here we report the incorporation of both SONICC and two-photon excited fluorescence (TPEF) into one imaging system that allows visualization of crystals as small as ~10 μm in their longest dimension. Using this system, we then documented an inverse correlation between the level of symmetry in examined crystals and the intensity of their SHG. Moreover, because of blue-green TPEF exhibited by most tested protein crystals, we also could identify and image SHG-silent protein crystals. Our experimental data suggest that the TPEF in protein crystals is mainly caused by the oxidation of tryptophan residues. Additionally, we found that unspecific fluorescent dyes are able to bind to lysozyme crystals and enhance their detection by TPEF. We finally confirmed that the observed fluorescence was generated by a two-photon rather than a three-photon process. The capability for imaging small protein crystals in turbid or opaque media with nondamaging infrared light in a single system makes the combination of SHG and intrinsic visible TPEF a powerful tool for nondestructive protein crystal identification and characterization during crystallization trials.

Three-dimensional Raman spectroscopic imaging of protein crystals deposited on a nanodroplet

Confocal Raman spectroscopic imaging has been used to find the location of protein crystals deposited in a nanodroplet. The depth of the protein crystal has been clearly identified by comparing the three-dimensional Raman spectroscopic images of the protein with those of water. Additionally, the low concentration region around a growing protein crystal in the nanodroplet was visualized using two-dimensional Raman spectroscopic imaging.

Nonlinear optical imaging of integral membrane protein crystals in lipidic mesophases.

Second-order nonlinear optical imaging of chiral crystals (SONICC) is explored for selective detection of integral membrane protein crystals grown in opaque and turbid environments. High turbidity is a hallmark of membrane protein crystallization due to the extensive use of detergent and/or lipids that often form various mesophases. Detection of crystals in such media by conventional optical methods (e.g., intrinsic UV fluorescence, birefringence, bright-field image analysis, etc.) is often complicated by optical scattering and by the small sizes of the crystals that routinely form. SONICC is shown to be well-suited for this application, by nature of its compatibility with imaging in scattering media and its high selectivity for protein crystals. Bright second harmonic generation (SHG) (up to 18 million counts/s) was observed from even relatively small crystals (5 mum) with a minimal background due to the surrounding lipid mesophase ( approximately 1 thousand counts/s). The low background nature of the resulting protein crystal images permitted the use of a relatively simple, particle counting analysis for preliminary scoring. Comparisons between a particle counting analysis of SONICC images and protocols based on the human expert analysis of conventional bright-field and birefringence images were performed.

Coherent X-ray diffractive imaging of protein crystals

The technique of coherent X-ray diffraction imaging (CXDI) has recently shown great promise for the study of inorganic nanocrystals. In this work the CXDI method has been applied to the study of micrometer-size protein crystals. Finely sampled diffraction patterns of single crystals were measured and iterative phase-retrieval algorithms were used to reconstruct the two-dimensional shape of the crystal. The density maps have limited reproducibility because of radiation damage, but show clear evidence for crystal facets. Qualitative analysis of a number of single-crystal diffraction peaks indicates the presence of inward surface contraction on 2 microm size crystals. A survey of several hundred diffraction patterns yielded a number of examples with dramatic single-sided streaks, for which a plausible model is constructed.