High-Throughput Molecular Imaging via Deep-Learning-Enabled Raman Spectroscopy.

Raman spectroscopy enables nondestructive, label-free imaging with unprecedented molecular contrast, but is limited by slow data acquisition, largely preventing high-throughput imaging applications. Here, we present a comprehensive framework for higher-throughput molecular imaging via deep-learning-enabled Raman spectroscopy, termed DeepeR, trained on a large data set of hyperspectral Raman images, with over 1.5 million spectra (400 h of acquisition) in total. We first perform denoising and reconstruction of low signal-to-noise ratio Raman molecular signatures via deep learning, with a 10× improvement in the mean-squared error over common Raman filtering methods. Next, we develop a neural network for robust 2-4× spatial super-resolution of hyperspectral Raman images that preserve molecular cellular information. Combining these approaches, we achieve Raman imaging speed-ups of up to 40-90×, enabling good-quality cellular imaging with a high-resolution, high signal-to-noise ratio in under 1 min. We further demonstrate Raman imaging speed-up of 160×, useful for lower resolution imaging applications such as the rapid screening of large areas or for spectral pathology. Finally, transfer learning is applied to extend DeepeR from cell to tissue-scale imaging. DeepeR provides a foundation that will enable a host of higher-throughput Raman spectroscopy and molecular imaging applications across biomedicine.

Multiplexed polarized hypodermic Raman needle probe for biostructural analysis of articular cartilage.

In this Letter, we report a multiplexed polarized hypodermic Raman needle probe for the biostructural analysis of articular cartilage. Using a custom-developed needle probe with a sapphire ball lens, we measure polarized Raman spectra of cartilage. By imaging two polarizations simultaneously on the charge-coupled device (CCD) and binning them separately, we capture both biochemical and structural tissue information in real time. Here, we demonstrate that polarized Raman spectroscopy can distinguish between different collagen fibril alignment orientations in a cartilage explant model system, supporting its capacity for diagnosing the hallmark collagen alignment changes occurring in the early stages of osteoarthritis (OA). Accordingly, this work shows that needle-based polarized Raman spectroscopy has great potential for the monitoring and diagnosis of early OA.

Structural and functional analysis of FIP2 binding to the endosome-localised Rab25 GTPase.

Rab small GTPases are the master regulators of intracellular trafficking in eukaryotes. They mediate spatial and temporal recruitment of effector proteins to distinct cellular compartments through GTP-induced changes in their conformation. Despite numerous structural studies, the molecular basis for Rab/effector specificity and subsequent biological activity remains poorly understood. Rab25, also known as Rab11c, which is epithelial-specific, has been heavily implicated in ovarian cancer development and independently appears to act as a tumour suppressor in the context of a distinct subset of carcinomas. Here, we show that Rab25 associates with FIP2 and can recruit this effector protein to endosomal membranes. We report the crystal structure of Rab25 in complex with the C-terminal region of FIP2, which consists of a central dimeric FIP2 coiled-coil that mediates a heterotetrameric Rab25-(FIP2)2-Rab25 complex. Thermodynamic analyses show that, despite a relatively conserved interface, FIP2 binds to Rab25 with an approximate 3-fold weaker affinity than to Rab11a. Reduced affinity is mainly associated with lower enthalpic gains for Rab25:FIP2 complex formation, and can be attributed to subtle differences in the conformations of switch 1 and switch 2. These cellular, structural and thermodynamic studies provide insight into the Rab11/Rab25 subfamily of small GTPases that regulate endosomal trafficking pathways in eukaryotes.

GRAB is a binding partner for the Rab11a and Rab11b GTPases.

Co-ordination of Rab GTPase function has emerged as a crucial mechanism in the control of intracellular trafficking processes in eukaryotic cells. Here, we show that GRAB/Rab3IL1 [guanine nucleotide exchange factor for Rab3A; RAB3A interacting protein (rabin3)-like 1], a protein that has previously be shown to act as a GEF (guanine nucleotide exchange factor) for Rab3a, Rab8a and Rab8b, is also a binding partner for Rab11a and Rab11b, but not the closely related Rab25 GTPase. We demonstrate that exogenous expression of Rab11a and Rab11b shift GRAB's distribution from the cytoplasm onto membranes. We find that the Rab11a/Rab11b-binding region of GRAB lies within its carboxy-terminus, a region distinct from its GEF domain and Rab3a-binding region. Finally, we describe a GRAB deletion mutant (GRABΔ223-228) that is deficient in Rab11-binding ability. These data identify GRAB as a dual Rab-binding protein that could potentially link Rab3 and Rab11 and/or Rab8 and Rab11-mediated intracellular trafficking processes.

Rab30 is required for the morphological integrity of the Golgi apparatus.

BACKGROUND INFORMATION: Rab GTPases are key coordinators of eukaryotic intracellular membrane trafficking. In their active states, Rabs localise to the cytoplasmic face of intracellular compartments where they regulate membrane trafficking processes. Many Rabs have been extensively characterised whereas others, such as Rab30, have to date received relatively little attention. RESULTS: Here, we demonstrate that Rab30 is primarily associated with the secretory pathway, displaying predominant localisation to the Golgi apparatus. We find by time-lapse microscopy and fluorescence recovery after photobleaching studies that Rab30 is rapidly and continuously recruited to the Golgi. We also show that Rab30 function is required for the morphological integrity of the Golgi. Finally, we demonstrate that inactivation of Rab30 does not impair anterograde or retrograde transport through the Golgi. CONCLUSIONS: Taken together, these data illustrate that Rab30 primarily localises to the Golgi apparatus and is required for the structural integrity of this organelle.

Spatial gradient consistency for unsupervised learning of hyperspectral demosaicking: application to surgical imaging.

PURPOSE: Hyperspectral imaging has the potential to improve intraoperative decision making if tissue characterisation is performed in real-time and with high-resolution. Hyperspectral snapshot mosaic sensors offer a promising approach due to their fast acquisition speed and compact size. However, a demosaicking algorithm is required to fully recover the spatial and spectral information of the snapshot images. Most state-of-the-art demosaicking algorithms require ground-truth training data with paired snapshot and high-resolution hyperspectral images, but such imagery pairs with the exact same scene are physically impossible to acquire in intraoperative settings. In this work, we present a fully unsupervised hyperspectral image demosaicking algorithm which only requires exemplar snapshot images for training purposes. METHODS: We regard hyperspectral demosaicking as an ill-posed linear inverse problem which we solve using a deep neural network. We take advantage of the spectral correlation occurring in natural scenes to design a novel inter spectral band regularisation term based on spatial gradient consistency. By combining our proposed term with standard regularisation techniques and exploiting a standard data fidelity term, we obtain an unsupervised loss function for training deep neural networks, which allows us to achieve real-time hyperspectral image demosaicking. RESULTS: Quantitative results on hyperspetral image datasets show that our unsupervised demosaicking approach can achieve similar performance to its supervised counter-part, and significantly outperform linear demosaicking. A qualitative user study on real snapshot hyperspectral surgical images confirms the results from the quantitative analysis. CONCLUSION: Our results suggest that the proposed unsupervised algorithm can achieve promising hyperspectral demosaicking in real-time thus advancing the suitability of the modality for intraoperative use.

Synthesis and evaluation of aromatic BDSF bioisosteres on biofilm formation and colistin sensitivity in pathogenic bacteria.

The diffusible signal factor family (DSF) of molecules play an important role in regulating intercellular communication, or quorum sensing, in several disease-causing bacteria. These messenger molecules, which are comprised of cis-unsaturated fatty acids, are involved in the regulation of biofilm formation, antibiotic tolerance, virulence and the control of bacterial resistance. We have previously demonstrated how olefinic N-acyl sulfonamide bioisosteric analogues of diffusible signal factor can reduce biofilm formation or enhance antibiotic sensitivity in a number of bacterial strains. This work describes the design and synthesis of a second generation of aromatic N-acyl sulfonamide bioisosteres. The impact of these compounds on biofilm production in Acinetobacter baumannii, Escherichia coli, Burkholderia multivorans, Burkholderia cepacia, Burkholderia cenocepacia, Pseudomonas aeruginosa and Stenotrophomonas maltophilia is evaluated, in addition to their effects on antibiotic tolerance. The ability of these molecules to increase survival rates on co-administration with colistin is also investigated using the Galleria infection model.

Synthetic white balancing for intra-operative hyperspectral imaging.

Purpose: Hyperspectral imaging shows promise for surgical applications to non-invasively provide spatially resolved, spectral information. For calibration purposes, a white reference image of a highly reflective Lambertian surface should be obtained under the same imaging conditions. Standard white references are not sterilizable and so are unsuitable for surgical environments. We demonstrate the necessity for in situ white references and address this by proposing a novel, sterile, synthetic reference construction algorithm. Approach: The use of references obtained at different distances and lighting conditions to the subject were examined. Spectral and color reconstructions were compared with standard measurements qualitatively and quantitatively, using ΔE and normalized RMSE, respectively. The algorithm forms a composite image from a video of a standard sterile ruler, whose imperfect reflectivity is compensated for. The reference is modeled as the product of independent spatial and spectral components, and a scalar factor accounting for gain, exposure, and light intensity. Evaluation of synthetic references against ideal but non-sterile references is performed using the same metrics alongside pixel-by-pixel errors. Finally, intraoperative integration is assessed though cadaveric experiments. Results: Improper white balancing leads to increases in all quantitative and qualitative errors. Synthetic references achieve median pixel-by-pixel errors lower than 6.5% and produce similar reconstructions and errors to an ideal reference. The algorithm integrated well into surgical workflow, achieving median pixel-by-pixel errors of 4.77% while maintaining good spectral and color reconstruction. Conclusions: We demonstrate the importance of in situ white referencing and present a novel synthetic referencing algorithm. This algorithm is suitable for surgery while maintaining the quality of classical data reconstruction.

Lightfield hyperspectral imaging in neuro-oncology surgery: an IDEAL 0 and 1 study.

Introduction: Hyperspectral imaging (HSI) has shown promise in the field of intra-operative imaging and tissue differentiation as it carries the capability to provide real-time information invisible to the naked eye whilst remaining label free. Previous iterations of intra-operative HSI systems have shown limitations, either due to carrying a large footprint limiting ease of use within the confines of a neurosurgical theater environment, having a slow image acquisition time, or by compromising spatial/spectral resolution in favor of improvements to the surgical workflow. Lightfield hyperspectral imaging is a novel technique that has the potential to facilitate video rate image acquisition whilst maintaining a high spectral resolution. Our pre-clinical and first-in-human studies (IDEAL 0 and 1, respectively) demonstrate the necessary steps leading to the first in-vivo use of a real-time lightfield hyperspectral system in neuro-oncology surgery. Methods: A lightfield hyperspectral camera (Cubert Ultris ×50) was integrated in a bespoke imaging system setup so that it could be safely adopted into the open neurosurgical workflow whilst maintaining sterility. Our system allowed the surgeon to capture in-vivo hyperspectral data (155 bands, 350-1,000 nm) at 1.5 Hz. Following successful implementation in a pre-clinical setup (IDEAL 0), our system was evaluated during brain tumor surgery in a single patient to remove a posterior fossa meningioma (IDEAL 1). Feedback from the theater team was analyzed and incorporated in a follow-up design aimed at implementing an IDEAL 2a study. Results: Focusing on our IDEAL 1 study results, hyperspectral information was acquired from the cerebellum and associated meningioma with minimal disruption to the neurosurgical workflow. To the best of our knowledge, this is the first demonstration of HSI acquisition with 100+ spectral bands at a frame rate over 1Hz in surgery. Discussion: This work demonstrated that a lightfield hyperspectral imaging system not only meets the design criteria and specifications outlined in an IDEAL-0 (pre-clinical) study, but also that it can translate into clinical practice as illustrated by a successful first in human study (IDEAL 1). This opens doors for further development and optimisation, given the increasing evidence that hyperspectral imaging can provide live, wide-field, and label-free intra-operative imaging and tissue differentiation.

Rab11-FIP3 links the Rab11 GTPase and cytoplasmic dynein to mediate transport to the endosomal-recycling compartment.

Several protein families control intracellular transport processes in eukaryotic cells. Here, we show that the Rab11 GTPase effector protein Rab11-FIP3 (henceforth, FIP3) directly interacts with the dynein light intermediate chain 1 (DLIC-1, gene symbol DYNC1LI1) subunit of the cytoplasmic dynein 1 motor protein complex. We show that Rab11a, FIP3 and DLIC-1 form a ternary complex and that DLIC-1 colocalises with endogenous FIP3 and Rab11a in A431 cells. We demonstrate that association between FIP3 and DLIC-1 at the cell periphery precedes minus-end-directed microtubule-based transport, that FIP3 recruits DLIC-1 onto membranes, and that knockdown of DLIC-1 inhibits pericentrosomal accumulation of key endosomal-recycling compartment (ERC) proteins. In addition, we demonstrate that expression of a DLIC-1-binding truncation mutant of FIP3 disrupts the ability of ERC proteins to accumulate pericentrosomally. On the basis of these and other data, we propose that FIP3 links the Rab11 GTPase and cytoplasmic dynein to mediate transport of material from peripheral sorting endosomes to the centrally located ERC.

Rab11 proteins in health and disease.

Comprising over 60 members, Rab proteins constitute the largest branch of the Ras superfamily of low-molecular-mass G-proteins. This protein family have been primarily implicated in various aspects of intracellular membrane trafficking processes. On the basis of distinct subfamily-specific sequence motifs, many Rabs have been grouped into subfamilies. The Rab11 GTPase subfamily comprises three members: Rab11a, Rab11b and Rab25/Rab11c, which, between them, have been demonstrated to bind more than 30 proteins. In the present paper, we review the function of the Rab11 subfamily. We describe their localization and primary functional roles within the cell and their implication, to date, in disease processes. We also summarize the protein machinery currently known to regulate or mediate their functions and the cargo molecules which they have been shown to transport.

The Rab family of proteins: 25 years on.

Intracellular membrane trafficking requires the complex interplay of several classes of trafficking proteins. Rab proteins, the largest subfamily of the Ras superfamily of small G-proteins, are central regulators of all aspects of intracellular trafficking processes including vesicle budding and uncoating, motility, tethering and fusion. In the present paper, we discuss the discovery, evolution and characterization of the Rab GTPase family. We examine their basic functional roles, their important structural features and the regulatory proteins which mediate Rab function. We speculate on outstanding issues in the field, such as the mechanisms of Rab membrane association and the co-ordinated interplay between distinct Rab proteins. Finally, we summarize the data implicating Rab proteins in an ever increasing number of diseases.

The role of endosomal-recycling in long-term potentiation.

Long-term potentiation (LTP) defines persistent increases in neurotransmission strength at synapses that are triggered by specific patterns of neuronal activity. LTP, the most widely accepted molecular model for learning, is best characterised at glutamatergic synapses on dendritic spines. In this context, LTP involves increases in dendritic spine size and the insertion of glutamate receptors into the post-synaptic spine membrane, which together boost post-synaptic responsiveness to neurotransmitters. In dendrites, the material required for LTP is sourced from an organelle termed the endosomal-recycling compartment (ERC), which is localised to the base of dendritic spines. When LTP is induced, material derived from the recycling compartment, which contains α-amino-3-hydroxy-5-methyl-4-isoxazole propionate-type glutamate receptors (AMPARs), is mobilised into dendritic spines feeding the increased need for receptors and membrane at the spine neck and head. In this review, we discuss the importance of endosomal-recycling and the role of key proteins which control these processes in the context of LTP.

Recent Developments in the Practical Application of Novel Carboxylic Acid Bioisosteres.

BACKGROUND: The carboxylic acid moiety is an important functional group which features in the pharmacophore of some 450 drugs. Unfortunately, some carboxylic acid-containing drugs have been withdrawn from market due to unforeseen toxicity issues. Other issues associated with the carboxylate moiety include reduced metabolic stability or limited passive diffusion across biological membranes. Medicinal chemists often turn to bioisosteres to circumvent such obstacles. OBJECTIVE: The aim of this review is to provide a summary of the various applications of novel carboxylic acid bioisosteres which have appeared in the literature since 2013. RESULTS: We have summarised the most recent developments in carboxylic acid bioisosterism. In particular, we focus on the changes in bioactivity, selectivity or physicochemical properties brought about by these substitutions, as well as the advantages and disadvantages of each isostere. CONCLUSION: The topics discussed herein highlight the continued interest in carboxylate bioisosteres. The development of novel carboxylic acid substitutes which display improved pharmacological profiles is a testament to the innovation and creativity required to overcome the challenges faced in modern drug design.

Hybrid confocal Raman endomicroscopy for morpho-chemical tissue characterization.

Confocal laser endomicroscopy (CLE) offers imaging of tissue microarchitecture and has emerged as a promising tool for in vivo clinical diagnosis of cancer across many organs. CLE, however, can show high inter-observer dependency and does not provide information about tissue molecular composition. In contrast, Raman spectroscopy is a label-free optical technique that provides detailed biomolecular compositional information but offers limited or no morphological information. Here we present a novel hybrid fiber-optic confocal Raman endomicroscopy system for morpho-chemical tissue imaging and analysis. The developed confocal endomicroscopy system is based on a novel detection scheme for rejecting Raman silica fiber interference permitting simultaneous CLE imaging and Raman spectral acquisition of tissues through a coherent fiber bundle. We show that this technique enables real-time microscopic visualization of tissue architecture as well as simultaneous pointwise label-free biomolecular characterization and fingerprinting of tissue paving the way for multimodal diagnostics at endoscopy.

Deep learning approach for hyperspectral image demosaicking, spectral correction and high-resolution RGB reconstruction.

Hyperspectral imaging is one of the most promising techniques for intraoperative tissue characterisation. Snapshot mosaic cameras, which can capture hyperspectral data in a single exposure, have the potential to make a real-time hyperspectral imaging system for surgical decision-making possible. However, optimal exploitation of the captured data requires solving an ill-posed demosaicking problem and applying additional spectral corrections. In this work, we propose a supervised learning-based image demosaicking algorithm for snapshot hyperspectral images. Due to the lack of publicly available medical images acquired with snapshot mosaic cameras, a synthetic image generation approach is proposed to simulate snapshot images from existing medical image datasets captured by high-resolution, but slow, hyperspectral imaging devices. Image reconstruction is achieved using convolutional neural networks for hyperspectral image super-resolution, followed by spectral correction using a sensor-specific calibration matrix. The results are evaluated both quantitatively and qualitatively, showing clear improvements in image quality compared to a baseline demosaicking method using linear interpolation. Moreover, the fast processing time of 45 ms of our algorithm to obtain super-resolved RGB or oxygenation saturation maps per image for a state-of-the-art snapshot mosaic camera demonstrates the potential for its seamless integration into real-time surgical hyperspectral imaging applications.

In vitro and in vivo investigation of a zonal microstructured scaffold for osteochondral defect repair.

Articular cartilage is comprised of zones that vary in architecture, extracellular matrix composition, and mechanical properties. Here, we designed and engineered a porous zonal microstructured scaffold from a single biocompatible polymer (poly [ϵ-caprolactone]) using multiple fabrication strategies: electrospinning, spherical porogen leaching, directional freezing, and melt electrowriting. With this approach we mimicked the zonal structure of articular cartilage and produced a stiffness gradient through the scaffold which aligns with the mechanics of the native tissue. Chondrocyte-seeded scaffolds accumulated extracellular matrix including glycosaminoglycans and collagen II over four weeks in vitro. This prompted us to further study the repair efficacy in a skeletally mature porcine model. Two osteochondral lesions were produced in the trochlear groove of 12 animals and repaired using four treatment conditions: (1) microstructured scaffold, (2) chondrocyte seeded microstructured scaffold, (3) MaioRegen™, and (4) empty defect. After 6 months the defect sites were harvested and analyzed using histology, micro computed tomography, and Raman microspectroscopy mapping. Overall, the scaffolds were retained in the defect space, repair quality was repeatable, and there was clear evidence of osteointegration. The repair quality of the microstructured scaffolds was not superior to the control based on histological scoring; however, the lower score was biased by the lack of histological staining due to the limited degradation of the implant at 6 months. Longer follow up studies (e.g., 1 yr) will be required to fully evaluate the efficacy of the microstructured scaffold. In conclusion, we found consistent scaffold retention, osteointegration, and prolonged degradation of the microstructured scaffold, which we propose may have beneficial effects for the long-term repair of osteochondral defects.

Rab GTPases and microtubule motors.

Rab proteins are a family of small GTPases which, since their initial identification in the late 1980s, have emerged as master regulators of all stages of intracellular trafficking processes in eukaryotic cells. Rabs cycle between distinct conformations that are dependent on their guanine-nucleotide-bound status. When active (GTP-bound), Rabs are distributed to the cytosolic face of specific membranous compartments where they recruit downstream effector proteins. Rab-effector complexes then execute precise intracellular trafficking steps, which, in many cases, include vesicle motility. Microtubule-based kinesin and cytoplasmic dynein motor complexes are prominent among the classes of known Rab effector proteins. Additionally, many Rabs associate with microtubule-based motors via effectors that act as adaptor molecules that can simultaneously associate with the GTP-bound Rab and specific motor complexes. Thus, through association with motor complexes, Rab proteins can allow for membrane association and directional movement of various vesicular cargos along the microtubule cytoskeleton. In this mini-review, we highlight the expanding repertoire of Rab/microtubule motor protein interactions, and, in doing so, present an outline of the multiplicity of transport processes which result from such interactions.

Dynein LIC1 localizes to the mitotic spindle and midbody and LIC2 localizes to spindle poles during cell division.

CD-1 (cytoplasmic dynein-1) is a multisubunit motor protein complex involved in intracellular trafficking and mitosis. The dynein LIC (light intermediate chain) subunits, LIC1 (DLIC-1, gene symbol DYNC1LI1) and LIC2 (DLIC-2, gene symbol DYNC1LI2), associate with the dynein HC (heavy chain) in a mutually exclusive manner and thus define distinct functional CD-1 complexes. Here, we analysed the mitotic distribution of LIC1 and LIC2. We found that from metaphase through anaphase, LIC1 localizes to the mitotic spindle and concentrates within the midbody during the abscission step of cytokinesis. Conversely, LIC2 strongly localizes to the spindle poles from prophase through telophase. These data suggest distinct functions for LIC1 and LIC2-containing CD-1 complexes during cell division.

Image-guided Raman spectroscopy probe-tracking for tumor margin delineation.

SIGNIFICANCE: Tumor detection and margin delineation are essential for successful tumor resection. However, postsurgical positive margin rates remain high for many cancers. Raman spectroscopy has shown promise as a highly accurate clinical spectroscopic diagnostic modality, but its margin delineation capabilities are severely limited by the need for pointwise application. AIM: We aim to extend Raman spectroscopic diagnostics and develop a multimodal computer vision-based diagnostic system capable of both the detection and identification of suspicious lesions and the precise delineation of disease margins. APPROACH: We first apply visual tracking of a Raman spectroscopic probe to achieve real-time tumor margin delineation. We then combine this system with protoporphyrin IX fluorescence imaging to achieve fluorescence-guided Raman spectroscopic margin delineation. RESULTS: Our system enables real-time Raman spectroscopic tumor margin delineation for both ex vivo human tumor biopsies and an in vivo tumor xenograft mouse model. We then further demonstrate that the addition of protoporphyrin IX fluorescence imaging enables fluorescence-guided Raman spectroscopic margin delineation in a tissue phantom model. CONCLUSIONS: Our image-guided Raman spectroscopic probe-tracking system enables tumor margin delineation and is compatible with both white light and fluorescence image guidance, demonstrating the potential for our system to be developed toward clinical tumor resection surgeries.