Yanhong Xue

Proteome profiling of exosomes derived from human primary and metastatic colorectal cancer cells reveal differential expression of key metastatic factors and signal transduction components

Exosomes are small extracellular 40–100 nm diameter membrane vesicles of late endosomal origin that can mediate intercellular transfer of RNAs and proteins to assist premetastatic niche formation. Using primary (SW480) and metastatic (SW620) human isogenic colorectal cancer cell lines we compared exosome protein profiles to yield valuable insights into metastatic factors and signaling molecules fundamental to tumor progression. Exosomes purified using OptiPrep™ density gradient fractionation were 40–100 nm in diameter, were of a buoyant density ∼1.09 g/mL, and displayed stereotypic exosomal markers TSG101, Alix, and CD63. A major finding was the selective enrichment of metastatic factors (MET, S100A8, S100A9, TNC), signal transduction molecules (EFNB2, JAG1, SRC, TNIK), and lipid raft and lipid raft‐associated components (CAV1, FLOT1, FLOT2, PROM1) in exosomes derived from metastatic SW620 cells. Additionally, using cryo‐electron microscopy, ultrastructural components in exosomes were identified. A key finding of this study was the detection and colocalization of protein complexes EPCAM‐CLDN7 and TNIK‐RAP2A in colorectal cancer cell exosomes. The selective enrichment of metastatic factors and signaling pathway components in metastatic colon cancer cell‐derived exosomes contributes to our understanding of the cross‐talk between tumor and stromal cells in the tumor microenvironment.

Ca2+ Triggers a Novel Clathrin-Independent but Actin-Dependent Fast Endocytosis in Pancreatic Beta Cells

The existence of clathrin‐independent recycling of secretory vesicles has been controversial. By combining patch‐clamp capacitance recording, optical methods and specific molecular interventions, we dissect two types of mechanistically different endocytosis in pancreatic β cells, both of which require GTP and dynamin. The fast one is a novel clathrin‐independent but actin‐dependent endocytosis that is triggered by high cytoplasmic Ca2+ concentration ([Ca2+]i). Large fluorescent dextran (10 nm in diameter) was able to be internalized by this pathway, indicating that it was not likely to be ‘kiss and run’. The slow endocytosis is a clathrin‐dependent process in which actin plays a complementary role. For the first time, we show that the rate constants for both types of endocytosis exhibit supralinear dependence on increase in [Ca2+]i. Compared with the slow endocytosis, higher [Ca2+]i level was required to fully accelerate the fast one, indicative of distinct Ca2+ sensors for different endocytosis. In the end, we show that physiologically relevant stimulation induces clathrin‐independent endocytosis in intact β cells, implying that it may contribute to the normal recycling of secretory vesicles in vivo.

Three-dimensional super-resolution protein localization correlated with vitrified cellular context.

We demonstrate the use of cryogenic super-resolution correlative light and electron microscopy (csCLEM) to precisely determine the spatial relationship between proteins and their native cellular structures. Several fluorescent proteins (FPs) were found to be photoswitchable and emitted far more photons under our cryogenic imaging condition, resulting in higher localization precision which is comparable to ambient super-resolution imaging. Vitrified specimens were prepared by high pressure freezing and cryo-sectioning to maintain a near-native state with better fluorescence preservation. A 2-3-fold improvement of resolution over the recent reports was achieved due to the photon budget performance of screening out Dronpa and optimized imaging conditions, even with thin sections which is at a disadvantage when calculate the structure resolution from label density. We extended csCLEM to mammalian cells by introducing cryo-sectioning and observed good correlation of a mitochondrial protein with the mitochondrial outer membrane at nanometer resolution in three dimensions.

Contribution of Different Mechanisms to Pancreatic Beta-cell Hyper-secretion in Non-obese Diabetic (NOD) Mice during Pre-diabetes

Background: Pre-diabetic islet hyper-secretion is crucial to the development of the disease but the mechanisms remain unknown. Results: We reveal dynamic changes in beta-cell mass and function in non-obese diabetic (NOD) mice of different ages. Conclusion: Beta-cell mass increase and individual beta-cell secretory ability enhancement contribute to islet hyperactivity at different stages. Significance: This may provide insights into alteration of beta-cell function during the disease progression. The development of insulin-dependent diabetes mellitus (IDDM) results from the selective destruction of pancreatic beta-cells. Both humans and spontaneous models of IDDM, such as NOD mice, have an extended pre-diabetic stage. Dynamic changes in beta-cell mass and function during pre-diabetes, such as insulin hyper-secretion, remain largely unknown. In this paper, we evaluated pre-diabetic female NOD mice at different ages (6, 10, and 14 weeks old) to illustrate alterations in beta-cell mass and function as disease progressed. We found an increase in beta-cell mass in 6-week-old NOD mice that may account for improved glucose tolerance in these mice. As NOD mice aged, beta-cell mass progressively reduced with increasing insulitis. In parallel, secretory ability of individual beta-cells was enhanced due to an increase in the size of slowly releasable pool (SRP) of vesicles. Moreover, expression of both SERCA2 and SERCA3 genes were progressively down-regulated, which facilitated depolarization-evoked secretion by prolonging Ca2+ elevation upon glucose stimulation. In summary, we propose that different mechanisms contribute to the insulin hyper-secretion at different ages of pre-diabetic NOD mice, which may provide some new ideas concerning the progression and management of type I diabetes.

Imaging beta-cell mass and function in situ and in vivo

Glucose-stimulated insulin secretion (GSIS) from pancreatic beta-cells is critical to the maintenance of blood glucose homeostasis in animals. Both decrease in pancreatic beta-cell mass and defects in beta-cell function contribute to the onset of diabetes, although the underlying mechanisms remain largely unknown. Molecular imaging techniques can help beta-cell study in a number of ways. High-resolution fluorescence imaging techniques provide novel insights into the fundamental mechanisms underlying GSIS in isolated beta-cells or in situ in pancreatic islets, and dynamic changes of beta-cell mass and function can be noninvasively monitored in vivo by imaging techniques such as positron emission tomography and single-photon emission computed tomography. All these techniques will contribute to the better understanding of the progression of diabetes and the search for the optimized therapeutic measures that reverse deficits in beta-cell mass and function.

HID-1 is required for homotypic fusion of immature secretory granules during maturation

Secretory granules, also known as dense core vesicles, are generated at the trans-Golgi network and undergo several maturation steps, including homotypic fusion of immature secretory granules (ISGs) and processing of prehormones to yield active peptides. The molecular mechanisms governing secretory granule maturation are largely unknown. Here, we investigate a highly conserved protein named HID-1 in a mouse model. A conditional knockout of HID-1 in pancreatic β cells leads to glucose intolerance and a remarkable increase in the serum proinsulin/insulin ratio caused by defective proinsulin processing. Large volume three-dimensional electron microscopy and immunofluorescence imaging reveal that ISGs are much more abundant in the absence of HID-1. We further demonstrate that HID-1 deficiency prevented secretory granule maturation by blocking homotypic fusion of immature secretory granules. Our data identify a novel player during the early maturation of immature secretory granules. DOI: http://dx.doi.org/10.7554/eLife.18134.001

Ultra-structural study of insulin granules in pancreatic β-cells of db/db mouse by scanning transmission electron microscopy tomography

Insulin granule trafficking is a key step in the secretion of glucose-stimulated insulin from pancreatic β-cells. The main feature of type 2 diabetes (T2D) is the failure of pancreatic β-cells to secrete sufficient amounts of insulin to maintain normal blood glucose levels. In this work, we developed and applied tomography based on scanning transmission electron microscopy (STEM) to image intact insulin granules in the β-cells of mouse pancreatic islets. Using three-dimensional (3D) reconstruction, we found decreases in both the number and the grey level of insulin granules in db/db mouse pancreatic β-cells. Moreover, insulin granules were closer to the plasma membrane in diabetic β-cells than in control cells. Thus, 3D ultra-structural tomography may provide new insights into the pathology of insulin secretion in T2D.

Alterations of the Ca2+ signaling pathway in pancreatic beta-cells isolated from db/db mice

Upon glucose elevation, pancreatic beta-cells secrete insulin in a Ca2+-dependent manner. In diabetic animal models, different aspects of the calcium signaling pathway in beta-cells are altered, but there is no consensus regarding their relative contributions to the development of beta-cell dysfunction. In this study, we compared the increase in cytosolic Ca2+ ([Ca2+]i) via Ca2+ influx, Ca2+ mobilization from endoplasmic reticulum (ER) calcium stores, and the removal of Ca2+ via multiple mechanisms in beta-cells from both diabetic db/db mice and non-diabetic C57BL/6J mice. We refined our previous quantitative model to describe the slow [Ca2+]i recovery after depolarization in beta-cells from db/db mice. According to the model, the activity levels of the two subtypes of the sarco-endoplasmic reticulum Ca2+-ATPase (SERCA) pump, SERCA2 and SERCA3, were severely down-regulated in diabetic cells to 65% and 0% of the levels in normal cells. This down-regulation may lead to a reduction in the Ca2+ concentration in the ER, a compensatory up-regulation of the plasma membrane Na+/Ca2+ exchanger (NCX) and a reduction in depolarization-evoked Ca2+ influx. As a result, the patterns of glucose-stimulated calcium oscillations were significantly different in db/db diabetic beta-cells compared with normal cells. Overall, quantifying the changes in the calcium signaling pathway in db/db diabetic beta-cells will aid in the development of a disease model that could provide insight into the adaptive transformations of beta-cell function during diabetes development.

Bulk-like endocytosis plays an important role in the recycling of insulin granules in pancreatic beta cells

Although bulk endocytosis has been found in a number of neuronal and endocrine cells, the molecular mechanism and physiological function of bulk endocytosis remain elusive. In pancreatic beta cells, we have observed bulk-like endocytosis evoked both by flash photolysis and trains of depolarization. Bulk-like endocytosis is a clathrin-independent process that is facilitated by enhanced extracellular Ca2+ entry and suppressed by the inhibition of dynamin function. Moreover, defects in bulk-like endocytosis are accompanied by hyperinsulinemia in primary beta cells dissociated from diabetic KKAy mice, which suggests that bulk-like endocytosis plays an important role in maintaining the exo-endocytosis balance and beta cell secretory capability.

A unified deep-learning network to accurately segment insulin granules of different animal models imaged under different electron microscopy methodologies

Insulin is important for body metabolism regulation and glucose homeostasis, and its dysregulation often leads to metabolic syndrome (MS) and diabetes. Insulin is normally stored in large dense-core vesicles (LDCVs) in pancreatic beta cells, and significant reductions in the number, size, gray level and density of insulin granules confer diabetes both in mice (Xue et al., 2012) and humans (Masini et al., 2012). Due to the difficulty of obtaining human islet samples, many works use mice as the animal model. However, the architecture of normal islets in humans differs significantly from that of rodents (Cabrera et al., 2006). Beta cells in the mouse islet core are surrounded by the mantle comprising of alpha and delta cells, whereas alpha, beta and delta cells are intermingled in human islets. The structural differences suggest a possible difference in islet function. In this sense, non-human primates such as rhesus monkeys are a better model, as their islets share a similar architecture with humans (Cabrera et al., 2006). The quantitative nature of the insulin granules within monkey islet beta cells, and whether they change during metabolic dysregulation remain to be explored. Under the electron microscope (EM), insulin granules are usually spherical organelles containing an electron dense-core separated from the surrounding membrane by a halo, with a size of ranged from 100–800 nm (MacDonald et al., 2006) in mouse beta cells. This number is estimated as ∼10,000 per beta cell. Because thin-section EM do not necessarily provide the correct spatial coordination of granules within one beta cell, recent years have witnessed the emergence of volumetric electron microscopy techniques such as electron tomography and focused ion beam scanning electron microscopy (FIB-SEM) (Briggman and Bock, 2012). For the first time, we collected three-dimensional images of pancreatic beta cells in wild type (WT) and MS rhesus monkeys with a FIB-SEM and manually annotated granules from a relatively small number of images. Because the morphological and structural natures of insulin granules are important for their optimal function, quantitative and automatic analysis of insulin granules in islets is important. Manually segmenting densely distributed LDCVs is a laborintensive task due to the big datasets brought by saturated and continuous sampling in the lateral and axial axes. Although several semi-automated segmentation methods for rodent LDCVs have been proposed (Diaz et al., 2010), they are built on time-consuming and human-designed features that cannot adapt to micrographs of different magnification and are extremely prone to errors for images with low signalto-noise ratios. Nevertheless, the machine learning field has witnessed a flourishing of “deep-learning” algorithms. Since AlexNet outperformed all other algorithms by a large margin in the ImageNet contest in 2012, a variety of deep-learning methods for image segmentation have been widely used, including the standard convolutional networks (CNN) (Van Valen et al., 2016), and fully convolutional networks (FCN) (Long, 2014). Recently, multi-scale features, dilated convolutions, context encoding, conditional random fields (CRFs) are incorporated to FCNs to improve spatial resolution, bringing more novel and complicated network structures such as Tiramisu (Jegou et al., 2017), Deeplab (Chen et al., 2018). However, different from natural images, we only have a small electron micrograph dataset annotated, in which insulin granules only occupy a small portion of the image. In order to prevent overfitting, we take concise deep-learning networks as the starting point, such as U-Net (Ronneberger et al., 2015). To automatically and precisely segment the insulin granules with different sizes and pattern features in EM images, we built and trained a multi-branch fully convolutional network (MFCN), which consists of three modules: a multi-scale inception module, a multi-branch sampling module, and a multi-scale ensemble module (Fig. 1). The “multi-scale” design was inspired by the naïve inception module (Szegedy et al., 2015), which uses different sizes kernels to extract both coarse and fine grained features. Although some papers have proposed that a stack of two 3 × 3 convolutional