Modeling Alzheimer's Disease Using Human Brain Organoids

Alzheimer's disease (AD) is the primary cause of dementia, to date. The urgent need to understand the biological and biochemical processes related to this condition, as well as the demand for reliable in vitro models for drug screening, has led to the development of novel techniques, among which stem cell methods are of utmost relevance for AD research, particularly the development of human brain organoids. Brain organoids are three-dimensional cellular aggregates derived from induced pluripotent stem cells (iPSCs) that recreate different neural cell interactions and tissue characteristics in culture. Here, we describe the protocol for the generation of brain organoids derived from AD patients and for the analysis of AD-derived pathology. AD organoids can recapitulate beta-amyloid and tau pathological features, making them a promising model for studying the molecular mechanisms underlying disease and for in vitro drug testing.

DYRK1A Regulates the Bidirectional Axonal Transport of APP in Human-Derived Neurons

Dyrk1a triplication in Down's syndrome and its overexpression in Alzheimer's disease suggest a role for increased DYRK1A activity in the abnormal metabolism of APP. Transport defects are early phenotypes in the progression of Alzheimer's disease, which lead to APP processing impairments. However, whether DYRK1A regulates the intracellular transport and delivery of APP in human neurons remains unknown. From a proteomic dataset of human cerebral organoids treated with harmine, a DYRK1A inhibitor, we found expression changes in protein clusters associated with the control of microtubule-based transport and in close interaction with the APP vesicle. Live imaging of APP axonal transport in human-derived neurons treated with harmine or overexpressing a dominant negative DYRK1A revealed a reduction in APP vesicle density and enhanced the stochastic behavior of retrograde vesicle transport. Moreover, harmine increased the fraction of slow segmental velocities and changed speed transitions supporting a DYRK1A-mediated effect in the exchange of active motor configuration. Contrarily, the overexpression of DYRK1A in human polarized neurons increased the axonal density of APP vesicles and enhanced the processivity of retrograde APP. In addition, increased DYRK1A activity induced faster retrograde segmental velocities together with significant changes in slow to fast anterograde and retrograde speed transitions, suggesting the facilitation of the active motor configuration. Our results highlight DYRK1A as a modulator of the axonal transport machinery driving APP intracellular distribution in neurons, and stress DYRK1A inhibition as a putative therapeutic intervention to restore APP axonal transport in Down's syndrome and Alzheimer's disease.SIGNIFICANCE STATEMENT Axonal transport defects are early events in the progression of neurodegenerative diseases, such as Alzheimer's disease. However, the molecular mechanisms underlying transport defects remain elusive. Dyrk1a kinase is triplicated in Down's syndrome and overexpressed in Alzheimer's disease, suggesting that DYRK1A dysfunction affects molecular pathways leading to early-onset neurodegeneration. Here, we show by live imaging of human-derived neurons that DYRK1A activity differentially regulates the intracellular trafficking of APP. Further, single-particle analysis revealed DYRK1A as a modulator of axonal transport and the configuration of active motors within the APP vesicle. Our work highlights DYRK1A as a regulator of APP axonal transport and metabolism, supporting DYRK1A inhibition as a therapeutic strategy to restore intracellular dynamics in Alzheimer's disease.

Nootropic effects of LSD: Behavioral, molecular and computational evidence

The therapeutic use of classical psychedelic substances such as d-lysergic acid diethylamide (LSD) surged in recent years. Studies in rodents suggest that these effects are produced by increased neural plasticity, including stimulation of the mTOR pathway, a key regulator of metabolism, plasticity, and aging. Could psychedelic-induced neural plasticity be harnessed to enhance cognition? Here we show that LSD treatment enhanced performance in a novel object recognition task in rats, and in a visuo-spatial memory task in humans. A proteomic analysis of human brain organoids showed that LSD affected metabolic pathways associated with neural plasticity, including mTOR. To gain insight into the relation of neural plasticity, aging and LSD-induced cognitive gains, we emulated the experiments in rats and humans with a neural network model of a cortico-hippocampal circuit. Using the baseline strength of plasticity as a proxy for age and assuming an increase in plasticity strength related to LSD dose, the simulations provided a good fit for the experimental data. Altogether, the results suggest that LSD has nootropic effects.

EhRho6 mediated actin degradation in Entamoeba histolytica is associated with compromised pathogenicity

Entamoeba histolytica causes amoebiasis which is a major health concern in developing countries. E. histolytica pathogenicity has been implicated to a large repertoire of small GTPases which switch between the inactive GDP bound state and the active GTP bound state with the help of guanine nucleotide exchange factors (GEFs) and GTPase activating protein (GAPs). Rho family of small GTPases are well known to modulate the actin cytoskeletal dynamics which plays a major role in E. histolytica pathogenicity. Here we report an atypical amoebic RhoGEF, and its preferred substrate EhRho6, which, upon overexpression abrogated the pathogenic behavior of the amoeba such as adhesion to host cell, monolayer destruction, erythrophagocytosis, and formation of actin dots. A causative immunoblot analysis revealed actin degradation in the EhRho6 overexpressing trophozoites that could be inhibited by blocking the amoebic proteasomal pathway. A careful analysis of the results from a previously published transcriptomics study, in conjunction with our observations, led to the identification of a clade of Rho GTPases in this pathogenic amoeba which we hypothesize to have implications during the amoebic encystation.

Inhibition of SARS-CoV-2 infection in human iPSC-derived cardiomyocytes by targeting the Sigma-1 receptor disrupts cytoarchitecture and beating

SARS-CoV-2 infects cardiac cells and causes heart dysfunction. Conditions such as myocarditis and arrhythmia have been reported in COVID-19 patients. The Sigma-1 receptor (S1R) is a ubiquitously expressed chaperone that plays a central role in cardiomyocyte function. S1R has been proposed as a therapeutic target because it may affect SARS-CoV-2 replication; however, the impact of the inhibition of S1R in human cardiomyocytes remains to be described. In this study, we investigated the consequences of S1R inhibition in iPSC-derived human cardiomyocytes (hiPSC-CM). SARS-CoV-2 infection in hiPSC-CM was productive and reduced cell survival. S1R inhibition decreased both the number of infected cells and viral particles after 48 hours. S1R inhibition also prevented the release of pro-inflammatory cytokines and cell death. Although the S1R antagonist NE-100 triggered those protective effects, it compromised cytoskeleton integrity by downregulating the expression of structural-related genes and reducing beating frequency. Our findings suggest that the detrimental effects of S1R inhibition in human cardiomyocytes' integrity may abrogate its therapeutic potential against COVID and should be carefully considered.

WIN 55,212-2 shows anti-inflammatory and survival properties in human iPSC-derived cardiomyocytes infected with SARS-CoV-2

Coronavirus disease 2019 (COVID-19) is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which can infect several organs, especially impacting respiratory capacity. Among the extrapulmonary manifestations of COVID-19 is myocardial injury, which is associated with a high risk of mortality. Myocardial injury, caused directly or indirectly by SARS-CoV-2 infection, can be triggered by inflammatory processes that lead to damage to the heart tissue. Since one of the hallmarks of severe COVID-19 is the "cytokine storm", strategies to control inflammation caused by SARS-CoV-2 infection have been considered. Cannabinoids are known to have anti-inflammatory properties by negatively modulating the release of pro-inflammatory cytokines. Herein, we investigated the effects of the cannabinoid agonist WIN 55,212-2 (WIN) in human iPSC-derived cardiomyocytes (hiPSC-CMs) infected with SARS-CoV-2. WIN did not modify angiotensin-converting enzyme II protein levels, nor reduced viral infection and replication in hiPSC-CMs. On the other hand, WIN reduced the levels of interleukins six, eight, 18 and tumor necrosis factor-alpha (TNF-α) released by infected cells, and attenuated cytotoxic damage measured by the release of lactate dehydrogenase (LDH). Our findings suggest that cannabinoids should be further explored as a complementary therapeutic tool for reducing inflammation in COVID-19 patients.

The Age of Brain Organoids: Tailoring Cell Identity and Functionality for Normal Brain Development and Disease Modeling

Over the past years, brain development has been investigated in rodent models, which were particularly relevant to establish the role of specific genes in this process. However, the cytoarchitectonic features, which determine neuronal network formation complexity, are unique to humans. This implies that the developmental program of the human brain and neurological disorders can only partly be reproduced in rodents. Advancement in the study of the human brain surged with cultures of human brain tissue in the lab, generated from induced pluripotent cells reprogrammed from human somatic tissue. These cultures, termed brain organoids, offer an invaluable model for the study of the human brain. Brain organoids reproduce the cytoarchitecture of the cortex and can develop multiple brain regions and cell types. Integration of functional activity of neural cells within brain organoids with genetic, cellular, and morphological data in a comprehensive model for human development and disease is key to advance in the field. Because the functional activity of neural cells within brain organoids relies on cell repertoire and time in culture, here, we review data supporting the gradual formation of complex neural networks in light of cell maturity within brain organoids. In this context, we discuss how the technology behind brain organoids brought advances in understanding neurodevelopmental, pathogen-induced, and neurodegenerative diseases.

The Age of Brain Organoids: Tailoring Cell Identity and Functionality for Normal Brain Development and Disease Modeling

Over the past years, brain development has been investigated in rodent models, which were particularly relevant to establish the role of specific genes in this process. However, the cytoarchitectonic features, which determine neuronal network formation complexity, are unique to humans. This implies that the developmental program of the human brain and neurological disorders can only partly be reproduced in rodents. Advancement in the study of the human brain surged with cultures of human brain tissue in the lab, generated from induced pluripotent cells reprogrammed from human somatic tissue. These cultures, termed brain organoids, offer an invaluable model for the study of the human brain. Brain organoids reproduce the cytoarchitecture of the cortex and can develop multiple brain regions and cell types. Integration of functional activity of neural cells within brain organoids with genetic, cellular, and morphological data in a comprehensive model for human development and disease is key to advance in the field. Because the functional activity of neural cells within brain organoids relies on cell repertoire and time in culture, here, we review data supporting the gradual formation of complex neural networks in light of cell maturity within brain organoids. In this context, we discuss how the technology behind brain organoids brought advances in understanding neurodevelopmental, pathogen-induced, and neurodegenerative diseases.

Non-permissive SARS-CoV-2 infection in human neurospheres

Coronavirus disease 2019 (COVID-19) was initially described as a viral infection of the respiratory tract. It is now known, however, that several other organs are affected, including the brain. Neurological manifestations such as stroke, encephalitis, and psychiatric conditions have been reported in COVID-19 patients, but the neurotropic potential of the virus is still debated. Herein, we sought to investigate SARS-CoV-2 infection in human neural cells. We demonstrated that SARS-CoV-2 infection of neural tissue is non-permissive, however, it can elicit inflammatory response and cell damage. These findings add to the hypothesis that most of the neural damage caused by SARS-CoV-2 infection is due to a systemic inflammation leading to indirect harmful effects on the central nervous system despite the absence of local viral replication.

In vitro antiviral activity of the anti-HCV drugs daclatasvir and sofosbuvir against SARS-CoV-2, the aetiological agent of COVID-19

BACKGROUND

Current approaches of drug repurposing against COVID-19 have not proven overwhelmingly successful and the SARS-CoV-2 pandemic continues to cause major global mortality. SARS-CoV-2 nsp12, its RNA polymerase, shares homology in the nucleotide uptake channel with the HCV orthologue enzyme NS5B. Besides, HCV enzyme NS5A has pleiotropic activities, such as RNA binding, that are shared with various SARS-CoV-2 proteins. Thus, anti-HCV NS5B and NS5A inhibitors, like sofosbuvir and daclatasvir, respectively, could be endowed with anti-SARS-CoV-2 activity.

METHODS

SARS-CoV-2-infected Vero cells, HuH-7 cells, Calu-3 cells, neural stem cells and monocytes were used to investigate the effects of daclatasvir and sofosbuvir. In silico and cell-free based assays were performed with SARS-CoV-2 RNA and nsp12 to better comprehend the mechanism of inhibition of the investigated compounds. A physiologically based pharmacokinetic model was generated to estimate daclatasvir's dose and schedule to maximize the probability of success for COVID-19.

RESULTS

Daclatasvir inhibited SARS-CoV-2 replication in Vero, HuH-7 and Calu-3 cells, with potencies of 0.8, 0.6 and 1.1 μM, respectively. Although less potent than daclatasvir, sofosbuvir alone and combined with daclatasvir inhibited replication in Calu-3 cells. Sofosbuvir and daclatasvir prevented virus-induced neuronal apoptosis and release of cytokine storm-related inflammatory mediators, respectively. Sofosbuvir inhibited RNA synthesis by chain termination and daclatasvir targeted the folding of secondary RNA structures in the SARS-CoV-2 genome. Concentrations required for partial daclatasvir in vitro activity are achieved in plasma at Cmax after administration of the approved dose to humans.

CONCLUSIONS

Daclatasvir, alone or in combination with sofosbuvir, at higher doses than used against HCV, may be further fostered as an anti-COVID-19 therapy.

Non-permissive SARS-CoV-2 infection in human neurospheres

Coronavirus disease 2019 (COVID-19) was initially described as a viral infection of the respiratory tract. It is now known, however, that several other organs are affected, including the brain. Neurological manifestations such as stroke, encephalitis, and psychiatric conditions have been reported in COVID-19 patients, but the neurotropic potential of the virus is still debated. Herein, we sought to investigate SARS-CoV-2 infection in human neural cells. We demonstrated that SARS-CoV-2 infection of neural tissue is non-permissive, however, it can elicit inflammatory response and cell damage. These findings add to the hypothesis that most of the neural damage caused by SARS-CoV-2 infection is due to a systemic inflammation leading to indirect harmful effects on the central nervous system despite the absence of local viral replication.

Quantitative profiling of axonal guidance proteins during the differentiation of human neurospheres

Axon guidance is required for the establishment of brain circuits. Whether much of the molecular basis of axon guidance is known from animal models, the molecular machinery coordinating axon growth and pathfinding in humans remains to be elucidated. The use of induced pluripotent stem cells (iPSC) from human donors has revolutionized in vitro studies of the human brain. iPSC can be differentiated into neuronal stem cells which can be used to generate neural tissue-like cultures, known as neurospheres, that reproduce, in many aspects, the cell types and molecules present in the brain. Here, we analyzed quantitative changes in the proteome of neurospheres during differentiation. Relative quantification was performed at early time points during differentiation using iTRAQ-based labeling and LC-MS/MS analysis. We identified 6438 proteins, from which 433 were downregulated and 479 were upregulated during differentiation. We show that human neurospheres have a molecular profile that correlates to the fetal brain. During differentiation, upregulated pathways are related to neuronal development and differentiation, cell adhesion, and axonal guidance whereas cell proliferation pathways were downregulated. We developed a functional assay to check for neurite outgrowth in neurospheres and confirmed that neurite outgrowth potential is increased after 10 days of differentiation and is enhanced by increasing cyclic AMP levels. The proteins identified here represent a resource to monitor neurosphere differentiation and coupled to the neurite outgrowth assay can be used to functionally explore neurological disorders using human neurospheres as a model.

Single-cell proteomics: A treasure trove in neurobiology

Single-cell analysis came to change the way we look at cell populations. RNA sequencing of single cells allowed us to appreciate the diversity of cell types in the human brain in an unprecedented manner and its power to reveal cell-type specific changes in cell populations has just begun to be explored. In this context, looking at the proteome of single cells promises to bring functional information and contribute to completing the picture. The potential of single cell proteome, in developing a better understanding of the intricate connections between the very diverse cell populations in the brain, is huge. Whereas early approaches to address single-cell proteome have identified hundreds of proteins, today, techniques combining isobaric labelling and LC-MS can lead to the identification of thousands of proteins. In this review, we describe methods which have been used to identify and quantify proteins from single cells and propose that the application of isobaric labeling and label-free quantitative proteomics approach for single-cell analysis is ready to provide useful information for the neurobiology field.

Zika virus infection leads to mitochondrial failure, oxidative stress and DNA damage in human iPSC-derived astrocytes

Zika virus (ZIKV) has been extensively studied since it was linked to congenital malformations, and recent research has revealed that astrocytes are targets of ZIKV. However, the consequences of ZIKV infection, especially to this cell type, remain largely unknown, particularly considering integrative studies aiming to understand the crosstalk among key cellular mechanisms and fates involved in the neurotoxicity of the virus. Here, the consequences of ZIKV infection in iPSC-derived astrocytes are presented. Our results show ROS imbalance, mitochondrial defects and DNA breakage, which have been previously linked to neurological disorders. We have also detected glial reactivity, also present in mice and in post-mortem brains from infected neonates from the Northeast of Brazil. Given the role of glia in the developing brain, these findings may help to explain the observed effects in congenital Zika syndrome related to neuronal loss and motor deficit.

Computational fluid dynamic analysis of physical forces playing a role in brain organoid cultures in two different multiplex platforms

Background

Organoid cultivation in suspension culture requires agitation at low shear stress to allow for nutrient diffusion, which preserves tissue structure. Multiplex systems for organoid cultivation have been proposed, but whether they meet similar shear stress parameters as the regularly used spinner flask and its correlation with the successful generation of brain organoids has not been determined.

Results

Here we used computational fluid dynamics (CFD) to simulate two multiplex culture conditions: steering plates on an orbital shaker and the use of a previously described bioreactor. The bioreactor had low speed and high shear stress regions that may affect cell aggregate growth, depending on volume, whereas the computed variables of the steering plates were closer to those of the spinning flask.

Conclusion

Our protocol improves the initial steps of the standard brain organoid formation, and the produced organoids displayed regionalized brain structures, including retinal pigmented cells. Overall, we conclude that suspension culture on orbital steering plates is a cost-effective practical alternative to previously described platforms for the cultivation of brain organoids for research and multiplex testing.

Electronic supplementary material

The online version of this article (10.1186/s12861-019-0183-y) contains supplementary material, which is available to authorized users.

Retrograde transport of Akt by a neuronal Rab5-APPL1 endosome

Long-distance axonal trafficking plays a critical role in neuronal function and transport defects have been linked to neurodegenerative disorders. Various lines of evidence suggest that the small GTPase Rab5 plays a role in neuronal signaling via early endosomal transport. Here, we characterized the motility of Rab5 endosomes in primary cultures of mouse hippocampal pyramidal cells by live-cell imaging and showed that they exhibit bi-directional long-range motility in axons, with a strong bias toward retrograde transport. Characterization of key Rab5 effectors revealed that endogenous Rabankyrin-5, Rabenosyn-5 and APPL1 are all present in axons. Further analysis of APPL1-positive endosomes showed that, similar to Rab5-endosomes, they display more frequent long-range retrograde than anterograde movement, with the endosomal levels of APPL1 correlated with faster retrograde movement. Interestingly, APPL1-endosomes transport the neurotrophin receptor TrkB and mediate retrograde axonal transport of the kinase Akt1. FRET analysis revealed that APPL1 and Akt1 interact in an endocytosis-dependent manner. We conclude that Rab5-APPL1 endosomes exhibit the hallmarks of axonal signaling endosomes to transport Akt1 in hippocampal pyramidal cells.

Quantitative proteomic analysis identifies proteins and pathways related to neuronal development in differentiated SH-SY5Y neuroblastoma cells

•

Differentiation analysis of SH-SY5Y cells with iTRAQ strategy is proposed.

•

Differentiated SH-SY5Y cells are more appropriated as a neuronal model.

•

Upregulated proteins are mainly related to ECM-interaction and apoptosis.

•

Proteins to explore as differentiation markers: AGRN, EMILIM-1, AIFM, STMN1.

SH-SY5Y neuroblastoma cells are susceptible to differentiation using retinoic acid (RA) and brain-derived neurotrophic factor (BDNF), providing a model of neuronal differentiation. We compared SH-SY5Y cells proteome before and after RA/BDNF treatment using iTRAQ and phosphopeptide enrichment strategies. We identified 5587 proteins, 366 of them with differential abundance. Differentiated cells expressed proteins related to neuronal development, and, undifferentiated cells expressed proteins involved in cell proliferation. Interactive network covered focal adhesion, cytoskeleton dynamics and neurodegenerative diseases processes and regulation of mitogen-activated protein kinase-related signaling pathways; key proteins involved in those processes might be explored as markers for neuronal differentiation.

Application of shotgun proteomics for discovery-driven protein-protein interaction

Affinity purification of protein complexes and identification of co-purified proteins by mass spectrometry is a powerful method to discover novel protein-protein interactions. Application of this method to the study of biological systems often requires the ability to process a large number of samples. Hence, there is great need to generate proteomic workflows compatible with large-scale studies. The major goal of this protocol is to present a fast, reliable, and scalable method to characterize protein complexes by mass spectrometry to overcome the limitations of conventional geLC-MS/MS or MudPIT protocols. This method was successfully employed for the discovery and characterization of novel protein complexes in cultured yeast, mammalian cells, and mice.

The endosomal protein Appl1 mediates Akt substrate specificity and cell survival in vertebrate development

During development of multicellular organisms, cells respond to extracellular cues through nonlinear signal transduction cascades whose principal components have been identified. Nevertheless, the molecular mechanisms underlying specificity of cellular responses remain poorly understood. Spatial distribution of signaling proteins may contribute to signaling specificity. Here, we tested this hypothesis by investigating the role of the Rab5 effector Appl1, an endosomal protein that interacts with transmembrane receptors and Akt. We show that in zebrafish, Appl1 regulates Akt activity and substrate specificity, controlling GSK-3beta but not TSC2. Consistent with this pattern, Appl1 is selectively required for cell survival, most critically in highly expressing tissues. Remarkably, Appl1 function requires its endosomal localization. Indeed, Akt and GSK-3beta, but not TSC2, dynamically associate with Appl1 endosomes upon growth factor stimulation. We propose that partitioning of Akt and selected effectors onto endosomal compartments represents a key mechanism contributing to the specificity of signal transduction in vertebrate development.

TPA-induced signal transduction: a link between PKC and EGFR signaling modulates the assembly of intercellular junctions in Caco-2 cells

Recent studies suggest that signal transduction may have an important role in the development and regulation of the metastatic phenotype. Here, we investigated the role of the epidermal growth factor receptor (EGFR), and protein kinase C (PKC), in the process of reassembly of cadherin-dependent cell-cell adhesion of Caco-2 cells. We used chemical activation of PKC and EGFR with 12- O-tetradecanoylphorbol-13-acetate (TPA), a tumor-promoting agent, pretreatment with protein kinase inhibitors and subcellular fractionation to analyze the effect of the phorbol ester on the redistribution of junctional proteins. Transepithelial resistance (TER), electron microscopy and immunofluorescence analyses were also carried out. Activation with TPA resulted in disassembly of adherens junctions (AJs), but the tight junction (TJ) structure and function remained unaltered. TPA affected E-cadherin levels. In Caco-2 cells at day 2 of culture, when most E-cadherin is not associated with the cytoskeleton, a decrease in the level of this protein was observed as soon as 6 h after TPA addition. However, at day 5 of culture, the major effect observed after 6 h of treatment was a translocation of the protein from the Triton-insoluble to the -soluble fraction. On the other hand, TPA did not significantly affect the E-cadherin-associated proteins alpha and beta-catenins. Potent specific EGFR inhibitors, such as PD153035 and Tyrphostin 25, as well as Calphostin C, an inhibitor of PKC, significantly blocked the effect of TPA on AJs. Furthermore, inhibition of the TPA effect by the PD98059 MAPK inhibitor suggests that activation of this kinase was the final event in the modulation of cadherin-dependent cell-cell adhesion. Pretreatment of cell monolayers with Calphostin C before EGF treatment, one of the ligands of EGFR, blocked the redistribution of E-cadherin caused by EGF. Based on these results, we conclude that both EGFR and PKC activation are involved in TPA-induced cell signaling for modulation of cadherin-dependent cell-cell adhesion and cell shape in Caco-2 cells.

Assembly and functional analysis of tight junction in a colon adenocarcinoma cell line: effect of glucose depletion

We describe morphologic and biochemical changes in the colonic epithelial HCT-116 cell line following depletion of glucose from the culture medium. Cultured cells under permissive differentiation conditions (inosine-supplemented glucose-free medium) exhibited, after confluence, an enterocytic differentiation, in contrast to cells grown under standard culture conditions, where they remain in an undifferentiated state. The differentiated phenotype was characterized by the presence of a monolayer of polarized cells displaying an apical tight junction, and by the presence of alkaline phosphatase, a well known brush border marker. We demonstrated that the formed tight junctions were functional using the following criteria: a) labeling of the junctions with antibodies recognizing the tight juntion proteins occludin and ZO-1, as observed by immunofluorescence and immunoblotting analysis; b) characteristic organization of the tight junction strands, as observed in freeze-fracture replicas; c) increase ofthe transepithelial resistance across the monolayer; d) not permeation of the ruthenium red stain across the tight junction, and e) presence of the hyperphosphorylated form of occludin.