Deep Domain Adversarial Learning for Species-Agnostic Classification of Histologic Subtypes of Osteosarcoma.

Osteosarcomas (OSs) are aggressive bone tumors with many divergent histologic patterns. During pathology review, OSs are subtyped based on the predominant histologic pattern; however, tumors often demonstrate multiple patterns. This high tumor heterogeneity coupled with scarcity of samples compared with other tumor types render histology-based prognosis of OSs challenging. To combat lower case numbers in humans, dogs with spontaneous OSs have been suggested as a model species. Herein, a convolutional neural network was adversarially trained to classify distinct histologic patterns of OS in humans using mostly canine OS data during training. Adversarial training improved domain adaption of a histologic subtype classifier from canines to humans, achieving an average multiclass F1 score of 0.77 (95% CI, 0.74-0.79) and 0.80 (95% CI, 0.78-0.81) when compared with the ground truth in canines and humans, respectively. Finally, this trained model, when used to characterize the histologic landscape of 306 canine OSs, uncovered distinct clusters with markedly different clinical responses to standard-of-care therapy.

Predicting Molecular Subtype and Survival of Rhabdomyosarcoma Patients Using Deep Learning of H&E Images: A Report from the Children's Oncology Group.

PURPOSE: Rhabdomyosarcoma (RMS) is an aggressive soft-tissue sarcoma, which primarily occurs in children and young adults. We previously reported specific genomic alterations in RMS, which strongly correlated with survival; however, predicting these mutations or high-risk disease at diagnosis remains a significant challenge. In this study, we utilized convolutional neural networks (CNN) to learn histologic features associated with driver mutations and outcome using hematoxylin and eosin (H&E) images of RMS. EXPERIMENTAL DESIGN: Digital whole slide H&E images were collected from clinically annotated diagnostic tumor samples from 321 patients with RMS enrolled in Children's Oncology Group (COG) trials (1998-2017). Patches were extracted and fed into deep learning CNNs to learn features associated with mutations and relative event-free survival risk. The performance of the trained models was evaluated against independent test sample data (n = 136) or holdout test data. RESULTS: The trained CNN could accurately classify alveolar RMS, a high-risk subtype associated with PAX3/7-FOXO1 fusion genes, with an ROC of 0.85 on an independent test dataset. CNN models trained on mutationally-annotated samples identified tumors with RAS pathway with a ROC of 0.67, and high-risk mutations in MYOD1 or TP53 with a ROC of 0.97 and 0.63, respectively. Remarkably, CNN models were superior in predicting event-free and overall survival compared with current molecular-clinical risk stratification. CONCLUSIONS: This study demonstrates that high-risk features, including those associated with certain mutations, can be readily identified at diagnosis using deep learning. CNNs are a powerful tool for diagnostic and prognostic prediction of rhabdomyosarcoma, which will be tested in prospective COG clinical trials.

Artificial Intelligence-based Tumor Segmentation in Mouse Models of Lung Adenocarcinoma.

BACKGROUND: Mouse models are highly effective for studying the pathophysiology of lung adenocarcinoma and evaluating new treatment strategies. Treatment efficacy is primarily determined by the total tumor burden measured on excised tumor specimens. The measurement process is time-consuming and prone to human errors. To address this issue, we developed a novel deep learning model to segment lung tumor foci on digitally scanned hematoxylin and eosin (H&E) histology slides. METHODS: Digital slides of 239 mice from 9 experimental cohorts were split into training (n=137), validation (n=37), and testing cohorts (n=65). Image patches of 500×500 pixels were extracted from 5× and 10× magnifications, along with binary masks of expert annotations representing ground-truth tumor regions. Deep learning models utilizing DeepLabV3+ and UNet architectures were trained for binary segmentation of tumor foci under varying stain normalization conditions. The performance of algorithm segmentation was assessed by Dice Coefficient, and detection was evaluated by sensitivity and positive-predictive value (PPV). RESULTS: The best model on patch-based validation was DeepLabV3+ using a Resnet-50 backbone, which achieved Dice 0.890 and 0.873 on validation and testing cohort, respectively. This result corresponded to 91.3 Sensitivity and 51.0 PPV in the validation cohort and 93.7 Sensitivity and 51.4 PPV in the testing cohort. False positives could be reduced 10-fold with thresholding artificial intelligence (AI) predicted output by area, without negative impact on Dice Coefficient. Evaluation at various stain normalization strategies did not demonstrate improvement from the baseline model. CONCLUSIONS: A robust AI-based algorithm for detecting and segmenting lung tumor foci in the pre-clinical mouse models was developed. The output of this algorithm is compatible with open-source software that researchers commonly use.

MoNuSAC2020: A Multi-Organ Nuclei Segmentation and Classification Challenge.

Detecting various types of cells in and around the tumor matrix holds a special significance in characterizing the tumor micro-environment for cancer prognostication and research. Automating the tasks of detecting, segmenting, and classifying nuclei can free up the pathologists' time for higher value tasks and reduce errors due to fatigue and subjectivity. To encourage the computer vision research community to develop and test algorithms for these tasks, we prepared a large and diverse dataset of nucleus boundary annotations and class labels. The dataset has over 46,000 nuclei from 37 hospitals, 71 patients, four organs, and four nucleus types. We also organized a challenge around this dataset as a satellite event at the International Symposium on Biomedical Imaging (ISBI) in April 2020. The challenge saw a wide participation from across the world, and the top methods were able to match inter-human concordance for the challenge metric. In this paper, we summarize the dataset and the key findings of the challenge, including the commonalities and differences between the methods developed by various participants. We have released the MoNuSAC2020 dataset to the public.

Multiresolution Application of Artificial Intelligence in Digital Pathology for Prediction of Positive Lymph Nodes From Primary Tumors in Bladder Cancer.

PURPOSE: To develop an artificial intelligence (AI)-based model for identifying patients with lymph node (LN) metastasis based on digital evaluation of primary tumors and train the model using cystectomy specimens available from The Cancer Genome Atlas (TCGA) Project; patients from our institution were included for validation of the leave-out test cohort. METHODS: In all, 307 patients were identified for inclusion in the study (TCGA, n = 294; in-house, n = 13). Deep learning models were trained from image patches at 2.5×, 5×, 10×, and 20× magnifications, and spatially resolved prediction maps were combined with microenvironment (lymphocyte infiltration) features to derive a final patient-level AI score (probability of LN metastasis). Training and validation included 219 patients (training, n = 146; validation, n = 73); 89 patients (TCGA, n = 75; in-house, n = 13) were reserved as an independent testing set. Multivariable logistic regression models for predicting LN status based on clinicopathologic features alone and a combined model with AI score were fit to training and validation sets. RESULTS: Several patients were determined to have positive LN metastasis in TCGA (n = 105; 35.7%) and in-house (n = 3; 23.1%) cohorts. A clinicopathologic model that considered using factors such as age, T stage, and lymphovascular invasion demonstrated an area under the curve (AUC) of 0.755 (95% CI, 0.680 to 0.831) in the training and validation cohorts compared with the cross validation of the AI score (likelihood of positive LNs), which achieved an AUC of 0.866 (95% CI, 0.812 to 0.920; P = .021). Performance in the test cohort was similar, with a clinicopathologic model AUC of 0.678 (95% CI, 0.554 to 0.802) and an AI score of 0.784 (95% CI, 0.702 to 0.896; P = .21). In addition, the AI score remained significant after adjusting for clinicopathologic variables (P = 1.08 × 10-9), and the combined model significantly outperformed clinicopathologic features alone in the test cohort with an AUC of 0.807 (95% CI, 0.702 to 0.912; P = .047). CONCLUSION: Patients who are at higher risk of having positive LNs during cystectomy can be identified on primary tumor samples using novel AI-based methodologies applied to digital hematoxylin and eosin-stained slides.

Spatial density and diversity of architectural histology in prostate cancer: influence on diffusion weighted magnetic resonance imaging.

BACKGROUND: To assess the influence of specific histopathologic patterns on MRI diffusion characteristics by performing rigorous whole-mount/imaging registration and correlating histologic architectures observed in prostate cancer with diffusion characteristics in prostate MRIs. METHODS: Fifty-two whole-mount pathology blocks from 15 patients who underwent multiparametric MRI (mpMRI) at a single institution prior to radical prostatectomy were retrospectively analyzed. Regions containing individual morphologic patterns (N=21 patterns, including variations of cribriforming, expansile sheets, single cells, patterns of early intraluminal complexity, and mucin rupture patterns) were digitally annotated by an expert genitourinary pathologist. Distinct tumor foci on each slide were also assigned a Gleason grade and scored as having any high-risk histologic pattern. Digital sections were aligned to MRI using a patient-specific mold and registered using local mean weighted piecewise transformation based on anatomic control points. Density and presence of morphological patterns was correlated to apparent diffusion coefficient (ADC) signal intensity using mixed effects model accounting for nested intra-foci, intra-patient correlation. Influence of intra-tumoral heterogeneity was assessed by affinity propagation clustering (APC) of morphology features and correlated to foci- and cluster-level ADC metrics. RESULTS: One hundred eleven distinct tumor foci were evaluated. Beta diversity, reflecting average morphology representation across inter- and intra-foci areas, demonstrated higher intra-tumor diversity within high-risk foci (P<0.05). ADC signal demonstrated an inverse correlation with foci-level Gleason grade (P>0.05), which was strengthened in cluster-level analysis for intra-foci regions containing high-risk morphologies (P=0.017). In voxel-based analysis, dense regions demonstrate lower ADC, but the presence and density for each morphology influenced ADC independently (ANOVA P<0.001). CONCLUSIONS: Architectural features influence ADC characteristics of MRI, with more complex tumors having lower ADC values regulated by presence and density of specific morphologies.

Apical periurethral transition zone lesions: MRI and histology findings.

PURPOSE: Apical periurethral transition zone (TZ) cancers can pose unique problems for surgery and radiation therapy. Here, we describe the appearance of such cancers on multiparametric MRI (mpMRI) and correlate this with histopathology derived from MRI-targeted biopsy. MATERIALS AND METHODS: Between May 2011 and January 2019, a total of 4381 consecutive patients underwent 3 T mpMRI. Of these, 53 patients with 58 apical periurethral TZ lesions underwent TRUS/MRI fusion-guided biopsy and 12-core systematic TRUS-guided biopsy. Correlation was made with patient age, PSA, PSA density, whole prostate volume, and Gleason scores. RESULTS: A total 53 men (median age 68 years, median PSA 7.94 ng/ml) were identified as having at least one apical periurethral TZ lesion on mpMRI and 5 (9%) patients had more than one apical periurethral lesion. Thus, 58 lesions were identified in 53 patients. Of these 37/53 patients (69%) and 40/58 lesions were positive at biopsy for prostate cancer. Seven were diagnosed by 12-core systematic TRUS-guided biopsy and 34 were diagnosed by TRUS/MRI fusion-guided biopsy. Gleason score was ≥ 3 + 4 in 34/58 (58%) lesions. CONCLUSION: Identification of apical periurethral TZ prostate cancers is important to help guide surgical and radiation therapy as these tumors are adjacent to critical structures. Because of the tendency to undersample the periurethral zone during TRUS biopsy, MRI-guided biopsy is particularly helpful for detecting apical periurethral TZ prostate cancers many of which prove to be clinically significant.

Telepathology at the Armed Forces Institute of Pathology: A Retrospective Review of Consultations From 1996 to 1997.

CONTEXT: - Telepathology is the practice of pathology at a distance, transmitting images using telecommunication methods for second opinion and/or diagnostic assistance, or for educational purposes. It may be the only means of consultation for some pathologists. OBJECTIVE: - To retrospectively review and evaluate a subset of telepathology consultations from June 1996 to March 1997, and to determine the concordance between the telepathology diagnosis of the contributor and pathologists at the Armed Forces Institute of Pathology (AFIP), Washington, District of Columbia, as well as the concordance between the telepathology diagnosis and the glass slide diagnosis, when available. DESIGN: - Photocopies of de-identified telepathology reports from the AFIP during a 15-month period between June 1996 and March 1997 were reviewed. Contributor versus telepathology diagnosis was graded as 1 (complete agreement), 2 (partial agreement), 3 (disagreement; usually a diagnosis of benign versus malignant), and deferred. Data were analyzed using descriptive statistical methods. RESULTS: - Of the 262 cases, 194 (74%) were in complete agreement with the contributor's diagnosis, 34 of 262 (13%) were in minor disagreement, and 21 of 262 (8%) were in major disagreement. Diagnoses were deferred in 5% (13 of 262) of cases. CONCLUSIONS: - Using commercial off-the-shelf technology and despite telecommunication challenges during that time, the AFIP demonstrated that telepathology could be conducted reliably.

Evaluation of Android Smartphones for Telepathology.

BACKGROUND: In the year 2014, Android smartphones accounted for one-third of mobile connections globally but are predicted to increase to two-thirds by 2020. In developing countries, where teleconsultations can benefit health-care providers most, the ratio is even higher. This study compared the use of two Android phones, an 8 megapixel (MP) and a 16 MP phone, for capturing microscopic images. METHOD: The Android phones were used to capture images and videos of a gastrointestinal biopsy teaching set of referred cases from the Armed Forces Institute of Pathology (AFIP). The acquired images and videos were reviewed online by two pathologists for image quality, adequacy for diagnosis, usefulness of video overviews, and confidence in diagnosis, on a 5-point Likert scale. RESULTS: The results show higher means in a 5-point Likert scale for the 8 MP versus the 16 MP phone that were statistically significant in adequacy of images (4.0 vs. 3.75) for rendering diagnosis and for agreement with the reference diagnosis (2.33 vs. 2.07). Although the quality of images was found higher in the 16 MP phone (3.8 vs. 3.65), these were not statistically significant. Adding video images of the entire specimen was found to be useful for evaluating the slides (combined mean, 4.0). CONCLUSION: For telepathology and other image dependent practices in developing countries, Android phones could be a useful tool for capturing images.

Hepatoid adenocarcinoma of the lung metastasizing to the tonsil.

Hepatoid adenocarcinoma (HAC) is a rare tumor that typically originates in gastrointestinal tissue, such as the stomach, but can also occasionally originate in the lung. The majority of HACs express α-fetoprotein (AFP) on tumor cells, and serum AFP can be used as a marker of response to treatment or disease progression. HAC has a poor prognosis, although early diagnosis and aggressive treatment can result in long-term survival. To the best of our knowledge, this is the first reported case of HAC metastasizing to the tonsil. Irradiation using intensity-modulated radiation therapy (IMRT) to 30 Gray (Gy) achieved a significant, and durable tumor response. IMRT can be considered for local control of HAC at other metastatic sites of disease.

Clinical Characteristics of Connective Tissue Nevi in Tuberous Sclerosis Complex With Special Emphasis on Shagreen Patches.

Importance: Patients with tuberous sclerosis complex (TSC) frequently develop collagenous connective tissue nevi. The prototypical lesion is a large shagreen patch located on the lower back, but some patients only manifest small collagenomas or have lesions elsewhere on the body. The ability to recognize these variable presentations can be important for the diagnosis of TSC. Objective: To describe the clinical characteristics of connective tissue nevi on the trunk and extremities of patients with tuberous sclerosis complex. Design, Setting, and Participants: A retrospective analysis of patient medical records and skin photography was performed; 104 adult patients with TSC were enrolled in an observational cohort study that was enriched for those with pulmonary lymphangioleiomyomatosis, and was therefore composed mostly of women (99 women, 5 men). All patients included were examined at the National Institutes of Health (NIH) in Bethesda, Maryland, from 1998 to 2013. Connective tissue nevi were categorized per anatomic location and size. Lesions less than 1 cm in diameter were termed collagenomas. Shagreen patches were characterized as small (1 to <4 cm), medium (4 to <8 cm), and large (≥8 cm). Main Outcome and Measures: Frequency, anatomic location, size, and histological appearance of connective tissue nevi in patients with TSC. Results: Overall, 58 of 104 patients (median [range] age, 42 [19-70] years) with TSC (56%) had at least 1 connective tissue nevus on the trunk or thighs; of these, 28 of 58 patients (48%) had a solitary lesion, and 30 of 58 patients (52%) had 2 or more lesions. Overall, 120 lesions from 55 patients were classified by size; 46 lesions (38%) were collagenomas; 39 lesions (32%) were small shagreen patches; 21 lesions (18%), medium shagreen patches; and 14 lesions (12%), large shagreen patches. The distribution of lesions was 9% (n = 11), upper back; 29% (n = 35), middle back; 51% (n = 61), lower back; and 11% (n = 13), other locations. All 26 shagreen patches that were analyzed histopathologically had coarse collagen fibers and 24 of 26 stained with Miller elastic stain had decreased elastic fibers. On immunoblot analysis, fibroblasts grown from shagreen patches expressed higher levels of phosphorylated ribosomal protein S6 than paired fibroblasts from normal-appearing skin. Conclusions and Relevance: Tuberous sclerosis complex-related connective tissue nevi are not limited to the lower back, and occasionally present on the central or upper back, buttocks, or thighs. Elastic fibers are typically decreased. Recognition of these variable presentations can be important for TSC diagnosis.

Lipopolysaccharide stimulates platelets through an IL-1ß autocrine loop.

LPS activates platelets through TLR4, aiding productive sepsis, with stimulated splicing and translation of stored heteronuclear pro-IL-1ß RNA. Although the IL-1R type 1 (IL-1R1) receptor for IL-1 shares downstream components with the TLR4 receptor, platelets are not known to express IL-1R1, nor are they known to respond to this cytokine. We show by flow cytometry and Western blotting that platelets express IL-1R1, and that IL-1ß and IL-1α stimulate heteronuclear I-1ß splicing and translation of the newly made mRNA in platelets. Platelets also respond to the IL-1ß they make, which is exclusively associated with shed microparticles. Specific blockade of IL-1R1 with IL-1R antagonist suppressed platelet stimulation by IL-1, so IL-1ß stimulates its own synthesis in an autocrine signaling loop. Strikingly, IL-1R antagonist inhibition, pharmacologic or genetic suppression of pro-IL-1ß processing to active cytokine by caspase-1, or blockade of de novo protein synthesis also blocked LPS-induced IL-1ß mRNA production. Robust stimulation of platelets by LPS therefore also required IL-1ß amplification. Activated platelets made IL-1ß in vivo as IL-1ß rapidly accumulated in occluded murine carotid arteries by posttranscriptional RNA splicing unique to platelets. We conclude that IL-1ß is a platelet agonist, that IL-1ß acts through an autocrine stimulatory loop, that an IL-1ß autocrine loop is required to amplify platelet activation by LPS, and that platelets immobilized in occlusive thrombi are activated over time to produce IL-1ß. IL-1 is a new platelet agonist that promotes its own synthesis, connecting thrombosis with immunity.

Lipopolysaccharide signaling without a nucleus: kinase cascades stimulate platelet shedding of proinflammatory IL-1ß-rich microparticles.

Platelets contain unspliced heteronuclear IL-1ß RNA, which is rapidly spliced and translated upon activation. LPS is a superior agonist for this atypical platelet response, but how LPS induces proinflammatory cytokine production in anucleate cells lacking NF-κB is unknown. Platelets express functional TLR4, and stimulation by LPS induced rapid splicing, translation, and secretion of mature IL-1ß after caspase-1 processing. LPS stimulated microparticle shedding, and secreted IL-1ß was exclusively present in these particles. Microparticles from LPS-stimulated platelets induced VCAM-1 production by cultured human endothelial cells, and blockade of endothelial IL-1ß receptor with IL-1 receptor antagonist completely suppressed endothelial activation. Splicing was posttranscriptional as the SR kinase inhibitor TG003 blocked IL-1ß RNA production by platelets, but not by monocytes, and was dependent on exogenous CD14--a property of platelets. We used a combination of small-molecule inhibitors, cell-penetrating chimeric peptide inhibitors, and gene-targeted animals to show splicing required MyD88 and TIRAP, and IRAK1/4, Akt, and JNK phosphorylation and activation. Traf6 couples MyD88 to the Akt pathway and, remarkably, a Traf6 interacting peptide-antennapedia chimera was more effective than LPS in stimulating IL-1ß splicing. The Traf6 chimera did not, however, stimulate microparticle shedding, nor was IL-1ß released. We conclude LPS-induced kinase cascades are sufficient to alter cellular responses, that three signals emanate from platelet TLR4, and that Akt and JNK activation are sufficient to initiate posttranscriptional splicing while another event couples microparticle shedding to TLR4 activation. Platelets contribute to the inflammatory response to LPS through production of microparticles that promote endothelial cell activation.

Lipopolysaccharide is a direct agonist for platelet RNA splicing.

Platelets express TLR4 receptors, but its ligand LPS does not directly activate thrombotic functions nor, obviously, transcription by these anucleate cells. Platelets, however, store information that changes their phenotype over a few hours in the form of unprocessed RNA transcripts. We show even low concentrations of LPS in the presence of soluble CD14 initiated splicing of unprocessed IL-1beta RNA, with translation and accumulation of IL-1beta protein. LPS was a more robust agonist for this response than thrombin. Platelets also contained cyclooxygenase-2 pre-mRNA, which also was spliced and translated after LPS stimulation. Flow cytometry and immunocytochemistry of platelets extensively purified by negative immunodepletion showed platelets contained IL-1beta, and quantitative assessment of white blood cell contamination by CD14 real time PCR confirms that leukocytes were not the IL-1beta source, nor were they required for platelet stimulation. LPS did not initiate rapid platelet responses, but over time did prime platelet aggregation to soluble agonists, induced actin rearrangement, and initiated granule secretion with P-selectin expression that resulted the coating of quiescent leukocytes with activated platelets. LPS is a direct agonist for platelets that allows these cells to directly participate in the innate immune response to bacteria.

Myosin II regulates the shape of three-dimensional intestinal epithelial cysts.

The development of luminal organs begins with the formation of spherical cysts composed of a single layer of epithelial cells. Using a model three-dimensional cell culture, this study examines the role of a cytoskeletal motor, myosin II, in cyst formation. Caco-2 and SK-CO15 intestinal epithelial cells were embedded into Matrigel, and myosin II was inhibited by blebbistatin or siRNA-mediated knockdown. Whereas control cells formed spherical cysts with a smooth surface, inhibition of myosin II induced the outgrowth of F-actin-rich surface protrusions. The development of these protrusions was abrogated after inhibition of F-actin polymerization or of phospholipase C (PLC) activity, as well as after overexpression of a dominant-negative ADF/cofilin. Surface protrusions were enriched in microtubules and their formation was prevented by microtubule depolymerization. Myosin II inhibition caused a loss of peripheral F-actin bundles and a submembranous extension of cortical microtubules. Our findings suggest that inhibition of myosin II eliminates the cortical F-actin barrier, allowing microtubules to reach and activate PLC at the plasma membrane. PLC-dependent stimulation of ADF/cofilin creates actin-filament barbed ends and promotes the outgrowth of F-actin-rich protrusions. We conclude that myosin II regulates the spherical shape of epithelial cysts by controlling actin polymerization at the cyst surface.

A key claudin extracellular loop domain is critical for epithelial barrier integrity.

Intercellular tight junctions (TJs) regulate epithelial barrier properties. Claudins are major structural constituents of TJs and belong to a large family of tetra-spanning membrane proteins that have two predicted extracellular loops (ELs). Given that claudin-1 is widely expressed in epithelia, we further defined the role of its EL domains in determining TJ function. The effects of several claudin-1 EL mimetic peptides on epithelial barrier structure and function were examined. Incubation of model human intestinal epithelial cells with a 27-amino acid peptide corresponding to a portion of the first EL domain (Cldn-1(53-80)) reversibly interfered with epithelial barrier function by inducing the rearrangement of key TJ proteins: occludin, claudin-1, junctional adhesion molecule-A, and zonula occludens-1. Cldn-1(53-80) associated with both claudin-1 and occludin, suggesting both the direct interference with the ability of these proteins to assemble into functional TJs and their close interaction under physiological conditions. These effects were specific for Cldn-1(53-80), because peptides corresponding to other claudin-1 EL domains failed to influence TJ function. Furthermore, the oral administration of Cldn-1(53-80) to rats increased paracellular gastric permeability. Thus, the identification of a critical claudin-1 EL motif, Cldn-1(53-80), capable of regulating TJ structure and function, offers a useful adjunct to treatments that require drug delivery across an epithelial barrier.

Organized migration of epithelial cells requires control of adhesion and protrusion through Rho kinase effectors.

Migration of epithelial cell sheets, a process involving F-actin restructuring through Rho family GTPases, is both physiologically and pathophysiologically important. Our objective was to clarify the mechanisms whereby the downstream RhoA effector Rho-associated coil-coil-forming kinase (ROCK) influences coordinated epithelial cell motility. Although cells exposed to a pharmacological ROCK inhibitor (Y-27632) exhibited increased spreading in wound closure assays, they failed to migrate in a cohesive manner. Two main phenomena were implicated: the formation of aberrant protrusions at the migrating front and the basal accumulation of F-actin aggregates. Aggregates reflected increased membrane affiliation and detergent insolubility of the actin-binding protein ezrin and enhanced coassociation of ezrin with the membrane protein CD44. While F-actin aggregation following ROCK inhibition was recapitulated by inhibiting myosin light chain (MLC) phosphorylation with the MLC kinase inhibitor ML-7, the latter did not influence protrusiveness and, in fact, significantly decreased cell migration. Our results suggest that excessive protrusiveness downstream of ROCK inhibition reflects an influence of ROCK on F-actin stability via LIM kinase 1 (LIMK-1), which phosphorylates and inactivates cofilin. Y-27632 reduced the levels of both active LIMK-1 and inactive cofilin (phospho forms), and expression of a dominant negative LIMK-1 mutant stimulated leading edge protrusiveness. Furthermore, Y-27632-induced protrusions were partially reversed by overexpression of LIMK-1 to restore cofilin phosphorylation. In summary, our results provide new evidence suggesting that adhesive and protrusive events involved in organized epithelial motility downstream of ROCK are separately coordinated through the phosphorylation of (respectively) MLC and cofilin.

Multiple protein interactions involving proposed extracellular loop domains of the tight junction protein occludin.

Occludin is a tetraspan integral membrane protein in epithelial and endothelial tight junction (TJ) structures that is projected to have two extracellular loops. We have used peptides emulating central regions of human occludin's first and second loops, termed O-A:101-121 and O-B:210-228, respectively, to examine potential molecular interactions between these two regions of occludin and other TJ proteins. A superficial biophysical assessment of A:101-121 and O-B:210-228 showed them to have dissimilar solution conformation characteristics. Although O-A:101-121 failed to strongly interact with protein components of the human epithelial intestinal cell line T84, O-B:210-228 selectively associated with occludin, claudin-one and the junctional adhesion molecule (JAM)-A. Further, the presence of O-B:210-228, but not O-A:101-121, impeded the recovery of functional TJ structures. A scrambled peptide sequences of O-B:210-228 failed to influence TJ assembly. These studies demonstrate distinct properties for these two extracellular segments of the occludin protein and provide an improved understanding of how specific domains of occludin may interact with proteins present at TJ structures.