Diagnosis of diabetic kidney disease in whole slide images via AI-driven quantification of pathological indicators.

Diagnosis of diabetic kidney disease (DKD) mainly relies on screening the morphological variations and internal lesions of glomeruli from pathological kidney biopsy. The prominent pathological alterations of glomeruli for DKD include glomerular hypertrophy and nodular mesangial sclerosis. However, the qualitative judgment of these alterations is inaccurate and inconstant due to the intra- and inter-subject variability of pathologists. It is necessary to design artificial intelligence (AI) methods for accurate quantification of these pathological alterations and outcome prediction of DKD. In this work, we present an AI-driven framework to quantify the volume of glomeruli and degree of nodular mesangial sclerosis, respectively, based on an instance segmentation module and a novel weakly supervised Macro-Micro Aggregation (MMA) module. Subsequently, we construct classic machine learning models to predict the degree of DKD based on three selected pathological indicators via factor analysis. These corresponding modules are trained and tested on a total of 281 whole slide images (WSIs) digitized from two hospitals with different scanners. Our designed AI framework achieved inspiring results with 0.926 mIoU for glomerulus segmentation, and 0.899 F1 score for glomerulus classification in the external testing dataset. Meantime, the visualized results of the MMA module could reflect the location of the lesions. The performance of predicting disease achieved the F1 score of 0.917, which further proved the effectiveness of our AI-driven quantification of pathological indicators. Additionally, the interpretation of the machine learning model with the SHAP method showed similar accordance with the development of DKD in pathology. In conclusion, the proposed auxiliary diagnostic technologies have the feasibility for quantitative analysis of glomerular pathological tissues and alterations in DKD. Pathological quantitative indicators will also make it more convenient to provide doctors with assistance in clinical practice.

Accurate Kidney Pathological Image Classification Method Based on Deep Learning and Multi-Modal Fusion Method with Application to Membranous Nephropathy.

Membranous nephropathy is one of the most prevalent conditions responsible for nephrotic syndrome in adults. It is clinically nonspecific and mainly diagnosed by kidney biopsy pathology, with three prevalent techniques: light microscopy, electron microscopy, and immunofluorescence microscopy. Manual observation of glomeruli one by one under the microscope is very time-consuming, and there are certain differences in the observation results between physicians. This study makes use of whole-slide images scanned by a light microscope as well as immunofluorescence images to classify patients with membranous nephropathy. The framework mainly includes a glomerular segmentation module, a confidence coefficient extraction module, and a multi-modal fusion module. This framework first identifies and segments the glomerulus from whole-slide images and immunofluorescence images, and then a glomerular classifier is trained to extract the features of each glomerulus. The results are then combined to produce the final diagnosis. The results of the experiments show that the F1-score of image classification results obtained by combining two kinds of features, which can reach 97.32%, is higher than those obtained by using only light-microscopy-observed images or immunofluorescent images, which reach 92.76% and 93.20%, respectively. Experiments demonstrate that considering both WSIs and immunofluorescence images is effective in improving the diagnosis of membranous nephropathy.

Molecular profiling of gene fusions in soft tissue sarcomas by Ion AmpliSeqTM: a study of 35 cases.

Background: The accurate diagnosis of sarcoma can be difficult as there are over 70 different subtypes. While molecular profiling in soft tissue sarcoma (STS) has gradually been incorporated into routine diagnostics, conventional methods such as fluorescence in situ hybridization (FISH), reverse transcriptase-PCR (RT-PCR), and Sanger sequencing have several drawbacks. By allowing simultaneous analysis of multiple targets and increasing sequencing depth to achieve ultra-sensitivity, next-generation sequencing (NGS) can not only detect common genetic abnormalities without prior assumptions but also identify uncommon or even new variants. Methods: In this study, the applicability of NGS in assessing STS using the Ion Torrent Proton was evaluated and compared with other methods. A cohort of 35 tissue specimens from STS patients, including alveolar soft-part sarcoma (ASPS), Ewing's sarcoma (ES), synovial sarcoma (SS), dermatofibrosarcoma protuberans (DFSP), and myxoid liposarcoma (MLPS) patients, were subjected to NGS by an Ion AmpliSeqTM Custom panel. Results: A proportion of 97.14% (34/35) were successfully conducted to detect gene fusion positive events and met all criteria for good quality. The concordance between NGS and conventional techniques was 94.12% (32/34). NGS also showed superior results, as Sanger sequencing and FISH in two cases were false negatives, demonstrating the excellent diagnostic utility of NGS for translocation detection in STS. Conclusions: The results in this study show the potential for NGS to aid in diagnosis and clinical monitoring of STS and warrant additional studies in larger cohorts.

Classification of renal biopsy direct immunofluorescence image using multiple attention convolutional neural network.

BACKGROUND AND OBJECTIVES: Direct immunofluorescence (DIF) is an important medical evaluation tool for renal pathology. In the DIF images, the deposition appearances and locations of immunoglobulin on glomeruli involve immunological characteristics of glomerulonephritis and thus can be used to aid in the identification of glomerulonephritis disease. Manual classification to such deposition patterns is time consuming and may lead to significant inter and intra operator variances. We wanted to automate the identification and fusion of deposition location and deposition appearance to assist physicians in achieving immunofluorescence reporting. METHODS: In this paper, we propose a framework that consists of a pre-segmentation module and a classification module for automatically segmenting glomerulus object and classifying the deposition pattern of immunoglobulin on glomerulus object. For the pre-segmentation module, the glomerulus object is segmented out from the acquired DIF images using a segmentation network, which excludes other tissues and makes the classification module focus on the glomerulus. For the classification module, two branches of classifying deposition region and appearance, respectively, are formed by using multiple attentions convolutional neural network (MANet) based on the segmented images, and the classification results of the two pre-trained classification networks are fused with labels. RESULTS: Experimental results show that the proposed framework achieves a high classification performance with an accuracy of 98% and 95% in terms of deposition region and appearance, respectively. The label fusion of deposition appearance and deposition classification is achieved with high accuracy based on well-trained classification. CONCLUSIONS: The data show that automated and accurate patterned immunofluorescence report generation is achieved, which can effectively help improve the diagnosis of autoimmune kidney disease.

Inhibition of NLRP3 inflammasome ameliorates podocyte damage by suppressing lipid accumulation in diabetic nephropathy.

BACKGROUND: Nucleotide leukin-rich polypeptide 3 (NLRP3) inflammasome is documented as a potent target for treating metabolic diseases and inflammatory disorders. Our recent work demonstrated that inhibition of NLRP3 inflammasome activation inhibits renal inflammation and fibrosis in diabetic nephropathy. This study was to investigate the effect of NLRP3 inflammasome on podocyte injury and the underlying mechanism in diabetic nephropathy. METHODS: In vivo, db/db mice were treated with MCC950, a NLRP3 inflammasome specific inhibitor. NLRP3 knockout (NKO) mice were induced to diabetes by intraperitoneal injections of streptozotocin (STZ). We assessed renal function, albuminuria, podocyte injury and glomerular lipid accumulation in diabetic mice. In vitro, apoptosis, cytoskeleton change, lipid accumulation, NF-κB p65 activation and reactive oxygen species (ROS) generation were evaluated in podocytes interfered with NLRP3 siRNA or MCC950 under high glucose (HG) conditions. In addition, the effect and mechanism of IL-1ß on lipid accumulation was explored in podocytes exposed to normal glucose (NG) or HG. RESULTS: MCC950 treatment improved renal function, attenuated albuminuria, mesangial expansion, podocyte loss, as well as glomerular lipid accumulation in db/db mice. The diabetes-induced podocyte loss and glomerular lipid accumulation were reversed in NLRP3 knockout mice. The increased expression of sterol regulatory element-binding protein1 (SREBP1) and SREBP2, and decreased expression of ATP-binding cassette A1 (ABCA1) in podocytes were reversed by MCC950 treatment or NLRP3 knockout in diabetic mice. In vitro, NLRP3 siRNA or MCC950 treatment markedly inhibited HG-induced apoptosis, cytoskeleton change, lipid accumulation, NF-κB p65 activation, and mitochondrial ROS production in cultured podocytes. In addition, BAY11-7082 or tempol treatment inhibited HG-induced lipid accumulation in podocytes. Moreover, exposure of IL-1ß to podocytes induced lipid accumulation, NF-κB p65 activation and mitochondrial ROS generation. CONCLUSION: Inhibition of NLRP3 inflammasome protects against podocyte damage through suppression of lipid accumulation in diabetic nephropathy. IL-1ß/ROS/NF-κB p65 mediates diabetes-associated lipid accumulation in podocytes. The suppression of NLRP3 inflammasome activation may be an effective therapeutic approach to diabetic nephropathy.

Prognostic significance of TOP2A in non-small cell lung cancer revealed by bioinformatic analysis.

BACKGROUND: Lung cancer has been a common malignant tumor with a leading cause of morbidity and mortality, current molecular targets are woefully lacking comparing to the highly progressive cancer. The study is designed to identify new prognostic predictors and potential gene targets based on bioinformatic analysis of Gene Expression Omnibus (GEO) database. METHODS: Four cDNA expression profiles GSE19188, GSE101929, GSE18842 and GSE33532 were chosen from GEO database to analyze the differently expressed genes (DEGs) between non-small cell lung cancer (NSCLC) and normal lung tissues. After the DEGs functions were analyzed, the protein-protein interaction network (PPI) of DEGs were constructed, and the core gene in the network which has high connectivity degree with other genes was identified. We analyzed the association of the gene with the development of NSCLC as well as its prognosis. Lastly we explored the conceivable signaling mechanism of the gene regulation during the development of NSCLC. RESULTS: A total of 92 up regulated and 214 down regulated DEGs were shared in four cDNA expression profiles. Based on their PPI network, TOP2A was connected with most of other genes and was selected for further analysis. Kaplan-Meier overall survival analysis (OS) revealed that TOP2A was associated with worse NSCLC patients survival. And both GEPIA analysis and immunohistochemistry experiment (IHC) confirmed that TOP2A was aberrant gain of expression in cancer comparing to normal tissues. The clinical significance of TOP2A and probable signaling pathways it involved in were further explored, and a positive correlation between TOP2A and TPX2 expression was found in lung cancer tissues. CONCLUSION: Using bioinformatic analysis, we revealed that TOP2A could be adopted as a prognostic indicator of NSCLC and it potentially regulate cancer development through co-work with TPX2. However, more detailed experiments are needed to clarify its drug target role in clinical medical use.

SRT1720 retards renal fibrosis via inhibition of HIF1α /GLUT1 in diabetic nephropathy.

Renal fibrosis is a major pathological characteristic of diabetic nephropathy (DN). Reportedly, increased SIRT1 expression played a renal protective role in animal models of DN. This study was designed to elucidate the molecular mechanisms underlying the protective effects of SRT1720, an SIRT1 activator, against diabetes-induced renal fibrosis. Type 2 diabetic mice (db/db) were treated with SRT1720 (50 mg/kg/d) by gavage for 10 weeks. Renal proximal tubular epithelial cells (HK-2 cells) were treated with high glucose (HG, 30 mM) in the presence or absence of SRT1720 (2.5 µM) for 48 h. We observed that impaired SIRT1 expression and activity were restored by SRT1720 administration in db/db mice as well as in HG-treated HK-2 cells. Moreover, SRT1720 administration improved renal function, attenuated glomerular hypertrophy, mesangial expansion, glomerulosclerosis and interstitial fibrosis, and inhibited TGFB1 and CTGF expressions and nuclear factor κB (NF-κB) activation in db/db mice. Similarly, HG-induced epithelial-to-mesenchymal transformation (EMT), and collagen IV and fibronectin expressions were inhibited in SRT1720 treated HK-2 cells. Mechanistic studies demonstrated that SRT1720 suppressed HIF1α, GLUT1 and SNAIL expressions both in vivo and in vitro. Furthermore, Hif1α or Glut1 knockdown effectively abrogated HG-induced EMT and collagen IV and fibronectin expressions in HK-2 cells. These findings suggest that SRT1720 prevented diabetes-induced renal fibrosis via the SIRT1/HIF1α/GLUT1/SNAIL pathway.

NLRP3 deficiency ameliorates renal inflammation and fibrosis in diabetic mice.

Diabetic nephropathy (DN) is the leading cause of end-stage renal disease. Activation of the nucleotide binding and oligomerization domain-like receptor family pyrin domain-containing 3 (NLRP3) inflammasome has been reported in diabetic kidney, yet the potential role of NLRP3 inflammasome in DN is not well known. In this study, we explored the role of NLRP3 inflammasome on inflammation and fibrosis in diabetic kidney using NLRP3 knockout mice. Renal expression of NLRP3, caspase-1 p10, interleukin-18 (IL-18) and cleaved IL-1ß was increased in diabetic wild-type (WT) mice at 24 weeks. NLRP3 knockout (KO) improved renal function, attenuated glomerular hypertrophy, glomerulosclerosis, mesangial expansion, interstitial fibrosis, inflammation and expression of TGF-ß1 and connective tissue growth factor (CTGF), as well as the activation of Smad3 in kidneys of STZ-induced diabetic mice. In addition, NLRP3 KO inhibited expression of thioredoxin-interacting protein (TXNIP) and NADPH oxidase 4 (Nox4) and superoxide production in diabetic kidneys. The diabetes-induced increase in urinary level of 8-hydroxydeoxyguanosine (8-OHdG) was attenuated in NLRP3 KO mice. In vitro experiments, using HK-2 cells, revealed that high glucose (HG)-mediated expression of TXNIP and Nox4 was inhibited by transfection with NLRP3 shRNA plasmid or antioxidant tempol treatment. Silencing of the NLRP3 resulted in reduced generation of reactive oxygen species (ROS) in HK-2 cells under HG conditions. Furthermore, we also found exposure of IL-1ß to HK-2 cells induced ROS generation and expression of TXNIP and Nox4. Taken together, inhibition of NLRP3 inflammasome activation inhibits renal inflammation and fibrosis at least in part via suppression of oxidative stress in diabetic nephropathy.

Thioredoxin-interacting protein deficiency ameliorates kidney inflammation and fibrosis in mice with unilateral ureteral obstruction.

Thioredoxin-interacting protein (TXNIP) is associated with inflammation, tubulointerstitial fibrosis, and oxidative stress in diabetic kidney disease, yet the potential role of TXNIP in nondiabetic renal injury is not well known. This study aimed to investigate the effect of TXNIP on renal injury by creating a unilateral ureteral obstruction (UUO) model in TXNIP knockout (TKO) mice. We performed sham or UUO surgery in 8-week-old TXNIP KO male mice and age and sex-matched wild-type (WT) mice. Animals were killed at 3, 5, 7, or 14 days after surgery, and renal tissues were obtained for RNA, protein, and other analysis. Our results show that the expression of TXNIP was increased in a time-dependent manner in the ligated kidneys. TXNIP deletion reduced renal fibrosis, apoptosis, α-SMA, TGF-ß1 and CTGF expression, and activation of Smad3, p38 MAPK, and ERK1/2 in UUO kidneys. We also found UUO-induced renal F4/80+ macrophage infiltration, MCP-1 expression and activation of NF-κB and NLRP3 inflammasome were attenuated in TKO mice. Furthermore, our study revealed that TXNIP deficiency inhibited the expression of 8-OHdG, heme oxygenase-1 (HO-1) and NADPH oxidase 4 (Nox4) in UUO kidney. In summary, our study suggests that TXNIP plays a key role in the renal inflammation and fibrosis induced by UUO. Inhibition of TXNIP may be a strategy to slow the progression of chronic kidney diseases.

Intermedin is upregulated and attenuates renal fibrosis by inhibition of oxidative stress in rats with unilateral ureteral obstruction.

AIM: Transforming growth factor-ß1 (TGF-ß1) plays a pivotal role in the progression of renal fibrosis. Reactive oxygen species mediate profibrotic action of TGF-ß1. Intermedin (IMD) has been shown to inhibit oxidative stress, but its role in renal fibrosis remains unclear. Here, we investigated the effects of IMD on renal fibrosis in a rat model of unilateral ureteral obstruction (UUO). METHODS: The expression of IMD and its receptors, calcitonin receptor-like receptor (CRLR) and receptor activity-modifying proteins (RAMP1/2/3), in the obstructed kidney was detected by real-time polymerase chain reaction (PCR), western blotting and immunohistochemistry. To evaluate the effects of IMD on renal fibrosis, we locally overexpressed exogenous IMD in the obstructed kidney using an ultrasound-microbubble-mediated delivery system. Renal fibrosis was determined by Masson trichrome staining. The expression of TGF-ß1, connective tissue growth factor (CTGF), α-smooth muscle actin (α-SMA) and fibronectin was measured. Smad2/3 activation and macrophage infiltration were evaluated. We also studied oxidative stress by measuring superoxide dismutase (SOD) activity and malondialdehyde (MDA) content. RESULTS: mRNA and protein expression of IMD increased after UUO. CRLR, RAMP1, RAMP2 and RAMP3 were also induced by ureteral obstruction. IMD overexpression remarkably attenuated UUO-induced tubular injury and blunted fibrotic response as shown by decreased interstitial collagen deposition and downregulation of fibronectin. Macrophage infiltration, α-SMA and CTGF upregulation caused by UUO were all relieved by IMD, whereas TGF-ß1 upregulation and Smad2/3 activation were not affected. Meanwhile, we noted increased oxidative stress in obstruction, which was also attenuated by IMD gene delivery. CONCLUSIONS: Our results indicate that IMD is upregulated after UUO. IMD plays a protective role in renal fibrosis via its antioxidant effects.