GNNGL-PPI: multi-category prediction of protein-protein interactions using graph neural networks based on global graphs and local subgraphs.

Most proteins exert their functions by interacting with other proteins, making the identification of protein-protein interactions (PPI) crucial for understanding biological activities, pathological mechanisms, and clinical therapies. Developing effective and reliable computational methods for predicting PPI can significantly reduce the time-consuming and labor-intensive associated traditional biological experiments. However, accurately identifying the specific categories of protein-protein interactions and improving the prediction accuracy of the computational methods remain dual challenges. To tackle these challenges, we proposed a novel graph neural network method called GNNGL-PPI for multi-category prediction of PPI based on global graphs and local subgraphs. GNNGL-PPI consisted of two main components: using Graph Isomorphism Network (GIN) to extract global graph features from PPI network graph, and employing GIN As Kernel (GIN-AK) to extract local subgraph features from the subgraphs of protein vertices. Additionally, considering the imbalanced distribution of samples in each category within the benchmark datasets, we introduced an Asymmetric Loss (ASL) function to further enhance the predictive performance of the method. Through evaluations on six benchmark test sets formed by three different dataset partitioning algorithms (Random, BFS, DFS), GNNGL-PPI outperformed the state-of-the-art multi-category prediction methods of PPI, as measured by the comprehensive performance evaluation metric F1-measure. Furthermore, interpretability analysis confirmed the effectiveness of GNNGL-PPI as a reliable multi-category prediction method for predicting protein-protein interactions.

[Acute heart failure in a neonate]. / æ–°ç”Ÿå„¿æ€¥æ€§å¿ƒåŠŸèƒ½ä¸å ¨1例.

The male patient, one day old, was admitted to the hospital due to hypoglycemia accompanied by apnea appearing six hours after birth. The patient had transient hypoglycemia early after birth, and acute heart failure suddenly occurred on the eighth day after birth. Laboratory tests showed significantly reduced levels of adrenocorticotropic hormone and cortisol, and pituitary magnetic resonance imaging was normal. Genetic testing results showed that the patient had probably pathogenic compound heterozygous mutations of the TBX19 gene (c.917-2A>G+c.608C>T), inherited respectively from the parents. The patient was conclusively diagnosed with congenital isolated adrenocorticotropic hormone deficiency caused by mutation of the TBX19 gene. Upon initiating hydrocortisone replacement therapy, cardiac function rapidly returned to normal. After being discharged, the patient continued with the hydrocortisone replacement therapy. By the 18-month follow-up, the patient was growing and developing well. In neonates, unexplained acute heart failure requires caution for possible endocrine hereditary metabolic diseases, and timely cortisol testing and genetic testing should be conducted.

A comprehensive review of the recent advances on predicting drug-target affinity based on deep learning.

Accurate calculation of drug-target affinity (DTA) is crucial for various applications in the pharmaceutical industry, including drug screening, design, and repurposing. However, traditional machine learning methods for calculating DTA often lack accuracy, posing a significant challenge in accurately predicting DTA. Fortunately, deep learning has emerged as a promising approach in computational biology, leading to the development of various deep learning-based methods for DTA prediction. To support researchers in developing novel and highly precision methods, we have provided a comprehensive review of recent advances in predicting DTA using deep learning. We firstly conducted a statistical analysis of commonly used public datasets, providing essential information and introducing the used fields of these datasets. We further explored the common representations of sequences and structures of drugs and targets. These analyses served as the foundation for constructing DTA prediction methods based on deep learning. Next, we focused on explaining how deep learning models, such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Transformer, and Graph Neural Networks (GNNs), were effectively employed in specific DTA prediction methods. We highlighted the unique advantages and applications of these models in the context of DTA prediction. Finally, we conducted a performance analysis of multiple state-of-the-art methods for predicting DTA based on deep learning. The comprehensive review aimed to help researchers understand the shortcomings and advantages of existing methods, and further develop high-precision DTA prediction tool to promote the development of drug discovery.

Drug-Online: an online platform for drug-target interaction, affinity, and binding sites identification using deep learning.

BACKGROUND: Accurately identifying drug-target interaction (DTI), affinity (DTA), and binding sites (DTS) is crucial for drug screening, repositioning, and design, as well as for understanding the functions of target. Although there are a few online platforms based on deep learning for drug-target interaction, affinity, and binding sites identification, there is currently no integrated online platforms for all three aspects. RESULTS: Our solution, the novel integrated online platform Drug-Online, has been developed to facilitate drug screening, target identification, and understanding the functions of target in a progressive manner of "interaction-affinity-binding sites". Drug-Online platform consists of three parts: the first part uses the drug-target interaction identification method MGraphDTA, based on graph neural networks (GNN) and convolutional neural networks (CNN), to identify whether there is a drug-target interaction. If an interaction is identified, the second part employs the drug-target affinity identification method MMDTA, also based on GNN and CNN, to calculate the strength of drug-target interaction, i.e., affinity. Finally, the third part identifies drug-target binding sites, i.e., pockets. The method pt-lm-gnn used in this part is also based on GNN. CONCLUSIONS: Drug-Online is a reliable online platform that integrates drug-target interaction, affinity, and binding sites identification. It is freely available via the Internet at http://39.106.7.26:8000/Drug-Online/ .

Recent Advances on Selenium Nutrition and Keshan Disease.

Keshan disease (KD) is a type of endemic cardiomyopathy with an unknown cause. It is primarily found in areas in China with low selenium levels, from northeast to southwest. The nutritional biogeochemical etiology hypothesis suggests that selenium deficiency is a major factor in KD development. Selenium is important in removing free radicals and protecting cells and tissues from peroxide-induced damage. Thus, low environmental selenium may affect the selenium level within the human body, and selenium level differences are commonly observed between healthy people in KD and nonKD areas. From the 1970s to the 1990s, China successfully reduced KD incidence in endemic KD areas through a selenium supplementation program. After years of implementing prevention and control measures, the selenium level of the population in the KD areas has gradually increased, and the prevalence of KD in China has remained low and stable in recent years. Currently, the pathogenesis of KD remains vague, and the effect of selenium supplementation on the prognosis of KD still needs further study. This paper comprehensively reviews selenium deficiency and its connection to KD. Thus, this study aims to offer novel ideas and directions to effectively prevent and treat KD in light of the current situation.

Coregulatory effects of multiple histone modifications in key ferroptosis-related genes for lung adenocarcinoma.

Aim: The present study was designed to investigate the coregulatory effects of multiple histone modifications (HMs) on gene expression in lung adenocarcinoma (LUAD). Materials & methods: Ten histones for LUAD were analyzed using ChIP-seq and RNA-seq data. An innovative computational method is proposed to quantify the coregulatory effects of multiple HMs on gene expression to identify strong coregulatory genes and regions. This method was applied to explore the coregulatory mechanisms of key ferroptosis-related genes in LUAD. Results: Nine strong coregulatory regions were identified for six ferroptosis-related genes with diverse coregulatory patterns (CA9, PGD, CDKN2A, PML, OTUB1 and NFE2L2). Conclusion: This quantitative method could be used to identify important HM coregulatory genes and regions that may be epigenetic regulatory targets in cancers.

[Peripherally inserted central venous catheter-related thrombosis in a neonate]. / æ–°ç”Ÿå„¿ç»å¤–å‘¨é™è„‰ç©¿åˆºçš„ä¸­å¿ƒé™è„‰å¯¼ç®¡ç›¸å ³è¡€æ “1例.

The female infant in this case study was admitted to the hospital 4 hours after birth due to preterm birth and respiratory distress. On the third day after birth, peripherally inserted central venous catheter (PICC) catheterization was performed. On day 42, thrombus was found at the entrance of the right atrium from the inferior vena cava during a cardiac ultrasound, and it was considered to be related to PICC placement. Low-molecular-weight heparin and urokinase were given. After two weeks of treatment, ultrasonic monitoring showed thrombus shrinkage. No bleeding or pulmonary embolism occurred during the treatment. The patient discharged after improvement. This article mainly introduces a multidisciplinary team approach to diagnosis and treatment of PICC-related thrombosis in neonates.

The complete chloroplast genome of Nymphaea atrans (Surrey Wilfrid Laurance Jacobs, 1992: Nymphaeaceae).

Nymphaea atrans belongs to the subgenus Anecphya and displays varied flower colors over successive days. Because of its excellent ornamental characteristics, this species is widely cultivated in waterscape gardens worldwide. Here, we have sequenced the complete chloroplast genome of N. atrans. The whole genome size is 160,990 bp in length with four subregions: large and small single-copy regions of 90,879 and 19,699 bp, respectively, separated by a pair of inverted repeat regions of 25,206 bp each. A total of 126 genes were annotated, including 82 coding genes, eight ribosomal RNAs, and 36 transfer RNAs. The total GC content of the complete genome was 39%. Phylogenetic analysis showed that N. atrans was closely related to N. immutabilis. In this study, we have provided the chloroplast genome of N. atrans as a valuable resource for further phylogenetic analysis of Nymphaea species.

Spatial Heterogeneity and Influencing Factors of HFRS Epidemics in Rural and Urban Areas: A Study in Guanzhong Plain of Shaanxi Province, China.

Objective: The Guanzhong Plain of Shaanxi Province is a severely afflicted hemorrhagic fever with renal syndrome (HFRS) epidemic area, while HFRS prevalence has decreased in most epidemic areas in China. Little information is available regarding the leading fine-scale influencing factors in this highly HFRS-concentrated area and the roles of natural environmental and socioeconomic factors. To investigate this, two regions in the Guanzhong Plain, that is, the Chang'an District and Hu County, with similar geographical environments, different levels of economic development, and high epidemic prevalence, were chosen as representative areas of the HFRS epidemic. Methods: Maximum entropy models were constructed based on HFRS cases and fine-scale influencing factors, including meteorological, natural environmental, and socioeconomic factors, from 2014 to 2016. Results: More than 95% of the HFRS cases in the study area were located in the northern plains, which has an altitude of less than 800 m, with topography contributed 84.1% of the impact on the spatial differentiation of the HFRS epidemic. In the northern plains, precipitation and population density jointly affected the spatial differentiation of the HFRS epidemic, with contribution rates of 60.7% and 28.0%, respectively. By comparing the influencing factors of the northern plains of Chang'an District and Hu County, we found that precipitation and the normalized difference vegetation index (NDVI) dominated the HFRS epidemic in the relatively developed Chang'an District, while land-use type, temperature, precipitation and population density dominated the HFRS epidemic in the relatively undeveloped Hu County. Conclusion: Topography was the primary key factor for HFRS prevalence in the Chang'an District and Hu County, and the spatial differentiation of HFRS was dominated by precipitation and population density in the northern plains. Compared with the influencing factors of the relatively developed Chang'an District, the developing Hu County was more affected by socioeconomic factors. When formulating targeted HFRS epidemic prevention and control strategies in the targeted areas, it is crucial to consider the local economic development state and combine natural environmental factors, including the meteorological environment and vegetation coverage.

Prevalence/potential risk factors for motoric cognitive risk and its relationship to falls in elderly Chinese people: a cross-sectional study.

BACKGROUND AND PURPOSE: Motoric cognitive risk syndrome (MCR) is characterized by slow walking speed and subjective memory complaints (SMCs). This study investigated the prevalence and potential risk factors of MCR and its association with falls in Chinese community-dwelling older adults. METHODS: The analysis was based on data from the Rugao Longevity and Aging Study (RuLAS). MCR was defined as the presence of both SMCs and slow walking speed in participants free of major neurocognitive disorders. SMCs were determined according to a positive answer to the question 'Do you feel you have more problems with memory than most?' in the 15-item Geriatric Depression Scale. Slow walking speed was defined as one standard deviation or more below the mean value for patients' age and sex. Data on falls were derived from a standardized questionnaire. RESULTS: The prevalence of SMCs, slow walking speed and MCR in the RuLAS cohort (N = 1592) was 51.9%, 15.6% and 8.3%, respectively. After adjusting for other covariates, an occupation of farming (odds ratio [OR] 2.358, 95% confidence interval [CI] 1.007-5.521, p = 0.048), history of cerebrovascular disease (OR 2.215, 95% CI 1.032-4.752, p = 0.041) and hospitalization (OR 2.008, 95% CI 1.120-3.602, p = 0.019) were risk factors for MCR. Binary logistic regression analysis indicated that the risk of falls was increased by MCR (OR 1.547, 95% CI 1.009-2.371), SMC (OR 1.308, 95% CI 1.003-1.707) and slow walking speed (OR 1.442, 95% CI 1.030-2.017). CONCLUSIONS: Early identification of potential risk factors of MCR can prevent the occurrence of adverse health events such as falls in the elderly.

Engineering of stepwise-targeting chitosan oligosaccharide conjugate for the treatment of acute kidney injury.

Acute kidney injury (AKI) is a common and serious clinical syndrome of acute renal dysfunction in a short period. One of therapeutic interventions for AKI is to reduce ROS massively generated in the mitochondria and then ameliorate cell damage and apoptosis induced by oxidative stress. In this study, stepwise-targeting chitosan oligosaccharide, triphenyl phosphine-low molecular weight chitosan-curcumin (TPP-LMWC-CUR, TLC), was constructed for sepsis-induced AKI via removing excessive ROS in renal tubular epithelial cells. Benefiting from good water solubility and low molecular weight, TLC was rapidly and preferentially distributed in the renal tissues and then specifically internalized by tubular epithelium cells via interaction between Megalin receptor and LMWC. The intracellular TLC could further delivery CUR to mitochondria due to high buffering capacity of LMWC and delocalized positive charges of TPP. Both in vitro and in vivo pharmacodynamic results demonstrated the enhanced therapeutic effect of TLC in the treatment of AKI.

Targeted next-generation sequencing identified a novel variant of SOX10 in a Chinese family with Waardenburg syndrome type 2.

OBJECTIVE: Waardenburg syndrome type 2 (WS2) is an autosomal dominant syndrome, characterized by bright blue eyes, hearing loss, and depigmented patches of hair and skin. It exhibits high phenotypic and genetic heterogeneity. We explored the molecular etiology in a Chinese family with WS2. METHODS: We recruited a three-generation family with three affected members. Medical history was obtained from all family members who underwent detailed physical examinations and audiology tests. Genomic DNA was extracted from peripheral blood of each individual, and 139 candidate genes associated with hearing loss were sequenced using Illumina HiSeq 2000 (Illumina Inc., San Diego, CA, USA) and verified by Sanger sequencing. RESULTS: Genetic evaluation revealed a novel nonsense heterozygous variant, NM_006941.4: c.342G>A (p.Trp114Ter) in exon 2 of the SOX10 gene in the three affected patients; no unaffected family member carried the variation. We did not detect the variation in 500 Chinese individuals with normal hearing or in 122 unrelated Chinese families with hearing loss, suggesting that it was specific to our patients. CONCLUSIONS: We identified a novel heterozygous nonsense variation in a family with syndromic hearing loss and WS2. Our findings expand the pathogenic spectrum and strengthen the clinical diagnostic role of SOX10 in patients with WS2.

The mutation frequencies of GJB2, GJB3, SLC26A4 and MT-RNR1 of patients with severe to profound sensorineural hearing loss in northwest China.

OBJECTIVE: To expose the spectrum and frequency of GJB2, GJB3, SLC26A4 and MT-RNR1 in northwest China and to investigate the underlying causative genes in patients without common mutations. METHODS: We analyzed the mutation screening results of GJB2, GJB3, SLC26A4 and MT-RNR1 in 398 unrelated severe-to-profound probands with bilateral, symmetrical sensorineural hearing loss. Subsequently, we selected 10 probands with a significant family history of inherited hearing loss (HL) that did not have the above four common gene mutations to perform next-generation sequencing (NGS) of 139 known deafness genes, followed by co-segregation analysis of all available family members. RESULTS: Among the 398 patients, 69 (17.34%) had the biallelic GJB2 gene mutations, and the most common mutations were c.235delC, c.109G>A and c.299_300delAT, with allele frequencies of 12.31%, 3.38% and 3.89%, respectively. A total of 63 (15.83%) cases with biallelic SLC26A4 mutations were detected, and the most common pathogenic alleles were c.919-2A>G, c.2168A>G and c.1174A>T, with allele frequencies of 9.17%, 2.26% and 0.88%, respectively. Mitochondrial gene mutations were detected in 9 (2.26%) patients, with 5 cases of mitochondrial DNA (mtDNA) m.1555A>G mutation and 4 cases of mtDNA m.1095T>C mutation. In 10 probands with a clear family history of HL, NGS showed two novel pathogenic variants in 2 families, including c.4129C>T/c.3268C>T in LOXHD1, c.334G>A/c.2968G>T in CDH23. Sanger sequencing confirmed that these variants segregated with the HL in each family. CONCLUSIONS: Our results showed that GJB2 and SLC26A4 were the two major HL-causing genes in northwest China. The most common mutation alleles in GJB2 were c.235delC, c.109G>A and c.299_300delAT, and those in SLC26A4 were c.919-2A>G, c.2168A>G and c.1174A>T. In addition, both genes and their loci can be used as the first selection of deafness gene screening. Additionally, for patients who did not have mutations of these common genes, NGS provided an efficient diagnosis for increasing known deafness genes.

Efficacy of anti-T-lymphocyte globulin-Fresenius as an induction agent in deceased-donor renal transplantation: A cohort study.

Anti-T-lymphocyte globulin (ATG) is frequently used in the induction regimen of renal transplantation, but its dose has not been standardized. In the present study, the efficacy of different ATG-Fresenius (ATG-F) doses was assessed in recipients of renal transplantation. A total of 131 adult recipients of renal transplantation who received ATG-F induction between August 2015 and July 2018 were included. The incidence of biopsy-confirmed acute rejection, graft function, as well as graft and patient survival within 12 months post-transplant, was assessed, and adverse events, including hematologic and infection-associated side effects, were compared between patients receiving a cumulative ATG-F dose of <7 or ≥7 mg/kg. The incidence of biopsy-confirmed acute rejection was similar between patients receiving cumulative doses of <7 and ≥7 mg/kg (7.5 vs. 4.7%, P=0.766). The incidence of infection within 12 months was lower in the ATG-F <7 mg/kg group compared with that in the ≥7 mg/kg group (26.9 vs. 50.0%, P=0.006), but the incidence of pneumonia did not differ between the ATG-F <7 and ≥7 mg/kg groups (10.4 vs. 20.3%, P=0.117). The incidence of urinary infection was higher in the ≥7 mg/kg group than in the <7 mg/kg group (20.4 vs. 7.46%, P=0.033), while the extent and duration of anemia and lymphopenia was similar between groups. There was no difference in graft function, delayed graft function, as well as overall and graft survival between the groups. In conclusion, a moderate reduction in the cumulative ATG-F dose was not associated with an increased risk of acute rejection, while the risk of infection was reduced. Optimization of the ATG-F dose for induction may facilitate the reduction of the risk of infection without compromising the induction efficacy in renal transplant recipients.

In vitro corrosion of pure Mg in phosphate buffer solution-Influences of isoelectric point and molecular structure of amino acids.

Influences of proteins on degradation of magnesium alloys are of great significance but not well understood. In particular the roles of amino acids, the basic unit of proteins in regulating the progress of biodegradation of magnesium based materials remain unclear. This study aims to investigate the impacts of alanine, glutamic acid and lysine on degradation of pure magnesium in phosphate buffer solution through SEM, XPS, FTIR, potentiodynamic polarisation curves, electrochemical impedance spectroscopy and immersion tests. The changed contents of amino acids in solutions were detected by UV-vis spectrophotometer. Results demonstrate that the charges of the selected amino acids imposed significant contribution to suppressing the degradation of pure magnesium in phosphate buffer solution. The presence of amino acids led to the formation of phosphate-based corrosion products, increasing free corrosion potential, and reduction in corrosion current density and solution pH depending on their isoelectric points and molecular structures. A plausible corrosion mechanism organised by amino acids on pure magnesium was proposed.

Preparation of Ethyl Cellulose Microspheres for Sustained Release of Sodium Bicarbonate.

Sustained release of thermal-instable and water-soluble drugs with low molecule weight is a challenge. In this study, sodium bicarbonate was encapsulated in ethyl cellulose microspheres by a novel solid-in-oil-in-oil (S/O/O) emulsification method using acetonitrile/soybean oil as new solvent pairs. Properties of the microspheres such as size, recovery rate, morphology, drug content, and drug release behavior were evaluated to investigate the suitable preparation techniques. In the case of that the ratio of the internal and external oil phase was 1: 9, Tween 80 as a stabilizer resulted in the highest drug content (2.68%) and a good spherical shape of microspheres. After the ratio increased to 1: 4, the microspheres using Tween 80 as the stabilizer also had high drug content (1.96%) and exhibited a sustained release behavior, with 70% of drug released within 12 h and a sustained release of more than 40 h. Otherwise, different emulsification temperatures at which acetonitrile was evaporated could influence the drug release behaviour of microspheres obtained. This novel method is a potential and effective method to achieve the encapsulation and the sustained release of thermal-instable and water-soluble drugs with low molecule weight.

A new physical-cognitive scale for assessment of frailty in Chinese Han elderly.

Objective: To develop a physical- cognitive scale for assessment of frailty and compare the clinical features between the new scale and the conventional Fried criteria. Methods: 1757 individuals aged 70-84 were analyzed. Participants reporting three or more Fried phenotypes were grouped as frail patients (FP) whereas others as non-frail (NF). A score of Hasegawa's dementia scale (HDS-R) higher than 21.5 were classified as non-cognitive impairment group (NCI) group. By combining the cognitive and frailty criteria, participants manifesting three or more positive components out of the six were categorized into the Physical-cognitive frailty group (Pc-F) while others into non- Pc-F (Pc-NF). Results: Of all the participants, 46.7% (820) were males and 53.3% (937) were females. The mean age was 75.33 ± 3.90. 10.1% (178/1757) were evaluated as FP patients. The prevalence of CI was 53.2%; CI was much higher in the frail group (77.0%) than in the non-frail group (50.5%). Based on the new Pc-F scale, 163 out of 1579 NF participants were identified as Pc-F, and the prevalence of Pc-F reached 19.4% (341/1757). In the Pc-F group, there are more females, patients of advanced age, diabetes, stroke, CHD, CKD, metabolic syndrome, and high hs-CRP. Within the Pc-F group, patients with CI showed a higher incidence of exhaustion, low activity, weakness, and slowness than those without CI. Conclusions: Our study revealed a significantly worse status in frail participants with CI than without. Our new scale shows a stronger correlation between frailty and complications than the classic phenotype.

Long non-coding RNA Hottip modulates high-glucose-induced inflammation and ECM accumulation through miR-455-3p/WNT2B in mouse mesangial cells.

Long non-coding RNAs (lncRNAs) play important roles in the pathogenesis of various diseases, including diabetic nephropathy (DN). However, the detailed mechanism is still largely unknown. High-glucose treated SV40-MES13 cells was used to mimic diabetic nephropathy in vitro. qRT-PCR was introduced to measure Hottip, collagen type I (Col. I), collagen type IV (Col. IV), fibronectin (FN), PAI-1, miR-455-3p and Wnt2B, IL-6, TNF-α mRNA level. Ellisa was used to examine the expression level of IL-6, TNF-α in the cell culture medium. Western blotting was employed to detect the protein level of Col. I, Col. IV, FN, PAI-1, Wnt2B, ß-catenin and cyclin D1. Cell viability was examined by MTT assay, luciferase reporter assay were used to determine the relationship between Hottip, miR-455-3p and Wnt2B. In the results, Hottip and Wnt2B was upregulated in db/db DN mice and high-glucose treated mouse mesangial cells (MMCs) while miR-455-3p was downregulated. High glucose treatment could enhance cell proliferation, and inflammation, increase fibrosis-related protein expression and active Wnt2B/ß-catenin/cyclin D1 pathway, while Hottip silencing reversed all the effects caused by high-glucose treatment. miR-455-3p was a sponge target of Hottip while Wnt2B was a downstream target of miR-445-3p. miR-445-3p inhibitor could suppress the effect of Hottip knockdown in cell proliferation, inflammation and fibrosis-related protein expression. Our data supported lncRNA Hottip/miR-455-3p/Wnt2B axis plays an important role in cell proliferation, inflammation, and extracellular matrix (ECM) accumulation in diabetic nephropathy.

CD44-targeted hyaluronic acid-curcumin prodrug protects renal tubular epithelial cell survival from oxidative stress damage.

Based on the abnormally increased expression of CD44 receptors on renal tubule epithelial cells during ischemia/reperfusion-induced acute kidney injury (AKI), we developed a hyaluronic acid-curcumin (HA-CUR) polymeric prodrug targeting to epithelial cells and then relieving oxidative stress damages. The water solubility of HA-CUR was significantly enhanced and approximately 27-fold higher than that of CUR. Cellular uptake test showed HA-CUR was preferably internalized by H2O2-pretreated tubular epithelial (HK-2) cells compared with free CUR benefiting from the specific binding between HA and CD44 receptors. Biodistribution results further demonstrated the increased accumulation of HA-CUR in kidneys with 13.9-fold higher than that of free CUR. Pharmacodynamic studies indicated HA-CUR effectively ameliorated AKI, and the exact mechanism was that HA-CUR protected renal tubule epithelial cells from oxidative stress damage via inhibiting PtdIns3K-AKT-mTOR signaling pathway. Taken together, this study provides a new therapeutic strategy for the treatment of AKI based on the pathogenesis of the disease.