1.Quality Evaluation of Lycii Cortex and Roasted Lycii Cortex Based on Fingerprint and Content Determination
Yihuan WU ; Wenli ZENG ; Xuemei QIN ; Zongxin SHI ; Chengcheng HUANG ; Yuntao DAI
Chinese Journal of Experimental Traditional Medical Formulae 2025;31(2):165-172
ObjectiveTo establish fingerprint profiles and a quantitative determination method for Lycii Cortex, providing a scientific basis for the formulation of quality standards for Lycii Cortex and its roasted products. MethodsHigh performance liquid chromatography(HPLC) was developed for the quantitative method for determining kukoamine B in Lycii Cortex and its roasted products on an Alphasil XD-C18 CH column(4.6 mm×250 mm, 5 μm). HPLC fingerprint profiles were established for 10 batches of Lycii Cortex and its roasted products, and ultra-performance liquid chromatography-quadrupole-time-of-flight mass spectrometry(UPLC-Q-TOF-MS) was used to identify the common peaks based on reference standards, literature and MS information. Quality evaluation indicators included yield of decoction pieces, appearance properties, content of kukoamine B, and fingerprint profiles. The temperature and time of the roasting process were investigated to select the optimal preparation process, which was then verified. Additionally, chemical pattern recognition was combined to assess the differences in the chemical composition of Lycii Cortex before and after roasting, as well as among samples from different origins. ResultsQuantitative analysis indicated that the contents of kukoamine B in Lycii Cortex and its roasted products were 0.35%-5.51% and 0.24%-4.15%, respectively. The transfer rate of kukoamine B was 58.6%-78.9% after roasting. The fingerprint profile analysis demonstrated that the method established in this study effectively separated kukoamine B from other components in the samples and distinctly differentiated it from its impurity peak, cis-N-caffeoylputrescine. The HPLC fingerprint profiles of Lycii Cortex and its roasted products showed high similarity(all above 0.95), with 7 common peaks identified and five common components, including kukoamine B, cis-N-caffeoylputrescine, N-coumaroyl tyramine, feruloyltyramine, and glucosyringic acid, confirmed. Process optimization confirmed that baking at 110 ℃ for 20 min was a stable and feasible method for roasting Lycii Cortex. Principal component analysis and cluster analysis showed that there was little difference in the chemical composition between raw and roasted Lycii Cortex, but the quality of Lycii Cortex from different origins differed greatly. ConclusionThis study successfully established the fingerprint profiles and a quantitative method for the effective component kukoamine B in Lycii Cortex and roasted Lycii Cortex. The qualitative and quantitative analyses clarified that the impact of the roasting process on the chemical composition of Lycii Cortex was less significant than the variations due to its geographical origin. The findings of this study offer a reference for the development of quality evaluation methods and the establishment of quality standards for Lycii Cortex and its processed products.
2.Analysis and forecast of the disease burden of schistosomiasis in China from 1992 to 2030
Kai LIN ; Chenhuan ZHANG ; Zhendong XU ; Xuemei LI ; Renzhan HUANG ; Yawen LIU ; Haihang YU ; Lisi GU
Chinese Journal of Schistosomiasis Control 2025;37(1):24-34
Objective To analyze the trends in the disease burden of schistosomiasis in China from 1992 to 2021, and to project the disease burden of schistosomiasis in China from 2022 to 2030, so as to provide insights into the elimination of schistosomiasis in China. Methods The prevalence, age-standardized prevalence, disability-adjusted life year (DALYs) rate and age-standardized DALYs rate of schistosomiasis, as well as the years lost due to disability (YLDs) rate and age-standardized YLDs rate of anemia attributable to Schistosoma infections in China, the world and different socio-demographic index (SDI) regions were captured from the Global Burden of Disease Study 2021 (GBD 2021) data resources, and the trends in the disease burden due to schistosomiasis were evaluated with estimated annual percentage change (EAPC) and its 95% confidence interval (CI). In addition, the age, period and cohort effects on the prevalence of schistosomiasis were examined in China using an age-period-cohort (APC) model, and the disease burden of schistosomiasis was predicted in China from 2022 to 2030 using a Bayesian age-period-cohort (BAPC) model. Results The age-standardized prevalence and DALYs rate of schistosomiasis, and the age-standardized YLDs rate of anemia attributable to Schistosoma infections were 761.32/105, 5.55/105 and 0.38/105 in China in 2021. These rates were all lower than the global levels (1 914.30/105, 21.90/105 and 3.36/105, respectively), as well as those in the medium SDI regions (1 413.61/105, 12.10/105 and 1.93/105, respectively), low-medium SDI regions (2 461.03/105, 26.81/105 and 4.48/105, respectively), and low SDI regions (5 832.77/105, 94.48/105 and 10.65/105, respectively), but higher than those in the high SDI regions (59.47/105, 0.49/105 and 0.05/105, respectively) and high-medium SDI regions (123.11/105, 1.20/105 and 0.12/105, respectively). The prevalence and DALYs rate of schistosomiasis were higher among men (820.79/105 and 5.86/105, respectively) than among women (697.96/105 and 5.23/105, respectively) in China in 2021, while the YLDs rate of anemia attributable to Schistosoma infections was higher among women (0.66/105) than among men (0.12/105). The prevalence of schistosomiasis peaked at ages of 30 to 34 years among both men and women, while the DALYs rate of schistosomiasis peaked among men at ages of 15 to 19 years and among women at ages of 20 to 24 years. The age-standardized prevalence of schistosomiasis showed a moderate decline in China from 1992 to 2021 relative to different SDI regions [EAPC = -1.51%, 95% CI: (-1.65%, -1.38%)], while the age-standardized DALYs rate [EAPC = -3.61%, 95% CI: (-3.90%, -3.33%)] and age-standardized YLDs rate of anemia attributable to Schistosoma infections [EAPC = -4.16%, 95% CI: (-4.38%, -3.94%)] appeared the fastest decline in China from1992 to 2021 relative to different SDI regions. APC modeling showed age, period, and cohort effects on the trends in the prevalence of schistosomiasis in China from 1992 to 2021, and the prevalence of schistosomiasis appeared a rise followed by decline with age, and reduced with period and cohort. BAPC modeling revealed that the age-standardized prevalence and age-standardized DALYs rate of schistosomiasis, and age-standardized YLDs rate of anemia attributable to Schistosoma infections all appeared a tendency towards a decline in China from 2022 to 2030, which reduced to 722.72/105 [95% CI: (538.74/105, 906.68/105)], 5.19/105 [95% CI: (3.54/105, 6.84/105)] and 0.30/105 [95% CI: (0.21/105, 0.39/105)] in 2030, respectively. Conclusions The disease burden of schistosomiasis appeared a tendency towards a decline in China from 1992 to 2021, and is projected to appear a tendency towards a decline from 2022 to 2030. There are age, period and cohort effects on the prevalence of schistosomiasis in China. Precision schistosomiasis control is required with adaptations to current prevalence and elimination needs.
3.Genetic Correlation and Mendelian Randomization Analysis Revealed an Unidirectional Causal Relationship Between Left Caudal Middle Frontal Surface Area and Cigarette Consumption
Hongcheng XIE ; Anlin WANG ; Minglan YU ; Tingting WANG ; Xuemei LIANG ; Rongfang HE ; Chaohua HUANG ; Wei LEI ; Jing CHEN ; Youguo TAN ; Kezhi LIU ; Bo XIANG
Psychiatry Investigation 2025;22(3):279-286
Objective:
Previous studies have discovered a correlation between cigarette smoking and cortical thickness and surface area, but the causal relationship remains unclear. The objective of this investigation is to scrutinize the causal association between them.
Methods:
To derive summary statistics from a genome-wide association study (GWAS) on cortical thickness, surface area, and four smoking behaviors: 1) age of initiation of regular smoking (AgeSmk); 2) smoking initiation (SmkInit); 3) smoking cessation (SmkCes); 4) cigarettes per day (CigDay). Linkage disequilibrium score regression (LDSC) was employed to examine genetic association analysis. Furthermore, for traits with significant genetic associations, Mendelian randomization (MR) analyses were conducted.
Results:
The LDSC analysis revealed nominal genetic correlations between AgeSmk and right precentral surface area, left caudal anterior cingulate surface area, left cuneus surface area, left inferior parietal surface area, and right caudal anterior cingulate thickness, as well as between CigDay and left caudal middle frontal surface area, between SmkCes and left entorhinal thickness, and between SmkInit and left rostral anterior cingulate surface area, right rostral anterior cingulate thickness, and right superior frontal thickness (rg=-0.36–0.29, p<0.05). MR analysis showed a unidirectional causal association between left caudal middle frontal surface area and CigDay (βIVW=0.056, pBonferroni=2×10-4).
Conclusion
Left caudal middle frontal surface area has the potential to serve as a significant predictor of smoking behavior.
4.Genetic Correlation and Mendelian Randomization Analysis Revealed an Unidirectional Causal Relationship Between Left Caudal Middle Frontal Surface Area and Cigarette Consumption
Hongcheng XIE ; Anlin WANG ; Minglan YU ; Tingting WANG ; Xuemei LIANG ; Rongfang HE ; Chaohua HUANG ; Wei LEI ; Jing CHEN ; Youguo TAN ; Kezhi LIU ; Bo XIANG
Psychiatry Investigation 2025;22(3):279-286
Objective:
Previous studies have discovered a correlation between cigarette smoking and cortical thickness and surface area, but the causal relationship remains unclear. The objective of this investigation is to scrutinize the causal association between them.
Methods:
To derive summary statistics from a genome-wide association study (GWAS) on cortical thickness, surface area, and four smoking behaviors: 1) age of initiation of regular smoking (AgeSmk); 2) smoking initiation (SmkInit); 3) smoking cessation (SmkCes); 4) cigarettes per day (CigDay). Linkage disequilibrium score regression (LDSC) was employed to examine genetic association analysis. Furthermore, for traits with significant genetic associations, Mendelian randomization (MR) analyses were conducted.
Results:
The LDSC analysis revealed nominal genetic correlations between AgeSmk and right precentral surface area, left caudal anterior cingulate surface area, left cuneus surface area, left inferior parietal surface area, and right caudal anterior cingulate thickness, as well as between CigDay and left caudal middle frontal surface area, between SmkCes and left entorhinal thickness, and between SmkInit and left rostral anterior cingulate surface area, right rostral anterior cingulate thickness, and right superior frontal thickness (rg=-0.36–0.29, p<0.05). MR analysis showed a unidirectional causal association between left caudal middle frontal surface area and CigDay (βIVW=0.056, pBonferroni=2×10-4).
Conclusion
Left caudal middle frontal surface area has the potential to serve as a significant predictor of smoking behavior.
5.Genetic Correlation and Mendelian Randomization Analysis Revealed an Unidirectional Causal Relationship Between Left Caudal Middle Frontal Surface Area and Cigarette Consumption
Hongcheng XIE ; Anlin WANG ; Minglan YU ; Tingting WANG ; Xuemei LIANG ; Rongfang HE ; Chaohua HUANG ; Wei LEI ; Jing CHEN ; Youguo TAN ; Kezhi LIU ; Bo XIANG
Psychiatry Investigation 2025;22(3):279-286
Objective:
Previous studies have discovered a correlation between cigarette smoking and cortical thickness and surface area, but the causal relationship remains unclear. The objective of this investigation is to scrutinize the causal association between them.
Methods:
To derive summary statistics from a genome-wide association study (GWAS) on cortical thickness, surface area, and four smoking behaviors: 1) age of initiation of regular smoking (AgeSmk); 2) smoking initiation (SmkInit); 3) smoking cessation (SmkCes); 4) cigarettes per day (CigDay). Linkage disequilibrium score regression (LDSC) was employed to examine genetic association analysis. Furthermore, for traits with significant genetic associations, Mendelian randomization (MR) analyses were conducted.
Results:
The LDSC analysis revealed nominal genetic correlations between AgeSmk and right precentral surface area, left caudal anterior cingulate surface area, left cuneus surface area, left inferior parietal surface area, and right caudal anterior cingulate thickness, as well as between CigDay and left caudal middle frontal surface area, between SmkCes and left entorhinal thickness, and between SmkInit and left rostral anterior cingulate surface area, right rostral anterior cingulate thickness, and right superior frontal thickness (rg=-0.36–0.29, p<0.05). MR analysis showed a unidirectional causal association between left caudal middle frontal surface area and CigDay (βIVW=0.056, pBonferroni=2×10-4).
Conclusion
Left caudal middle frontal surface area has the potential to serve as a significant predictor of smoking behavior.
6.Genetic Correlation and Mendelian Randomization Analysis Revealed an Unidirectional Causal Relationship Between Left Caudal Middle Frontal Surface Area and Cigarette Consumption
Hongcheng XIE ; Anlin WANG ; Minglan YU ; Tingting WANG ; Xuemei LIANG ; Rongfang HE ; Chaohua HUANG ; Wei LEI ; Jing CHEN ; Youguo TAN ; Kezhi LIU ; Bo XIANG
Psychiatry Investigation 2025;22(3):279-286
Objective:
Previous studies have discovered a correlation between cigarette smoking and cortical thickness and surface area, but the causal relationship remains unclear. The objective of this investigation is to scrutinize the causal association between them.
Methods:
To derive summary statistics from a genome-wide association study (GWAS) on cortical thickness, surface area, and four smoking behaviors: 1) age of initiation of regular smoking (AgeSmk); 2) smoking initiation (SmkInit); 3) smoking cessation (SmkCes); 4) cigarettes per day (CigDay). Linkage disequilibrium score regression (LDSC) was employed to examine genetic association analysis. Furthermore, for traits with significant genetic associations, Mendelian randomization (MR) analyses were conducted.
Results:
The LDSC analysis revealed nominal genetic correlations between AgeSmk and right precentral surface area, left caudal anterior cingulate surface area, left cuneus surface area, left inferior parietal surface area, and right caudal anterior cingulate thickness, as well as between CigDay and left caudal middle frontal surface area, between SmkCes and left entorhinal thickness, and between SmkInit and left rostral anterior cingulate surface area, right rostral anterior cingulate thickness, and right superior frontal thickness (rg=-0.36–0.29, p<0.05). MR analysis showed a unidirectional causal association between left caudal middle frontal surface area and CigDay (βIVW=0.056, pBonferroni=2×10-4).
Conclusion
Left caudal middle frontal surface area has the potential to serve as a significant predictor of smoking behavior.
7.Genetic Correlation and Mendelian Randomization Analysis Revealed an Unidirectional Causal Relationship Between Left Caudal Middle Frontal Surface Area and Cigarette Consumption
Hongcheng XIE ; Anlin WANG ; Minglan YU ; Tingting WANG ; Xuemei LIANG ; Rongfang HE ; Chaohua HUANG ; Wei LEI ; Jing CHEN ; Youguo TAN ; Kezhi LIU ; Bo XIANG
Psychiatry Investigation 2025;22(3):279-286
Objective:
Previous studies have discovered a correlation between cigarette smoking and cortical thickness and surface area, but the causal relationship remains unclear. The objective of this investigation is to scrutinize the causal association between them.
Methods:
To derive summary statistics from a genome-wide association study (GWAS) on cortical thickness, surface area, and four smoking behaviors: 1) age of initiation of regular smoking (AgeSmk); 2) smoking initiation (SmkInit); 3) smoking cessation (SmkCes); 4) cigarettes per day (CigDay). Linkage disequilibrium score regression (LDSC) was employed to examine genetic association analysis. Furthermore, for traits with significant genetic associations, Mendelian randomization (MR) analyses were conducted.
Results:
The LDSC analysis revealed nominal genetic correlations between AgeSmk and right precentral surface area, left caudal anterior cingulate surface area, left cuneus surface area, left inferior parietal surface area, and right caudal anterior cingulate thickness, as well as between CigDay and left caudal middle frontal surface area, between SmkCes and left entorhinal thickness, and between SmkInit and left rostral anterior cingulate surface area, right rostral anterior cingulate thickness, and right superior frontal thickness (rg=-0.36–0.29, p<0.05). MR analysis showed a unidirectional causal association between left caudal middle frontal surface area and CigDay (βIVW=0.056, pBonferroni=2×10-4).
Conclusion
Left caudal middle frontal surface area has the potential to serve as a significant predictor of smoking behavior.
8.Research progress on the chemical constituents,pharmacological mechanisms and clinical application of Jiegeng decoction
Yun HUANG ; Shunwang HUANG ; Jinwei QIAO ; Qian XU ; Xiaoming GAO ; Xuemei BAO ; Manqin YANG ; Ruonan XIE ; Ming CAI
China Pharmacy 2025;36(18):2348-2352
Jiegeng decoction is a classic prescription composed of two Chinese medicinal herbs: Platycodon grandiflorum and Glycyrrhiza uralensis. It has the efficacy of diffusing lung qi, resolving phlegm, relieving sore throat and discharging pus, and is commonly used in the treatment of respiratory diseases such as cough and pharyngodynia. This article reviews the chemical components, pharmacological mechanisms and clinical applications of Jiegeng decoction. It was found that Jiegeng decoction contains triterpenoid saponins, flavonoids, glycosides, acids, and other components, with platycodin D, platycodin D2, glycyrrhizic acid, glycyrrhetinic acid, liquiritin, etc., serving as the main active pharmaceutical ingredients. Jiegeng decoction and its chemical constituents exert anti-inflammatory effects by inhibiting signaling pathways such as nuclear factor-κB and mitogen- activated protein kinases, and demonstrate anti-tumor activities through mechanisms like modulating the tumor immune microenvironment and promoting cancer cell apoptosis. Additionally, it exhibits various pharmacological actions including antibacterial, antiviral, and antioxidant effects. Clinically, Jiegeng decoction, its modified prescription and compound combinations are widely used in the treatment of respiratory diseases such as cough, pneumonia, and pharyngitis, as well as digestive system disorders like constipation.
9.Determination and Mechanism Prediction of Potential Active Ingredients in Erdongtang Based on UHPLC-QqQ-MS and Network Pharmacology
Xueyuan WANG ; Baoxin LI ; Shougang SHI ; Zhengjun HUANG ; Yuntao DAI ; Xuemei QIN
Chinese Journal of Experimental Traditional Medical Formulae 2024;30(5):28-34
ObjectiveTo establish an ultra-high performance liquid chromatography-tandem triple quadrupole mass spectrometry(UHPLC-QqQ-MS) for determination of the active ingredients in Erdongtang, and to predict the targets and pathways of anti-insulin resistance action of this formula. MethodThe analysis was performed on an ACQUITY UPLC BEH C18 column(2.1 mm×100 mm, 1.7 μm) with the mobile phase of 0.1% formic acid aqueous solution(A)-acetonitrile(B) for gradient elution(0-3 min, 90%-87%A; 3-6 min, 87%-86%A; 6-9 min, 86%-83%A; 9-11 min, 83%-75%A; 11-18 min, 75%-70%A; 18-19 min, 70%-52%A; 19-22 min, 52%A; 22-25 min, 52%-5%A; 25-27 min, 5%-90%A; 27-30 min, 90%A). The contents of active ingredients in Erdongtang was detected by electrospray ionization(ESI) and multiple reaction monitoring(MRM) mode under positive and negative ion modes. On this basis, network pharmacology was applied to predict the targets and pathways of Erdongtang exerting anti-insulin resistance effect. ResultThe 20 active ingredients in Erdongtang showed good linear relationships within a certain mass concentration range, and the precision, stability, repeatability and recovery rate were good. The results of determination showed that the ingredients with high content in 15 batches of samples were baicalein(1 259.39-1 635.78 mg·L-1), baicalin(1 078.37-1 411.52 mg·L-1), the ingredients with medium content were mangiferin(148.59-217.04 mg·L-1), timosaponin BⅡ(245.10-604.89 mg·L-1), quercetin-3-O-glucuronide(89.30-423.26 mg·L-1), rutin(46.91-1 553.61 mg·L-1), glycyrrhizic acid(55.97-391.47 mg·L-1), neomangiferin(37.45-127.03 mg·L-1), nuciferine(0.89-63.48 mg·L-1), hyperoside(6.96-136.78 mg·L-1), liquiritin(30.89-122.78 mg·L-1), liquiritigenin(26.64-110.67 mg·L-1), protodioscin(58.57-284.26 mg·L-1), the ingredients with low content were wogonin(7.16-20.74 mg·L-1), pseudoprotodioscin(5.49-22.96 mg·L-1), ginsenoside Rb1(7.31-23.87 mg·L-1), ginsenoside Rg1(10.78-28.33 mg·L-1), ginsenoside Re(7.78-24.76 mg·L-1), ophiopogonin D(2.08-4.29 mg·L-1), methylophiopogonanone A(0.74-1.67 mg·L-1). The results of network pharmacology indicated that the mechanism of anti-insulin resistance exerted by Erdongtang might be related to the phosphatidylinositol 3-kinase/protein kinase B(PI3K/Akt) signaling pathway. ConclusionThe established UHPLC-QqQ-MS has the advantages of simple sample processing, strong exclusivity and high sensitivity, and can simultaneously determine the contents of the main ingredients from seven herbs in Erdongtang, which can lay the foundation for the development of Erdongtang compound preparations. The results of the network pharmacology can provide a reference for the mechanism study of Erdongtang in the treatment of type 2 diabetes mellitus.
10.Analysis on the Effects of PPARγ Gene-silenced Human Bone Marrow Stromal Cells on Hematopoietic Function in Mice with Bone Marrow Suppression
Xuemei WANG ; Chunlan HUANG ; Tiejun ZHOU
Acta Medicinae Universitatis Scientiae et Technologiae Huazhong 2024;53(1):6-12
Objective To investigate the effects of peroxisome proliferator activated receptor-gamma(PPARγ)gene silen-cing in human bone marrow stromal cells(HS-5)on hematopoietic function in bone marrow-suppressed mice,and to explore the potential mechanisms involved.Methods A bone marrow-suppressed mouse model was established by whole-body X-ray irradi-ation.Two hours after modeling,the mice were randomly divided into three groups:experimental group(intravenous injection of PPARγ RNAi-interfered HS-5 cells through the tail vein),control group(intravenous injection of PPARγ RNAi-uninterfered HS-5 cells through the tail vein),and blank group(intravenous injection of an equal amount of saline through the tail vein),with 5 mice in each group.Peripheral blood routine tests were performed before,24 hours after,1 week after,and 2 weeks after radio-therapy.In vitro osteogenic and adipogenic induction was performed in cells,and the cells were divided into experimental group(PPARγ RNAi-interfered HS-5 cells),control group(PPARγ-uninterfered HS-5 cells),and blank group(HS-5 cells without os-teogenic/adipogenic induction).Osteogenic/adipogenic staining was observed.The effects of PPARγ gene-silenced HS-5 cells on mouse bone marrow hematopoietic stem cells(HSCs)were detected by CCK-8 proliferation assay.The groups included experi-mental group(PPARγ RNAi-interfered HS-5 cells were co-cultured with mouse HSCs after 3 days of osteogenic induction dif-ferentiation),positive control group(HS-5 cells treated with 50 μmol/L PPARγ inhibitor were co-cultured with mouse HSCs af-ter 3 days of osteogenic induction differentiation),negative control group(PPARγ RNAi-uninterfered HS-5 cells were co-cul-tured with mouse HSCs after 3 days of osteogenic induction differentiation),and blank group(Mouse HSCs were cultured alone without co-culturing with HS-5 cells).Results After radiotherapy,the hematological parameters of mice in each group showed a decreasing trend initially,and then increased.One week after radiotherapy,there were significant differences in platelet and white blood cell levels among the three groups(experimental group>control group>blank group,all P<0.05).Two weeks after radiotherapy,there were significant differences in the percentage of adipocyte vacuole area among the three groups(experi-mental group<control group<blank group,all P<0.05).Pearson correlation analysis showed a negative correlation between hematological parameters and PPARγ expression levels(all P<0.05),as well as a negative correlation between hematological parameters and the percentage of adipocyte vacuole area(all P<0.05).After in vitro osteogenic/adipogenic induction differenti-ation,compared to the control group,the experimental group showed a significantly lower proportion of orange-red cells and a significantly higher proportion of red calcium nodules.After 3 days of osteogenic induction differentiation,the experimental group,positive control group,and negative control group of human bone marrow stromal cells were co-cultured with mouse HSCs,while HSCs were solely cultured in the blank group.The results showed that after 24 h,48 h and 72 h of co-culture,the A values of mouse HSC cells in the experimental group and positive control group were higher than those in the negative control group and blank group(all P<0.05).Conclusion Silencing of the PPARγ gene in HS-5 cells implanted into bone marrow-sup-pressed mice contributes to enhanced hematopoietic function in mice.After interference and silencing of the PPARγ gene,the os-teogenic differentiation ability of HS-5 cells is enhanced,while the adipogenic differentiation ability is weakened.Furthermore,osteogenic-induced HS-5 cells can further enhance the proliferation capacity of mouse HSCs.

Result Analysis
Print
Save
E-mail