Using Deep Learning in a Monocentric Study to Characterize Maternal Immune Environment for Predicting Pregnancy Outcomes in the Recurrent Reproductive Failure Patients.

Recurrent reproductive failure (RRF), such as recurrent pregnancy loss and repeated implantation failure, is characterized by complex etiologies and particularly associated with diverse maternal factors. It is currently believed that RRF is closely associated with the maternal environment, which is, in turn, affected by complex immune factors. Without the use of automated tools, it is often difficult to assess the interaction and synergistic effects of the various immune factors on the pregnancy outcome. As a result, the application of Artificial Intelligence (A.I.) has been explored in the field of assisted reproductive technology (ART). In this study, we reviewed studies on the use of A.I. to develop prediction models for pregnancy outcomes of patients who underwent ART treatment. A limited amount of models based on genetic markers or common indices have been established for prediction of pregnancy outcome of patients with RRF. In this study, we applied A.I. to analyze the medical information of patients with RRF, including immune indicators. The entire clinical samples set (561 samples) was divided into two sets: 90% of the set was used for training and 10% for testing. Different data panels were established to predict pregnancy outcomes at four different gestational nodes, including biochemical pregnancy, clinical pregnancy, ongoing pregnancy, and live birth, respectively. The prediction models of pregnancy outcomes were established using sparse coding, based on six data panels: basic patient characteristics, hormone levels, autoantibodies, peripheral immunology, endometrial immunology, and embryo parameters. The six data panels covered 64 variables. In terms of biochemical pregnancy prediction, the area under curve (AUC) using the endometrial immunology panel was the largest (AUC = 0.766, accuracy: 73.0%). The AUC using the autoantibodies panel was the largest in predicting clinical pregnancy (AUC = 0.688, accuracy: 78.4%), ongoing pregnancy (AUC = 0.802, accuracy: 75.0%), and live birth (AUC = 0.909, accuracy: 89.7%). Combining the data panels did not significantly enhance the effect on prediction of all the four pregnancy outcomes. These results give us a new insight on reproductive immunology and establish the basis for assisting clinicians to plan more precise and personalized diagnosis and treatment for patients with RRF.

NKG2D as a Cell Surface Marker on γδ-T Cells for Predicting Pregnancy Outcomes in Patients With Unexplained Repeated Implantation Failure.

Maternal immune tolerance to semi-allogeneic fetus is essential for a successful implantation and pregnancy. Growing evidence indicated that low cytotoxic activity of γδ-T cells, which is mediated by activation and inhibitory receptors, is important for establishment of maternal immune tolerant microenvironment. However, the correlation between receptors on peripheral blood γδ-T cells, such as NKG2D, CD158a, and CD158b, and pregnancy outcome in patients with unexplained repeated implantation failure (uRIF) remains unclear. In this study, the association between the expression level of these receptors and pregnancy outcome in patients with uRIF was investigated. Thirty-eight women with uRIF were enrolled and divided into two groups: successful group and failed group, according to the pregnancy outcome on different gestational periods. The percentage of NKG2D+ γδ-T cells in lymphocytes was significantly higher in uRIF patients who had failed clinical pregnancy in subsequent cycle, compared with those who had successful clinical pregnancy. However, there were no differences about the frequencies of CD158a+ and CD158b+ γδ-T cells between the successful and failed groups. The receiver operating characteristic curve exhibited that the optimal cut-off value of NKG2D+ γδ-T cells was 3.24%, with 92.3% sensitivity and 66.7% specificity in predicting clinical pregnancy failure in uRIF patients. The patients with uRIF were further divided into two groups, group 1 (NKG2D+ γδ-T cells <3.24%) and group 2 (NKG2D+ γδ-T cells ≥3.24%), based on the cut-off value. The live birth rate of patients in the group 1 and group 2 were 61.5 and 28.0%, respectively. Kaplan-Meier survival curve further suggested that the frequency of NKG2D+ γδ-T cells in lymphocytes negatively correlated with live birth rate in patients with uRIF. In conclusion, our study demonstrated that the frequency of peripheral blood NKG2D+ γδ-T cells among lymphocytes is a potential predictor for pregnancy outcome in uRIF patients.

Granzyme B-expressing γδ-T and NK cells as a predictor of clinical pregnancy failure in patients with unexplained repeated implantation failure.

The limited cytotoxicity of immune cells facilitates a successful establishment of pregnancy. However, the association between cytotoxic granules and unexplained repeated implantation failure (uRIF) remains unkown. Twenty-one fertile controls and 54 patients with uRIF were included in this study. The pregnancy outcomes were monitored at different gestational periods. The peripheral blood lymphocytes were detected using specific monoclonal antibodies by flow cytometry. The percentage of perforin+ (Pfr+), granzyme B+ (GrB+), or granulysin+ (Gnly+) lymphocytes was not significantly different among fertile controls, uRIF patients with successful pregnancy outcomes, and uRIF patients with pregnancy failure. The percentage of GrB+ γδ-T cells in lymphocytes was markedly higher in uRIF patients with implantation failure and clinical pregnancy failure than that in uRIF patients with a corresponding successful pregnancy outcome. A four-tier risk model showed that the risk of suffering clinical pregnancy failure in uRIF patients among high risk tier (83.3 %), normal risk tier (65.0 %) and low risk tier (39.1 %) was elevated by 2-4 fold compared with uRIF patients among lowest risk tier (20.0 %). In addition, the percentage of GrB+ NK cells in lymphocytes tended to decrease in uRIF patients with pregnancy failure. The AUC of the combined indicator with GrB+ γδ-T cells and GrB+ NK cells was increased than that of GrB+ γδ-T cells and GrB+ NK cells for predicting clinical pregnancy failure. In conclusion, the frequency of GrB-expressing γδ-T and NK cells in peripheral blood could serve as a predictor of clinical pregnancy failure in patients with uRIF.

The low cytotoxic activity of peripheral blood NK cells may relate to unexplained recurrent miscarriage.

PROBLEM: Unexplained recurrent miscarriage (uRM) is defined as two or more spontaneous abortions prior to 20 weeks of gestation with unknown etiology. Peripheral blood natural killer (pNK) cells contact with the villus and exert important role in normal pregnancy. However, it is still controversial about the association between pNK cytotoxicity and uRM, and the underlying mechanism remains unknown so far. METHOD OF STUDY: In this study, we aim to compare the percentage, immunophenotype, and function of pNK cells between patients with uRM and fertile controls. The peripheral blood was collected from 49 patients with uRM and 11 fertile women in their middle luteal phase of the menstrual cycle. pNK cells were co-cultured with K562 cells at different cell ratios to measure the cytotoxicity. The percentage of CD3- CD56+ , CD3- CD56bright , and CD3- CD56dim pNK was analyzed by flow cytometry and quantified to evaluate the expression of cytotoxic granules (granzyme B, granulysin, and perforin), and the cell surface receptors related to pNK cell cytotoxicity (NKG2D, NKp30, NKp46, CD158a, and CD158b) were also detected. RESULTS: The general linear model analysis showed that pNK cell cytotoxicity in patients with uRM was significantly lower than that in fertile controls. In addition, the ratios of NKG2D/CD158a, NKp30/CD158a, and NKp46/CD158a in CD3- CD56bright pNK subsets were significantly lower in uRM group than that in fertile control. The logistical regression analysis showed that the reduced NKp30/CD158a, NKp46/CD158a ratios in CD3- CD56bright pNK subsets were significantly associated with uRM. CONCLUSION: Our results suggested that a low pNK cytotoxicity, which is mediated by inhibitory signals, might be associated with uRM.

The number and cytotoxicity and the expression of cytotoxicity-related molecules in peripheral natural killer (NK) cells do not predict the repeated implantation failure (RIF) for the in vitro fertilization patients.

Natural killer (NK) cells are thought to play a key role in the successful establishment of a pregnancy by facilitating immunological adaptation of the semi-allogeneic developing embryo. The aim of this study was to explore the cell number, immunophenotypic characteristics, and activities of peripheral blood NK cells in women with repeated implantation failure (RIF). Peripheral blood was obtained from 27 women with RIF and 11 healthy, fertile controls during the middle luteal phase of the menstrual cycle. CD3- CD56+ NK cells were quantified and analyzed by flow cytometry for the expression of cytolytic molecules (granzyme B, granulysin, and perforin) as well as cell surface receptors responsible for NK cell activation or inhibition (NKG2D, NKp30, NKp46, CD158a, CD158b). NK cytotoxicity was measured at three effector-to-target cell ratios. Women with RIF and fertile controls did not differ significantly in the percentage of circulating CD3-CD56+ NK cells, or in the proportions of these cells that expressed granzyme B, granulysin, or perforin. The two groups also did not differ significantly in the proportions of NK cells expressing the receptors NKG2D, NKp30, NKp46, CD158a or CD158b. General linear model analysis showed that NK cytotoxicity increased with effector-to-target cell ratio. However, NK cytotoxicity did not differ significantly between patients with RIF and fertile controls. These results suggest that RIF is not associated with significant alterations in the number or function of peripheral blood NK cells.

Chronic hepatitis B infection alters peripheral immune response in women with reproductive failure.

PROBLEM: Does hepatitis B infection affect peripheral blood immune response in women with reproductive failure? METHOD OF STUDY: Two hundred and twenty-seven women, including 10 HBsAg+ HBeAg+ , 27 HBsAg+ HBeAg- hepatitis sero-positive women, and 190 women without HBV infection, formed the study population. Their peripheral immune responses containing lymphocyte subsets, cytokine production, expression of cell surface markers and intracellular toxicity molecules, and pregnancy outcomes were retrospectively compared. RESULTS: Comparing with HBsAg+ HBeAg- carriers and HBsAg- group, HBsAg+ HBeAg+ group had lower rates of CD3+ CD4+ helper T cells (31.7% vs 38.0% and 36.8%, P < 0.05, respectively), but higher frequency of CD19+ B cells (17.8% vs 14.0% and 13.2%, P < 0.05 and P < 0.01, respectively). NK cells in HBsAg+ HBeAg+ patients showed lower cytotoxic activity than that in two other groups (P < 0.05). Comparing with HBsAg- patients, HBsAg+ HBeAg+ group exhibited decreased expression of the activating receptor NKG2D (56.2% vs 66.1%, P < 0.05), as well as reduced expression of granzyme B (54.8% vs 70.5%, P < 0.05), perforin (49.9% vs 65.0%, P < 0.05), and granulysin (52.0% vs 67.9%, P < 0.01). Generally, a higher clinical pregnancy rate (85.7% vs 56.9%) and higher early miscarriage rate (33.3% vs 20.3%) were noticed in HBsAg+ HBeAg+ group than HBsAg- group. CONCLUSION: Chronic HBV infection alters peripheral immune responses by upregulating B-cell frequency, decreasing CD3+ CD4+ helper T cells, and decreasing peripheral NK function and toxicity. These may influence pregnancy outcome on HBV-infected patients, and the pathogenesis of HBV infection on pregnancy outcome deserves to be further studied.

RGD/TAT-functionalized chitosan-graft-PEI-PEG gene nanovector for sustained delivery of NT-3 for potential application in neural regeneration.

In this study, we prepared a multifunctional gene delivery nanovector containing a chitosan (CS) backbone and polyethylenimine (PEI) arms with arginine-glycine-aspartate (RGD)/twin-arginine translocation (TAT) conjugated via polyethylene glycol (PEG). Branched PEI, with a molecular weight of 2000 Da, was used to achieve a balance between biocompatibility and transfection efficiency, whereas RGD/TAT peptides were conjugated for enhanced targeting ability and cellular uptake. Synthesis of the copolymers was confirmed by characterizing the chemical structure with 1H nuclear magnetic resonance and Fourier Transform Infrared Spectroscopy (FTIR). The nanovector was biocompatible with cells and showed excellent capability for DNA condensation; the resulting complexes with DNA were well-formed, and possessed small particle size and reasonable positive charge. Higher gene transfection efficiency, compared to that achieved with PEI (25 kDa), was confirmed in tumor (HeLa cells) and normal cells (293T and NIH 3T3 cells). More importantly, the cells transfected with the chitosan-graft-PEI-PEG/pCMV-EGFP-Ntf3 complex produced sustained neurotrophin-3 with a linear increase in cumulative concentration, which induced neuronal differentiation of neural stem cell and promoted neurite outgrowth. These findings suggested that our multifunctional copolymers might be ideal nanovectors for engineering cells via gene transfection, and could potentially be applied in tumor therapy and regenerative medicine. STATEMENT OF SIGNIFICANCE: We successfully prepared a multifunctional gene delivery nanovector containing branched PEI with a molecular weight of 2000 Da to balance between biocompatibility and transfection efficiency, and RGD/TAT peptides for enhanced targeting ability and cellular uptake. The well-formed CPPP/DNA complexes of small particle size and reasonable positive charges potentially enhanced gene transfection in both tumor and normal cells. More importantly, the CPPP/pCMV-EGFP-Ntf3 complex-transfected 293T cells could produce sustained NT-3 with a constant ratio, which induced neuron differentiation of NSC and promoted neurite outgrowth. Therefore, our study provided an effective strategy for producing neurotrophins by engineering cells with gene delivery, which deserved wide investigation and potential application in regenerative medicine.

Self-assembly behaviors of molecular designer functional RADA16-I peptides: influence of motifs, pH, and assembly time.

In the current study, we present three designer self-assembling peptides (SAPs) by appending RADA 16-I with epitopes IKVAV, RGD, and YIGSR, which have different net charges and amphiphilic properties at neutral pH. The self-assembly of the designer SAPs is intensively investigated as a function of pH, canion type, and assembly time. The morphologies of the designer SAPs were studied by atomic force microscope. The secondary structure was investigated by circular dichroism. The dynamic viscoelasticity of designer SAP solutions was examined during titration with different alkaline reagents. Our study indicated that both electrostatic and hydrophilic/hydrophobic interactions of the motifs exhibited influences on the self-assembly, consequentially affecting the fiber morphologies and rheological properties. Moreover, NaOH induced a quicker assembly/reassembly of the designer SAPs than Tris because of its strong ionic strength. Therefore, our study gained comprehensive insight into the self-assembling mechanism as references for developing RADA 16-I-based functional SAPs.

Fabrication of carbon nanotube nanocomposites via layer-by-layer assembly and evaluation in biomedical application.

AIM: To investigate the influence of ion pairing of carboxylated multiwalled carbon nanotubes (MWCNT-COOH) and polycations on the layer-by-layer assembly of nanocomposites and their biocompatibility for biomedical applications. MATERIALS & METHODS: Strong polycation poly(dimethyldiallylammonium chloride), and weak polycations chitosan and polyethyleneimine were selected to assembly with MWCNT-COOH. The MWCNT-COOH/polycation nanocomposites were analyzed by their physicochemical, electrical properties and biocompatibility. RESULTS: The ion pairing of CNTs/polyelectrocytes played a critical role in the layer-by-layer assembly. Strong interactions between MWCNTs and poly(dimethyldiallylammonium chloride) produced thicker nanocomposites with rougher surfaces, higher MWCNT mass and better conductance. All the MWCNT multilayered nanocomposites were of good biocompatibility. CONCLUSION: The MWCNT multilayered nanocomposites hold high potential for biomedical applications.

Cytotoxic and adhesion-associated response of NIH-3T3 fibroblasts to COOH-functionalized multi-walled carbon nanotubes.

As novel, promising, man-made nanomaterials with extraordinary properties, carbon nanotubes have been attracting massive attention in regenerative medicine. However, published reports on their potential cytotoxic effects are not concordant and are even conflicting. In the current study, the cytotoxic effects of carboxyl-modified multi-walled carbon nanotubes (COOH-MWCNTs), as well as their influences on the cell adhesion of NIH-3T3 fibroblasts, were thoroughly investigated. Live/dead cell viability assay and cell counting kit-8 assay both indicated that the viability of the NIH-3T3 cells exposed to COOH-MWCNTs in the culture medium was dependent on the latter's concentration. Cell viability increased at COOH-MWCNT concentrations below 50 µg ml(-1) and then decreased with increasing concentration. Scanning electron microscopy and immunofluorescent staining of the NIH-3T3 cells revealed that the cells were well adherent to the substrate after exposure to the COOH-MWCNTs for 48 h. Western blot demonstrated that COOH-MWCNT exposure enhanced the expression of adhesion-associated proteins compared with normal cells, peaking at an intermediate concentration. Our study showed that the cytotoxicity of COOH-MWCNTs, as well as their effects on NIH-3T3 fibroblast adhesion, was dose dependent. Therefore, COOH-MWCNT concentrations in the cell culture medium should be considered in the biomedical application of COOH-MWCNTs.

Functional Self-Assembling Peptide Nanofiber Hydrogels Designed for Nerve Degeneration.

Self-assembling peptide (SAP) RADA16-I (Ac-(RADA)4-CONH2) has been suffering from a main drawback associated with low pH, which damages cells and host tissues upon direct exposure. In this study, we presented a strategy to prepare nanofiber hydrogels from two designer SAPs at neutral pH. RADA16-I was appended with functional motifs containing cell adhesion peptide RGD and neurite outgrowth peptide IKVAV. The two SAPs were specially designed to have opposite net charges at neutral pH, the combination of which created a nanofiber hydrogel (-IKVAV/-RGD) characterized by significantly higher G' than G″ in a viscoelasticity examination. Circular dichroism, Fourier transform infrared spectroscopy, and Raman measurements were performed to investigate the secondary structure of the designer SAPs, indicating that both the hydrophobic/hydrophilic properties and electrostatic interactions of the functional motifs play an important role in the self-assembling behavior of the designer SAPs. The neural progenitor cells (NPCs)/stem cells (NSCs) fully embedded in the 3D-IKVAV/-RGD nanofiber hydrogel survived, whereas those embedded within the RADA 16-I hydrogel hardly survived. Moreover, the -IKVAV/-RGD nanofiber hydrogel supported NPC/NSC neuron and astrocyte differentiation in a 3D environment without adding extra growth factors. Studies of three nerve injury models, including sciatic nerve defect, intracerebral hemorrhage, and spinal cord transection, indicated that the designer -IKVAV/-RGD nanofiber hydrogel provided a more permissive environment for nerve regeneration than the RADA 16-I hydrogel. Therefore, we reported a new mechanism that might be beneficial for the synthesis of SAPs for in vitro 3D cell culture and nerve regeneration.

Nano-Engineered Environment for Nerve Regeneration: Scaffolds, Functional Molecules and Stem Cells.

One of the most complex systems in the human body is the nervous system, which is divided into the central and peripheral nervous systems. The regeneration of the CNS is a complex and challenging biological phenomenon hindered by the low regenerative capacity of neurons and the prohibition factors in response to nerve injuries. To date, no effective approach can achieve complete recovery and fully restore the functions of the nervous system once it has been damaged. Developments in neuroscience have identified properties of the local environment with a critical role in nerve regeneration. Advances in biomaterials and biomedical engineering have explored new approaches of constructing permissive environments for nerve regeneration, thereby enabling optimism with regard to nerve-injury treatment. This article reviews recent progress in nanoengineered environments for aiding nerveinjury repair and regeneration, including nanofibrous scaffolds, functional molecules, and stem cells.