Preparation, Properties and Therapeutic Effect of a TPL Nanoparticle Thermosensitive Gel for Intra-Articular Injection.

Most injectable preparations for the articular cavity are solution-type preparations that are frequently administered because of rapid elimination. In this study, triptolide (TPL), an effective ingredient in the treatment of rheumatoid arthritis (RA), was prepared in the form of a nanoparticle thermosensitive gel (TPL-NS-Gel). The particle size distribution and gel structure were investigated by TEM, laser particle size analysis and laser capture microdissection. The effect of the nanoparticle carrier material PLGA on the phase transition temperature was investigated by 1H variable temperature NMR and DSC. The tissue distribution, pharmacokinetic behavior, four inflammatory factors and therapeutic effect were determined in a rat RA model. The results suggested that PLGA increased the gel phase transition temperature. The drug concentration of the TPL-NS-Gel group in joint tissues was higher than that in other tissues at different time points, and the retention time was longer than that of the TPL-NS group. After 24 days of administration, TPL-NS-Gel significantly improved the joint swelling and stiffness of the rat models, and the improvement degree was better than that of the TPL-NS group. TPL-NS-Gel significantly decreased the levels of hs-CRP, IL-1, IL-6 and TNF-α in serum and joint fluid. There was a significant difference between the TPL-NS-Gel and TPL-NS groups on Day 24 (p < 0.05). Pathological section results showed that inflammatory cell infiltration was lower in the TPL-NS-Gel group, and no other obvious histological changes were observed. Upon articular injection, the TPL-NS-Gel prolonged drug release, reduced the drug concentration outside the articular tissue and improved the therapeutic effect in a rat RA model. The TPL-NS-Gel can be used as a new type of sustained-release preparation for articular injection.

Thiol-Aldehyde Polycondensation for Bio-based Adaptable and Degradable Phenolic Polymers.

Designing sustainable materials with tunable mechanical properties, intrinsic degradability, and recyclability from renewable biomass through a mild process has become vital in polymer science. Traditional phenolic resins are generally considered to be not degradable or recyclable. Here we report the design and synthesis of linear and network structured phenolic polymers using facile polycondensation between natural aldehyde-bearing phenolic compounds and polymercaptans. Linear phenolic products are amorphous with Tg between -9 °C and 12 °C. Cross-linked networks from vanillin and its di-aldehyde derivative exhibited excellent mechanical strength between 6-64 MPa. The connecting dithioacetals are associatively adaptable strong bonds and susceptible to degradation in oxidative conditions to regenerate vanillin. These results highlight the potential of biobased sustainable phenolic polymers with recyclability and selective degradation, as a complement to the traditional phenol-formaldehyde resins.

Class-Wise Denoising for Robust Learning Under Label Noise.

Label noise is ubiquitous in many real-world scenarios which often misleads training algorithm and brings about the degraded classification performance. Therefore, many approaches have been proposed to correct the loss function given corrupted labels to combat such label noise. Among them, a trend of works achieve this goal by unbiasedly estimating the data centroid, which plays an important role in constructing an unbiased risk estimator for minimization. However, they usually handle the noisy labels in different classes all at once, so the local information inherited by each class is ignored which often leads to unsatisfactory performance. To address this defect, this paper presents a novel robust learning algorithm dubbed "Class-Wise Denoising" (CWD), which tackles the noisy labels in a class-wise way to ease the entire noise correction task. Specifically, two virtual auxiliary sets are respectively constructed by presuming that the positive and negative labels in the training set are clean, so the original false-negative labels and false-positive ones are tackled separately. As a result, an improved centroid estimator can be designed which helps to yield more accurate risk estimator. Theoretically, we prove that: 1) the variance in centroid estimation can often be reduced by our CWD when compared with existing methods with unbiased centroid estimator; and 2) the performance of CWD trained on the noisy set will converge to that of the optimal classifier trained on the clean set with a convergence rate [Formula: see text] where n is the number of the training examples. These sound theoretical properties critically enable our CWD to produce the improved classification performance under label noise, which is also demonstrated by the comparisons with ten representative state-of-the-art methods on a variety of benchmark datasets.