Many organic ligands were synthesized to recognize G-quadruplexes. However, different kinds of G-quadruplexes (G4s) possess different structures and functions. Therefore, selective recognition of certain types of G4s is important for the study of G4s. In this paper, a novel cyanine dye, 3-(2-(4-vinylpyridine))-6-(2-((1-(4-sulfobutyl))-3,3-dimethyl-2-vinylbenz[e]indole)-9-ethyl-carbazole (9E PBIC), composed of benzindole and carbazole was designed and synthesised. The studies on UV-vis and fluorescence properties of the dye with different DNA forms showed that the dye exhibits almost no fluorescence under aqueous buffer conditions, but it increased over 100 fold in the presence of c-myc G4 and 10-30 fold in the presence of other G4s, while little in the presence of single/double-stranded DNA, indicating that it has excellent selectivity to c-myc 2345 G4. For the binding studies the dye is interacted with the c-myc 2345 G-quadruplex by using the end-stack binding model. It can be said that the dye is an excellent targeting fluorescent probe for c-myc G-quadruplexes.

In the current study, the lipid-shell and polymer-core hybrid nanoparticles (lpNPs) modified by Arg–Gly–Asp(RGD) peptide, loaded with curcumin (Cur), were developed by emulsification-solvent volatilization method. The RGD-modified hybrid nanoparticles (RGD–lpNPs) could overcome the poor water solubility of Cur to meet the requirement of intravenous administration and tumor active targeting. The obtained optimal RGD-lpNPs, composed of PLGA (poly(lactic-co-glycolic acid))–mPEG (methoxyl poly(ethylene- glycol)), RGD–polyethylene glycol (PEG)–cholesterol (Chol) copolymers and lipids, had good entrapment efficiency, submicron size and negatively neutral surface charge. The core-shell structure of RGD–lpNPs was verified by TEM. Cytotoxicity analysis demonstrated that the RGD–lpNPs encapsulated Cur retained potent anti-tumor effects. Flow cytometry analysis revealed the cellular uptake of Cur encapsulated in the RGD–lpNPs was increased for human umbilical vein endothelial cells (HUVEC). Furthermore, Cur loaded RGD–lpNPs were more effective in inhibiting tumor growth in a subcutaneous B16 melanoma tumor model. The results of immunofluorescent and immunohistochemical studies by Cur loaded RGD–lpNPs therapies indicated that more apoptotic cells, fewer microvessels, and fewer proliferation-positive cells were observed. In conclusion, RGD–lpNPs encapsulating Cur were developed with enhanced anti-tumor activity in melanoma, and Cur loaded RGD–lpNPs represent an excellent tumor targeted formulation of Cur which might be an attractive candidate for cancer therapy.

AI-based protein structure prediction pipelines, such as AlphaFold2, have achieved near-experimental accuracy. These advanced pipelines mainly rely on Multiple Sequence Alignments (MSAs) and templates as inputs to learn the co-evolution information from the homologous sequences. Nonetheless, searching MSAs and templates from protein databases is time-consuming, usually taking dozens of minutes. Consequently, we attempt to explore the limits of fast protein structure prediction by using only primary sequences of proteins. HelixFold-Single is proposed to combine a largescale protein language model with the superior geometric learning capability of AlphaFold2. Our proposed method, HelixFold-Single, first pre-trains a large-scale protein language model (PLM) with thousands of millions of primary sequences utilizing the self-supervised learning paradigm, which will be used as an alternative to MSAs and templates for learning the co-evolution information. Then, by combining the pre-trained PLM and the essential components of AlphaFold2, we obtain an end-to-end differentiable model to predict the 3D coordinates of atoms from only the primary sequence. HelixFold-Single is validated in datasets CASP14 and CAMEO, achieving competitive accuracy with the MSA-based methods on the targets with large homologous families. Furthermore, HelixFold-Single consumes much less time than the mainstream pipelines for protein structure prediction, demonstrating its potential in tasks requiring many predictions. The code of HelixFold-Single is available at https://github.com/PaddlePaddle/PaddleHelix/tree/ dev/apps/protein_folding/helixfold-single, and we also provide stable web services on https://paddlehelix.baidu.com/app/drug/protein-single/forecast.

AI-based protein structure prediction pipelines, such as AlphaFold2, have achieved near-experimental accuracy. These advanced pipelines mainly rely on Multiple Sequence Alignments (MSAs) as inputs to learn the co-evolution information from the homologous sequences. Nonetheless, searching MSAs from protein databases is time-consuming, usually taking dozens of minutes. Consequently, we attempt to explore the limits of fast protein structure prediction by using only primary sequences of proteins. HelixFold-Single is proposed to combine a large-scale protein language model with the superior geometric learning capability of AlphaFold2. Our proposed method, HelixFold-Single, first pre-trains a large-scale protein language model (PLM) with thousands of millions of primary sequences utilizing the self-supervised learning paradigm, which will be used as an alternative to MSAs for learning the co-evolution information. Then, by combining the pre-trained PLM and the essential components of AlphaFold2, we obtain an end-to-end differentiable model to predict the 3D coordinates of atoms from only the primary sequence. HelixFold-Single is validated in datasets CASP14 and CAMEO, achieving competitive accuracy with the MSA-based methods on the targets with large homologous families. Furthermore, HelixFold-Single consumes much less time than the mainstream pipelines for protein structure prediction, demonstrating its potential in tasks requiring many predictions. The code of HelixFold-Single is available at https://github.com/PaddlePaddle/PaddleHelix/tree/dev/apps/protein_folding/helixfold-single, and we also provide stable web services on https://paddlehelix.baidu.com/app/drug/protein-single/forecast.

No abstract