Search

A multi-modal 3D object-detection method, based on data from cameras and LiDAR, has become a subject of research interest. PointPainting proposes a method for improving point-cloud-based 3D object detectors using semantic information from RGB images. However, this method still needs to improve on the following two complications: first, there are faulty parts in the image semantic segmentation results, leading to false detections. Second, the commonly used anchor assigner only considers the intersection over union (IoU) between the anchors and ground truth boxes, meaning that some anchors contain few target LiDAR points assigned as positive anchors. In this paper, three improvements are suggested to address these complications. Specifically, a novel weighting strategy is proposed for each anchor in the classification loss. This enables the detector to pay more attention to anchors containing inaccurate semantic information. Then, SegIoU, which incorporates semantic information, instead of IoU, is proposed for the anchor assignment. SegIoU measures the similarity of the semantic information between each anchor and ground truth box, avoiding the defective anchor assignments mentioned above. In addition, a dual-attention module is introduced to enhance the voxelized point cloud. The experiments demonstrate that the proposed modules obtained significant improvements in various methods, consisting of single-stage PointPillars, two-stage SECOND-IoU, anchor-base SECOND, and an anchor-free CenterPoint on the KITTI dataset.

Hedgehog (Hh) signalling plays essential roles in regulating embryonic development and contributes to tumour initiation, growth and progression in multiple cancers. The detailed mechanism by which Hh signalling participates in tumour growth warrants thorough study, although several downstream target genes have been identified. Herein, a set of novel targets of Hh signalling was identified in multiple types of tumour cells via RNA-Seq analysis. Among these targets, the expression regulation and oncogenic function of the extracellular matrix component biglycan (BGN) were investigated. Further investigation verified that Hh signalling activates the expression of BGN via the transcription factor Gli2, which directly binds to the promoter region of BGN. Functional assays revealed that BGN facilitates tumour cell growth and proliferation in colorectal cancer (CRC) cells, and xenograft assays confirmed that BGN also promotes tumour growth . Moreover, analysis of clinical CRC samples showed that both the protein and mRNA levels of BGN are increased in CRC tissues compared to those in adjacent tissues, and higher expression of BGN is correlated with poorer prognosis of CRC patients, further confirming the function of BGN in CRC. Taken together, aberrantly activated Hh signalling increases the expression of BGN, possibly regulates the extracellular matrix, and thereby promotes tumour growth in CRC.

Continuous frames of point-cloud-based object detection is a new research direction. Currently, most research studies fuse multi-frame point clouds using concatenation-based methods. The method aligns different frames by using information on GPS, IMU, etc. However, this fusion method can only align static objects and not moving objects. In this paper, we proposed a non-local-based multi-scale feature fusion method, which can handle both moving and static objects without GPS- and IMU-based registrations. Considering that non-local methods are resource-consuming, we proposed a novel simplified non-local block based on the sparsity of the point cloud. By filtering out empty units, memory consumption decreased by 99.93%. In addition, triple attention is adopted to enhance the key information on the object and suppresses background noise, further benefiting non-local-based feature fusion methods. Finally, we verify the method based on PointPillars and CenterPoint. Experimental results show that the mAP of the proposed method improved by 3.9% and 4.1% in mAP compared with concatenation-based fusion modules, PointPillars-2 and CenterPoint-2, respectively. In addition, the proposed network outperforms powerful 3D-VID by 1.2% in mAP.

A hydroxynitrile lyase (HNL) from the millipede Chamberlinius hualienensis has high potential for industrial use in the synthesis of cyanohydrins. However, obtaining sufficient amounts of millipedes is difficult, and the production of the Chamberlinius hualienensis HNL (ChuaHNL) in E. coli has not been very successful. Therefore, we investigated the conditions required for high-yield heterologous production of this enzyme using Pichia pastoris. When we employed P. pastoris to express His-ChuaHNL, the yield was very low (22.6 ± 3.8 U/L culture). Hence, we investigated the effects of ChuaHNL codon optimization and the co-production of two protein disulfide isomerases (PDIs) [from P. pastoris (PpPDI) and C. hualienensis (ChuaPDI1, ChuaPDI2)] on His-ChuaHNL production. The productivity of His-ChuaHNL was increased approximately 140 times per unit culture to 3170 ± 144.7 U/L by the co-expression of codon-optimized ChuaHNL and PpPDI. Moreover, we revealed that the N-glycosylation on ChuaHNL had a large effect on the stability, enzyme secretion, and catalytic properties of ChuaHNL in P. pastoris. This study demonstrates an economical and efficient approach for the production of HNL, and the data show that glycosylation has a large effect on the enzyme properties and the P. pastoris expression system.

We determined the X‐ray crystallographic structure of a unique hydroxynitrile lyase from the cyanogenic millipede, Chamberlinius hualienensis, and elucidated the reaction mechanism. We revealed that it belongs to the lipocalins, a family of proteins active in the transport small hydrophobic molecules. It is one of a few lipocalins with enzyme activities. This addition expands the evolutionary relationship of the enzyme to a new family of proteins., Hydroxynitrile lyases (HNLs) catalyze the cleavage of cyanohydrin into cyanide and the corresponding aldehyde or ketone. Moreover, they catalyze the synthesis of cyanohydrin in the reverse reaction, utilized in industry for preparation of enantiomeric pure pharmaceutical ingredients and fine chemicals. We discovered a new HNL from the cyanogenic millipede, Chamberlinius hualienensis. The enzyme displays several features including a new primary structure, high stability, and the highest specific activity in (R)‐mandelonitrile ((R)‐MAN) synthesis (7420 U·mg−1) among the reported HNLs. In this study, we elucidated the crystal structure and reaction mechanism of natural ChuaHNL in ligand‐free form and its complexes with acetate, cyanide ion, and inhibitors (thiocyanate or iodoacetate) at 1.6, 1.5, 2.1, 1.55, and 1.55 Å resolutions, respectively. The structure of ChuaHNL revealed that it belongs to the lipocalin superfamily, despite low amino acid sequence identity. The docking model of (R)‐MAN with ChuaHNL suggested that the hydroxyl group forms hydrogen bonds with R38 and K117, and the nitrile group forms hydrogen bonds with R38 and Y103. The mutational analysis showed the importance of these residues in the enzymatic reaction. From these results, we propose that K117 acts as a base to abstract a proton from the hydroxyl group of cyanohydrins and R38 acts as an acid to donate a proton to the cyanide ion during the cleavage reaction of cyanohydrins. The reverse mechanism would occur during the cyanohydrin synthesis. (Photo: Dr. Yuko Ishida) Databases Structural data are available in PDB database under the accession numbers 6JHC, 6KFA, 6KFB, 6KFC, and 6KFD.

Methanol-inducible gene promoters in methanol-utilizing yeasts are used in high-level heterologous gene expression systems. Generally, expression of methanol-inducible genes is completely repressed by the presence of glucose. In this study we identified the MIG1 gene in Candida boidinii, which encodes a homologue of the glucose repressor Mig1p of Saccharomyces cerevisiae. Disruption of the CbMIG1 gene had no growth effect on various carbon sources. Activation of the methanol-inducible AOD1 gene, which encodes alcohol oxidase, was increased in the early stage of methanol induction when cells of the CbMIG1-disrupted strain were transferred from glucose medium to methanol medium. Furthermore, CbMig1p tagged with yellow fluorescent protein was primarily localized in the nucleus of glucose-grown cells, but was diffuse in the cytosol of methanol-grown cells. This cytosolic diffusion in methanol-grown cells occurred in a CbMsn5p-dependent manner. These results suggest that CbMig1p is involved in negative regulation of methanol-inducible gene expression in C. boidinii. Copyright © 2012 John Wiley & Sons, Ltd.

Methylotrophic yeasts, which can utilize methanol as sole carbon and energy source, are exposed to two toxic metabolic intermediates, formaldehyde and hydrogen peroxide, during growth on methanol. Here we report that Msn5p, an importin-β family nuclear exporter, participated in the formaldehyde resistance mechanism but not in the hydrogen peroxide resistance mechanism in Candida boidinii. Disruption of the MSN5 gene in this yeast caused retardation of growth on formaldehyde-generating growth substrates such as methanol and methylamine, but the expression levels of the methanol-metabolizing enzymes did not fall. The Msn5p-depleted strain was sensitive to formaldehyde but not to hydrogen peroxide. Furthermore, a yellow fluorescent protein-tagged Msn5p was diffuse in the cytoplasm of C. boidinii when the cells were treated with high concentrations of formaldehyde or ethanol, but was predominantly associated with the nuclei following treatment with hydrogen peroxide.