Search

DNA ploidy analysis of cells is an automation technique applied in pathological diagnosis. It is important for this technique to classify various nuclei images accurately. However, the lack of overlapping nuclei images in training data (imbalanced training data) results in low recognition rates of overlapping nuclei images. To solve this problem, a new method which synthesizes overlapping nuclei images with single-nuclei images is proposed. Firstly, sample selection is employed to make the synthesized samples representative. Secondly, random functions are used to control the rotation angles of the nucleus and the distance between the centroids of the nucleus, increasing the sample diversity. Then, the Lambert-Beer law is applied to reassign the pixels of overlapping parts, thus making the synthesized samples quite close to the real ones. Finally, all synthesized samples are added to the training sets for classifier training. The experimental results show that images synthesized by this method can solve the data set imbalance problem and improve the recognition rate of DNA ploidy analysis systems.

β-1,3-glucanosyltransferases play essential roles in cell wall biosynthesis in yeast. Kluyveromyces lactis has six putative β-1,3-glucanosyltransferase genes. KlGAS1-1 and KlGAS1-2 are homologs of Saccharomyces cerevisiae gene GAS1. RT-qPCR indicated the transcription level of KlGAS1-1 was significantly reduced while heterologous protein (thermostable xylanase B) secretion was enhanced during medium optimization. To evaluate if these two events were related, and to improve xylanase B secretion in K. lactis, we constructed KlGAS1-1 and KlGAS1-2 single deletion strains and double deletion strain, respectively. KlGAS1-1 gene deletion resulted in the highest xylanase B activity among the three mutants. Only the double deletion strain showed morphology similar to that of the GAS1 deletion mutant in S. cerevisiae. The two single deletion strains differed in terms of cell wall thickness and xylanase B secretion. Transcription levels of β-1,3-glucanosyltransferase genes and genes related to protein secretion and transport were assayed. The β-1,3-glucanosyltransferase genes displayed transcription complementation in the cell wall synthesis process. KlGAS1-1 and KlGAS1-2 affected transcription levels of secretion- and transport-related genes. Differences in protein secretion ratio among the three deletion strains were associated with changes of transcription levels of secretion- and transport-related genes. Our findings indicate that KlGAS1-1 deletion is an effective tool for enhancing industrial-scale heterologous protein secretion in K. lactis.

An extremely thermostable xylanase gene, xynB, from hyperthermophilic bacterium Thermotoga maritima MSB8 was successful expressed in Kluyveromyces lactis. Response surface methodology (RSM) was applied to optimize medium components for production of XynB secreted by the recombinant K. lactis. Secretion level (102 mg/L) and enzyme activity (49 U/ml) of XynB in the optimized medium (yeast extract, lactose, and urea; YLU) were much higher than those (56 mg/L, 16 U/ml) in original medium (yeast extract, lactose, and peptone; YLP). It was also observed that the secretory efficiency of mature XynB was improved by the YLU medium. mRNA levels of 13 characterized secretion-related genes between K. lactis cultured in YLP and YLU were detected using semi-quantitative RT-PCR method. It was found that unfolded protein response (UPR) related genes such as ero1, hac1, and kar2 were up-regulated in K. lactis cultured in YLU. Therefore, nutrient ingredient, especially nitrogen source had a significant influence on the XynB secretory efficiency in the host K. lactis.