Genome-wide identification and expression analysis of GL2-interacting-repressor (GIR) genes during cotton fiber and fuzz development.

MAIN CONCLUSION: GL2-interacting-repressor (GIR) family members may contribute to fiber/fuzz formation via a newly discovered unique pathway in Gossypium arboreum. There are similarities between cotton fiber development and the formation of trichomes and root hairs. The GL2-interacting-repressors (GIRs) are crucial regulators of root hair and trichome formation. The GaFzl gene, annotated as GaGIR1, is negatively associated with trichome development and fuzz initiation. However, there is relatively little available information regarding the other GIR genes in cotton, especially regarding their effects on cotton fiber development. In this study, 21 GIR family genes were identified in the diploid cotton species Gossypium arboreum; these genes were divided into three groups. The GIR genes were characterized in terms of their phylogenetic relationships, structures, chromosomal distribution and evolutionary dynamics. These GIR genes were revealed to be unequally distributed on 12 chromosomes in the diploid cotton genome, with no GIR gene detected on Ga06. The cis-acting elements in the promoter regions were predicted to be responsive to light, phytohormones, defense activities and stress. The transcriptomic data and qRT-PCR results revealed that most GIR genes were not differentially expressed between the wild-type control and the fuzzless mutant line. Moreover, 14 of 21 family genes were expressed at high levels, indicating these genes may play important roles during fiber development and fuzz formation. Furthermore, Ga01G0231 was predominantly expressed in root samples, suggestive of a role in root hair formation rather than in fuzz initiation and development. The results of this study have enhanced our understanding of the GIR genes and their potential utility for improving cotton fiber through breeding.

Gene Expression Correlation Analysis Reveals MYC-NAC Regulatory Network in Cotton Pigment Gland Development.

Plant NAC (NAM, ATAF1/2, and CUC2) family is involved in various development processes including Programmed Cell Death (PCD) associated development. However, the relationship between NAC family and PCD-associated cotton pigment gland development is largely unknown. In this study, we identified 150, 153 and 299 NAC genes in newly updated genome sequences of G. arboreum, G. raimondii and G. hirsutum, respectively. All NAC genes were divided into 8 groups by the phylogenetic analysis and most of them were conserved during cotton evolution. Using the vital regulator of gland formation GhMYC2-like as bait, expression correlation analysis screened out 6 NAC genes which were low-expressed in glandless cotton and high-expressed in glanded cotton. These 6 NAC genes acted downstream of GhMYC2-like and were induced by MeJA. Silencing CGF1(Cotton Gland Formation1), another MYC-coding gene, caused almost glandless phenotype and down-regulated expression of GhMYC2-like and the 6 NAC genes, indicating a MYC-NAC regulatory network in gland development. In addition, predicted regulatory mechanism showed that the 6 NAC genes were possibly regulated by light, various phytohormones and transcription factors as well as miRNAs. The interaction network and DNA binding sites of the 6 NAC transcription factors were also predicted. These results laid the foundation for further study of gland-related genes and gland development regulatory network.

Improved reconstruction and comparative analysis of chromosome 12 to rectify Mis-assemblies in Gossypium arboreum.

BACKGROUND: Genome sequencing technologies have been improved at an exponential pace but precise chromosome-scale genome assembly still remains a great challenge. The draft genome of cultivated G. arboreum was sequenced and assembled with shotgun sequencing approach, however, it contains several misassemblies. To address this issue, we generated an improved reassembly of G. arboreum chromosome 12 using genetic mapping and reference-assisted approaches and evaluated this reconstruction by comparing with homologous chromosomes of G. raimondii and G. hirsutum. RESULTS: In this study, we generated a high quality assembly of the 94.64 Mb length of G. arboreum chromosome 12 (A_A12) which comprised of 144 scaffolds and contained 3361 protein coding genes. Evaluation of results using syntenic and collinear analysis of reconstructed G. arboreum chromosome A_A12 with its homologous chromosomes of G. raimondii (D_D08) and G. hirsutum (AD_A12 and AD_D12) confirmed the significant improved quality of current reassembly as compared to previous one. We found major misassemblies in previously assembled chromosome 12 (A_Ca9) of G. arboreum particularly in anchoring and orienting of scaffolds into a pseudo-chromosome. Further, homologous chromosomes 12 of G. raimondii (D_D08) and G. arboreum (A_A12) contained almost equal number of transcription factor (TF) related genes, and showed good collinear relationship with each other. As well, a higher rate of gene loss was found in corresponding homologous chromosomes of tetraploid (AD_A12 and AD_D12) than diploid (A_A12 and D_D08) cotton, signifying that gene loss is likely a continuing process in chromosomal evolution of tetraploid cotton. CONCLUSION: This study offers a more accurate strategy to correct misassemblies in sequenced draft genomes of cotton which will provide further insights towards its genome organization.

Fine mapping and identification of the fuzzless gene GaFzl in DPL972 (Gossypium arboreum).

KEY MESSAGE: The fuzzless gene GaFzl was fine mapped to a 70-kb region containing a GIR1 gene, Cotton_A_11941, responsible for the fuzzless trait in Gossypium arboreum DPL972. Cotton fiber is the most important natural textile resource. The fuzzless mutant DPL972 (Gossypium arboreum) provides a useful germplasm resource to explore the molecular mechanism underlying fiber and fuzz initiation and development. In our previous research, the fuzzless gene in DPL972 was identified as a single dominant gene and named GaFzl. In the present study, we fine mapped this gene using F2 and BC1 populations. By combining traditional map-based cloning and next-generation sequencing, we mapped GaFzl to a 70-kb region containing seven annotated genes. RNA-Sequencing and re-sequencing analysis narrowed these candidates to two differentially expressed genes, Cotton_A_11941 and Cotton_A_11942. Sequence alignment uncovered no variation in coding or promoter regions of Cotton_A_11942 between DPL971 and DPL972, whereas two single-base mutations in the promoter region and a TTG insertion in the coding region were detected in Cotton_A_11941 in DPL972. Cotton_A_11941 encoding a homologous gene of GIR1 (GLABRA2-interacting repressor) in Arabidopsis thaliana is thus the candidate gene most likely responsible for the fuzzless trait in DPL972. Our findings should lead to a better understanding of cotton fuzz formation, thereby accelerating marker-assisted selection during cotton breeding.

Correction to: Fine mapping and identification of the fuzzless gene GaFzl in DPL972 (Gossypium arboreum).

Unfortunately, Figs. 5 and 6 were interchanged in the results section. Figures should swap positions, whereas the legends should stay in the given order.

Light emitting properties of Si+ self-ion implanted silicon-on-insulator from visible to infrared band.

The photoluminescence (PL) properties of silicon-on-insulator (SOI) samples, modified by the Si+ self-ion-implantation (SII) into Si thin film followed by annealing, have been well investigated. The well-known W-line can also be observed in SII SOI samples, its emitting behavior and structural evolution have been discussed in this article. The parallel PL pattern trend and the similar changes of temperature-dependent intensity suggest that luminescence center of I1 and I2 peaks located in the near-infrared band originates from the same interstitial-clusters (InCs). The PL peak at 1.762 eV can be ascribed to the quantum confinement (QC) from small-sized Si nanocrystals. Based on the electron spin resonance (ESR) experiments and the variation of normalized PL intensities at different annealing temperature (TA), the neutral oxygen vacancy (NOV) [O3≡Si-Si≡O3] is proposed to be responsible for the blue emission of P2 and P3 peaks, whose intensity can be restrained by the existence of the paramagnetic E1' defects [O3≡Si+]. The density of E1' defect is found to reduce with the increase of annealing temperature (TA). Our results provide a useful method to identify the origin of luminescence centers and pave a way for the application of new type optical defects on silicon based light emitting devices (LEDs).

Fine mapping and candidate gene analysis of the dominant glandless gene Gl 2 (e) in cotton (Gossypium spp.).

KEY MESSAGE: Dominant glandless gene Gl 2 (e) was fine-mapped to a 15 kb region containing one candidate gene encoding an MYC transcription factor, sequence and expression level of the gene were analyzed. Cottonseed product is an excellent source of oil and protein. However, this nutrition source is greatly limited in utilization by the toxic gossypol in pigment glands. It is reported that the Gl 2 (e) gene could effectively inhibit the formation of the pigment glands. Here, three F2 populations were constructed using two pairs of near isogenic lines (NILs), which differ nearly only by the gland trait, for fine mapping of Gl 2 (e) . DNA markers were identified from recently developed cotton genome sequence. The Gl 2 (e) gene was located within a 15-kb genomic interval between two markers CS2 and CS4 on chromosome 12. Only one gene was identified in the genomic interval as the candidate for Gl 2 (e) which encodes a family member of MYC transcription factor with 475-amino acids. Unexpectedly, the results of expression analysis indicated that the MYC gene expresses in glanded lines while almost does not express in glandless lines. These results suggest that the MYC gene probably serves as a vital positive regulator in the organogenesis pathway of pigment gland, and low expression of this gene will not launch the downstream pathway of pigment gland formation. This is the first pigment gland-related gene identification in cotton and will facilitate the research on glandless trait, cotton MYC proteins and low-gossypol cotton breeding.

Development of chromosome-specific markers with high polymorphism for allotetraploid cotton based on genome-wide characterization of simple sequence repeats in diploid cottons (Gossypium arboreum L. and Gossypium raimondii Ulbrich).

BACKGROUND: Tetraploid cotton contains two sets of homologous chromosomes, the At- and Dt-subgenomes. Consequently, many markers in cotton were mapped to multiple positions during linkage genetic map construction, posing a challenge to anchoring linkage groups and mapping economically-important genes to particular chromosomes. Chromosome-specific markers could solve this problem. Recently, the genomes of two diploid species were sequenced whose progenitors were putative contributors of the At- and Dt-subgenomes to tetraploid cotton. These sequences provide a powerful tool for developing chromosome-specific markers given the high level of synteny among tetraploid and diploid cotton genomes. In this study, simple sequence repeats (SSRs) on each chromosome in the two diploid genomes were characterized. Chromosome-specific SSRs were developed by comparative analysis and proved to distinguish chromosomes. RESULTS: A total of 200,744 and 142,409 SSRs were detected on the 13 chromosomes of Gossypium arboreum L. and Gossypium raimondii Ulbrich, respectively. Chromosome-specific SSRs were obtained by comparing SSR flanking sequences from each chromosome with those from the other 25 chromosomes. The average was 7,996 per chromosome. To confirm their chromosome specificity, these SSRs were used to distinguish two homologous chromosomes in tetraploid cotton through linkage group construction. The chromosome-specific SSRs and previously-reported chromosome markers were grouped together, and no marker mapped to another homologous chromosome, proving that the chromosome-specific SSRs were unique and could distinguish homologous chromosomes in tetraploid cotton. Because longer dinucleotide AT-rich repeats were the most polymorphic in previous reports, the SSRs on each chromosome were sorted by motif type and repeat length for convenient selection. The primer sequences of all chromosome-specific SSRs were also made publicly available. CONCLUSION: Chromosome-specific SSRs are efficient tools for chromosome identification by anchoring linkage groups to particular chromosomes during genetic mapping and are especially useful in mapping of qualitative-trait genes or quantitative trait loci with just a few markers. The SSRs reported here will facilitate a number of genetic and genomic studies in cotton, including construction of high-density genetic maps, positional gene cloning, fingerprinting, and genetic diversity and comparative evolutionary analyses among Gossypium species.

Learning an Invariant and Equivariant Network for Weakly Supervised Object Detection.

Weakly Supervised Object Detection (WSOD) is of increasing importance in the community of computer vision as its extensive applications and low manual cost. Most of the advanced WSOD approaches build upon an indefinite and quality-agnostic framework, leading to unstable and incomplete object detectors. This paper attributes these issues to the process of inconsistent learning for object variations and the unawareness of localization quality and constructs a novel end-to-end Invariant and Equivariant Network (IENet). It is implemented with a flexible multi-branch online refinement, to be naturally more comprehensive-perceptive against various objects. Specifically, IENet first performs label propagation from the predicted instances to their transformed ones in a progressive manner, achieving affine-invariant learning. Meanwhile, IENet also naturally utilizes rotation-equivariant learning as a pretext task and derives an instance-level rotation-equivariant branch to be aware of the localization quality. With affine-invariance learning and rotation-equivariant learning, IENet urges consistent and holistic feature learning for WSOD without additional annotations. On the challenging datasets of both natural scenes and aerial scenes, we substantially boost WSOD to new state-of-the-art performance. The codes have been released at: https://github.com/XiaoxFeng/IENet.

Learning to Assess Image Quality Like an Observer.

Human observers are the ultimate receivers and evaluators of the image visual information and have powerful perception ability of visual quality with short-term global perception and long-term regional observation. Thus, it is natural to design an image quality assessment (IQA) computational model to act like an observer for accurately predicting the human perception of image quality. Inspired by this, here, we propose a novel observer-like network (OLN) to perform IQA by jointly considering the global glimpsing information and local scanning information. Specifically, the OLN consists of a global distortion perception (GDP) module and a local distortion observation (LDO) module. The GDP module is designed to mimic the observer's global perception of image quality through performing classification of images' distortion categories and levels. Simultaneously, to simulate the human local observation behavior, the LDO module attempts to gather the long-term regional observation information of the distorted images by continuously tracing the human scanpath in the observer-like scanning manner. By leveraging the bilinear pooling layer to collaborate the short-term global perception with the long-term regional observation, our network precisely predicts the quality scores of distorted images, such as human observers. Comprehensive experiments on the public datasets powerfully demonstrate that the proposed OLN achieves state-of-the-art performance.

P-CNN: Part-Based Convolutional Neural Networks for Fine-Grained Visual Categorization.

This paper proposes an end-to-end fine-grained visual categorization system, termed Part-based Convolutional Neural Network (P-CNN), which consists of three modules. The first module is a Squeeze-and-Excitation (SE) block, which learns to recalibrate channel-wise feature responses by emphasizing informative channels and suppressing less useful ones. The second module is a Part Localization Network (PLN) used to locate distinctive object parts, through which a bank of convolutional filters are learned as discriminative part detectors. Thus, a group of informative parts can be discovered by convolving the feature maps with each part detector. The third module is a Part Classification Network (PCN) that has two streams. The first stream classifies each individual object part into image-level categories. The second stream concatenates part features and global feature into a joint feature for the final classification. In order to learn powerful part features and boost the joint feature capability, we propose a Duplex Focal Loss used for metric learning and part classification, which focuses on training hard examples. We further merge PLN and PCN into a unified network for an end-to-end training process via a simple training technique. Comprehensive experiments and comparisons with state-of-the-art methods on three benchmark datasets demonstrate the effectiveness of our proposed method.

Weighted Gene Co-Expression Network Analysis Reveals Hub Genes Contributing to Fuzz Development in Gossypium arboreum.

Fuzzless mutants are ideal materials to decipher the regulatory network and mechanism underlying fuzz initiation and formation. In this study, we utilized two Gossypium arboreum accessions differing in fuzz characteristics to explore expression pattern differences and discriminate genes involved in fuzz development using RNA sequencing. Gene ontology (GO) analysis was conducted and found that DEGs were mainly enriched in the regulation of transcription, metabolic processes and oxidation-reduction-related processes. Weighted gene co-expression network analysis discerned the MEmagenta module highly associated with a fuzz/fuzzless trait, which included a total of 50 hub genes differentially expressed between two materials. GaFZ, which negatively regulates trichome and fuzz formation, was found involved in MEmagenta cluster1. In addition, twenty-eight hub genes in MEmagenta cluster1 were significantly up-regulated and expressed in fuzzless mutant DPL972. It is noteworthy that Ga04G1219 and Ga04G1240, which, respectively, encode Fasciclin-like arabinogalactan protein 18(FLA18) and transport protein, showed remarkable differences of expression level and implied that they may be involved in protein glycosylation to regulate fuzz formation and development. This module and hub genes identified in this study will provide new insights on fiber and fuzz formation and be useful for the molecular design breeding of cotton genetic improvement.

Study on Well Spacing Optimization in a Tight Sandstone Gas Reservoir Based on Dynamic Analysis.

Compared to the shale gas and coalbed methane in China, tight gas has been recently considered as a priority in the exploration and exploitation of unconventional gas resources. In the development of a tight gas field, how to enhance the gas recovery is a prevalent topic. Unlike the conventional gas reservoir, the ultimate gas recovery is not only determined by the geological characteristics but is also affected by other factors such as well drainage area and well spacing design. For tight sandstone reservoirs, the gas recovery can be improved by increasing the drainage area. Moreover, the well drainage area is closely associated with well spacing. Therefore, effective drainage area estimation and well spacing optimization are essential aspects for tight gas exploitation. In this paper, a new optimization workflow is established, which combined dynamic analysis and numerical simulation techniques. First, through interference well test results and production data dynamic analysis, the total gas production can be expressed and predicted. Then the well density can be optimized by the economic evaluation method. Meanwhile, a numerical model is built up to determine the optimal well spacing. This new optimization workflow can provide guidance to the operators of tight gas fields where the interference well test results are available and several years of production data are collected. Furthermore, in the case of the Sulige gas field, the single well drainage area is estimated and the optimal well pattern is obtained by the established approach. The results indicate that the well pattern of 500 m × 600 m is most reasonable for the pilot gas field.

Glutathione S-transferases GhGSTF1 and GhGSTF2 involved in the anthocyanin accumulation in Gossypium hirsutum L.

The glutathione S-transferases (GSTs) are important enzymes of secondary metabolism in plants. In this study, two putative GSTs, GhGSTF1 and GhGSTF2, were identified as anthocyanin-related GSTs by the transcriptome data of the leaves of Gossypium hirsutum L. TM-1 and T586. The quantitative real-time PCR showed that GhGSTF1 and GhGSTF2 were highly expressed in red leaves and stems of Gossypium hirsutum L. T586. Orthologous genes of GhGSTF2 in two Gossypium barbadense L. 3-79 and Xinhai21 contain bases deletion in N-terminal (GbGSTF2a) and C-terminal (GbGSTF2b) respectively. Among which, GhGSTF1 and GhGSTF2 can restore pigmentation in hypocotyls of Arabidopsis thaliana mutant tt19-7 while GbGSTF2a and GbGSTF2b cannot. Furthermore, in vitro assays showed the recombinant GhGSTF1 and GhGSTF2 had Glutathione S-transferase activities. Fluorescence quenching assays showed that Cya could obviously quench the fluorescence of GhGSTF1, GhGSTF2, GbGSTF2a and GbGSTF2b to lower levels as compared to C3G. Moreover, the transient dual-luciferase assays showed that the promoters of GhGSTF1 and GhGSTF2 could be activated by GhPAP1D at different levels. GUS staining assays showed that their promoters have different activities to light. This study indicated that GhGSTF1 and GhGSTF2 play important roles in anthocyanin accumulation and the regulatory mechanism of anthocyanin accumulation in allotetraploid Gossypium are complicated.

Comparative Transcriptomic Analysis to Identify the Genes Related to Delayed Gland Morphogenesis in Gossypium bickii.

Cotton is one of the major industrial crops that supply natural fibers and oil for industries. This study was conducted to understand the mechanism of delayed gland morphogenesis in seeds of Gossypium bickii. In this study, we compared glandless seeds of G. bickii with glanded seeds of Gossypium arboreum. High-throughput sequencing technology was used to explore and classify the expression patterns of gland-related genes in seeds and seedlings of cotton plants. Approximately 131.33 Gigabases of raw data from 12 RNA sequencing samples with three biological replicates were generated. A total of 7196 differentially-expressed genes (DEGs) were identified in all transcriptome data. Among them, 3396 genes were found up-regulated and 3480 genes were down-regulated. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations were performed to identify different functions between genes unique to glandless imbibed seeds and glanded seedlings. Co-expression network analysis revealed four modules that were identified as highly associated with the development of glandless seeds. Here the hub genes in each module were identified by weighted gene co-expression network analysis (WGCNA). In total, we have selected 13 genes involved in transcription factors, protein and MYB-related functions, that were differentially expressed in transcriptomic data and validated by quantitative reverse-transcription polymerase chain reaction (qRT-PCR). These selected genes may play an important role for delayed gland morphogenesis. Our study provides comprehensive insight into the key genes related to glandless traits of seeds and plants, and can be further exploited by functional and molecular studies.

Genome-wide comparative analysis of RNA-binding Glycine-rich protein family genes between Gossypium arboreum and Gossypium raimondii.

RB-GRP (RNA-binding Glycine-rich protein gene) family belongs to the fourth subfamily of the GRP (Glycine-rich protein gene) superfamily, which plays a great role in plant growth and development, as well as in abiotic stresses response, while it has not been identified in cotton. Here, we identified 37 and 32 RB-GRPs from two cotton species (Gossypium arboreum and Gossypium raimondii, respectively), which were divided into four distinct subfamilies based on the presence of additional motifs and the arrangement of the glycine repeats. The distribution of RB-GRPs was nonrandom and uneven among the chromosomes both in two cotton species. The expansion of RB-GRP gene family between two cultivars was mainly attributed to segmental and tandem duplication events indicated by synteny analysis, and the tandem duplicated genes were mapped into homologous collinear blocks, indicated that they shared a common ancestral gene in both species. Furthermore, most RB-GRPs in two cotton species undergone stronger negative selective pressure by evolutionary analysis of RB-GRP orthologous genes. Meanwhile, RB-GRPs participated in different abiotic stresses (Abscisic acid, salt and Polyethylene glycol) responses and tissues at different developmental stages between two cotton species were showed by gene expression analysis. This research would provide insight into the evolution and function of the RB-GRPs in Gossypium species.

Transcriptome analysis reveals long noncoding RNAs involved in fiber development in cotton (Gossypium arboreum).

Long noncoding RNAs (lncRNAs) play important roles in various biological regulatory processes in yeast, mammals, and plants. However, no systematic identification of lncRNAs has been reported in Gossypium arboreum. In this study, the strand-specific RNA sequencing (ssRNA-seq) of samples from cotton fibers and leaves was performed, and lncRNAs involved in fiber initiation and elongation processes were systematically identified and analyzed. We identified 5,996 lncRNAs, of which 3,510 and 2,486 can be classified as long intergenic noncoding RNAs (lincRNAs) and natural antisense transcripts (lncNAT), respectively. LincRNAs and lncNATs are similar in many aspects, but have some differences in exon number, exon length, and transcript length. Expression analysis revealed that 51.9% of lincRNAs and 54.5% of lncNATs transcripts were preferentially expressed at one stage of fiber development, and were significantly highly expressed than protein-coding transcripts (21.7%). During the fiber and rapid elongation stages, rapid and dynamic changes in lncRNAs may contribute to fiber development in cotton. This work describes a set of lncRNAs that are involved in fiber development. The characterization and expression analysis of lncRNAs will facilitate future studies on their roles in fiber development in cotton.