The other side of the coin: systemic effects of Serendipita indica root colonization on development of sedentary plant-parasitic nematodes in Arabidopsis thaliana.

MAIN CONCLUSION: Upon systemic S. indica colonization in split-root system cyst and root-knot nematodes benefit from endophyte-triggered carbon allocation and altered defense responses what significantly facilitates their development in A. thaliana. Serendipita indica is an endophytic fungus that establishes mutualistic relationships with different plants including Arabidopsis thaliana. It enhances host's growth and resistance to different abiotic and biotic stresses such as infestation by the cyst nematode Heterodera schachtii (CN). In this work, we show that S. indica also triggers similar direct reduction in development of the root-knot nematode Meloidogyne javanica (RKN) in A. thaliana. Further, to mimick the natural situation occurring frequently in soil where roots are unequally colonized by endophytes we used an in vitro split-root system with one half of A. thaliana root inoculated with S. indica and the other half infected with CN or RKN, respectively. Interestingly, in contrast to direct effects, systemic effects led to an increase in number of both nematodes. To elucidate this phenomenon, we focused on sugar metabolism and defense responses in systemic non-colonized roots of plants colonized by S. indica. We analyzed the expression of several SUSs and INVs as well as defense-related genes and measured sugar pools. The results show a significant downregulation of PDF1.2 as well as slightly increased sucrose levels in the non-colonized half of the root in three-chamber dish. Thus, we speculate that, in contrast to direct effects, both nematode species benefit from endophyte-triggered carbon allocation and altered defense responses in the systemic part of the root, which promotes their development. With this work, we highlight the complexity of this multilayered tripartite relationship and deliver new insights into sugar metabolism and plant defense responses during S. indica-nematode-plant interaction.

Serendipita indica changes host sugar and defense status in Arabidopsis thaliana: cooperation or exploitation?

MAIN CONCLUSION: Manipulation of sugar metabolism upon S. indica root colonization triggers changes in sugar pools and defense responses in A. thaliana. Serendipita indica is an endophytic fungus that establishes mutualistic relationships with many different plants including important crops as well as the model plant A. thaliana. Successful root colonization typically results in growth promotion and enhanced tolerance against various biotic and abiotic stresses. The fungus delivers phosphorus to the host and receives in exchange carbohydrates. There are hints that S. indica prefers hexoses, glucose, and fructose, products of saccharose cleavage driven by invertases (INVs) and sucrose synthases (SUSs). Carbohydrate metabolism in this interaction, however, remains still widely unexplored. Therefore, in this work, the sugar pools as well as the expression of SUSs and cytosolic INVs in plants colonized by S. indica were analyzed. Using sus1/2/3/4 and cinv1/2 mutants the importance of these genes for the induction of growth promotion and proper root colonization was demonstrated. Furthermore, the expression of several defense-related marker genes in both multiple mutants in comparison to the wild-type plants was determined. Our results show that in colonized A. thaliana plants S. indica manipulates the sugar metabolism by altering the expression of host's INV and SUS and modulates both the sugar pools and plant defense in its favor. We conclude that the interaction A. thaliana-S. indica is a balancing act between cooperation and exploitation, in which sugar metabolism plays a crucial role. Small changes in this mechanism can lead to severe disruption resulting in the lack of growth promotion or altered colonization rate.

Deep Metric Learning with BIER: Boosting Independent Embeddings Robustly.

Learning similarity functions between image pairs with deep neural networks yields highly correlated activations of embeddings. In this work, we show how to improve the robustness of such embeddings by exploiting the independence within ensembles. To this end, we divide the last embedding layer of a deep network into an embedding ensemble and formulate the task of training this ensemble as an online gradient boosting problem. Each learner receives a reweighted training sample from the previous learners. Further, we propose two loss functions which increase the diversity in our ensemble. These loss functions can be applied either for weight initialization or during training. Together, our contributions leverage large embedding sizes more effectively by significantly reducing correlation of the embedding and consequently increase retrieval accuracy of the embedding. Our method works with any differentiable loss function and does not introduce any additional parameters during test time. We evaluate our metric learning method on image retrieval tasks and show that it improves over state-of-the-art methods on the CUB-200-2011, Cars-196, Stanford Online Products, In-Shop Clothes Retrieval and VehicleID datasets. Therefore, our findings suggest that by dividing deep networks at the end into several smaller and diverse networks, we can significantly reduce overfitting.

Does iron reduction control the release of dissolved organic carbon and phosphate at catchment scales? Need for a joint research effort.

Increasing concentrations of dissolved iron and DOC are likely linked to decreasing nitrogen depositon.

Unexpected release of phosphate and organic carbon to streams linked to declining nitrogen depositions.

Reductions in emissions have successfully led to a regional decline in atmospheric nitrogen depositions over the past 20 years. By analyzing long-term data from 110 mountainous streams draining into German drinking water reservoirs, nitrate concentrations indeed declined in the majority of catchments. Furthermore, our meta-analysis indicates that the declining nitrate levels are linked to the release of dissolved iron to streams likely due to a reductive dissolution of iron(III) minerals in riparian wetland soils. This dissolution process mobilized adsorbed compounds, such as phosphate, dissolved organic carbon and arsenic, resulting in concentration increases in the streams and higher inputs to receiving drinking water reservoirs. Reductive mobilization was most significant in catchments with stream nitrate concentrations <6 mg L-1 . Here, nitrate, as a competing electron acceptor, was too low in concentration to inhibit microbial iron(III) reduction. Consequently, observed trends were strongest in forested catchments, where nitrate concentrations were unaffected by agricultural and urban sources and which were therefore sensitive to reductions of atmospheric nitrogen depositions. We conclude that there is strong evidence that the decline in nitrogen deposition toward pre-industrial conditions lowers the redox buffer in riparian soils, destabilizing formerly fixed problematic compounds, and results in serious implications for water quality.

Identification and expression analyses of putative sesquiterpene synthase genes in Phormidium sp. and prevalence of geoA-like genes in a drinking water reservoir.

The occurrence of taste and odor problems in drinking water supplies is a widespread phenomenon. From a Saxonian water reservoir we isolated a cyanobacterial species which was classified as Phormidium sp. Under laboratory conditions it produced an earthy-musty smell due to the synthesis of geosmin. The only genes shown to be involved in geosmin biosynthesis are cyc2 and geoA of Streptomyces. Based on the alignment of Cyc2 with a putative sesquiterpene synthase of Nostoc punctiforme, a degenerate primer pair was designed. By PCR, we could amplify two similar genes in Phormidium sp., which we named geoA1 and geoA2. Their expression was studied by reverse transcription-PCR. This revealed that both genes are expressed at 20 degrees C and a light-dark cycle of 12 h. Expression was not detectable at the end of a 24-h dark period. To analyze the prevalence of geoA1 and geoA2 in samples from the phytobenthos, we generated PCR fragments with the same degenerate primer pair. Fifty-five different sequences that might represent geoA variants were obtained. The GC content ranged from 42% to 67%, suggesting that taxonomically very different bacteria might contain such genes.