Indoleamine 2,3-dioxygenase and regulatory dendritic cells contribute to the allograft protection induced by infusion of donor-specific splenic stromal cells

It has been reported that splenic stromal cells(SSCs)are capable of directly supporting the development of CD11c ^(lo)CD45RB^(+) IL-10-producing dendritic cells(DCs)from lineage-negative c-kit^(+) progenitor cells in the absence of exogenous cytokines.In vitro,DCs that differentiate on stromal cells suppress mixed leukocyte reaction responses and induce primary alloreactive CD4^(+) T cells to differentiate into IL-10-producing Tr1 cells.However,the precise mechanisms by which these SSCs exert their regulatory functions in vivo remain undefined.Furthermore,their possible contribution to the development of allograft transplantation tolerance has yet to be examined.Here,we have used both murine skin and cardiac allograft transplantation models to explore whether in vivo alloresponses can be regulated by infusion with donor-derived SSCs and to investigate the possible mechanisms by which SSCs exert regulatory effects to prevent allograft rejection.We show that intravenous SSC infusion prolonged murine skin allograft survival.The prolonged graft survival is associated with augmentation of the generation of regulatory DC subsets and CD4^(+) CD25^(+) Foxp3^(+) regulatory T cells(Tregs),as well as upregulation of the production of suppressive cytokines IL-10 and transforming growth factor(TGF)-b.Moreover,we found that indoleamine 2,3-dioxygenase and SSC-derived regulatory DCs contribute to allograft protection by infusion of donor-specific SSCs.Our data suggest that donor-derived SSCs could be used as a therapeutic target to promote transplantation tolerance.

A causal convolutional neural network for multi-subject motion modeling and generation

Inspired by the success of WaveNet in multi-subject speech synthesis,we propose a novel neural network based on causal convolutions for multi-subject motion modeling and generation.The network can capture the intrinsic characteristics of the motion of different subjects,such as the influence of skeleton scale variation on motion style.Moreover,after fine-tuning the network using a small motion dataset for a novel skeleton that is not included in the training dataset,it is able to synthesize high-quality motions with a personalized style for the novel skeleton.The experimental results demonstrate that our network can model the intrinsic characteristics of motions well and can be applied to various motion modeling and synthesis tasks.