No abstract

Anti-DNA antibodies play an essential role in the pathogenesis of lupus nephritis. Mammalian DNA alone, however, is poorly immunogenic. We speculated that the antigenic trigger for the production of human nephritogenic anti-DNA antibodies is a non-DNA substance. The cDNA library from peripheral blood lymphocytes (PBLs) of a patient with active lupus nephritis was screened using the single-chain Fv of a human monoclonal nephritogenic O-81 anti-DNA antibody in a two-hybrid system. A clone containing the gene of an endoplasmic reticulum (ER) stress-inducible protein, Herp, was obtained: The O-81 antibody bound to recombinant Herp protein synthesized by Escherichia coli. Immunization with Herp elicited both anti-double-stranded DNA (anti-dsDNA) and anti-single-stranded (anti-ssDNA) antibodies in BALB/c mice and formed deposits of IgG in renal glomeruli. Anti-DNA antibodies purified from SLE sera bound to Herp. Moreover, anti-Herp antibodies showed specific binding to DNA. Herp was spontaneously expressed in PBLs of patients with active SLE, but not in PBLs of healthy subjects. These results imply that an inducible intracellular self-protein represents a candidate trigger for human nephritogenic anti-DNA autoantibodies. Any cell stress causing ER stress, such as viral infection, ultraviolet radiation, and chemicals, might be responsible for anti-DNA antibody production via Herp.

This paper addresses the development of a three-fingered hand and discusses a strategy for grasping unknown objects by groping. In the former part of the paper, the design of the developed robot hand is presented. The robot hand has thumb, index finger and middle finger with a total of eight degrees-of-freedom. The location of thumb is designed considering opposability of human hand. Eightyseven touch sensors are distributed over the surface of palm and fingers of the robot hand. In the latter part of the paper, a strategy for grasping unknown objects by groping using the developed hand is discussed. Groping is a kind of "active sensing." When the system does not have any models of objects to grasp, active sensing becomes inevitable. The aim of the groping is to find a grasping configuration for unknown objects. A can, a ball, a cone, a plate, and a cube are chosen for the unknown objects in the experiments. Experimental results demonstrate the capability of the strategy for grasping unknown objects. This strategy needs neither models of objects nor complicated computation, and therefore, useful especially for assembly tasks in the real world.

No abstract

No abstract