Epitope topography of agonist antibodies to the checkpoint inhibitory receptor BTLA

B and T lymphocyte attenuator (BTLA) is an attractive target for a new class of therapeutics that attempt to rebalance the immune system by agonizing checkpoint inhibitory receptors (CIRs). Herpesvirus entry mediator (HVEM) binds BTLA in both trans- and cis-orientations. We report here the development and structural characterization of three humanized BTLA agonist antibodies, 22B3, 25F7, and 23C8. We determined the crystal structures of the antibody-BTLA complexes, showing that these antibodies bind distinct and non-overlapping epitopes of BTLA. While all three antibodies activate BTLA, 22B3 mimics HVEM binding to BTLA and shows the strongest agonistic activity in functional cell assays and in an imiquimod-induced mouse model of psoriasis. 22B3 is also capable of modulating HVEM signaling through the BTLA-HVEM cis-interaction. The data obtained from crystal structures, biochemical assays, and functional studies provide a mechanistic model of HVEM and BTLA organization on the cell surface and informed the discovery of a highly active BTLA agonist.

Time course of disassociation of bone formation signals with bone mass and bone strength in sclerostin antibody treated ovariectomized rats

Sclerostin antibodies increase bone mass by stimulating bone formation. However, human and animal studies show that bone formation increases transiently and returns to pre-treatment level despite ongoing antibody treatment. To understand its mechanism of action, we studied the time course of bone formation, correlating the rate and extent of accrual of bone mass and strength after sclerostin antibody treatment. Ovariectomized (OVX) rats were treated with a sclerostin-antibody (Scle-ab) at 20mg/kg sc once weekly and sacrificed at baseline and 2, 3, 4, 6, and 8weeks post-treatment. In Scle-ab treated rats, serum PINP and OCN rapidly increased at week 1, peaked around week 3, and returned to OVX control levels by week 6. Transcript analyses from the distal femur revealed an early increase in bone formation followed by a sustained decrease in bone resorption genes. Lumbar vertebral (LV) osteoblast surface increased 88% by week 2, and bone formation rate (BFR/BS) increased 138% by week 4. Both parameters were below OVX control by week 8. Bone formation was primarily a result of modeling based formation. Endocortical and periosteal BFR/BS peaked around week 4 at 313% and 585% of OVX control, respectively. BFR/BS then declined but remained higher than OVX control on both surfaces through week 8. Histomorphometric analyses showed LV-BV/TV did not further increase after week 4, while BMD continued to increase at LV, mid femur (MF), and femoral neck (FN) through week 8. Biomechanical tests showed a similar improvement in bone strength through 8weeks in MF and FN, but bone strength plateaued between weeks 6 and 8 for LV. Our data suggest that bone formation with Scle-ab treatment is rapid and modeling formation dominated in OVX rats. Although transient, the bone formation response persists longer in cortical than trabecular bone.

Dicyclohexylcarbodiimide inhibits proton pumping in ubiquinol:cytochrome c oxidoreductase of Rhodobacter sphaeroides and binds to aspartate-187 of cytochrome b

In recent studies we reported that dicyclohexylcarbodiimide (DCCD) inhibited proton translocation in ubiquinol:cytochrome c oxidoreductase (cytochrome bc1 complex) from yeast mitochondria where it was bound to aspartate-160 of cytochrome b. In the current study, we report that DCCD and its fluorescent analogue, N-cyclohexyl-N'-[4-(dimethylamino)naphthyl]-carbodiimide (NCD-4), inhibit 50-60% proton pumping in the cytochrome bc1 complex of the bacterium Rhodobacter sphaeroides with a 20% inhibition of electron transfer activity. Radioactive DCCD is bound exclusively to cytochrome b at aspartate-187, which is located at the C-terminal region of the CD loop connecting membrane-spanning helices C and D of cytochrome b. Fluorescent studies with NCD-4 revealed that aspartate-187 is located in a mildly hydrophobic pocket in the bc1 complex at a distance of 2-3 A from the surface of the membrane.

The role of the membrane-spanning and extra-membranous regions of the iron-sulfur protein in its assembly into the cytochrome bc1 complex of yeast mitochondria

The assembly of six deletion mutants of the Rieske iron-sulfur protein into the cytochrome bc1 complex was investigated by immunoprecipitation from detergent-solubilized mitochondria with specific antisera against either the iron-sulfur protein or the intact cytochrome bc1 complex. After import, the mutant proteins lacking residues 41-55 or 66-78, located at the membrane-spanning region of the protein, and residues 182-196 located at the C-terminus of the protein, were assembled in vitro into the bc1 complex approximately 50% as effectively as the wild type iron-sulfur protein suggesting that these regions of the iron-sulfur protein may not be critical for the assembly. By contrast, only trace amounts of the mutant proteins lacking residues 80-95, 122-135, 138-153 located in the extra-membranous region of the iron-sulfur protein were assembled into the bc1 complex. After import in vitro into mitochondria isolated from a cytochrome b-deficient yeast strain, the mutants lacking residues 41-55 and 182-196 were assembled as efficiently as the wild type; however, the mutants lacking residues 55-66 and 66-78 were assembled less efficiently in the absence of cytochrome b suggesting that the hydrophobic membrane-spanning region, residues 55-78, of the iron-sulfur protein, may interact with cytochrome b during the assembly of the bc1 complex.