Platelet factor 4 (PF4), a secreted platelet protein, alleviates concanavalin A-induced immunosuppression in mice. We now rind that activity also resides in (i) the C-terminal tridecapeptide of PF4 (P13S), (ii) an analog of this in which arginine replaces the lysine residues and in which the last two amino acids are absent, (iii) the C-terminal 18 amino acids of low-affmity platelet factor 4, which is very similar to P13S, and (iv) peptide fragments of P13S that contain only 5-9 amino acids. P13S treated with fluorescamine to derivatize the free amino groups retained immunoregulatory activity but did not bind to heparin-agarose. The N-terminal and middle portions of PF4, polylysine, protamine, and three unrelated peptides were inactive in this assay.Platelet factor 4 (PF4), a 7800-kDa secretable platelet protein with anti-heparin and chemotactic activity (1-3), alleviates immunosuppression induced in mice by injection of concanavalin A (Con A) or y-irradiated reticulum sarcoma cells (4, 5). Earlier results (5) showed that it was active at 0.2 ,ug per mouse and secretable from as few as 2 x 107 human platelets. Recent results have shown activity at 0.05 gg.Recombinant PF4 with additional amino acids at each end of the molecule and with aspartic acid instead of asparagine at position 47 is also active (6). We now report studies on the immunoregulatory activity of fragments of PF4. MATERIALS AND METHODSSource of Peptides. The amino acid sequences of the PF4 molecule and other peptides studied are shown in Fig. 1. T. Maione (Repligen, Cambridge, MA) donated recombinant preparations of (i) PF4 (rPF4) identical to the native molecule, (ii) the C-terminal 41 amino acids (designated C41), (iii) the N-terminal 29 amino acids (designated N29) of PF4, and (iv) synthetic P13S, the C-terminal 13 amino acids of PF4. Low-affinity PF4 (LA-PF4) was prepared from extracts of fresh platelets and was separated from PF4, ,8-thromboglobulin, and platelet basic protein by chromatography on CMSephadex (7). Peptides H18, H22, and H24 were derived from reduced S-pyridylethyl-modified native PF4 (900 ug) by digestion with 60 ,ug of protease isolated from Staphylococcus aureus V8 (Pierce) at 370C for 3 hr in 0.05 M sodium phosphate (pH 7.8). The digest was fractionated on a Vydac wide-pore C18 reverse-phase column (The Separations Group, Hesperia, CA) equilibrated with 0.1% trifluoroacetic acid and was eluted with a gradient of 0-55% acetonitrile in the same solvent. Identity of the collected peaks was established by automated N-terminal sequencing.The C-terminal octadecapeptide of LA-PF4, designated P18D, and a second preparation of P13S were synthesized by Peninsula Laboratories and were provided by Stefan Niewiarowski (Temple University). Peptides unrelated to l 2 3 4 5

SDS-polyacrylamide gel electrophoresis was used to study the effects of the thiol inhibitor monobromobimane (MB), EDTA, and prostaglandin E1 (PGE1) on the formation and composition of the platelet cytoskeletal core (Triton-insoluble residue) and its association with glycoprotein (GP) IIIa. Stimulation or aggregation of platelets in response to ADP or thrombin increased the amount of Triton-insoluble myosin. Aggregation resulted in incorporation of [125I]GP IIIa and a new band at about 210 kDa into the cytoskeletal core. EDTA and PGE1 caused little disaggregation of platelets that were aggregated in PRP with ADP and that had secreted the contents of their granules. In contrast to EDTA, PGE1 decreased the amount of Triton-insoluble residue and its association with GP IIIa. MB added after ADP-induced aggregation caused an increase in the amount of cytoskeletal core despite marked disaggregation and a substantial decrease in core-associated GP IIIa. With aspirin-treated platelets that had not secreted, EDTA, PGE1, and MB all caused disaggregation and loss of cytoskeletal GP IIIa. MB diminished, but did not reverse, thrombin-induced aggregation of washed platelets and arrested GP IIIa incorporation into the cytoskeletal core. Concanavalin A (Con A) cross-links glycoproteins on a single platelet and induces incorporation of GP IIIa into the Triton-insoluble residue in the absence of platelet aggregation. This induction was not inhibited by MB, although this reagent, as well as aspirin, inhibited Con A-induced secretion. Since GP IIIa incorporation caused by ADP-induced aggregation differs from that caused by Con A in its susceptibility to MB, it seems unlikely that thiol groups are directly involved in the association of GP IIIa with the cytoskeletal core.