Does a first fibrinolytic episode change the second? A test proposal for the LL-37, plasmin, antiplasmin and PAI-1 sequence

1. Summary and the central test Four well-studied proteins sit around the same event: LL-37 (the human cathelicidin alarm peptide), plasmin, alpha2-antiplasmin and PAI-1. Some pairs have been studied; for others, no work was found. The four have not been measured together over time in any disease. This proposal sets out one hypothesis about how they might behave as a sequence, and a small set of experiments that could show it is wrong. The hypothesis in brief: after an insult the sequence resolves to a residual state, and that state sets the starting conditions for the next insult. PAI-1 is proposed to act on two timescales: as a brake within a cycle, and as a carrier of state between cycles. If so, a second matched insult should not look like the first, and the difference should depend on PAI-1. If cycle 2 looks identical to cycle 1, the idea fails, and a great deal of the proposed mechanism goes with it. The central test. Give a first insult, a washout longer than the measured relaxation time (otherwise carry-over cannot be separated from incomplete relaxation), then a second matched insult. Use five groups: untreated two-insult; PAI-1 blockade during the washout only, cleared before the second insult so that acute inhibition is not what is measured; vehicle or unrelated-inhibitor control in the washout; PAI-1 manipulation during the first insult; and a naive group given the insult once at the cycle-2 time point, which separates carry-over from elapsed time and ageing. Measure whole trajectories of LL-37, plasmin activity, PAI-1 and fibrin or matrix state, not endpoints. Falsified if cycle 2 is indistinguishable from the naive response, or PAI-1 blockade changes nothing. A cycle-2 difference that persists under PAI-1 blockade points to ordinary priming or tolerance, not to this mechanism. PAI-1 blockade can itself impair repair (Maquerlot 2006, in vitro), so a cycle-2 change after blockade could be impaired healing rather than removal of a retained state. Include a repair readout in the blockade groups and confirm the blocker does not delay cycle-1 repair, or add a second PAI-1 manipulation by a different route (genetic, or a different inhibitor class). A realistic first step is this test in a single well-characterised system, with LL-37 measured by a validated assay. A negative result there removes most of the proposal.