The mechanism of aggregation of the coat protein of tobacco mosaic virus. A comparative study on vulgare and mutant proteins.

The aggregation behaviour of tobacco mosaic virus (TMV) protein vulgare was compared to that of two mutants, A 14 and Ni725, with amino acid exchanges localized in the coat protein at posi­ tions 107 (Thr -*• Met, in Ni725) and 129 (lie -> Thr, in both mutants). This behaviour, as meas­ ured by sedimentation, hydrogen ion titration, light-scattering, and near-UV absorption difference and circular dichroism (CD) spectroscopy, differs characteristically both in the range of the Aprotein (pH 8) and near neutrality, whereas nuclear magnetic resonance (NMR) and far-UV CD point at only subtle, or no structural differences between the three strains. Near pH 8, the A-proteins of both mutants sediment nearly exclusively as 8 S aggregates, under conditions where vulgare protein forms a 4S /8S mixture (two-layer and three-layer aggregates, Vogel etal., Z. Natunorsch., 1979); in conditions where vulgare 4S aggregates dominate, both mutants sediment as a 4S /8S mixture. The average molecular weights of the 8S proteins corre­ spond to 12 (vulgare) to 15 (mutants) subunits. Near neutrality both mutants titrate and polyme­ rize more cooperatively than vulgare protein; additionally, the pK(app.) of Ni 725 is shifted up­ wards, due to the higher a-helix forming potential o f Met against Thr (pos. 107). Both mutants form large aggregates (> 200 S) of obviously helical conformation, by the uptake of one proton per subunit, whereas 20 S-disks constituting, under the same conditions, the stable entities in vul­ gare protein, are made only in minor amounts. These large mutant aggregates are remarkably more stable than the vulgare “overshoot” aggregates which transiently, too, may approach s-values and turbidities similar to the mutant aggregates; conformational changes, observed prior or in parallel to the formation of vulgare overshoot and disk aggregates, are significantly retarded in the large mutant aggregates. — Raising the ionic strength seems the only way to form mutant disks and stacks of disks (2 0 -3 0 S) comparable to vulgare, pointing to the different pathways of disk formation, either at neutral pH or high ionic strength. — Evidence is given that the 8S aggregates of both mutant and vulgare proteins may behave similar in aggregation, the differences mainly being inserted by the 4S (two-layer) aggregates present in vulgare protein, which near neutrality seem responsible for the direct formation of (two-layer) disks. The non-conservative exchange in po­ sition 129, altering the environment of Trp residues (52+17?), should weaken the “extended saltbridge system” (“pairing”) observed between the two layers of the disk (Bloomer et al., Nature, 1978). A competition is suggested between the strength of this pairing, and the binding of a third layer, regulating the mode of aggregation to two-layer, to three-layer, and to higher aggregates; this is corroborated by comparison with published results on temperature-sensitive (ts I) mutants and chemically modified proteins. — To explain the effects of residue 129 on the titration of the protein we suggest a mechanical analogy, made up of a balance between the charge and state of the “carboxyle cage” (Stubbs et al., Nature, 1977), as regulatory site, and the strength of the saltbridge system, about 25 A apart, as the site active in the process o f polymerization; a signal exerted on one end of the system may be transmitted to the other end by the rigid a-helical system in between. — This mechanism remains open to discussion and further investigations.