|
Severing
of actin filaments by ADF/cofilins |
| Severing
was attribute to ADF/cofilins from the earliest reports to explain the
effects of the family on the viscosity of actin filaments. |
 |
Figure
1.
This was the first direct visualization of filament
severing by any actin binding protein. Actin filaments were immobilised on a nitrocellulose
coated coverslip via enzymatically inactive myosin II from Acanthamoeba. Actophorin,
an ADF/cofilin also from Acanthamoeba was flowed through at time 0, and images of the
ensuing severing events recorded through computer enhanced video microscopy.
|
| Severing Actin Filaments: From Maciver
et al. (1991) J.Cell Biol. 115, 1611-1620 |
Cofilin is a small actin binding protein that is abundant in eukaryotic cells. Cofilin binds G-actin at high pH values and F-actin in lower pH within the physiological range, and causes rapid rearrangements of actin when the pH is altered. Early studies on members of the ADF/cofilin family proposed that these proteins decreased the viscosity of actin filament either by increasing the off rates from the filament ends or by severing filaments. We and others have shown that cofilins increase the off rate from the pointed end, but actual filament severing by cofilins has been doubted. We believe that severing does in fact take place in vitro at
least, and that the pH-induced depolymerisation observed is a severing event at every position that was previously bound by cofilin at low pH. A full account of our argument in favour of a severing mechanism can be found in, Maciver,
S.K. (1998) Curr.Opin.Cell Biol. 19; 140-144. |
 |
| In 1991 we reported our observation of severing of actin filaments by fluorescence microscopy (see above). This data firmly convinced us that actophorin (a
cofilin from Acanthamoeba) severed actin filaments. In addition, we postulated a molecular mechanism based on the assumption that some of the actin : actin
contacts within the filament broke spontaneously (Figure above). Each actin monomer in a filament is held in place by four actin-actin contacts, two weak
"diagonal bonds" and two stronger "longitudinal bonds". We hypothesise that when longitudinal bonds spontaneously break as the filaments flex in thermal motion,
cofilin is now able to bind between sub-domain 1 of one actin subunit and sub-domain 2 of its longitudinally associated neighbour. In agreement with this notion,
several pieces of evidence show that the cofilins bind to the lower half of sub-domain 1 of actin. We hypothesise that once bound at this site, cofilins prevent the
actin filament re-forming the longitudinal bond. This makes it more likely that the other bonds will break at this point to sever the filament as the cofilins uniquely
alter the twist of the bound actin filament. This altered twist places extra tension on the actin-actin bonds around it. By viewing the filament from the barbed end,
each cofilin causes a clockwise rotation of the longitudinally associated subunit. This means that in a cofilin bound filament the cross over distance is 3/4 that of
normal actin filaments. In this way the extreme co-operative nature of cofilin binding to filaments is explained as each bound cofilin propagates the
torsional distortion along the filament |
|
|
|
