The kinetochore uncoiled

Raymond Zinkowski planned to change that. When he joined Brinkley's lab at the University of Alabama, Birmingham, as a grad student in 1986, “only a handful of labs were seriously investigating the centromeric region of chromosomes,” says Zinkowski. “It just looked like a parking place for a kinetochore handle to move chromosomes.”

Zinkowski became interested when studies using autoantibodies from scleroderma patients identified kinetochore-associated proteins (Earnshaw and Rothfield, 1985). Brinkley already held a notion that the kinetochore might be organized as repeat subunits. The Indian muntjac, a small Asian deer, had been identified as the mammal with the lowest number of chromosomes—a mere seven (Wurster and Benirschke, 1970). These cells with very few but large chromosomes gave Brinkley an easy target for examining kinetochore structure. He found what appeared to be compound, segmented structures. These may have evolved from fusions of the 46 chromosomes of an almost identical cousin, the Chinese muntjac (Brinkley et al., 1984). Also, Indian muntjac prekinetochores stained with the scleroderma autoantibodies in interphase showed more than seven discrete spots arranged in threadlike arrays.

Zinkowski treated G1 muntjac cells with a combination of hydroxyurea and caffeine, which breaks up the chromatin and induces cells to enter mitosis prematurely. The “pulverized chromatin” wandered from the spindle but the multiple kinetochores still made functional attachments to the spindle. What's more, the seven Indian muntjac kinetochores now appeared as 80–100 kinetochore subunits (Zinkowski et al., 1991).

In addition, when Brinkley and Zinkowski stretched out metaphase centromeres by hypotonic treatment and then stained kinetochore proteins using IF, they saw a repeating subunit pattern. It was similar to the less-organized interphase staining pattern.

The staining patterns, together with the kinetochore subunits found in G1 cells, argued for discrete, preformed protein subunits. Those subunits, the authors concluded, were separated by stretches of centromeric DNA that came together into coils as chromosomes condensed.

The model went against the current thinking that centromeres recruited kinetochore proteins into a solid disc structure after condensation (Jokelainen, 1967; Ris and Witt, 1981; Rattner, 1986). Brinkley says the model holds up in other stretched chromosome studies (Haaf and Ward, 1994) and in studies of budding yeast and worm centromeric proteins that appear to form subunits before chromosome condensation is complete (Meluh and Koshland, 1997; Moore et al., 1999).