arget
cell by molecular motors. According to this hypothesis, dynein motors anchored at
the T cell interface with the target “reel in” the centrosome
http://ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1000260#article1.body1.sec1.p2

At the time when this paper of ours was in revision, an article appeared that presented molecular-biological evidence of the dynein-driven mechanism which we model here. Here is the reference that we regret came to our attention too late to be added "in proof": Martin-Cofreces et al., J. Cell Biol., 182:951-962, 2008. In essence (as it relates to what is modeled here), the cited paper reports experiments with overexpression of dynamitin, and with dynein heavy chain RNA interference, in which the centrosome polarization was inhibited. We are glad to see this experimental substantiation of the main assumption of our computational model, which had heretofore been a postulate supported by comparatively indirect evidence (work from the Poenie lab, as cited in the present paper), but at odds with the "negative result" in dynamitin experiments previously attempted in the Burkhardt lab (unpublished as such but briefly mentioned in Cannon and Burkhardt, Immunol. Rev., 186:90-99, 2002).

Since this new paper lends such crucial support to our modeling presented here, I feel obligigated to notice some aspects of these new results that appear less clear to me. While the statistics in the new paper by Martin-Cofreces is most impressive, many of the micrographs reveal morphology in the genetically manipulated cells that is different from the "control" cells. And specifically, this experimentally induced morphology looks "unpromising" for the centrosome polarization in light of our previous computational analysis of the effects of cell-body compactization on centrosome positioning (Arkhipov and Maly, Phys. Biol. 3:209, 2006; Baratt et al., PLoS ONE, 3:e3861, 2008). So, in figure 1F of Martin-Cofreces et al., the control cell ("M1") develops a wide synapse with the body of the cell closely apposed to it. This is the cell shape that energetically favors juxtaposition of the centrosome to the synapse (in a sense similar to the "favored" molecular conformations in organic chemistry), according to our previous computational analysis cited above. The first cell overexpressing dynamitin ("M2") in the discussed figure develops a far smaller synaptic area, while the second cell overexpressing dynamitin ("M3") develops a wide synapse, but the main cell body remains separated from it by an extended "stalk". These morphologies are in my experience comparatively unusual for mature T-cell conjugates, and they should not favor the functional centrosome orientation for quasi-elastic, energetic reasons analyzed in our structural-optimality models cited above. Further in Fig. 2 of Martin-Cofreces et al., panel A shows a much smaller area of contact with the target in the "experimental" cell than in the "control" cell. The complex morphologies of ternary conjugates in panel B are somewhat difficult to interpret for me. In panel C, again, the "control" cell has a wide area of contact with the target and a closely apposed main cell body, while the "experimental" cell has an unusually elongated body. I.e. this figure, too, raises the same questions: is the distinctive morphology of the cells with disrupted dynein function a consequence, or, as our cited computational analysis may suggest, rather the cause of the lack of proper centrosome positioning in them?

We apologize for the delayed reading of this groundbreaking paper by Martin-Cofreces et al., and for failure to add a reference to it at the final stages of this manuscript's revision and proofreading.