On 2017 Oct 10, L David Sibley commented:

We appreciate the posting of the comment by Dr. Meissner and also the primary data provided by the Kursula group (Kumpula et al., Kumpula EP, 2017). Although we agree that there are differences in the conclusions, we feel that these stem primarily from differences in methodology and biological substrates. As such, it would be premature from this new study to make sweeping statements about the polymerization behavior “apicomplexan” actins.

First, we note that our study used Toxoplasma gondii actin (TgACT1) Skillman KM, 2013, while Kursula group studied Plasmodium actin, PfACT1 more specifically. We agree that the two actins are somewhat similar, and so it would be surprising if they had an intrinsically different mechanism of polymerization. However, our previous studies have noted differences between these substrates and found that polymerization of PfACT1 is more efficient than that of TgACT1 as shown by light scattering and microscopy Skillman KM, 2011.

Second, Kumpula et al., suggest that sedimentation may not be a reliable method for monitoring polymerization kinetics of apicomplexan actins due to the formation of small oliogmeric species that do not fully sediment at 100,000g or even 450,000g. We are well aware of this limitation, and indeed it is also shown in our paper by several methods Gershon D, 1990. We considered the formation of heterogenous small oligomers in the experiments we performed, including in the simulations and determination of rate constants. Importantly all of our findings show that a simple isodesmic model can fit the data for assembly and turnover. These findings further predict that if there is a cooperative component, it is exceedingly small relative to conventional actins. Moreover, our conclusion that TgACT1 polymerizes by an isodesmic mechanism was not based solely on sedimentation but was supported by light scattering assays, which demonstrate a lack of a lag period associated with a normal critical concentration (Cc) Skillman KM, 2013.

Third, Kumpula et al., used a modified pyrene assay to monitor the “kinetics’ of filament formation. We agree that the pyrene assay is normally well suited for studying kinetics, but may not be ideal for monitoring thermodynamics associated with the intrinsic polymerization mechanism. Kumpula et al., report that at low concentrations of PfACT they detect evidence for a Cc, based on plotting the fluorescence signal vs. actin concentration. However, under these conditions where pyrene labeling was done at sub-stochiometric levels, the lack of a signal below a particular concentration may be due to lack of sensitivity, rather than an actual Cc. It would be important to determine whether such a Cc could also be detected by intrinsic tryptophan quenching, which may be more sensitive. We would also like to point out that unlike these two assays that attempt to monitor the Cc at very low concentration of protein, the estimate of Cc from sedimentation assays is based on behavior across a wide range of concentrations and thus it is more robust to small fluctuations, variability in protein levels, or assay differences.

Finally, we note than both our study and that of Kumpula used recombinant actin produced in insect cells. Hence, the previous suggestion that apicomplexan actins cannot be functionally expressed heterologously Olshina MA, 2016 is incorrect and not the reason for the differences described in polymerization behavior. Rather, it is clear that recombinant actins produced in an appropriate manner exhibit many interesting functional properties. Further studies may reveal to what extent these features are shared vs. unique among divergent actins.

On 2017 Sep 23, Markus Meissner commented:

A recent report from the Kursula group demonstrated that the sedimentation assays used in Skillmann et al., 2013 are unlikely to obtain reliable results for apicomplexan actins. In this study the Kursula group (see: https://www.nature.com/articles/s41598-017-11330-w) used a protocol based on pyrene labelling and demonstrated that polymerisation of apicomplexan (Plasmodium) actin is occurring in a cooperative manner, meaning it requires a critical concentration. Surprisingly, the Cc is very similar to canonical rabbit actin. Furthermore, it was shown that shorter filament lengths result from higher depolymerisation rate. In conclusion, it appears that -like all other known eukaryotic actins- apicomplexan actin polymerises in a cooperative manner, requiring a nucleation reaction (see Kumpula et al., 2017).

On 2017 Sep 23, Markus Meissner commented:

A recent report from the Kursula group demonstrated that the sedimentation assays used in Skillmann et al., 2013 are unlikely to obtain reliable results for apicomplexan actins. In this study the Kursula group (see: https://www.nature.com/articles/s41598-017-11330-w) used a protocol based on pyrene labelling and demonstrated that polymerisation of apicomplexan (Plasmodium) actin is occurring in a cooperative manner, meaning it requires a critical concentration. Surprisingly, the Cc is very similar to canonical rabbit actin. Furthermore, it was shown that shorter filament lengths result from higher depolymerisation rate. In conclusion, it appears that -like all other known eukaryotic actins- apicomplexan actin polymerises in a cooperative manner, requiring a nucleation reaction (see Kumpula et al., 2017).

On 2017 Oct 10, L David Sibley commented:

We appreciate the posting of the comment by Dr. Meissner and also the primary data provided by the Kursula group (Kumpula et al., Kumpula EP, 2017). Although we agree that there are differences in the conclusions, we feel that these stem primarily from differences in methodology and biological substrates. As such, it would be premature from this new study to make sweeping statements about the polymerization behavior “apicomplexan” actins.

First, we note that our study used Toxoplasma gondii actin (TgACT1) Skillman KM, 2013, while Kursula group studied Plasmodium actin, PfACT1 more specifically. We agree that the two actins are somewhat similar, and so it would be surprising if they had an intrinsically different mechanism of polymerization. However, our previous studies have noted differences between these substrates and found that polymerization of PfACT1 is more efficient than that of TgACT1 as shown by light scattering and microscopy Skillman KM, 2011.

Second, Kumpula et al., suggest that sedimentation may not be a reliable method for monitoring polymerization kinetics of apicomplexan actins due to the formation of small oliogmeric species that do not fully sediment at 100,000g or even 450,000g. We are well aware of this limitation, and indeed it is also shown in our paper by several methods Gershon D, 1990. We considered the formation of heterogenous small oligomers in the experiments we performed, including in the simulations and determination of rate constants. Importantly all of our findings show that a simple isodesmic model can fit the data for assembly and turnover. These findings further predict that if there is a cooperative component, it is exceedingly small relative to conventional actins. Moreover, our conclusion that TgACT1 polymerizes by an isodesmic mechanism was not based solely on sedimentation but was supported by light scattering assays, which demonstrate a lack of a lag period associated with a normal critical concentration (Cc) Skillman KM, 2013.

Third, Kumpula et al., used a modified pyrene assay to monitor the “kinetics’ of filament formation. We agree that the pyrene assay is normally well suited for studying kinetics, but may not be ideal for monitoring thermodynamics associated with the intrinsic polymerization mechanism. Kumpula et al., report that at low concentrations of PfACT they detect evidence for a Cc, based on plotting the fluorescence signal vs. actin concentration. However, under these conditions where pyrene labeling was done at sub-stochiometric levels, the lack of a signal below a particular concentration may be due to lack of sensitivity, rather than an actual Cc. It would be important to determine whether such a Cc could also be detected by intrinsic tryptophan quenching, which may be more sensitive. We would also like to point out that unlike these two assays that attempt to monitor the Cc at very low concentration of protein, the estimate of Cc from sedimentation assays is based on behavior across a wide range of concentrations and thus it is more robust to small fluctuations, variability in protein levels, or assay differences.

Finally, we note than both our study and that of Kumpula used recombinant actin produced in insect cells. Hence, the previous suggestion that apicomplexan actins cannot be functionally expressed heterologously Olshina MA, 2016 is incorrect and not the reason for the differences described in polymerization behavior. Rather, it is clear that recombinant actins produced in an appropriate manner exhibit many interesting functional properties. Further studies may reveal to what extent these features are shared vs. unique among divergent actins.