3. Experimental results and discussion 57
3.15. Discussion 96
3.15.1. Why do we observe different ring peeling behaviour with
We set out to investigate the effect of reducing actin turnover during fission yeast cytokinesis. To do so we employed a number of previously characterised mutants of the S. pombe actin severing protein Adf1 [66,93], whilst also attempting to use Jasplakinolide treatment of WT fission yeast cells to stabilise actin in contracting rings. From our experiments, we saw that contracting rings in S. pombe adf1 mutant cells (A)$
(B)$
Figure$16:$Effect$of$blocking$ring$constric:on$in$Adf1<M3$cells$
(A) Kymograph-and-montage-of-peeling-behaviour-in-rings-in-adf19M3-cps19191-cells-which-have-been-blocked-at-
36°C.-Faint-actomyosin-bundles-(asterisks)-can-be-observed-peeling-off-towards-an-actomyosin-aggregate-in-the- middle-of-the-ring.-
(B) Kymographs-of-adf19M3-cps19191-myp2Δ-rings,-made-from-two-perpendicular-views-of-the-ring-(as-indicated-in-
Figure-5A)-showing-the-absence-of-an-actomyosin-aggregate-in-the-middle-of-the-ring.-
(C) Kymograph-and-montage-of-ring-contracJon-in-adf19M3-bgs1+-cell-at-36°C,-which-also-shows-the-presence-of-the-
actomyosin-aggregate-seen-in-rings-in-adf19M3-cps1.191-cells.-
Scale-bars-in-montages/single-images-are-2-μm.-Scale-bars-in-kymographs-are-2-μm-and-5-minutes.- (C)$ adf19M3-cps19191-Rlc19mNG-(36°C)/ * * adf19M3-cps19191-myp2Δ-Rlc19mNG-(36°C)/ adf19M3-bgs1+-Rlc19mNG-(36°C)/ Δt-=-1-min- Δt-=-40s-
Figure 3.14: Effect of blocking ring constriction in Adf1-‐M3 cells. (A) Kymograph and montage of peeling behaviour in rings in adf1-‐M3 cps1-‐191 cells
which have been blocked at 36°C. Faint actomyosin bundles (asterisks) can be observed peeling off towards an actomyosin aggregate in the middle of the ring. (B) Kymographs of adf1-‐M3 cps1-‐191 myp2Δ rings, made from two perpendicular views
of the ring (as indicated in Figure 3.5A) showing the absence of an actomyosin aggregate in the middle of the ring.
(C) Kymograph and montage of ring contraction in adf1-‐M3 bgs1+ cell at 36°C, which also shows the presence of the actomyosin aggregate seen in rings in adf1-‐M3 cps1-‐
191 cells.
Scale bars in montages/single images are 2 μm. Scale bars in kymographs are 2 μm and 5 minutes.
displayed a peeling phenotype (Figure 3.1B, Figure 3.1C, Figure 3.2B). This phenotype was dynamic, with multiple peeling events occurring during cytokinesis, and in adf1-‐M2 and adf1-‐M3 cells the peeling bundles moved in a back-‐and-‐forth manner across the AMR (Figure 3.3E). When we treated S. Japonicus cells with Jasplakinolide, we also saw the peeling of actomyosin bundles away from the ring (Figure 3.12C). However this was a much more static phenotype, with only a single peeling event occurring directly after treatment with the drug, and with the peeled bundle remaining in the centre of the ring until the main ring contracted inwards to meet it.
This difference can perhaps be explained by the observation that peeling occurs across almost the entire circumference of the ring in S. japonicus (Figure 3.12C), whereas in S. pombe the peeling originates only from a small arc of the ring (Figure 3.1B, Figure 3.1C). This might mean that, in S. pombe, once the bundle is peeled off it is able to be pulled in to the opposite side of the ring through its attachment points, whilst because the bundles in S. japonicus peel off from everywhere it is not left with any attachment points to the ring, so the bundle cannot be reeled in. However, the question then becomes why there is a difference in the peeling locations between the two organisms. Perhaps there are some unknown structural differences between the AMRs in each organism, or perhaps Adf1 has an unknown additional role and/or Jasplakinolide has an unknown additional effect, which causes this difference in behaviour. Then there is also the question of why the peeling phenotype of rings in adf1-‐M2 and adf1-‐M3 cells is different to what we observed in adf1-‐1 cells. It is probably unlikely that this is due to differing degrees of severity of these mutations, since the adf1-‐M2 mutation induced a similar relative change in the ring contraction rate compared to the adf1-‐1 mutation at 30°C (Figure 3.1A, Figure 3.2A), potentially indicating that these alleles are similar with regards to their severity. Whilst the Adf1-‐ M2 and Adf1-‐M3 proteins have been biochemically characterised [66], this has not been carried out for Adf1-‐1 [93]. Therefore, we do not know whether the phenotype of the adf1-‐1 mutation is caused by a reduced F-‐
actin binding affinity, a reduced actin severing rate, a mixture of both, or by some other factor. Since the Adf1-‐M2 and Adf1-‐M3 proteins were found to have reduced actin binding and actin severing [66], it is possible that the Adf1-‐1 protein may only experience a reduction in one of these, which could then affect the exact nature of the observed peeling phenotype. We shall discuss this further in the next section.
Going further, one may also wonder why the ring peeling phenotypes in adf1-‐M2 and adf1-‐M3 cells are so similar, even though they are completely different mutations in the adf1 gene. Both Adf1-‐M2 and Adf1-‐M3 were previously found to have reduced actin binding and actin severing, and their behaviour only differed in the degree to which these properties were reduced [66]. This is despite the fact that the adf1-‐ M2 mutations are located on the opposite side of the protein to the adf1-‐ M3 mutations, which are found in the actin-‐binding domain [66]. How these two sets of mutations lead to the same qualitative effects will be difficult to answer without, for example, performing molecular dynamics or protein folding simulations to investigate their allosteric effect. However, the fact that these two mutations produce the same qualitative effects on the properties of the protein would suggest that the subsequent phenotypes would also be similar.
3.15.2. What is the cause of ring peeling in adf1 mutant cells?