Comparison between contraction-relaxation cycles

We examined the trajectories of landmarks in one dataset over six contraction- relaxation cycles (Fig. 5.9). The first four contractions were spontaneous (blue tones), and the final two followed application of 10nM oxytocin (red tones). We computed the quantitative tracking measures in Table 5.1 for each of the six cycles (Figs. 5.10, 5.11 and 5.12). The landmark trajectories were produced directly us- ing our motion-tracking algorithm, which refreshes after each contraction-relaxation cycle. Therefore while many landmarks are tracked over each cycle, there is some variability in landmark selection between cycles. All landmarks were used in the distributions shown in Figs. 5.10, 5.11 and 5.12 (top rows); for statistical analysis only the landmarks tracked over all cycles were used (bottom rows).

We found that the total distance and net distance travelled by landmarks over one cycle increased significantly following the application of oxytocin (Fig. 5.10A,

Figure 5.9:Trajectories of landmarks tracked over six contraction-relaxation cycles. (A) Raw fluorescence image of a Fluo-4 loaded myometrial slice. (B) Trajectories of all landmarks tracked over six contraction-relaxation cycles, colour-coded according to cycle number. Note that the as the algorithm refreshes after each contraction-relaxation cycle, there is variability between landmarks identified for tracking over each cycle. (C) Change in baseline fluorescence (∆F) averaged over entire tissue slice, colour-coded as in (B). Note that spontaneous contractions are shown in blue tones and oxytocin-induced contractions are shown in red tones.

B). This is in agreement with the protracted motion trajectories seen in Fig. 5.3 fol- lowing oxytocin application. This is further supported by a decrease in the stochas- ticity ratio following application of oxytocin; a higher stochasticity ratio measured over a contraction-relaxation cycle represents a more direct trajectory. The total distance travelled by landmarks during the spontaneous contraction immediately preceding oxytocin application is significantly less than the distance travelled during the first three spontaneous contractions. The net distance travelled by landmarks during this cycle is significantly greater than during the first three cycles, indicating that the end position of the landmarks is further away from their starting point. To- gether these results suggest that this tissue has not fully relaxed in the fourth cycle when the application of oxytocin induces the next contraction. The confinement ratio (Fig. 5.11A) when taken over an entire contraction-relaxation cycle provides another indication of how far a landmark’s end point is to its start point; the higher the ratio, the greater the distance between the two points. The confinement ratio for the fourth spontaneous contraction is significantly higher than for the first three spontaneous contractions, further supporting that the tissue does not completely re- lax during this cycle. This result supports previous experiments showing incomplete tissue relaxation following oxytocin application (Shmygol et al., 2006).

The total distance travelled by landmarks during the first oxytocin-induced contraction is significantly greater than the distance travelled during the second oxytocin-induced contraction (Fig. 5.10A). This is not surprising, as the contraction- relaxation cycle lasts longer in the first oxytocin-induced cycle than the second (see ∆F plot in Fig. 5.9C). The area under the ∆F curve is much greater for the first oxytocin-induced contraction, and this is associated with the increase in [Ca2+]iand greater force behind oxytocin-induced contractions (Nakao et al., 1997). The net distance travelled in the first oxytocin-induced contraction is higher than in the sec- ond and the confinement ratio for the first cycle is higher (Figs .5.10, 5.11). This could be due to experimental conditions (e.g., tissue slippage) or it may represent a biophysical property of oxytocin-induced contractions. The tissue in the first con- traction may have not relaxed back to its pre-contraction state, or the extra force generated during this contraction may deform the tissue slightly. There is also sig- nificant variation between the net distances and the confinement ratios in the first three spontaneous contractions; further analysis would be required to investigate whether this variation is found within other datasets.

There is no significant change in the mean curvilinear speed between all spontaneous contractions and oxytocin-induced contractions or within each group (Fig. 5.12A). This would suggest that the increase in force generated by oxytocin-

Figure 5.10: Increase in distance travelled by landmarks during oxytocin-induced contractionsStatistics showing (A) total distance (B) net distance and (C) maximum dis- tance travelled during six contraction-relaxation cycles. (A, B, C) Top: Distributions of landmark distances over six contraction-relaxation cycles. Bottom: Comparison of dis- tance travelled by the same 36 landmarks (left) between first four spontaneous contraction- relaxation cycles; (middle) between two contraction-relaxation cycles following application of 10nM OT; and (right) between all spontaneous cycles and all oxytocin-induced cycles. *P <0.05, **P <0.01, ***P <0.001, Wilcoxon rank-sum test.

Figure 5.11:Landmark trajectories are less direct and end further away from starting point in oxytocin-induced contraction-relaxation cyclesStatistics showing (A) confine- ment ratio and (B) stochasticity ratio for landmarks during six contraction-relaxation cycles. (A, B, C)Top: Distributions of landmark ratios over six contraction-relaxation cycles. Bot- tom:Comparison of ratios by the same 36 landmarks (left) between first four spontaneous contraction-relaxation cycles; (middle) between two contraction-relaxation cycles follow- ing application of 10nM OT; and (right) between all spontaneous cycles and all oxytocin- induced cycles. *P <0.05, **P <0.01, ***P <0.001, Wilcoxon rank-sum test.

Figure 5.12: No change in average speed of landmarks between spontaneous and oxytocin-induced contractionsStatistics showing (A) mean curvilinear speed and (B) mean straight-line speed for landmarks during six contraction-relaxation cycles. (A, B, C)Top: Distributions of landmark speeds over six contraction-relaxation cycles. Bottom:

Comparison of speeds by the same 36 landmarks (left) between first four spontaneous contraction-relaxation cycles; (middle) between two contraction-relaxation cycles follow- ing application of 10nM OT; and (right) between all spontaneous cycles and all oxytocin- induced cycles. *P <0.05, **P <0.01, ***P <0.001, Wilcoxon rank-sum test.

Measure Definition

Instantaneous speed vi =d(pi−1,pi)/∆t

Instantaneous confinement ratio rcon,i=d(p1,pi)/ ∑i−1

j=1d(pj,pj+1)

Table 5.2: Instantaneous tracking measuresThe pointpi = (xi, yi)is the location of a landmark at framei∈ {2 :N}and the distanced(pi,pj)between two pointspiandpj is taken as the Euclidean norm. ∆tis the time interval in seconds between two consecutive frames.

induced contractions occurs only via protraction rather than protraction and an in- crease in the velocity of the contraction. However, these results include the entire contraction-relaxation cycle, and an increase in velocity of contraction may there- fore not be detected. It would be interesting to apply quantitative measures to the contraction phase of the cycle only to elicit whether a change in velocity of contrac- tion is detected following application of oxytocin.