Action sequencing in the spontaneous swimming behavior of zebrafish larvae - implications for drug development

All motile organisms need to organize their motor output to obtain functional goals. In vertebrates, natural behaviors are generally composed of a relatively large set of motor components which in turn are combined into a rich repertoire of complex actions. It is therefore an experimental challenge to investigate the organizational principles of natural behaviors. Using the relatively simple locomotion pattern of 10 days old zebrafish larvae we have here characterized the basic organizational principles governing the swimming behavior. Our results show that transitions between different behavioral states can be described by a model combining a stochastic component with a control signal. By dividing swimming bouts into a limited number of categories, we show that similar types of swimming behavior as well as stand-stills between bouts were temporally clustered, indicating a basic level of action sequencing. Finally, we show that pharmacological manipulations known to induce alterations in the organization of motor behavior in mammals, mainly through basal ganglia interactions, have related effects in zebrafish larvae. This latter finding may be of specific relevance to the field of drug development given the growing importance of zebrafish larvae in phenotypic screening for novel drug candidates acting on central nervous system targets.

Control analyses showing that the result that bouts of the same type are often repeated (see Figure  31 4) is not sensitive to the exact choice of number of trajectory sample points or bout classes (colors 32 represent few (blue; n=2) to many (red; n=20) clusters and symbols for the three lines of each color 33 (*, +, o) represent 10/20/30 sample points, respectively. Black line denotes the data presented in 34 Figure 4C (20 samples and 15 bout clusters). A) Conditional probabilities of bout repetitions 35 estimated from observed data, B) Theoretical probabilities assuming no memory of previous bout 36 type, C) Difference between observed and theoretical probabilities (i.e. A-B), and D) The differences 37 observed in C, in terms of number of standard deviations (Z-score; estimated from variance in A). 38 Note that too few clusters makes it highly likely to repeat the same bout type by chance and that too 39 many clusters makes in unlikely to observe a sufficient number of events to compare observed data 40 to chance level. Importantly, for all selections of sample and cluster numbers identical bout types are 41 repeated above chance level. in Figure 2E-F. 53 54

Supplementary Table 1 A-E 55
Quantitative comparison of goodness-of-fit to experimental data for five different models (cf. Figure  56 2 and 3) in terms of Negative log-likelihood, Akaike information criterion (AIC) and Bayesian 57 information criterion (BIC)  Feature description for a given zebrafish larvae and a time interval.

Number of values Distribution of bout classes
The fraction of swim bouts that are of class X.

1-15 Distribution of second order bout classes chains
The fraction of pairs of consecutive swim bouts where the first is of class X and the second of class Y.

16-240
Distribution of inter-bout waiting times The fraction of waiting times that are among the k/N and (k+1)/N shortest waiting times (by the all-time waiting time distribution of the current zebrafish larvae). N is the number of bins and k goes from 0 to N-1. Here, N=9 is used.

241-249
Distribution of second, third and fourth order inter-bout waiting time chains The fraction of (two/three/four)-pairs of consecutive waiting times where the first is of type X, the second of type Y, etc. Here the waiting times are binned by N = 3.

250-366
Distribution of swim bout durations The fraction of swim bouts with durations between 0.05 seconds and 0.05( + 1) seconds, where k goes from 0 to 19. Longer swim bouts are placed in bin number 21.

367-387
Distribution of swim bout distances The fraction of swim bouts with distances between 0.1 mm and 0.1 ( + 1) mm, where goes from 0 to 19. Longer swim bouts are placed in bin number 21.

388-408
Distribution of swim bout cumulative turning The fraction of swim bouts with cumulative turning between 15 degrees and 15( + 1) degrees, where k goes from −12 to 11. Swim bouts with cumulative turning less than -180 degrees are placed in bin number 1 and swim bouts with cumulative turning greater than 180 degrees are placed in bin number 26.

409-434
Number of bouts per second The number of bouts per second. 435 81 82