Coupled Active Systems Encode an Emergent Hunting Behavior in the Unicellular Predator Lacrymaria olor

Abstract

Many single-celled protists use rapid morphology changes to perform fast animal-like behaviors. To understand how such behaviors are encoded, we analyzed the hunting dynamics of the predatory ciliate Lacrymaria olor, which locates and captures prey using the tip of a slender "neck" that can rapidly extend more than seven times its body length (500 μm from its body) and retract in seconds. By tracking single cells in real-time over hours and analyzing millions of sub-cellular postures, we find that these fast extension-contraction cycles underlie an emergent hunting behavior that comprehensively samples a broad area within the cell's reach. Although this behavior appears complex, we show that it arises naturally as alternating sub-cellular ciliary and contractile activities rearrange the cell's underlying helical cytoskeleton to extend or retract the neck. At short timescales, a retracting neck behaves like an elastic filament under load, such that compression activates a series of buckling modes that reorient the head and scramble its extensile trajectory. At longer timescales, the fundamental length of this filament can change, altering the location in space where these transitions occur. Coupling these fast and slow dynamics together, we present a simple model for how Lacrymaria samples the range of geometries and orientations needed to ensure dense stochastic sampling of the immediate environment when hunting to locate and strike at prey. More generally, coupling active mechanical and chemical signaling systems across different timescales may provide a general strategy by which mechanically encoded emergent cell behaviors can be understood or engineered.

Keywords: coupled active systems; emergent cellular dynamics; single cell behavior.

Figure 1.. The unicellular predator Lacrymaria olor uses ultrafast morphology dynamics to hunt.

(A) Stylized image of Lacrymaria cell in the act of striking prey. (B) Image stack (40×1 sec) of an actively hunting cell alongside a dormant cell. (C) Digitization scheme for tracking Lacrymaria subcellular anatomy and shape. (D) Visualization of the hunting behavior (head position and neck length) of a single cell over 90 minutes. Locations of the head are plotted in the lab reference frame and are color-coded by time. See also Figure S7, Video S1 and Video S2.

Figure 2.. During a hunting event, Lacrymaria comprehensively and efficiently samples points from within an expansive strike zone.

(A) Lab-reference frame sampling area plots derived from hunting events from 8 separate cells, color-coded by time. (B) Representative trace of the body displacement during a hunting event. See also Figure S1A. (C) Dynamics for unique sampling for 25 hunting events. (D) The data from (C), normalized to a theoretical maximum number of points that could be sampled based on the maximum neck length and the positions of the body during the event. (E) Scheme for aligning the cell’s sampling behavior to the reference frame of the cell body to produce a “strike zone” of the points the cell can strike from a fixed position. (F) Strike zone plots for 4 cells of different indicated sizes. See also Figure S1C. (G) Correlation between size of the cell’s sampling area mean neck length. (H) Average strike zone plot obtained by normalizing all strike zone plots to their mean neck length. The dotted line shows the range expected for a circle of radius L_max centered at the base of the cell neck. The solid line shows the ellipse of best fit to the boundary of the empirically derived average sampling area. (I) Autocorrelation plots for extensile and lateral components of sampling, derived from 25 cells (grey dots) and globally fit to an exponential (red line).

Figure 3.. The head and neck take turns applying ciliary and contractile forces that drive the motion of the head and deform the structure of the neck.

(A) Head velocity distribution of a representative cell oriented to the tip of the cell neck. Forward velocities are purple; reverse velocities are grey. (B) DIC images of the cell head and neck while moving in the forward or reverse direction. (C) Diagram illustrating how the head and neck take turns driving the extension and retraction of the head. (D) Head speed distributions from 25 separate hunting events. (E) Neck length distributions from 8 separate hunting events. (F) Correlation between neck length and head speed. (G) Length dynamics from a 5-minute window of a representative hunting event, with a zoom in showing a 10-second window. (H) The moving average (15s window) for the neck length (L₀) from (G) in red superimposed on the overall dynamics in grey. (I) The fast dynamics (ΔL₀) that remain after subtracting off the slow-moving average. (J) Length distribution for the entire hunting event (grey) and the contribution of L0 to the variation (red). (K) as in (J) but with the contribution of ΔL₀ to the observed length distribution in purple. See also Figure S2 and Video S3.

Figure 4.. Lacrymaria’s neck and body are supported by distinct geometries of a continuous helical cytoskeleton which can undergo two modes of rearrangement.

(A) Maximum z-projection of two Lacrymaria cells (digitally juxtaposed) immunostained for acetylated tubulin. (B) Maximum z-projection of an extended Lacrymaria cell immunostained for centrin. (C) Maximum z-projection of an extended Lacrymaria neck immunostained for acetylated tubulin (cyan) and centrin (magenta). (D) Schematic of the helical cellular architecture of Lacrymaria that supports the morphological rearrangements that occur during hunting events. (E) Schematic depicting two distinct modes by which the cytoskeleton can rearrange to change the observed length of Lacrymaria’s neck. A slow mechanism which converts a segment of cytoskeleton in the body configuration into one in the neck configuration; and a fast mechanism which rearranges the existing pool of cytoskeleton already in the neck configuration. (F) The initial rates of length change measured during the early slow initiation period of a hunting event (n=10 events). (G) The initial rates of length change measured during the early slow termination period of a hunting event (n=10 events). (H) Box and whisker plots for the distribution of L₀ for 25 hunting events. (I) Distribution (points) and mean (line) extensile or contractile displacement about L₀ (ΔL₀) as a function of the magnitude of L₀. See also Figure S3 and Video S4.

Figure 5.. At short timescales, Lacrymaria’s neck behaves like an elastic filament under compressive or extensile loading.

(A) Physical analogy between Lacrymaria’s neck and an elastic filament of length L₀ under load. (B) The first four eigenshape normal modes in the harmonic series obtained from PCA analysis (globally-fit shapes shown in color; shapes from individual cells in grey). (C) Fraction of strong shape mode occurrence (>2.5 amplitude) as a function of displacement from L₀ for 25 hunting events (grey), with sigmoid fits and associated parameters shown (color) for the first and second shape modes. (D) Average amplitude of the first four shape modes as a function of L₀. The distribution of accessible L₀ values is shown as a grey histogram. (E) Extension behavior of the head speed (pink) and neck length (grey) derived from 101 pulls that we extracted from the data. Fits to an exponential for head speed (red) and neck length (black) are shown. (F) Length (displacement from L₀) and shape dynamics (blue: first mode; orange: second mode) during 1 minute of hunting activity. See also Figure S4 and Video S5.

Figure 6.. Compression of the neck scrambles the direction of the head’s extensile trajectory, expanding its sampling area.

(A) 3 superimposed DIC images of a cell during a cycle of retraction, compression and extension; and a schematic depicting the changes in trajectory. (B) Distribution of reorientation angles arising from neck shape obtained by pooling data from 25 hunting events. The associated range of tip angles is also depicted. (C) A plot of the tip reorientation angle as a function of amplitude for the first four eigenshapes. (D) Autocorrelation functions for length (grey) or amplitudes of each eigenshape mode (colors) during search (points) and associated exponential fit (line). (E) Strike zones for a representative hunting event recolored by the average amplitude of the first or second eigenshape mode of the neck when sampling that point. (F) Experimentally observed patterns of shape mode usage that facilitate sub-behaviors such as reorientation, whipping, or steering.

Figure 7.. Coupling active systems across fast and slow timescales encodes Lacrymaria’s emergent sampling behavior.

(A) Proposed model: fast cycles of compression and extension trigger buckling and sample trajectories about L₀; slower variations in L₀ change the location in space about which trajectories are sampled. (B) Sampling area from a simple simulation of (A) based on geometry and the empirical length and shape distributions. The simulated sampling area was fit to an ellipse (red). (C) Schematic showing that cobalt chloride treatment disrupts Lacrymaria’s rapid cycling. (D) Representative length dynamics for a single cell over the course of 2 hours. (E) Representative length and shape dynamics for a single hunting event from (D). (F) Representative point sampling behavior for a cobalt treated cell, showing the points visited over 5 minutes color coded by time. (G) Autocorrelation functions for extensile sampling pre (black) and post (red) cobalt treatment and associated integral timescale (n=10). (H) As in (G) but for lateral sampling. See also Figure S5, Figure S6, Video S6 and Video S7.