Foraging mechanisms in excavate flagellates shed light on the functional ecology of early eukaryotes

Abstract

The phagotrophic flagellates described as "typical excavates" have been hypothesized to be morphologically similar to the Last Eukaryotic Common Ancestor and understanding the functional ecology of excavates may therefore help shed light on the ecology of these early eukaryotes. Typical excavates are characterized by a posterior flagellum equipped with a vane that beats in a ventral groove. Here, we combined flow visualization and observations of prey capture in representatives of the three clades of excavates with computational fluid dynamic modeling, to understand the functional significance of this cell architecture. We record substantial differences amongst species in the orientation of the vane and the beat plane of the posterior flagellum. Clearance rate magnitudes estimated from flow visualization and modeling are both like that of other similarly sized flagellates. The interaction between a vaned flagellum beating in a confinement is modeled to produce a very efficient feeding current at low energy costs, irrespective of the beat plane and vane orientation and of all other morphological variations. Given this predicted uniformity of function, we suggest that the foraging systems of typical excavates studied here may be good proxies to understand those potentially used by our distant ancestors more than 1 billion years ago.

Keywords: clearance rate; early eukaryotic evolution; feeding current; prey capture; vane-bearing flagella.

Fig. 1.

(A) Summary tree of eukaryotes, showing current inferences that “excavates” represent extremely deeply branching lineages on either side of the root of eukaryotes, consistent with a common ancestry for the “typical excavate” cell architecture (orange) and an excavate-like LECA (red ball). Phylogeny after (5), with weakly supported branches collapsed and the two major clades containing most eukaryotes (Amorphea + CRuMs; Diaphoretickes) cartooned for clarity. The tree is rooted after (6). Some estimates of eukaryote phylogeny place Metamonada adjacent to Discoba [dashed yellow line; (7)], and one recent analysis suggests that the eukaryote root lies within metamonads [yellow ball; (8)]. These alternatives are also consistent with an excavate-like ancestry for all/virtually all extant eukaryotes. (B–F) Schematic representations of typical excavate species: (B) Jakoba libera, (C) Reclinomonas americana, (D) Kipferlia bialata, (E) Carpediemonas membranifera, and (F) Malawimonas californiensis. While foraging, J. libera, R. americana, and K. bialata are attached to the surface (bottom line); instead, C. membranifera and M. californiensis skid on the substrate (wavy background). Note that C. membranifera has a short third vane (orange). Reference scale bar, 5 µm.

Fig. 2.

Three modes of posterior flagellum kinematics inside the ventral groove. Schematic representations of mid-transversal cross-section views of the ventral groove (green, teal, and purple shadows) and the position of the flagellum (blue circle) equipped with a vane/s (yellow) between the first half of a beat cycle (panels on the Left) and the second half of the cycle (panels on the Right). The orange arrows indicate the direction of the moving flagellum. The right-side wall of the cell is on the right-side of the viewer. The cross-sections correspond to the observations of (A and B) J. libera and R. americana, (C and D) K. bialata; (E and F) C. membranifera, and M. californiensis. See animations in Movie S5.

Fig. 3.

Example particle tracks for 4 species: (A) J. libera, (B) R. americana, (C) K. bialata, and (D) C. membranifera,. Tracer particles are positioned (dots) at a frequency of 60 Hz (Except R. americana: 50 Hz), and the distance between positions thus indicates flow speed. Far from the flagellate, particle tracks are dominated by Brownian motion. The color indicates the direction of the flow, from yellow through red to blue.

Fig. 4.

CFD cases demonstrating the influence of various morphological aspects and surface proximity. In each panel, the case is first described (number of vanes, width of vane, asymmetry of groove, the presence of a surface dorsally (−5 µm) or ventrally (+5 µm) to the cell, and the presence of an active anterior and extended posterior flagellum). The anterior end of the cell is to the Right. Streamlines represent the averaged flow field, with sections omitted where flow velocity is below 2 µm s⁻¹ (corresponding to a Peclet number of ~1 for 0.5 µm passive prey particles). This threshold indicates where the advective feeding current overcomes the diffusive Brownian motion of passive prey particles. Q is the estimated clearance rate and P the estimated power consumption.

Fig. 5.

Effect of confinement on the feeding current generated by a (vaned) flagellum. A snapshot of the flow during the flagellum beating cycle (A, Top) shows that the vaned flagellum pushes the flow against the wall of the groove, directing the flow posteriorly (green arrow). This interaction, during the complete beat cycle, results in a relatively strong averaged flow through the groove (A, Bottom). In the absence of the groove (B, Top), the vaned flagellum pushes the flow in free space where only a small component of such flow is directed downward (green arrow), while most of the flow is directed sideways (red arrow). The oscillating sideways flows cancel out during the beat cycle, resulting in a weak average flow (B, Bottom). The clearance rate (Q) increases with the width of the vane (W), but most so for a flagellum within a groove (C, Top). Clearance rate (Q) vs. power consumption (P) (C, Bottom). Comparing clearance rates when adding vanes versus increasing the beat frequency. The data points represent simulation results for vane widths of 0.4 µm, 0.5 µm, 0.7 µm, 0.8 µm (1-vane configuration), and the last point 0.7µm (2-vane configuration). The dashed line illustrates the effect of increasing the beat frequency given by $Q=Q_{\mathit{naked}}*\sqrt{P/P_{\mathit{naked}}}$.