Functional Morphology of Diatom Frustule Microstructures: Hydrodynamic Control of Brownian Particle Diffusion and Advection
Total Page:16
File Type:pdf, Size:1020Kb
AQUATIC MICROBIAL ECOLOGY Vol. 24: 287–295, 2001 Published July 18 Aquat Microb Ecol Functional morphology of diatom frustule microstructures: hydrodynamic control of Brownian particle diffusion and advection Michelle S. Hale*, James G. Mitchell School of Biological Sciences, Flinders University, PO Box 2100, Adelaide, South Australia 5001, Australia ABSTRACT: While the intricate frustules of diatoms have been well described, understanding of function lags behind that of form. Here, we show that in the presence and absence of tangential flows of up to 800 µm s–1, diatom surface microtopographies control the diffusion and advection of sub- micrometre particles across their surfaces. With no flow, particles were localised on the solid areas of diatom frustules and had reduced diffusion compared to diffusion over flat glass. The magnitude of the effect was dependent on the ratio of the bead radius to the areolae radius (a/Ro). Under flow, par- ticles were deflected from the direction of flow by up to 170°, with over 60% of particles shifted more than 20° from the direction of flow. These results suggests that diatom frustules act as particle sorting structures, determining which particles reach the cell membrane and its receptors. This effect has important implications for nutrient uptake and fouling of cells by colloids and particulates, particu- larly when the particles are much smaller than the areolae. Variations in the hydrodynamic effects of different frustule microstructures on the diffusion and advection of Brownian particles may help explain the diverse range of frustule morphologies observed amongst diatoms. KEY WORDS: Diatom frustules · Particle diffusion · Brownian motion · Functional morphology Resale or republication not permitted without written consent of the publisher INTRODUCTION (Round et al. 1990). The species-specific size, arrange- ment and structure of areolae have been well de- Since the 19th century, the porous, lattice-like silica scribed for the 10 000 to 12 000 known species (Hasle & frustules that encase diatoms have inspired morpholo- Syvertson 1996) and morphology also varies within gists to carry out detailed descriptions of their form some species (Babanazarova et al. 1996). (Greville 1859). Frustules of centric diatoms exhibit Despite the diversity of frustule structures known to surface structures ranging from molecular to micro- exist, understanding of frustule function is poor. Frus- scopic, including radial rows of hexagonal chambers, tules have been postulated as a defence mechanism called areolae (Fig. 1). Each chamber has an outer against attacks by invasive protozoans (Kühn 1997). wall, exposed to the external environment, and an Other studies examining frustule function have focused inner wall, close to the cell membrane. It is usual that on the long spines produced by some species, which one of the walls is perforated by a large round hole are believed to deter predators (Fogg & Thake 1987) (foramen), while the other contains a porous plate, and either reduce or increase sinking rates (Smayda called a sieve plate (van der Hoek et al. 1995). The flux 1970). While areolae are recognised as the logical loca- of nutrients and exudates across the cell membrane tion for nutrient flux (Round et al. 1990), there have occurs at the base of each areola (the inner wall) been no adequate explanations for the morphological diversity observed among coexisting diatoms. Diatoms are exposed to a range of living and non- *E-mail: [email protected] living Brownian particles. In the ocean, they include © Inter-Research 2001 288 Aquat Microb Ecol 24: 287–295, 2001 patterns can play a role in the behaviour of Brownian particles by altering rates of advection and diffusion close to surfaces (Dinsmore et al. 1996, Duke & Austin 1998, Chou et al. 1999). This suggests that particle-surface interactions may alter the behaviour of Brownian particles close to frustule microstructures such as areolae. However, particle-surface inter- actions remain poorly understood and can produce unpredictable or inexplica- ble results (Feitosa & Mesquita 1991). While considerable progress has been made for solutions, fluid-particle systems still present significant technical chal- lenges, particularly for seawater where mixtures of complex molecules are com- mon. Previous experiments and models used flat surfaces and the controlling interactions were short-range van der Waals force and longer-range electrosta- tic force (Song & Elimelech 1995, Faibish et al. 1998). In cell-fluid systems, ranging from blood to the ocean, high salt con- centrations and membranes or other coat- ings reduce the Debye-Huckel length for Fig. 1. (A) Schematic representation of frustule structure of a centric diatom, electrostatic force to less than a few nano- showing the main features of areolae, including the hexagonal structure of each chamber and the presence of a porous sieve plate. Structure and metres (Kaplan et al. 1994). This leaves arrangement of areolae within the frustules of centric diatoms are species surface-induced drag on Brownian parti- specific. Scanning electron micrographs of cleaned centric diatom frustules cles as the dominant process at supra- show the areolation patterns of (B, C) Thalassiosira eccentrica and (D) Cos- nanometre distances from the surface. cinodiscus sp. Areolae are arranged in regular patterns across the frustule surface (B). In species such as T. eccentrica, areolae open to the external For diatoms, the presence of a frustule environment (outside) by foramen and open to towards the cell membrane excludes the formation of structures such (inside) by a sieve plate. The structure of Coscinodiscus sp. areolae is as cilia and flagella, and the uptake of reversed, with the foramen opening to the inside, and the frustule opening to large particles. The absence of flagella the outside by a recessed sieve plate and cilia, combined with strong ionic screening by seawater, suggests that nutrient molecules (~1014 ml–1), colloids (5 nm to 2 µm sorting, breakup or dispersal of these particles occur and ~107 to 109 ml–1; Wells & Goldberg 1994), viruses through drag interactions with the rigid, patterned, (20 to 200 nm; Ackerman & Dubow 1987) ~107 ml–1; surface microstructures. The absence of particle con- Hennes & Suttle 1995), and bacteria (0.2 to 1 µm and trol processes exhibited by other phytoplankton, cou- ~106 ml–1; Ducklow & Shiah 1993). Given that they are pled with the geometric regularity of frustules, make continually bathed in this array and concentration of diatoms ideal test systems for examination and experi- particles, frustules are remarkably clean surfaces. mentation of microstructure-induced particle control Without exterior moving parts or an external layer to by biological surfaces. Diatoms are technically as well continually shed, the mechanisms by which areolae as conceptually appealing, as most other biological remain open are unknown. At the micro- and nano- surfaces are membranaceous, delicate and do not pre- scales of single cells and their interactions, interpreta- sent the single sharp focal plane under light and elec- tions based solely on principles in physics have often tron microscopy that frustules do. been enlightening (Purcell 1977, Vogel 1983). In this Here we show that the microtopographies of diatom case however, principles from subdisciplines in physics frustules alter the diffusion and advection of sub- provide only hints and have not been applied to the micrometre particles, resulting in localised concentra- behaviour of particles at the cell-environment inter- tions of particles on ridged areas surrounding areolae. face. Recent advances in nanotechnology and micro- Variations in the hydrodynamic effects of surfaces on fluidics have shown that artificially constructed surface the diffusion and advection of Brownian particles may Hale & Mitchell: Functional morphology of diatom frustule microstructures 289 help explain the diverse range of frustule morpholo- tance coupling lenses to a Panasonic CCD camera gies observed amongst diatoms. It appears that passive (WV-BP550). Only the beads that were in focus were surface microstructures may control the diffusion of included in the analysis. Due to the depth of field of the particles near surfaces, hence helping to increase objective, this ensured that beads were no more than nutrient uptake by reducing areolae fouling. 1µm from the surface. With no flow, deviations from unbiased Brownian motion were quantified by analysing video footage of MATERIALS AND METHODS beads, frame by frame (1/25 s intervals) for 5 s. We measured the distance of beads from the centre of the Particle distributions. To examine the effects of are- ridged areas surrounding each areola, and calculated olae on particle movement, live and dead, acid-washed the proportion of time that beads spent on the ridged (Hasle & Fryxell 1970) cells of the diatoms Coscinodis- areas. The proportion of time that beads spent on the cus sp. and Thalassiosira eccentrica were attached to ridged areas surrounding the areolae was measured glass coverslips and placed in a small flow chamber. from 130 sequential frames of video footage, for each The chamber design was modified from those com- bead-surface combination. Experiments were repeated monly used in biofilm studies (Korber et al. 1990). Flow 5 times, each with a different diatom frustule to ac- cells were constructed from 2 perspex slides separated count for minor shape variation that occurs among the by a 2.5 mm thick o-ring. One slide had an ellipse cut individual frustules of one species. For control data, from it (major axis = 15 mm, minor axis = 10 mm) and a beads diffusing over a flat glass surface were viewed coverslip was placed over the hole and sealed against frame by frame, with a 2-dimensional diatom surface the o-ring by 4 small screws joining the perspex slides. structure of each species superimposed on the video Tubing was connected to two 23 G syringe needles, screen. This procedure ensured that observed localisa- which were pushed through the o-ring and into the tions of beads on ridged areas were due to the pres- chamber.