Surface-bound iron: a metal ion buffer in the marine brown alga Ectocarpus siliculosus?

Although the iron uptake and storage mechanisms of terrestrial/higher plants have been well studied, the corresponding systems in marine algae have received far less attention. Studies have shown that while some species of unicellular algae utilize unique mechanisms of iron uptake, many acquire iron through the same general mechanisms as higher plants. In contrast, the iron acquisition strategies of the multicellular macroalgae remain largely unknown. This is especially surprising since many of these organisms represent important ecological and evolutionary niches in the coastal marine environment. It has been well established in both laboratory and environmentally derived samples, that a large amount of iron can be ‘non-specifically’ adsorbed to the surface of marine algae. While this phenomenon is widely recognized and has prompted the development of experimental protocols to eliminate its contribution to iron uptake studies, its potential biological significance as a concentrated iron source for marine algae is only now being recognized. This study used an interdisciplinary array of techniques to explore the nature of the extensive and powerful iron binding on the surface of both laboratory and environmental samples of the marine brown alga Ectocarpus siliculosus and shows that some of this surface-bound iron is eventually internalized. It is proposed that the surface-binding properties of E. siliculosus allow it to function as a quasibiological metal ion ‘buffer’, allowing iron uptake under the widely varying external iron concentrations found in coastal marine environments.


Introduction
Iron acquisition by marine organisms is of high interest due to its crucial role in many biological processes of the oceanic biota. In particular, algae have received considerable attention for their important ecological position as primary producers with high intracellular iron requirements related to photosynthesis, respiration, and electron transport. Satisfying these demands is complicated by the low solubility of Fe(III) in oxic seawater and the fact that a large fraction of this limited iron is complexed by strong natural organic chelates (Rue and Bruland, 1995). Evidenced by phytoplankton bloom formation through iron fertilization in high-nutrient, low-chlorophyll regions, iron is now well known to limit primary productivity in certain oceanic regimes (Martin and Fitzwater, 1988).
While the iron uptake and storage mechanisms of terrestrial/higher plants have been well studied, the corresponding systems in marine algae have received far less attention, with phytoplankton attracting most of this interest. Studies have shown that while some species of coccolithophores, diatoms, and unicellular green algae utilize unique mechanisms of iron uptake (Morrissey and Bowler, 2012), many acquire iron through the same general mechanisms as higher plants (Sutak et al., 2012). The first of the two general mechanisms utilized by land plants (Strategy I) consists of a ferric chelate reductase in which transplasma membrane electron transport via flavohemoproteins reduces Fe(III) to Fe(II). Free ferrous iron is then taken up directly or reoxidized by a multicopper oxidase coupled to a high-affinity iron permease. Strategy II involves secretion by the plant cell, or associated bacteria, of high-affinity, iron-specific phytosiderophores. These phytosiderophores scavenge iron in the soil/extracellular milieu and are then transported back into the cell by sideroreceptors, after which iron dissociates (Morrissey and Bowler, 2012). However, while a phytosiderophore mechanism has never been confirmed in marine algae, it has been suggested to be present in the dinoflagellate Prorocentrum minimum (Trick et al., 1983) and the coccolithophore Emiliania huxleyi (Boye and van den Berg, 2000). Having adapted to various irondepleted environments, the cell-wall-containing unicellular eukaryote Saccharomyces cerevisiae has served as a model for investigations of these uptake strategies (Philpott, 2006). Genomic and iron uptake studies have confirmed the ability of yeast to exploit both of these strategies.
In contrast, the iron acquisition strategies of the multicellular macroalgae remain largely unknown. This is especially surprising since many of these organisms represent important ecological and evolutionary niches in the coastal marine environment. Together with the Stramenopile lineages of diatoms, golden-brown algae, and oomycetes, brown algae have evolved independently of other major photosynthetic lineages such as green plants and red algae and thus display many unique characteristics. Kelp forests and filamentous colonies dominate many coastal regions in terms of biomass and therefore play key roles in the natural product market (i.e. alginate and fucans), halogen biogeochemical cycle, and anthropogenic element remediation (Küpper et al., 1998;Davis et al., 2003;Raize et al., 2004;Usov and Bilan, 2009;Benvegnu and Sassi, 2010). Indeed, brown algal biomass is so prevalent in some areas that certain industries view it as a nuisance, prompting considerations for brown algae as a feedstock for biofuels (Bringezu et al., 2009;Vertès et al., 2010).
Ectocarpus siliculosus is a filamentous brown alga with a worldwide distribution along temperate coastlines, and is a nuisance as a 'fouling' organism on many man-made surfaces in the sea. It has some significant advantages as an experimental model and constitutes one of the best-studied seaweeds (Peters et al., 2004;Charrier et al., 2008). It can easily be cultivated in small volumes of seawater media both axenically and with associated bacteria, its entire lifecycle can be completed within a few months in culture (Müller et al., 2008), and many molecular tools are available, including mutant collections, microarrays, and proteomics data. It has also recently become the first seaweed of which the entire genome has been sequenced and thus offers unprecedented opportunities for study (Cock et al., 2010). Using a combination of genomic and spectroscopic approaches, this study group has recently shown that E. siliculosus acquires iron using a modified Strategy 1 reductive-oxidative mechanism, with iron storage likely involving a non-ferritin vacuole based system (Böttger et al., 2012).
While it seems likely that the iron that many marine algae take up from the environment, irrespective of its detailed internalization mechanism, arrives at the cell surface by diffusion, there is growing evidence for more 'active' means of concentrating this element prior to uptake. It has been well established in both laboratory and environmentally derived samples that a large amount of iron can be 'non-specifically' adsorbed to the surface of marine algae (Hutchins et al., 1999). This surface-bound iron may derive from simple electrostatic attraction between colloidal iron hydroxide particles and the cell surface, from de novo precipitation of iron hydroxo polymers from equilibrium solutions at the cell surface due to the increased surface pH relative to bulk seawater, or possibly from other mechanisms (Milligan et al., 2009). While this surface adsorption phenomenon is widely recognized and has prompted the development of experimental protocols to eliminate its contribution to iron uptake studies, its potential exploitation as a concentrated iron source for marine algae is only now being recognized (Hudson and Morel, 1989;Tovar-Sanchez et al., 2003). The extraordinary swarming behaviour of the related cyanobacterium Trichodesmium when confronted with particulate iron-containing dust is a spectacular example of active surface concentration of iron prior to actual uptake (Rubin et al., 2011). More recent examples of possible surface concentration of iron come from the diatoms, the alveolate Chromera velia, and others where in some cases there is evidence that the surface-bound iron is ultimately internalized (Sutak et al., 2010(Sutak et al., , 2012Ingall et al., 2013).
Since the cell wall of the brown algae is largely composed of anionic alginate polysaccharides, which account for 20-40% of their dry mass (Mabeau, 1987), it is scarcely surprising that the surface of these organisms should have considerable metalbinding properties. Sulphated polysaccharides are another structural polymer common to brown algae with metal-binding anionic functional groups. Indeed brown algae have been widely discussed for their potential in heavy metal biosorption applications (Davis et al., 2003).
The current study explores the nature of the extensive and powerful iron binding on the surface of both laboratory and environmental samples of E. siliculosus and examines its potential effects on the subsequent uptake mechanism. It is proposed that the surface-binding properties of E. siliculosus function as a sort of metal ion 'buffer' allowing iron uptake under the widely varying external iron concentrations found in coastal marine environments.

Biological material
For the initial XAS experiments at EMBL-DESY, E. siliculosus CCAP 1310/4-of which the genome has been sequenced recently (Cock et al., 2010)-was obtained from the Culture Collection of Algae and Protozoa (CCAP) at the Scottish Association for Marine Science and grown axenically in modified Provasoli-enriched (Andersen, 2005) seawater from the Firth of Lorne (west coast of Scotland) at 14 °C under a 12/12 light/dark photocycle. For later experiments, E. siliculosus EcSil NZ KU 1-3♂ (CCAP 1310/56) was grown axenically in modified Provasoli-enriched Scripps Pier seawater (SPSW) at 17 °C under a 12/12 light/dark photocycle. The iron content of SPSW was determined to be approximately 4 nM and is thus defined as the concentration for iron-limited growth. Prior to all experiments, Ectocarpus was starved for a period of 5-10 days under iron-limited conditions. Iron-replete conditions were obtained by adding Fe-EDTA to SPSW at 30 μM. Environmental samples of E. siliculosus were harvested from the shore near Dunstaffnage Marine Lab in Oban, Scotland.
Surface-bound iron studies 55 FeCl 3 radionuclide was obtained from Perkin-Elmer and used to prepare the 55 FeEDTA solution used as an iron source. Internalized iron was determined after all surface-bound Fe(III) species were removed from E. siliculosus using the titanium(III)citrate/EDTA procedure as described by Hudson and Morel (1989). Total iron taken up (both external and internal) was determined using samples not washed with the Ti-citrate EDTA reagent. Surface-bound iron was defined as the 55 Fe signal of cells not treated with Ti less the internalized iron signal of titanium washed replicates. For pulse-chase experiments, E. siliculosus cells were incubated in 100 ml of 30 μM radiolabled 55 FeEDTA SPSW medium for 6 h. Cells were then removed from the medium, washed with 5 ml low iron SPSW (~4 nM 56 Fe), and resuspended in 100 ml of either (a) ultra-low iron oligotrophic Pacific Ocean water (~0.1 nM 56 Fe) or (b) SPSW containing 30 μM 'cold' 56 FeEDTA. All samples were filtered on 5 μm Millipore polycarbonate filters and washed with 25 ml of artificial sea water (Amin et al., 2007). Thorough washing with both the titanium reagent and subsequently with artificial seawater is essential to eliminate artefacts caused by surface binding of Fe(III) and Fe(II), respectively. Filtered cells were dried under a vacuum and placed in scintillation vials with 1 ml sodium hypochlorite to bleach the chlorophyll and reduce quenching effects. Vials were then heated in a 55 °C water bath for 1 h and allowed to vent overnight at room temperature to allow chloride evaporation. Hionic Fluor liquid scintillation fluid (15 ml; Perkin-Elmer) was added to each of the vials which were incubated in the dark for at least 2 h to eliminate any background chemiluminesence. The 55 Fe taken up was measured on a Beckman-Coulter LS 6500 scintillation counter using the tritium channel. Total iron uptake per mg dry E. siliculosus cells was calculated based on specific activity, measured count rates, scintillation counting efficiency, and biomass measurements.

Binding constant and equilibrium studies
Cell-surface iron affinity is expressed as the equilibrium constant K′ eff : Fe T 3+ in reaction 1 (also known as Fe(III)') represents the total soluble and unbound aqueous iron (III)  ). Cell-surface iron affinity constants were estimated by measuring surface-bound iron (reaction 2) with varying concentrations of EDTA and varying pH while [Fe(III)] remained fixed at 30 μM: where FeEDTA' represents the total EDTA/iron(III) species (FeEDTA -+ FeHEDTA + Fe(OH)EDTA 2-) and EDTA' represents the total EDTA species (EDTA 4-+ HEDTA 3-+ H 2 EDTA 2-+ H 3 EDTA 1-+ H 4 EDTA + H 5 EDTA 1+ ) in equilibrium under these conditions. In the calculations, it was assumed that only the hydrated iron(III) species ( ) bind to the cell surface while neither EDTA' nor FeEDTA' do. Relatively constant values (± 0.5 log unit) were calculated for Fe T 3+ at a constant pH over a range of experimental conditions (10-18 mg cells exposed to 33-1500 μM EDTA in the presence of 30 μM Fe(III) at pH 8.7 ± 0.1), which confirms the validity of the assumptions and the computational approach. Details of the data analysis can be found in the Supplementary Material S1 (available at JXB online).
Energy-dispersive X-ray analysis Cells were fixed in a 0.1 M phosphate buffer solution containing 2% (w/v) paraformaldehyde, 1% (w/v) glutaraldehyde, and 1% (w/v) caffeine for 2 h. The fixed cells were then washed with 0.1 M phosphate buffer and dehydrated in successive ethanol baths of 30, 50, 75, 85, 95, and 100% (×3). The cells were then embedded in 1:1 (v/v) ethanol/LWR for 3 h followed by 100% LWR overnight in gelatin capsules under vacuum. Sections (3 μm) were cut on a Leica EMUC6 microtome and deposited on glass slides. Slides were coated with platinum in a Quorum Technologies Q150T ES sputter coater. Platinum-coated samples were analysed under high vacuum in a Quanta 450 FEG environmental scanning electron microscope equipped with an Oxford Instruments INCA energy-dispersive X-ray microanalysis system.

Transmission Mössbauer spectroscopy
For transmission Mössbauer spectroscopy (TMS), algal samples were either washed with the titanium citrate/EDTA reagent in order to remove adventitious iron from the algal surface, or they were used untreated. Cells were harvested by vacuum-assisted filtration. Pellets were weighed, transferred into Delrin Mössbauer sample holders, frozen in liquid nitrogen, and kept at this temperature until measurement except for overnight transport on dry ice. The Mössbauer spectra were recorded in the horizontal transmission geometry using a constant acceleration spectrometer operated in conjunction with a 512-channel analyser in the time-scale mode. The detector consisted of a proportional counter filled with argon/methane (9:1). The source was at room temperature and consisted of 1.4 GBq [ 57 Co] diffused in Rh foil (WissEl, Starnberg, Germany). The spectrometer was calibrated against α-iron at room temperature. For measurements at 77 K, samples were placed in a continuous-flow cryostat (Oxford Instruments). For measurements at 4.3K and 1.8 K, a helium bath cryostat (MD306, Oxford Instruments) was employed. Spectral data were transferred from the multichannel analyser to a PC for further analysis employing the public domain Vinda program on an Excel 2003 platform. Isomer shift δ, quadrupole splitting ΔEQ, magnetic hyperfine splitting B hf , line width Γ, and percentage of the total absorption area were obtained by least-squares fits of Lorentzian lines to the experimental spectra. All values are rounded to the last given digit. The isomers shifts, the quadrupole splitting, and the line width are given in mm s -1 . Hyperfine fields are given in Tesla. The relative area is given in parts per hundreds.

X-ray absorption spectroscopy
For X-ray absorption spectroscopy (XAS), the K-edge iron X-ray absorption spectrum was recorded at the beamline D2 of the EMBL Outstation Hamburg at DESY, Germany. The DORIS storage ring operated at 4.5 GeV with the positron beam current ranging from 145 to 80 mA. A Si(111) double-crystal monochromator scanned X-ray energies around Fe K-edge (6.9-7.85 KeV). Harmonic rejection was achieved by a focusing mirror (cut-off energy at 20.5 KeV) and a monochromator detuning to 50% of its peak intensity. The sample cells were mounted in a two-stage Displex cryostat and kept at about 20 K. The X-ray absorption spectra were recorded as Fe K α fluorescence spectra with a Canberra 13-element Ge solid-state detector. Data reduction, such as background removal, normalization, and extraction of the fine structure, was performed with KEMP2 (Korbas et al., 2006;Wellenreuther and Meyer-Klaucke, 2009) assuming a threshold energy E 0,Fe = 7120 eV. Sample integrity during exposure to synchrotron radiation was checked by monitoring the position and shape of the absorption edge on sequential scans. No changes were detectable.

Extended X-ray analysis of fine structure
For extended X-ray analysis of fine structure (EXAFS), a forward Fourier transformation was performed for a k-range from 2.1 to 10.0 Å -1 using a Hanning window function. The final overall fit of the model (vide infra) uses only two-legged scattering paths of Fe-O and Fe-[O]-C. The fits were performed in R-space using a Hanning-type window with r from 1 to 3.4 Å. For the least-square fits in R-space, a k-weight of 2 was used. All corrections and fits were performed by the Demeter 0.9.13 program package of Ravel and Newville (2005).

Kinetics and thermodynamics of surface-bound iron
Previous studies indicated that iron is internalized by Ectocarpus from FeEDTA in a time-and concentrationdependent manner (Böttger et al., 2012) with the soluble Fe(III) species known as Fe(III)′ as the likely biological substrate as opposed to the intact Fe(III)-EDTA complex. Internal iron was determinable only after extensive washing of the cells with the Ti-citrate-EDTA reagent described by Hudson and Morel (1989), as unwashed cells were found to bind a large amount of iron but in a time-independent manner indicative of rapid surface binding (Fig. 1). The amount of surface binding as compared to internalization was at least a 100-fold higher. While this surface binding of iron was not unexpected, it was initially dismissed as nonspecific and an experimental nuisance. However, upon further examination, it was found that the high level of surface binding persisted even when the iron in the growth media was presented in the form of highly stable EDTA or other chelates, which was inconsistent with weak nonspecific binding. Indeed, surface binding was not eliminated until the EDTA/Fe ratio approached 100:1, indicating that the stability of the Fe-surface interaction had to be greater than that of FeEDTA itself (Fig. 2). Measuring the surface binding from a solution of fixed [ 55 Fe] (1 μM) as a function of [EDTA] and pH allowed an estimation of an effective surface-binding constant K eff ' (reaction 1: see Supplementary Tables S1-S8 for detailed analysis). The values obtained were relatively constant ( K eff ' = 3-29 × 10 21 M -1 ) over a wide range (33-1500 μM) of excess EDTA, confirming uniform binding and reliability of the data set. At high pH (8.6-8.8), iron binding was very strong (log K eff ' = 21-22) while at acidic pH (4.8) the binding constant was drastically reduced (log K eff ' = 14), presumably due to protonation of the alginate carboxylate groups thought to be the major ironbinding moieties.  While it was clear that iron bound to the surface of E. siliculosus with high thermodynamic stability, its kinetic stability was undetermined. 55 Fe pulse experiments were thus conducted to investigate the effect on uptake after initial binding at the cell surface. Pulse-labelling the cell surface with 55 Fe followed by replacement of the media with 'iron-free' media (oligotrophic, open ocean water with a sub-nM concentration of iron) showed that approximately 50% of the surfacebound iron quickly re-entered solution while intracellular iron continued to increase with time (Fig. 3). However, ambiguity remained as to whether the increasing intracellular iron was derived directly from the surface-bound iron or from surface-bound iron that re-entered the solution before being taken up. The latter possibility was tested by a pulse-chase experiment in which an initial 55 FeEDTA pulse was followed by a chase of 'cold' 56 FeEDTA. Continued 55 Fe uptake following the 56 Fe chase would suggest that the Fe(III) taken up (internalized) was bound very tightly to the surface and did not re-enter the solution (and hence was unaffected by isotopic dilution). This, however, was not the case, as can be seen in Figs. 4 and 5 for Ti-washed and unwashed samples, respectively. The complete cessation of 55 Fe uptake upon isotopic dilution implied that surface-bound iron is not internalized directly; rather it re-enters the bulk solution.

Localization of iron on the surface
Microscopic analyses were performed to investigate the spatial distribution and uniformity of iron bound to the cell surface. The iron-specific stain Perls was lacking in intensity for optical microscopic analysis so the enhancing agent DAB was used in conjunction with Perls. This proved effective for detecting extracellular as well as intracellular iron (Fig. 6). It was clear that the majority of the iron in samples not washed with the Ti reagent was present on the surface with internal iron only visible in washed samples. The preponderance of surface-bound iron in unwashed samples was corroborated by energy-dispersive X-ray analysis (Fig. 7). Figure 8 depicts a scanning electron micrograph of an E. siliculosus cell alongside the corresponding energy-dispersive X-ray analysis of iron superimposed on the same field.

Nature of the surface-bound iron
TMS and XAS were used to determine more precise details of the surface iron and its surrounding ligands. The TMS spectra of a sample of Ectocarpus grown for 48 h exposed to 30 μM 57 FeEDTA and either washed with Ti-citrate reagent or unwashed can be seen in Fig. 9B and A, respectively. Essentially no signal was detectable in the washed sample indicating that, as expected (Böttger et al., 2012), the amount of iron internalized after short-term incubation was below the TMS detection limit. Therefore the entire signal seen in the unwashed sample represented external or surface-bound iron. The data for the unwashed sample could be fitted (Fig. 9A) with two doublets Fig. 3. Distribution of 55 Fe after exposure of Ectocarpus to a 30 μM solution of FeEDTA for 6 h, followed by resuspension of cells in iron-free media for either 24 or 168 h. Cell-surface iron represents the difference between samples before (total Fe) and after washing (internalized Fe) with the Ti-citrate-EDTA reagent. Data represent mean of three determinations.  exhibiting the parameters shown in Table 1. The main component (c2a + c2b) had parameters which are typical for polymeric Fe(III) octahedrally coordinated to primarily oxygen ligands and fitted into the range observed for carboxylate Fe(III) model complexes (Dziobkowski et al., 1981). The second and very minor component (~10% of total area) exhibited parameters that belong to an as-yet unknown species. The parameters obtained for the major component were similar to but distinct from (Δδ = 0.05 mm s -1 and different relaxation behaviour, as will be discussed further) those derived earlier (Böttger et al., 2012) for the major component of the internalized iron seen upon long-term exposure. The most significant difference was in the blocking temperature (Mørup, 2011) obtained from the temperature dependence of the TMS spectrum (Fig. 10), which   was above 4.2 K for the surface-bound iron as compared to below 1.8 K for that internalized. In the case of mineral species, this behaviour would normally suggest either a more crystalline structure and/or larger particle sizes.
In addition to TMS, this study employed XAS to probe the chemical nature of the external iron. The EXAFS spectrum for an unwashed sample of Ectocarpus is shown in Fig. 11. Based on the above Mössbauer data, the presence of a superparamagnetic polymeric Fe(III)-oxo system was indicated. Within this constraint, various models were tested. Models for species were constructed and coordinates were obtained after molecular mechanics energy minimization using the program Chem 3D. These coordinates were then used as input to the FEFF program to obtain the scattering paths. The best fits came from a hydroxo-carboxylato model (see Supplementary Model). The final fit of the EXAFS spectrum (Fig. 11)

Discussion
Until recently, adsorption of iron to the cell surface of marine algae has been viewed as an experimental artefact that complicates uptake studies. While it is indeed necessary to remove this signal when quantifying uptake, the potential biological significance of this surface-bound iron has generally been ignored. Such was the case in the previous study of E. siliculosus (Böttger et al., 2012). However in the process of reexamining the data, the current study group was intrigued by the fact that a high level of surface binding was occurring despite the fact that all the iron present in the media was in the form of the very stable FeEDTA chelate. Indeed surface binding persisted even when an excess of EDTA was present. Thus it appeared that the binding of iron to the surface of E. siliculosus is far stronger than that expected for a weak nonspecific electrostatic interaction and thus might have some biological significance. This possibility was supported by the estimate of the effective surface-binding constant ( K eff ' ) for Fe(III)' of 3-29 × 10 21 M -1 , which was comparable to the affinity of EDTA' for Fe(III)' at pH 8.7 (Supplementary Table  S8). The magnitude of the drop in affinity with decreasing pH was also comparable to that observed for EDTA, which is a polycarboxylic acid and was thus consistent with the binding of iron to the cell surface of Ectocarpus primarily by the carboxylate groups found in the alginates, sulphated Fig. 9. TMS at 77 K of samples of Ectocarpus grown before (A) and after (B) washing with Ti-citrate-EDTA reagent: dots, experimental data; dark grey line, Fe-carboxylate species; black line, unknown species; light grey line, the fit to the data using the parameters given in Table 1. polysaccharides, and pectins, which are the major cell-wall components in this class of organism. Histological staining and energy-dispersive X-ray analysis of both environmental and laboratory-derived samples of E. siliculosus confirmed the spatial uniformity and primarily surface binding of iron, as would be expected of a constitutive scaffold such as the cell wall. While the numerous subgroups and variable side chains of these components presented a challenge for the identification of the specific iron-binding site(s), this study used a combination of TMS and EXAFS spectroscopic studies to enable the determination of iron oxidation state, geometry, and ligation in the first and second coordination spheres. The results are entirely consistent with octahedral hydroxo-carboxylato ligation of the iron. Finally, the question remains: what is the biological significance (if any) of this strong surface binding of iron? While open-ocean-dissolved iron concentrations are virtually always extremely low (sub-nM), concentrations of this element in coastal regimes are far more variable and range from 0.1 to 300 nM (Chase et al., 2002). Thus organisms in coastal ecosystems have to contend with an inherently different set of acquisition problems as compared with organisms in oligotrophic open ocean environments. In particular, iron concentrations are subject to considerable variation depending on a variety of climatic conditions. Thus coastal organisms can be subjected to either very low or very high concentrations of iron in their environment. Surface binding of iron could serve as a metal ion buffer to assure organisms like Ectocarpus of an adequate supply of iron for metabolic needs in the face of widely varying external concentrations   11. Spectrum of extended X-ray analysis of fine structure in R-space (top) and q-space (bottom). Black dots, experimental data; light grey line, the fit; black line, Hanning-type windows. Fig. 10. TMS at different temperatures of a sample of Ectocarpus grown before washing with Ti-citrate-EDTA reagent: black dots, 1.8 K; light grey dots, 4.2 K; dark grey dots, 77 K; solid lines, the fit to the data using the parameters given in Table 1. of the element. In times of plenty the equilibrium would strongly favour iron bound to the surface, while in times of deprivation the equilibrium would shift to dissociation of the surface-bound iron into solution. This study's overall proposed mechanism for iron uptake in Ectocarpus, based on the data presented here and in Böttger et al. (2012), is shown in Fig. 12.
Since the data most closely support a model where iron uptake is from the solution phase rather than directly from the cell surface, such equilibria would assure an adequate solution iron concentration for efficient uptake only if the solution-phase iron derived from the surface was not free to diffuse away. However, in non-uniseriate plants, the cell-wall space (apoplast) between cells is interconnected and serves as a continuous matrix through which surface-bound nutrients could experience highly restricted diffusion despite the fact that it does not serve an obvious transport function. Maintaining a diverse cast of apoplastic polysaccharides with anionic ligands of considerable strength could be a strategy employed by E. siliculosus and likely by many other marine macroalgae to buffer the widely fluctuating coastal iron concentrations. Such a mechanism could be biologically important and allow these organisms to be competitive particularly in high-iron environments. Possible ecological evidence for this mechanism comes from the dramatic iron-induced phase shifts seen around some of the reefs on the Line Islands where corals dominate under natural oligotrophic conditions while macro and turf algae dominate near the site of recent shipwrecks releasing large amounts of iron into the shallow-water marine environment (Kelly et al., 2012).

SUPPLEMENTARY MATERIAL
Supplementary data are available at JXB online.
Supplementary Material S1. Surface iron-binding constant and equilibrium studies. Supplementary