Bacterial abundance in natural samples was determined by counting cells stained with DAPI (2 μg mL−1 in a 4 : 1 mixture of Citifluor-Vectashield for 10 min) by microscopy using image analysis. The abundance of A. macleodii was determined by flow cytometry (FACS Calibur; Marie et al., 1997).

The application of microautoradiography to investigate the activity of heterotrophic bacteria at a single-cell level requires that the cells associated with silver grains are representative of cells that actively incorporate the radioisotope used. In the case of iron, the complete elimination of extracellular iron is challenging, especially in aquatic environments, where iron is present as FeOx, which are easily adsorbed at the surface of biogenic or lithogenic particles. An efficient washing step of bacterial cells is therefore necessary to remove extracellular iron before exposure of the cells to the photographic Pirfenidone concentration emulsion. Several washing methods were proposed previously, and two of them were used for seawater samples. The Ti-citrate-EDTA method (Hudson & Morel, 1989) is based on the reduction of Fe (III) with TiCl3. In the case of oxalate-EDTA (Tovar-Sanchez et al., 2003), the dissolution is very likely due to a ligand-promoted process (Tang & Morel, 2006). In the present study, we tested the efficiency of Ti-citrate-EDTA, of oxalate-EDTA and of 0.2-μm-filtered seawater

(Fig. 1, steps a + d and c). An almost complete removal of extracellular Doxorubicin research buy Bay 11-7085 iron was only obtained with the Ti-citrate-EDTA reagent (96 ± 0.3%, n = 3). Oxalate-EDTA and 0.2-μm-filtered seawater were less efficient with 88 ± 1% (n = 3) and 76 ± 0.2% (n = 3) of 55Fe removed, respectively (data not shown). These results are consistent

with Hutchins et al. (1999) who report the removal of up to 97% of surface-adsorbed 55Fe using the Ti-citrate-EDTA solution for phytoplankton cells. Tang & Morel (2006) also concluded that the oxalate-EDTA wash is not as efficient as the Ti-citrate-EDTA wash in dissolving extracellular FeOx in phytoplankton cultures. When a washing step is applied, it is also important to assure that no intracellular iron release occurs due to the damage of the cell membrane. Contrasting results are reported for phytoplankton cultures. Tang & Morel (2006) did not observe any membrane damage when using Ti-citrate-EDTA and oxalate-EDTA. By contrast, other studies report that Ti-citrate-EDTA could produce leakage of the intracellular content (Sunda & Huntsman, 1995; Tovar-Sanchez et al., 2003). To our knowledge, only one study tested whether the wash protocol with Ti-citrate-EDTA alters the integrity of bacterial cell membranes, which could result in the release of intracellular Fe (Chase & Price, 1997). These authors tested the release of radioactivity after washing with Ti-citrate-EDTA using A. macleodii (jul88 strain) incubated concurrently with 14C-glucose and 55Fe.