POM 1 The commonly used eDNA stains we evaluated
The commonly used eDNA stains we evaluated were the intercalating dye propidium iodide, the highly sensitive stain PicoGreen commonly used in DNA quantification, the cyanine monomer SYTOX® Green, and the dye 7-hydroxy-9H-(1,3-dichloro-9,9-dimethylacridin-2-one) (DDAO) not only sold as a standard for ether-, ester-, and phosphate substrates, but also commonly used as an eDNA stain in biofilms since 2006 (). Additionally, we evaluated the lesser used cyanine monomer TO-PRO-3 and cyanine dimer TOTO-1, all at a final working concentration of 2μM. As counterstains we evaluated the DNA-binding cyanine dyes SYTO 9, SYTO 12, and SYTO 60, as well as the non-fluorescent cell-permanent acetomethoxy derivates of calcein (CAM-red and CAM-green) which become cleaved by esterases upon entering living cells, resulting in the entrapment of fluorescent and membrane-impermeant calcein within the cell. All counterstains were used at a working concentration of 1μM with the exception of SYTO 60, which was determined to function optimally at a concentration of 10μM. All stains were obtained from Life technologies™. Biofilms were all grown for 24h in Lysogeny broth at 37°C, with the exception of which was incubated at 30°C, and which was grown at 37°C in tryptic soy broth. Biofilms were grown in IBIDI μ-plates for optical microscopy (Cat no. 89621), and visualized in PBS using a Zeiss 700 confocal laser scanning microscope. Of the eDNA stains, TOTO-1, PicoGreen and SYTOX® Green allowed excellent visualization of eDNA biofilm structures (a, b and c). However, PicoGreen penetrates the cell membrane over time, eventually producing an erroneous green signal from all POM 1 (data not shown). The excellent visualization by these stains was not surprising, as they are among the most sensitive DNA-binding stains available. The DNA binding affinity of the cyanine dimers is several orders of magnitude higher than their parent monomer dyes, and they furthermore have low intrinsic fluorescence and high quantum yields. The DNA–TOTO-1 complex in particular is very stable (), and it has a high DNA binding affinity due to its four positive charges. Because TOTO-1 has no time constraints and displays excellent sensitivity, it is the most appropriate choice for visualization of eDNA. Propidium iodide, perhaps commonly used as an eDNA stain due to its role in the BacLight™ bacterial viability kits, was indeed an excellent stain for dead cells and dense eDNA structures (d), but was unable to visualize delicate eDNA structures such as those present in biofilms (). Therefore, we only recommend the use of propidium iodide when a fluorophore in the red spectrum is required, such as in conjunction with GFP-tagged cells. DDAO and TO-PRO-3 were not only unable to visualize eDNA structures such as those present in biofilms (e and f), but they also penetrated all cells, thereby producing erroneous results as to the location and abundance of eDNA. This is especially worrying due to the increasing popularity of DDAO as an eDNA stain. We urge caution if these stains are used for visualization of eDNA in biofilms. It has been suggested that combining several DNA stains can result in FRET effects (), where emission from one fluorophore excites another neighboring fluorophore when bound within a distance of 10nm (or 30 base pairs) on the same DNA molecule. It is therefore desirable for an eDNA counterstain to not bind DNA. To this end we evaluated the counterstains CAM-green and CAM-red, which do not bind DNA. The fluorescent spectra of these stains, however, produce cross-talk with eDNA stains TOTO-1 and propidium iodide, as well as fading appreciably within hours. Therefore, we recommend SYTO 60 and SYTO 9 as counterstains for TOTO-1 and propidium iodide, respectively. We have determined an optimal staining procedure for the visualization of eDNA in biofilms. Here we show that TOTO-1 for eDNA, combined with SYTO 60 as counterstain, can be used to produce high quality images of eDNA in a variety of Gram-negative and Gram-positive bacterial species (). Because this method offers significant improvement to the quality of eDNA images currently found in the scientific literature, we propose that this technique become the new standard for the visualization of eDNA in bacterial biofilms.