Panel A. Biofilm-inhibiting factors in conditioned medium. B. subtilis strain NCBI3610 was grown at 22 °C in 12-well plates in liquid biofilm-inducing medium for 3 or 8 days. Conditioned medium (500 ml) from an 8-day-old culture was concentrated on the C-18 column and eluted step-wise with methanol. Shown is the result of growing cells in fresh medium to which had been added 20 µl of the 25%, 35% or 40% methanol eluates.
Panel B. Norspermidine inhibits biofilm formation. Cells of NCBI3610 were grown in fresh medium containing PBS buffer (control), norspermidine (100 µM), morpholine (100 µM) HPLC-purified fatty acid (~100 µM), or spermidine (100 µM). Brighter images of the norspermidine-treated cell revealed cells near the bottom of the well.
Panel C. Detection of norspermidine. Pellicles were collected from 3- and 8-day-old cultures (100 ml) of the wild type (NCBI3610) and from an 8-day-old culture (100 ml) of a gabT mutant (IKG623). After mild sonication of the pellicles, cells were separated from extracellular material. (Other experiments showed that norspermidine is largely found in pellicles.) Norspermidine in the extracellular material was derivatized with Fmoc-Cl and the resulting Fmoc-norspermidine was detected using an Agilent LC/MS system. Fmoc-norspermidine was detectable in the old pellicle from wild type cells but not in the young or mutant pellicles. See also Figure S1, Panel B.
Panels D and E. Quanitification of the biofilm-inhibiting activity of norspermidine and spermidine). Pellicle formation of strain NCBI3610 was tested in the presence of the indicated concentrations of norspermidine (D) or spermidine (E). See also Figure S1, Panel A.

Panel A. Pellicle longevity. Shown are 7 day-old cultures of the wild type (WT), a mutant (ΔgbaT) blocked in norspermidine production (IKG623), a double mutant (ΔylmE ΔracX) blocked in D-amino acid production (IKG55) and a triple mutant (ΔgbaT ΔylmE ΔracX) blocked in the production of both (IKG625). See also Figure S2, Panel A.
Panel B. Preventing biofilm formation. Cells were grown for 3 days in medium containing as indicated D-tyrosine (D-Tyr), norspermidine, a mixture of D-tyrosine, D-methionine, D-leucine and D-tryptophan (D-aa) and the indicated combinations of amino acids and norspermidine at the indicated concentrations.
Panel C. Qunatifying pellicle breakdown. On the surface of 3 day-old pellicles were placed droplets (50 µl) containing buffer (PBS), a mixture of D-tyrosine, D-methionine, D-leucine and D-tryptophan each at final concentration of 12.5 µM, norspermidine at a final concentration of 50 µM, or, as in panel C, a combination of D-amino acids each at a concentration of only 2.5 µM and norspermidine at a concentration of 10 µM. After incubation for the indicated times, pellicle material and the medium were separated and each brought to a volume of 3 ml. After mild sonication, the OD₆₀₀ was determined for each sample. The % of disassembly represents the OD₆₀₀ of the medium as a percent of the sum of the OD₆₀₀ of the medium and the OD₆₀₀ of the pellicle.

Shown are phase contrast and fluorescence images of cells of the wild-type (WT; NCBI3610) harvested from pellicles grown in the presence or absence (untreated) of norspermidine (25 µM) or a high concentration of spermidine (1 mM). The cells were washed in PBS and stained for exopolysaccharide with a conjugate of concanavalin A with Texas-Red. See also Figures S2, Panel B, and Figure S3

Panel A. Dynamic light scattering. Listed are the average hydrodynamic radii of the exopolysaccharide as measured by dynamic light scattering. Exopolysaccharide was purified from pellicles. Light scattering was measured for exopolysaccharide alone as well as for exopolysaccharide that had been mixed with 0.75 mM norspermidine or with 0.75 mM spermidine. Shown are the results obtained in the absence of polyamine (black), in the presence of norspermidine (white), and in the presence of spermidine (grey) with exopolysaccharide at the indicated concentrations and pH. Error bars represent the standard deviation of polymer radii among the polymers in a single sample.
Panel B. Scanning Electron Microscopy Purified exopolysaccharide was dissolved in double distilled water at a final concentration of 10 mg/ml and mixed with either norspermidine or spermidine (0.75 mM final concentration). Samples were prepared as described in Experimental Procedures. Shown are three different magnifications of representative fields showing exopolysaccharide alone (EPS) and exopolysaccharide that had been mixed with norspermidine (EPS + norspermidine) or with spermidine (EPS + spermidine). See Figure S4 for controls showing little effect on growth or eps transcription.

Panel A shows compounds tested for biofilm-inhibiting activity.
Panel B shows the effect of the numbered compounds on pellicle formation by B. subtilis (NCBI3610). The compounds were tested at 200 µM.
Panels C and D show the results of modeling the interaction of norspermidine and spermidine with an acidic exopolysaccharide. Norspermidine binds via salt bridges between amino and carboxyl groups (dotted lines) in a clamp-like mode across the exopolysaccharide secondary structure of a disaccharide repeat. [α(1,6)Glc-β(1,3)GlcA]_n. Whereas norspermidine aligns well with the repeat, the spacing of amino groups of spermidine does not match the symmetric pattern of anionic side groups, implying weaker affinity. See supporting information for modelling of plausible interactions with other charged and non-charged polysaccharide structures. For tests of additional molecules see Figure S5 and Table S2.

Panel A shows the effect of the numbered compounds displayed in Figure 5A on the formation of submerged biofilms by S. aureus strain SCO1. The compounds were tested at 500 µM. Biofilm formation was visualized by crystal violet staining of submerged biofilms. Panel B shows quantification of the effects of norspermidine, norspermine, spermine and spermidine as measured by crystal violet staining (see Experimental procedures). See also Figure S6.

Panel A shows the effect of the numbered compounds shown in Figure 5A on submerged biofilm formation by E. coli strain MC4100. The compounds were tested at 500 µM. Biofilm formation was visualized by crystal violet staining of submerged biofilms. Panel B shows quantification of the effects of norspermidine, norspermine, spermine and spermidine as measured by crystal violet staining (see Experimental procedures). See also Figure S6.