Plasmonic probing of the adhesion strength of single microbial cells

Significance Studying interfacial dynamics of a single biological entity (cell, virus, or organelle) is critical for understanding microbial biofilm formation processes, developing biosensors, and designing biomaterials. However, the ability to measure the binding force between a single biological entity and surface remains a great challenge due to the hydrated feature of biological entity. In this paper, we present an optical imaging method that is able to measure the adhesion strength of single microbial cells. Unlike the existing methods with limited throughput, this method determines the adhesion strength of multiple individual cells simultaneously. This approach has the potential to contribute to a better understanding of the adhesion process at the microscopic scale and probe the biointerfaces.


Supplementary Notes
Thermodynamic considerations on the binding constant under different adhesion conditions.
Correlation between bacterial binding constant and dissociation constant.

Supplementary Videos
Video S1. Plasmonic imaging of the vertical fluctuation of single bacterial cell. Video acquired by a CCD camera at frame rate of 106 frames per second.
Video S2. Typical bacterial fluctuation behaviors at the single-cell level. Video acquired by a CCD camera at frame rate of 106 frames per second.
Bacterial strain and its growth condition. Sphingomonas wittichii RW1 (CICC10426) was purchased from China Center of Industrial Culture Collection (Beijing, China). S. wittichii RW1 was grown on nutrient agar plate containing agar (1.5 g/L), peptone (5 g/L), beef extract (3 g/L) and NaCl (5 g/L).(1) Single colony on the nutrient agar plate was transferred to the nutrient broth media containing peptone (5 g/L), beef extract (3 g/L), NaCl (5 g/L) and the culture was incubated at 30 °C until early stationary phase.
Then the culture was further diluted in the nutrient broth media adding 200 μg/mL streptomycin (dilution ratio of cell culture was 1:100) and incubated at 30 °C for about 24 h to its early stationary phase. The bacteria were then harvested by centrifugation (6000 g, 5 min), washed twice in KCl solution of different concentrations (0.5 mM, 5 mM or 50 mM), respectively. We re-suspended bacterial pellets in the KCl solution for the adhesion experiment at the corresponding ionic strength. These ionic strengths were selected to minimize the potential influence of ionic strength on the bacteria activity.
Setup of the plasmonic imaging system. The plasmonic interferometric imaging system was based on a total internal reflection microscope (Ti-E, Nikon, Japan). (Malvern Instruments Co, UK). The Zeta potential of sensing chips were measured using a SurPASS3 streaming potential analyzer (Anton Paar, Austria). All the above measurements were repeated at least six times.

Calculation of bacterium-surface interactions.
Considering the amplitudes and nonlinearities of potential curves, we can rule out the contribution of electrostatic and gravity potential energy in the fluctuation potential analysis (Fig. S3).
According to the X-DLVO theory, the interactions between colloids (bacteria) and infinite surface (sensing chips) can be decomposed into components of Lifshitz-van der Waals interaction (ΦvdW), Lewis acid-base interaction (ΦAB) and electrostatic interaction (ΦEL).
Lifshitz-van der Waals interaction can be calculated from the results of superficial characterization according to where A is Hamaker constant, r is the radius of bacteria. λ0 is characteristic wavelength (100 nm). h is the separation distance between bacteria and sensing chips. Hamaker constant is derived from Lifshitz-van der Waals free energy (ΔGLW) and minimum equilibrium cut-off distance (d0), given by Similarly, the Lewis acid-base interaction (ΦAB) and electrostatic interaction (ΦEL) are derived from the Lewis acid-base free energy (ΔGAB) and electrokinetic property of bacteria and sensing chips according to (3,4) (6) in which λ is the correlation length in water (0.6 nm); the φ1 and φ2 are the zeta potential of bacteria and sensing chips; 1/κ represents the Debye Length. k, T and e refer to Boltzmann constant, absolute temperature and electron charge, respectively.
Considering the SAMs on the chips, the actual minimum separation distance is 2.1 nm (the length of C12 carbon chain). (5) All other parameters used in the X-DLVO potential calculation were listed in Table S1, and the obtained X-DLVO potentials at actual minimum separation distance (Φh0 ) and primary energy minimum (Φmin1) were listed in Table S2. where I0 is the initial SPR intensity before the bacteria desorption, I is the SPR intensity during bacteria desorption at time t, and kd is the obtained apparent dissociation constant.

conditions.
To give an insight of the single-cell binding constant from a thermodynamic view of point, we calculated the adhesion Gibbs free energy changes of bacteria based on X-DLVO theory. There was a notable difference among the X-DLVO potential profiles in different adhesion conditions ( Fig. S10a-b for details). The actual minimum separation distance (Φh0) is associated with interfacial energy barrier in X-DLVO potential. We therefore correlated the actual minimum separation distance with the binding constant (Fig. S10c). The binding constant decayed exponentially with the increase in minimum separation distance. When the interfacial Gibbs free energy enhanced, the corresponding binding probability of EPS would drop exponentially, which resulted in the similar decrease of K.

Correlation between bacterial binding constant and dissociation constant.
To validate the reliability of the binding constant for quantifying bacteria adhesion strength, we also tested the bacteria desorption behavior under lateral flow (shear stress: 1.41 Pa). The desorption kinetics under different conditions exhibits noticeable different features ( Fig. S7a and b). We fitted the SPR intensity attenuations derived from the bacteria desorption and obtained the apparent dissociation constants. The dissociation constants of -COOH surfaces were much larger than those of -NH2 surfaces and declined sharply as the ionic strength increased (Fig. S7c), indicating that the binding between bacteria and surface was weak. The correlation between the dissociation constant and binding constant ( Fig. S7d; R = -0.872) confirmed that our method is reliable for adhesion strength quantification.