Background

Recently, collective motions of bacteria have attracted a lot of attention in the field of active matter/fluid (Rabani et al., 2013 ; Zhang, H. et al., 2010; Marchetti et al., 2013). Due to the self-driven motion of individual bacterium propelled by flagella, high density bacteria suspensions would fall into a non-equilibrium state and emerge collective structures such as vortices, jets ( Mendelson et al., 1999 ) or turbulences (Wensink et al., 2012 ) in a scale far larger than the size of an individual bacterium. As one of the most widely used model system for the study of bacteria collection motions (Copeland and Weibel, 2009), bacteria swarming refers to coordinated migrations of cells across wet solid agar surfaces with swirling patterns (Kearns, 2010). During swarming, motile cells are immersed in a thin layer of highly viscous fluid called “swarm fluid” (Wu and Berg, 2012). Physically, the interplay between the fluid medium and the stirring flagella provides the dynamic origins of the collective motion though far-field hydrodynamic interactions (Koch and Subramanian, 2011). Biologically, the fluid environment enhances the mixing and diffusion of oxygen, nutrition as well as the signaling molecules involved in Quorum sensing (Hardman et al., 1998 ) and chemotaxis (Taktikos et al., 2012 ). Although the swarm fluid is important for the generation of large-scale synergetic patterns, how it works and its relationships with fast moving cells remain unclear especially at the microscopic scale.

Single particle tracking has been regarded as a powerful technique to investigate the complex fluid (Burov et al., 2011 ). Previous studies have used micro-sized particles as single particle tracers to characterize the properties of swarm fluid. Wu et al. used 1~2 µm micro-bubbles produced by explosive transformation of the water insoluble surfactant Span-83 droplets as tracers to reveal the intensive matter transfer flow in the leading edge of the swarming colony (Wu et al., 2011). Zhang et al. fabricated MgO particles as tracer particles through burning magnesium ribbons. They found that 0.2 µm MgO particles only diffused normally within a small region (~4 µm²) as the swarm front approaches (Zhang, R. et al., 2010). However, these methods of producing tracer particles are complicated and difficult to control. In addition, the prepared gold nanorods are often uneven in size. On the other hand, micro-sized tracers with various sizes and materials would either inevitably collide with bacteria bodies or be trapped at the liquid-air surface, making the tracers’ motions incapable of revealing real dynamics of the fluid motions. Plasmonic gold nanorods (AuNRs) have long been used as nanotracers for long-duration observations in biological studies with high spatio-temporal resolution. The AuNRs have good photostability and low cytotoxicity. Moreover, the methods of preparing stable, uniform, and monodispersed AuNRs are very mature. Here, with Bacillus subtilis as a model strain, we introduced 40 x 84 nm AuNRs as tracers into the bacterial swarm fluid. The AuNRs tracers could move freely in the upper layer without obvious physical contacts with the cell bodies. Combined with high spatiotemporal resolution of imaging and tracking technique, the trajectories of multi-nanoparticles in the large field of view could be obtained. By detailed diffusion analysis of tracers’ trajectories, we found that the swarm fluid could transport the AuNRs to long distances in a super diffusive, Lévy mode (Figure 1). This result is consistent with previous studies that individual bacterium in swarming would exhibit characteristic motions of Lévy walk ( Ariel et al., 2015 ).

Figure 1. The gold nanorods move in high-efficiency on the upper fluid layer of the swarming bacteria. A. A captured image of the swarming bacteria and AuNRs. The brown rods are bacteria. The red spots which are indicated by white arrows are AuNRs. The scale bar is 10 μm. B. A typical trajectory illustrates that AuNRs move in a super-diffusive, Lévy mode with heterogeneous MSD local exponents.