Human host-associated microbial communities in body sites such as gut, skin, mouth, and urogenital track largely determine the disease states of patients (Clemente, Ursell, Parfrey, & Knight, 2012). The presence of specific bacterial genera provides not only a reliable biomarker but also pathological information for clinical diagnosis (Chassaing, Aitken, Gewirtz, & Vijay-Kumar, 2012). For examples, primary skin infections, such as impetigo (Bowen, Tong, Chatfield, & Carapetis, 2014), folliculitis (Shu Wei et al., 2019), and boils (O'Gara, 2017), are caused by a type of β-hemolytic and coryneform streptococci, known as Staphylococcus aureus, under conditions of eczema or insect bites (Sarkiri, Fox, Fratila-Apachitei, & Zadpoor, 2019). Another example is skin rash, which is commonly examined via standard skin test, e.g. scarping (Hahler, 2006), biopsy (Tarantola et al., 2013), allergy test (Balato, Balato, Di Costanzo, & Ayala, 2011) and blood test (Caubet et al., 2011) to determine the skin bacteria compositions in the heterogeneous bacteria community and thus identify a narrow-spectrum of antibiotics for the effective treatment. Therefore, identification of the responsible organisms in bacterial infection is the key to achieve the corresponding efficient treatment (Mooney, Galloway, & Riley, 2019), yet identifying the specific species is technically challenging in terms of manipulation of single bacteria cell, particularly in infections caused by more persistent bacteria and sophisticated bacterial communities (Russo et al., 2016).
Conventional single bacterial cell isolation techniques include dilution-to-extinction (Giovannoni & Stingl, 2007), single-cell micromanipulation (Sato et al., 2009), and flow cytometry (Wang, Hammes, Boon, Chami, & Egli, 2009). However, most of the existing methods offer a limited isolation yield or require at least a few microliters of the biosample volume, which could be too much to be extracted from the infection sites (Ishii, Tago, & Senoo, 2010). For example, dilution-to-extinction is hard to directly isolate rare cells and single-cell micromanipulation offers a very limited throughput. Though flow cytometry on the basis of fluorescence-activated cell sorting (FACS) can achieve a high-throughput single-cell isolation (Kim et al., 2013), it still requires a significant amount of the biosample volume (Burmølle, Hansen, Oregaard, & Sørensen, 2003) and indispensable procedures for cell staining which can alter cell properties and limit the more detailed live-cell analyses(Terekhov et al., 2018).
Microfluidics, owing to its capability of precise flow control and micro-particle manipulation, has garnered significant interests on cell isolation of the single cells. For examples, a deterministic ratchet for selective separation of particles with different sizes around 1 – 3 µm was reported by Loutherback et al. (Loutherback, Puchalla, Austin, & Sturm, 2009), yet its application for bacteria separation has not be demonstrated. By contrast, Chun et al. (C. H. Chen et al., 2011) reported a microfluidics-based FACS for isolating fluorescence-labeled Escherichia coli from a microbial sample. In this study single cells can be extracted by a piezoelectric actuator, however it can not separate the droplets for individual incubation and any downstream analysis afterward. In addition, microemulsion, which forms stable aqueous microdroplets in an immiscible buffer fluid, usually oil, has been demonstrated as a powerful tool for bacteria isolation and further analyses including incubation, DNA analysis, and drug screening (D. Chen et al., 2008; Dendukuri, Tsoi, Hatton, & Doyle, 2005; Edgar et al., 2009; Frenz et al., 2008; Koster et al., 2008; Shah, Kim, Agresti, Weitz, & Chu, 2008). For example, Eun et al. (Eun, Utada, Copeland, Takeuchi, & Weibel, 2011) presented the encapsulation of Escherichia coli in agarose particles and subsequently analysis by FACS, but the cell extraction for downstream inoculation and analysis was then technically difficult. In addition, Boedicker et al. (Boedicker, Li, Kline, & Ismagilov, 2008) presented a microemulsion system for rapid detection and drug susceptibility screening of bacteria in complex biological matrics. However, the isolated droplet cannot be extracted individually for single bacteria analysis, such as polymerase chain reaction (PCR) amplification and DNA sequencing. Moreover, Liu et al. (Liu, Kim, Lucchetta, Du, & Ismagilov, 2009) developed a microfluidic emulsion-based device for incubation of single bacteria isolated from a mixture of P. curdlanolyticus and E. coli. However, the operation configuration has not been optimized for the higher yield of single cell encapsulation and a microdroplet elimination scheme for the void and multi-bacteria droplets is required. Therefore, an integrated microfluidic system that achieves a high throughput yet efficient capture of bacteria at a single-cell level from a limited input of biosample and is highly compatible with various downstream analyses via controllable release is urgently needed yet largely unavailable.
Recently, we have developed a microfluidic strategy for sequential isolation of mammalian cells (10 – 20 µm) in aqueous biosamples, e.g. blood, in microfluidic based platform (Li et al., 2020; Tran, Kong, Hu, Marcos, & Lam, 2016) with a yield of >95%. Cells are sequentially captured in the micro-sieve structures, whose positions have been arranged carefully according to the deterministic lateral shift (Huang, Cox, Austin, & Sturm, 2004) for the maximum capture rate. Notably, on top of the additional functional capability offered by microemulsion, the cell-encapsulating microdroplets enlarge the physical scale (from sub-micron for bacteria to tens of micron for microdroplets) of the ‘particles’ to be manipulated. Hence, we hypothesized that integration of a microemulsion unit for encapsulation of bacteria in a droplet of which the working scale is similar to our previous device will extend its application for single-bacteria manipulation and subsequent genetic and functional analyses.
In this work, we developed a micro-device integrated with microemulsion and sequential micro-sieves for single-bacteria isolation and selective cell extraction. We optimized the system dimensions and operation parameters by characterizing the microdroplet generation, the sequential microdroplet trapping, the single-bacteria encapsulation in droplets and the microdroplet extraction. As a proof-of-concept, we demonstrate applicability of the micro-device with prepared skin bacteria and human skin biosample and achieved single-bacteria extraction from mulit-species bacterial communities and the following basic genetic analysis. We believe our integrated microfluidic system will demonstrate a useful and powerful tool for clinical diagnosis (Lei, 2012) of microbial infection and as well for screening potential drugs or antibiotics.