http://informahealthcare.com/bty ISSN: 0738-8551 (print), 1549-7801 (electronic) Crit Rev Biotechnol, Early Online: 1–13 ! 2014 Informa Healthcare USA, Inc. DOI: 10.3109/07388551.2014.899556

REVIEW ARTICLE

Measuring gene expression in single bacterial cells: recent advances in methods and micro-devices Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

Xu Shi1, Weimin Gao1, Jiangxin Wang1, Shih-Hui Chao1*, Weiwen Zhang2*, and Deirdre R. Meldrum1 1

Center for Biosignatures Discovery Automation, Biodesign Institute, Arizona State University, Tempe, AZ, USA and 2School of Chemical Engineering and Technology, Tianjin University, Tianjin, P.R. China Abstract

Keywords

Populations of bacterial cells that grow under the same conditions and/or environments are often considered to be uniform and thus can be described by ensemble average values of their physiologic, phenotypic, genotypic or other parameters. However, recent evidence suggests that cell-to-cell differences at the gene expression level could be an order of magnitude greater than previously thought even for isogenic bacterial populations. Such gene expression or transcriptional-level heterogeneity determines not only the fate of individual bacterial cells in a population but could also affect the ultimate fate of the population itself. Although techniques for single-cell gene expression measurement in eukaryotic cells have been successfully implemented for a decade or so, they have only recently become available for single bacterial cells. This is due to the difficulty of efficient lysis of most bacterial cells, as well as short half-life time (low stability) of bacterial mRNA. In this article, we review the recent progress and challenges associated with analyzing gene expression levels in single bacterial cells using various semi-quantitative and quantitative methods. In addition, a review of the recent progress in applying microfluidic devices to isolate single bacterial cells for gene expression analysis is also included.

Bacteria, gene expression, microfluidics, single cells, transcriptome

Introduction Microbiologists typically assume that bacterial cells growing under the same conditions are uniform populations (BrehmStecher & Johnson, 2004). Based on this assumption, microbiologists in the past measured bacterial populations and used ensemble average values to describe behavioral patterns of bacterial cells. While this approach has contributed significantly to our current understanding of microorganisms, it has been suggested that in a bacterial community, individual cells can exhibit significant cellto-cell differences that are an order of magnitude greater than previously thought (McAdams & Arkin, 1997; Spudich & Koshland, 1976). It is well documented that in bacterial populations, cells with distinct metabolic profiles, stress responses, and/or other specific biological activities are juxtaposed within a single populations (Macfarlane & Dillon, 2007). Therefore, profiling of whole population generates only average values that fail to represent the

Address for correspondence: Dr. Shih-Hui Chao, Center for Biosignatures Discovery Automation, Biodesign Institute, Arizona State University, PO. Box, 876501, Tempe, AZ 85287-6501, USA. E-mail: [email protected] Dr. Weiwen Zhang, School of Chemical Engineering and Technology, Tianjin University, Tianjin, 300072, P.R. China. E-mail: [email protected]

History Received 12 September 2012 Revised 13 December 2013 Accepted 13 January 2014 Published online 2 April 2014

sub-populations that have distinct functions in the community (An & Parsek, 2007). Recently, a surge of evidences found that isogenic populations of microorganisms show substantial cell-to-cell heterogeneity at both cellular and molecular levels (Becskei et al., 2005; Colman-Lerner et al., 2005; Golding et al., 2005; Kelly & Rahn, 1932; Kuang et al., 2004; Le et al., 2005; Lidstrom & Meldrum, 2003; Maloney & Rotman, 1973; Pedraza & van Oudenaarden, 2005; Rosenfeld et al., 2005; Siegele & Hu, 1997; Strovas & Lidstrom, 2009; Strovas et al., 2007). The mechanisms that contribute to this gene expression heterogeneity include intrinsic heterogeneity that exists in biology, noise in the transcription machinery, micro-scale chemical gradients, and genotypic variations because of mutations and selection in individual cells (Stewart & Franklin, 2008). The amplitude of such stochasticity, or noise, in gene expression is controlled by many factors, including transcription rate, regulatory dynamics, and other genetic factors of the cells (e.g. microRNA (miRNA), transposon) (Banerjee et al., 2004; Colman-Lerner et al., 2005; Newman et al., 2006; Pedraza & van Oudenaarden, 2005; Rosenfeld et al., 2005; Strovas et al., 2007). This transcriptional noise, once amplified, could offer the opportunity to generate and sustain heterogeneity at the cellular level in a clonal bacterial population. To gain a deeper insight into the intricacies of cellular diversity and its functional relevance, methodologies capable of measuring gene

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

2

X. Shi et al.

expression levels in single bacterial cells need to be developed. Another reason to pursue single bacterial cell analysis stems from the fact that the majority (499%) of bacterial species obtained from environmental samples cannot be cultured under laboratory conditions (Rajilic´-Stojanovic´ et al., 2007). Therefore, they are not accessible to conventional gene expression analysis methods that typically require large numbers (105 to 106) of cells. To circumvent the need for cultivation, the development and establishment of single-cell based gene expression profiling is important for our understanding of microbial diversity in various natural environments. With rapid development of microfluidics devices, new opportunities for the analysis of cell-to-cell heterogeneity (Wang & Bodovitz, 2010) have been opened up. Microfluidics, usually used interchangeably with the term ‘‘lab-on-a-chip’’, is a technology that is based on the manipulation of small amounts of liquid (typically in the range of nanoliters to picoliters) in confined volumes. This technology is especially advantageous to the manipulation and analysis at the single-cell level for the following reasons: (i) Individual cells can be precisely trapped, moved, and distributed individually in microscale channels; (ii) Isolated individual cells can be easily monitored in microchambers and dilution of the cellular contents is minimized thus increasing the sensitivity of a downstream analysis; and (iii) Highly parallel, fully automated multi-step operations can be implemented for high-throughput analyses resulting in significant time and cost savings. The application of microfluidic devices to study single eukaryotic cells has been a thriving field (Banaeiyan et al., 2013; Benı´tez et al., 2012; Beta & Bodenschatz, 2011; Clausell-Tormos et al., 2008; Meldrum & Holl, 2002; Molter et al., 2009, 2008; Roman et al., 2007; White et al., 2011; Zhu et al., 2012). While applying these technologies to microbiology research is at its early age, rapid developments have been witnessed in recent years. Thereby, a systematic review of current available gene expression technologies and tools is necessary. The significance of single-cell microbiology and some relevant technologies has been discussed in detail in several previous reviews (Brehm-Stecher & Johnson, 2004; Broude, 2011), so this review will focus primarily on the progress over the past several years in the development of miniature devices for individual bacterial cells and corresponding gene expression measurement methods. The review is organized following the sequential order of a typical single-cell analysis, i.e. single cell manipulation (isolation/capture) followed by single cell gene expression analysis.

Manipulation of single bacterial cells Although manipulating single eukaryotic cells has become common practice, manipulating single bacterial cells is significantly more challenging due to the following reasons: (i) the total volume of a typical bacterial cell is 2–3 orders of magnitude smaller than that of a typical eukaryotic cell (1–10 fL versus 1 pL); (ii) in addition to small dimensions, many bacteria are highly mobile (Bardy et al., 2003; Pijper & Discombe, 1946) making them even more difficult to

Crit Rev Biotechnol, Early Online: 1–13

manipulate. Several methods used to manipulate single bacterial cells are summarized below. Dilution-to-extinction The conventional dilution-to-extinction method utilizes serial dilution to isolate single bacterial cells into test tubes or wells on microtiter plates (Button et al., 1993; Rappe´ et al., 2002; Schut et al., 1993; Sizova et al., 2012). The same principle can be applied to microfluidic devices to load single cells into microscale reaction chambers (Boedicker et al., 2008). Cell occupancy of the wells follows the Poisson distribution and can be manipulated by controlling cell concentration in the media prior to loading into the microchambers. As the microchamber volumes are many orders of magnitude smaller than those of conventional analysis vials, the sample dilution is accordingly minimized. For example, if single cells were loaded in an array of 1-picoliter chambers, the resulting concentration in the microwells would be on the order of 1 cell/pL or 109 cell/mL, which is within the range of typical concentrations for bulk bacterial cultures (Sezonov et al., 2007). Two microfluidic approaches have utilized the dilution-toextinction approach to isolate single bacterial cells by seeding cells in microfabricated chambers or encapsulating cells in emulsion. An example for the former approach was a device developed by Ottesen et al. (2006) who isolated bacterial cells randomly from a complex environmental sample and then performed digital polymerase chain reaction (PCR) in order to identify new species. They randomly seeded cells from a diluted environmental sample on their device and obtained single-cell occupancy in 28% of the reaction chambers, while the rest of the chambers contained either multiple cells (6%) or were empty. Another approach is based on the encapsulation of individual bacterial cells in aqueous droplets (Eun et al., 2011; Guo et al., 2012; Lian et al., 2012; Shim et al., 2009). Eun et al. (2011) used a microfluidic flowfocusing nozzle to generate Escherichia coli-containing agarose microdroplets. After the agarose microdroplets had solidified E. coli cells were encapsulated in agarose microparticles for downstream incubation and analysis. Zeng et al. (2010) randomly seeded E. coli cells into droplets containing primer-adhered microspheres and real-time PCR reagents. Um et al. (2012) encapsulated single E. coli cells in a pico-liter volume droplet. They employed a secondary breakup of droplets, thereby reducing stochastic loading of the cells and increasing the single-cell occupancy from 30% to around 50%. Lin et al. (2009) introduced a new method by generating stationary droplets as reaction chambers. They loaded a diluted suspension of E. coli onto an array of oilcovered surface-adhering droplets that were spatially confined by oil through hydrophilic/hydrophobic patterns on the substrate. The number of randomly seeded E. coli cells in droplets followed the Poisson distribution. Gao et al. (2013) used this method to cultivate pure bacterial isolates from single cells in the environmental bacterial communities. This dilution-to-extinction method does not require special single-cell manipulation technologies or devices. As long as the cell concentration in the bulk medium is properly diluted, fast and relatively easy seeding of single cells in

DOI: 10.3109/07388551.2014.899556

microfabricated chambers or droplets can be achieved in a high throughput. The major drawbacks of this method are the random nature of the cell occupancy and the low efficiency of obtaining single cell occupancy while reducing the number of wells containing multiple and zero cells. Large numbers of empty compartments result in a waste of chemical reagents, reduced overall throughput, and the need to determine the number of cells in each well to discern reaction chambers containing single cell, multiple cells or empty.

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

Cell trapping In contrast to the dilution-to-extinction method, single cell trapping is a deterministic method to isolate single cells. Multiple traps can be implemented in a device, facilitating parallel measurements at a given time. Three types of singlecell trapping methods have been used in the past to isolate single bacterial cells from populations: Mechanical, hydrodynamic, and dielectrophoretic. In what follows, we discuss the details of these trapping methods. Mechanical trap Mechanical cell trapping is achieved by physical obstacles, barriers or side channels/chambers to hold or catch individual cells flowing through microfluidic channels (Furutani et al., 2010). Microscale U-shaped barriers (Di Carlo et al., 2006) have been utilized to trap mammalian cells. However, these barriers are inefficient for trapping single bacterial cells due to their small dimension. Gru¨nberger et al. (2012) applied a channel of 2 mm width and 1 mm height as a seeding inlet and successfully isolated single E. coli cells in the culturing chamber. More complicated microfluidic devices were also being used to trap single cells. For instance, Huang et al. (2007) used pneumatic valves to trap Synechococcus PCC 7942 cells and observed significant cell-to-cell heterogeneity in populations under nitrogen-depleted growth condition. Leung et al. (2012) designed an elaborate microfluidic device which could trap single bacterial cells, Salmonella typhimurium E. coli and different environmental samples, by using a peristaltic pump and combined with versatile downstream studies, such as culture and PCR analysis. Hydrodynamic trapping Hydrodynamic trapping is a non-contact cell trapping method. It relies on flow stagnation or microeddies (Lutz et al., 2006a; Tanyeri & Schroeder, 2013). Cells are trapped at steady streaming eddies generated by hydrodynamic forces (Lutz et al., 2006b). Compared to eukaryotic cells, applying hydrodynamic traps to small bacterial cells poses significant challenges since hydrodynamic forces are typically proportional to the surface areas of cells. However, (Tanyeri et al. 2010; Tanyeri & Schroeder, 2013) have demonstrated hydrodynamic trapping of as small as 100 nm particles which is even smaller than normal bacterial cells. They used hydrodynamic traps to achieve high accuracy of trapping and manipulation of single bacterial cells in a microfluid device. To trap single bacterial cells E. coli, their device produced a flow stagnation point in the center of two perpendicularly crossed flow channels. The drawback of this method is that

Gene expression in single bacterial cells

3

the instruments to achieve precise control over the flow to create flow stagnation and eddies are not easily accessible to conventional microbiology laboratory. The use of feedbackbased flow control may alleviate this problem although at the cost of increased complexity of the system. Dielectrophoretic (DEP) trap A polarizable particle, like cells, subjected to an unevenly distributed electric field generated by microfabricated electrodes will experience the DEP force. Applying DEP forces on bulk bacterial cells can be traced back to 1980s (Pohl et al., 1981). Peitz & Leeuwen (2010) first used the DEP force to trap single bacterial cells. They reported using DEP traps between parallel 10-mm electrodes to capture living E. coli K12 cells in a microfluidic channel. Fritzsch et al. (2013) applied a nDEP (negative DEP) force and singularized bacterial cells E. coli C600, Bacillus subtilis, 168, Corynebacterium glutamicum ATCC, 13032, by using a well-integrated microchannel and microelectrode labon-a-chip device. Arumugam et al. (2007) demonstrated the generation of DEP traps using vertically aligned carbon nanofibers such as nanoelectrodes. These nanoscale electrodes generated large DEP forces in a small region, ideal for trapping single bacterial cells. They successfully demonstrated an inexpensive and convenient way to produce DEP trap arrays with high-throughput. Although DEP can reliably and precisely trap single bacterial cells, it requires a tight integration of micro/nanoscale electrodes and driving circuits which increase the complexity of this method. Micromanipulation Micromanipulation is a precise method to isolate and manipulate single cells with a relatively low throughput, typically one cell at a time. There have been two types of micromanipulator for bacteria: mechanical and optical (also known as optical tweezers) micromanipulation. Micromanipulation has been applied to single bacterial cells since 1960s (Nossal et al., 1964; Wood, 1967). In mechanical micromanipulation, single cells are individually captured from a population and transferred using a micropipette (Anis et al., 2010, 2011; Ashida et al., 2010; Gao et al., 2011; Roeder et al., 2010; Shi et al., 2011; Teramoto et al., 2010; Tsang et al., 2006). The isolated single cells can be subsequently used for different applications such as cultivation or gene expression analysis. The microbial cells tested by micromanipulation included E. coli DH5a, B. subtilis 168, denitrifying bacteria, Listeria monocytogenes, Salmonella enterica, Synechocystis PCC, 6803, E. coli K12 and Caulobacter crescentus CB15. In optical micromanipulation (Altindal et al., 2011; Carmon & Feingold, 2011; Mirsaidov et al., 2008; Zhang & Liu, 2008), single cells, Vibrio alginolyticus and E. coli, are trapped and manipulated using highly focused laser beams. The foundation of optical traps, also known as optical tweezers, was developed by Ashkin et al. in the 1980s (Ashkin & Dziedzic, 1987; Ashkin et al., 1987). They demonstrated optical tweezers for trapping and manipulating single E. coli cells in media (Ashkin & Dziedzic, 1987; Ashkin et al., 1987). Block et al. (1989)

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

4

X. Shi et al.

Crit Rev Biotechnol, Early Online: 1–13

used optical tweezers to measure the mechanical properties of single E. coli and Streptococcus cells. This method is amenable to integration with transparent microfluidic devices as long as the device design is compatible with high numerical aperture optics to achieve steep intensity gradients around the target cells in the microchannels (Ku¨hn et al., 2009; Min et al., 2009; Xie et al., 2005). The early reports on the integration of optical tweezers and microfluidic devices date back to 2004 (Enger et al., 2004; Munce et al., 2004; Pamp et al., 2012). In these studies, trapped individual cells were transported by the laser beam to a compartment for subsequent culturing and/or analyses. Because the trapped cells are not exposed to the ambient environments during optical micromanipulation, sample contamination issues can be minimized. However, the application of mechanical micromanipulation is perhaps subjected to undesired stresses introduced by mechanical forces. While for optical micromanipulation, wavelength is an important consideration since cells can be light-damaged by intensive power of the light source. Near-infrared laser should be chosen which can reduce the photodamaging effects (Zhang & Liu, 2008). In addition, optical trapping could cause the local temperature rising (Rasmussen et al., 2008) and formation of reactive oxygen species (ROS) (Neuman et al., 1999) which may be harmful for biological samples. In addition, micromanipulation is usually labor-intensive and time consuming with the low throughput representing the major constraint to the method’s widespread usage in the bacterial research field. Cell sorting The purpose of cell sorting is to separate a heterogeneous mixture of biological cells into corresponding sub-populations, typically one cell at a time. In order to distinguish between the sub-populations of cells, the use of specific markers or stains and a sensitive detection method are required. Laser Induced Fluorescence (LIF) is recognized as one of the most sensitive and reliable detection methods. Fluorescence-activated Cell Sorting (FACS) is based upon the detection of laser-induced scattered light and/or fluorescence

signals emanating from the cell or fluorescent markers, respectively, and sorting of individual cells according to their scatter/fluorescence signatures. It provides fast, accurate and quantitative recording of fluorescence signals of individual cells as well as physical separation of cell populations of particular interest. FACS can be used as a very efficient way to separate individual cells. Its application to sort bacterial cells started in late 1990s (Baptista et al., 1999; Fuchs et al., 1996; Yi et al., 1998) and become popular for single cell isolation recently (Dupont et al., 2012; Woyke et al., 2009). The technology has been applied to many microbial strains, such as Enterococcus faecalis E. coli, Mycobacterium, uncultured SAR86 and unknown coastal water samples. This technology requires sophisticated devices and trained operators. Potentially inexpensive, chip-level FACS systems have been produced to circumvent these shortcomings. Fu et al. (1999) developed a microfabricated FACS device and demonstrated its effectiveness in sorting micrometer-sized latex beads and bacterial cells. Compared with the conventional FACS, chip-level FACS devices offer the advantages of: (i) integration with other chip-level analytical technologies, such as PCR or microarrays, (ii) incorporation of multiple cell sorters on a single chip for parallel processing, allowing increased throughput, and (iii) markedly lower reagent consumption and thus cost-effectiveness. One example is the microfluidic cytometer featuring 384 channels for parallel operation developed by Mckenna et al. (2009) for rare-cell screening. Their device was able to perform a genome-wide cDNA screening assay with statistically significant results on positive counts of only several dozen cells in a background of several million negatives. Although this device was not designed for bacterial cells, the principle should be the same for bacterial cell sorting and can potentially be extended to bacterial cell sorting in the near future. Table 1 compares several currently available single-cell isolation techniques and Figure 1 depicts schematic representations of these methods. In general, the selection of these techniques depends on the purpose, available resource and technical requirements of a study. The dilution-to-extinction

Table 1. Comparison of technologies for single bacterial cell isolation. Single cell isolation technology Dilution-to-extinction

Cell Trapping

Micromanipulation FACS a

Stochastic or deterministic 3

10 for stationary chambers (Ottesen et al., 2006); flowing microdroplet generating at4103 droplets per second (Beer et al., 2007); Ave. 30% single cell occupancy based on Poisson distribution Deterministic 103 wellsa (Furutani et al., 2010); Variable DEP trapsb (Peitz & Leeuwen, 2010; Fritzsch et al., 2013) Deterministic 102 cell/second (Shi et al., 2011) Deterministic 4104 cell/second (Ibrahim & van den Engh, 2003) Stochastic

Mechanical trapping. Hydrodynamic and DEP trapping.

b

Capacity

Operation complexity

Compatibility with microfluidic device

Selected references

Low

ˇ

Ottesen et al. 2006; Eun et al., 2011; Guo et al., 2012, Lian et al., 2012; Shim et al., 2009; Zeng et al. 2010; Um et al. 2012; Lin et al. 2009

Lowa/Highb

ˇ

Medium

–

Low

ˇ

Gru¨nberger et al. 2012; Huang et al. 2007; Tanyeri et al. 2010; Fritzsch et al. 2013; Arumugam et al. 2007 Ashida et al. 2010; Altindal et al. 2011 Dupont et al. 2012

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

DOI: 10.3109/07388551.2014.899556

Gene expression in single bacterial cells

Figure 1. Possible single bacterium isolation methods.

5

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

6

X. Shi et al.

method is easy to use and is more suitable for cell isolation from a pure culture or of most abundant microbes. The microdevice for this method is simple to design and construct, and it does not require a precise control of liquid manipulation. In the literatures, the common loading efficiency of the dilution-to-extinction is about 30% and the capacity depends on the number of chambers. These microfluidic devices with stationary chambers currently contain a few thousand chambers, and the dilution-to-extinction cell loading can be completed within one minute. The devices that use flowing microdroplets can generate more than 103 droplets per second. Therefore, the number of chambers (e.g. microdroplets) is determined by the duration of droplet generation. Mechanical trapping shares the low level of complexity with the dilution-to-extinction method, while hydrodynamic trapping and DEP trapping require precise liquid control to achieve accurate and reliable isolation of single cells thus increasing the overall cost and complexity of both techniques. Especially for hydrodynamic trapping, it can trap only a limited number of cells, which has become a major constraint for its application. The most reliable techniques for single-cell isolation are micromanipulation and FACS. However, the throughput of the mechanical micromanipulation-based approaches is relatively low, typically about one cell per a few minutes. The major limitation for using a mechanical micromanipulator in combination with a microfluidic device is that microchannels are usually sealed from the ambient environment preventing the pipette tip of the micromanipulator from accessing the samples. A key feature of the micromanipulation method is that it provides researchers with a means to precisely control the cell selection procedure. FACS is typically capable of single-cell separation with throughputs of up to 104 cells/second. The main principle of separating eukaryotic and prokaryotic cells in FACS is not much different. However, most isolation experiments using commercially available instruments are not with bacterial cells, and their operation would need further optimization (Lomas et al., 2011).

Gene expression analysis in single bacterial cells Several approaches have been utilized to address the gene expression heterogeneity of bacterial communities at the single cell level (Stewart & Franklin, 2008). The first method is to utilize reporter genes (Chalfie et al., 1994). The simple and sensitive enzymatic assays, for example, b-galactosidase and luciferase, have allowed detailed investigations of gene expression mechanisms. These reporter systems can be obtained through the construction of the relevant fusions between promoters of interest and the respective reporter genes. However, the main challenge of this approach is that not all bacterial species, especially the ones found in natural environments, are amenable to genetic manipulation. The second method is fluorescence in situ hybridization (FISH). FISH has been used effectively for assessing the diversity of species in nature. FISH targeting ribosomal RNA (rRNA) is a highly useful method for the phylogenetic identification of bacteria (Amann et al., 1995). However, its accuracy as a quantitative method for determining the expression levels of lowly expressed genes is questionable. The third method is

Crit Rev Biotechnol, Early Online: 1–13

in situ PCR combined with in situ reverse transcription (in situ RT-PCR) (Aoi, 2002). In situ RT-PCR was developed to amplify and detect functional genes and their expression levels inside a single bacterial cell. This is a very useful approach to characterize the genetic and phylogenetic properties of natural bacterial communities at the single-cell level. In the field of environmental microbiology, Hodson et al. (1995) developed an in situ PCR method for prokaryotic cells (bacteria) and detected its gene expression in a single Pseudomonas cell within a model microbial community. This in situ RT-PCR approach has been successfully applied to detect gene expression in many species (Lange et al., 2000; Tolker-Nielsen et al., 1997). In addition to these traditional methods, total transcript amplification (TTA) became available for single bacterial gene expression analysis combined with microarray or messenger RNA sequencing (mRNA-Seq) (Kang et al., 2011). Single-cell based imaging methods have been improved recently and provided useful information about not only the gene expression but also the spatial information about mRNA inside a single bacterium (Taniguchi et al., 2010). RT-qPCR based gene expression measurements in single bacterial cells Reverse-transcription quantitative PCR (RT-qPCR) is the most straightforward and scalable approach for gene expression analysis in single cells (Kubista et al., 2006; Nolan et al., 2006; Shi et al., 2013). Coupled with various cell sorting and collecting methods, several protocols have been published for gene expression analysis by RT-qPCR for single mammalian cells (Lindqvist et al., 2002; Taniguchi et al., 2009; Wacker et al., 2008). The most advanced protocol was published by Taniguchi et al. (2009) who used a quantitative PCR method featuring a reusable single-cell cDNA library immobilized on beads for measuring the expression of multiple cDNA targets (from several copies to several hundred thousand copies) in a single mammalian cell. Significant progress in gene expression profiling of a small number of bacterial cells was made in 2008. In combination with micro-dissection, Lenz et al. (2008) captured subsets of cells from the vertical strata within P. aeruginosa biofilms and quantified transcripts of mRNA and 16S rRNA using RT-qPCR. So far, very few publications have been reported for gene expression measurements in single bacterial cells using the RT-qPCR-based method. This is probably due to the technical challenges specific to the analysis of bacterial cells, including the fact that most bacterial cells are difficult to lyse efficiently, the short half-life of the bacterial mRNA as compared with those from eukaryotic cells, the small dimensions of bacterial cells as compared with mammalian cells (2–3 mm versus 10–20 mm) and, as a result, lower amounts of any given mRNA molecule. Attempts were made in the Center for Biosignatures Discovery Automation in the Biodesign Institute at Arizona State University (ASU), to overcome these issues by developing two methods based on a combination of SYBR Green and RT-qPCR to directly determine the gene expression levels in single bacterial cells (Gao et al., 2011) (Figure 2). The first method is a single-tube approach that allows the analysis of one gene from each

Gene expression in single bacterial cells

7

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

DOI: 10.3109/07388551.2014.899556

Figure 2. Overview of approaches available for single bacterium gene expression profiling.

bacterial cell. The procedure includes single cell picking with an automated micropipette manipulator, followed by a thermal cell lysis in water and one-step RT-qPCR. The second method features a two-stage protocol that consists of RNA isolation from a single bacterial cell and complimentary DNA (cDNA) synthesis in the first stage, and qPCR in the second stage which is possible to simultaneously determine the expression levels of multiple genes in single bacterial cells. The results demonstrated that the method is sensitive enough not only for measuring cellular responses at the

single-cell level, but also for revealing gene expression heterogeneity among bacterial cells. Furthermore, our results showed that the two-stage method can reproducibly measure multiple highly expressed genes from a single E. coli cell. This finding provides a foundation for the future development of a high-throughput, lab-on-a-chip methodology for wholegenome RT-qPCR of single bacterial cells. The TaqMan technique was demonstrated to be useful for single bacterial cell RT-qPCR analysis and can reach both sensitivity and reliability similar to that of the SYBR Green-based approach.

8

X. Shi et al.

Although probe/dye choice in RT-qPCR is usually not important for gene expression profiling, our study provided evidences showing either probe or dye based RT-qPCR could be equally applied at single cell level.

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

Total-transcriptome based gene expression measurements in single bacterial cells Currently, gene expression profiling for complex biological traits on a genomic scale depends on recent advances in highthroughput gene expression analysis technologies, such as DNA microarrays, Serial Analysis of Gene Expression (SAGE) and Next Generation Sequencing (NGS). With these techniques, it is possible to quantitatively investigate complex cellular processes systematically (Kitano, 2002). However, current technologies require relatively large quantities of the initial RNA in order to obtain reliable data. For example, several hundred nanograms to micrograms of total RNA is needed for transcriptome profiling, which corresponds to a sample size of more than, 10 000 eukaryotic cells. To address this issue, several successful attempts have been made in recent years to develop a TTA method for single eukaryotic cells and use either DNA microarray (Kurimoto et al., 2007) or mRNA sequencing (mRNA-Seq) (Tang et al., 2009) technologies to analyze the gene expression levels. TTA methodologies can be divided into two major categories based on amplification techniques: (i) PCR based mRNA Seq (Figure 2), including Smart-seq (Goetz & Trimarchi, 2012; Ramsko¨ld et al., 2012); and (ii) Non-PCR based amplification, such as T7-mediated transcription (Schneider et al., 2004; Wang et al., 2000), RiboSPIAÔ amplification (Single Primer Isothermal Amplification, SPIA) from NuGen (Singh et al., 2005) (Figure 2), and Multiple displacement amplification (MDA) (Spits et al., 2006) (Figure 2). The SMART-Seq which involves a reverse transcriptase enzyme from the Moloney murine leukemia virus has two important features: template switching and adding non-templated cytosine residues to the cDNA (Goetz & Trimarchi, 2012). It is a new single cell transcriptome technology and was claimed as a robust and reproducible method for full length of mRNAs (Ramsko¨ld et al., 2012). However, as poly(T) primer is required for this method, which is also true for most of other PCR-based method (Brady & Iscove, 1993; Tang et al., 2011, 2009), restricting those procedures to be expanded to single bacterial transcriptome amplification. With technique improvement, new strategies without the need of poly(T) primer may be introduced to academia soon so that more single bacterial cell applications may become possible. For T7-mediated methods that are nonPCR based techniques, poly(A) structures of mRNAs are still needed and thus make it unsuitable for single bacterial global gene expression profiling. While other non-PCR based amplification, like RiboSPIAÔ, it could potentially be used for whole transcriptome studies of single prokaryotic cells. It applies a chimeric RNA/DNA primer, RNase H and DNA polymerase to produce amplification of several thousand-fold from single-stranded DNA-based amplifiers (amplified products) (Figure 2). It is relatively fast (typically 6 h per TTA round), can be used with as low as picogram amounts of starting total RNA (Singh et al., 2005). Although NuGen has

Crit Rev Biotechnol, Early Online: 1–13

already introduced a new The Ovation Prokaryotic RNA-Seq System that required only around 500 ng total RNA input, no single-cell bacterial study has been reported so far. The first single bacterial transcriptome analysis using the ’29 polymerase MDA technique was reported recently by Kang et al. (2011) who used Burkholderia thailandensis cells exposed to 0.01% (w/v) of glyphosate, an antibacterial agent, for single-bacterium TTA. The amplified whole transcriptome from a single B. thailandensis cell was analyzed by means of a DNA microarray. The results showed lower fold-change bias (less than two-fold difference and Pearson correlation coefficient R  0.87–0.89) and drop-outs (4%–6% of 2842 detectable genes) as compared with the data obtained from non-amplified RNA samples. In addition, Sanger sequencing of 192 clones generated from the TTA product obtained from a single cell, with and without enrichment by eliminating rRNA and tRNA, detected only B. thailandensis sequences without contamination. Although the sensitivity and accuracy of the total transcripome analysis are, in general, lower than that of RT-qPCR, it can measure expression levels of several thousand genes simultaneously, a marked advantage that has not been replicated by other existing methods. However, the approach is very time-consuming (about 3 days for a round of TTA). Also, we noticed that the estimated total RNA in a single B. thailandensis cell is about two picograms, which is several orders of magnitude higher than the estimated total RNA amount of a typical bacterial cell, such as E. coli (Gao et al., 2011; Schmid et al., 2010). Therefore, a further evaluation of the method is needed in order to assess its feasibility for single bacterial cell studies. With further technology improvement, it should be possible to overcome some of the limitations like bias introduced by pre-amplification of cDNAs and extend single cell TTA technologies for the detection of bacterial cells which lack poly(A) tail in their mRNA (Tischler & Surani, 2013). Optical-based gene expression measurements in single bacterial cells Powerful methodologies based on reported probes and imaging allow for the visualization and validation of expression levels of specific mRNAs in both intact eukaryotic and prokaryotic cells (Tyagi, 2009). Because no intrinsically fluorescent RNA motifs exist, in vivo imaging of mRNA transcripts is less common than proteins. Instead, fluorescent proteins binding to specific RNA motifs (Bertrand et al., 1998; Calapez et al., 2002; Daigle & Ellenberg, 2007; Golding & Cox, 2004; Kerppola, 2006; Rackham & Brown, 2004; Valencia-Burton et al., 2007), sequence-specific oligonucleotide probes (Cardullo et al., 1988; Li et al., 2002; Morrison et al., 1989; Sando & Kool, 2002; Sixou et al., 1994), aptamer tagging (Babendure et al., 2003; Sando et al., 2007), rapid detection of miRNA by a silver nanocluster DNA probe (Yang & Vosch, 2011) and RNA mimics of green fluorescent protein (Paige et al., 2011) have been adopted for mRNA imaging. Such efforts, although still in their early stage, have already shed light on the RNA distribution and dynamics in living cells. We separated them into three categories and briefly described the strengths and limitations of these methods. Since no special requirement of mRNA

Gene expression in single bacterial cells

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

DOI: 10.3109/07388551.2014.899556

structure, all of these methods can be applied to gene expression analysis in single bacterial cells without technique hurdle. The first and most established method for imaging of intracellular RNA in live cells is tagging mRNA with a fluorescent protein (i.e. green fluorescent protein (GFP), yellow fluorescent protein (YFP), and red fluorescent protein (RFP)) (Bertrand et al., 1998; Calapez et al., 2002; Daigle & Ellenberg, 2007; Golding & Cox, 2004; Kerppola, 2006; Montero Llopis et al., 2010; Nevo-Dinur et al., 2011; Rackham & Brown, 2004; Valencia-Burton et al., 2007). To tag a specific target mRNA, one must fuse an RNA-binding protein to GFP and at the same time tag the 30 -untranslated region of the target mRNA with an RNA motif recognized by the RNA-binding protein. Bertrand et al. (1998) first introduced the MS2 coat protein-GFP approach for imaging mRNA dynamics in live cells. There are two components of this method. The first is the MS2 coat protein, a phage RNA-binding protein, expressed as a fusion with intact GFP. The second is the target ASH1 mRNA, which is tagged with multiple copies of MS2-binding motifs. When these two components are co-transformed and co-expressed in cells, MS2–GFP fusion proteins bind to their cognate motif on the mRNA and render it fluorescent (Figure 2). The major challenge in tagging with intact GFP is the need to distinguish bound GFP from unbound GFP, since GFP constructs are always fluorescent. This was recently accomplished by adopting the reconstruction of GFP, by splitting GFP into two non-fluorescent fragments. The two fragments are non-fluorescent until a pair of tags attached to each fragment recognizes the target mRNA and assembles the two split GFP fragments into a correctly folded and functional protein (Kerppola, 2006) (Figure 2). The MS2 coat protein and zip code-binding protein fused with split GFP fragments (Rackham & Brown, 2004) and split eIF4A domains fused with N- and C-terminal of GFP fragments (Valencia-Burton et al., 2007) have successfully been demonstrated as applications of the split GFP approach. However, a drawback of the split GFP tagging method is the high affinity of the two protein fragments to each other, making the binding difficult to reverse. This prevents the method from being utilized for imaging of fast dynamic processes (Magliery et al., 2005). The second approach is based on imaging of endogenous mRNAs using fluorescence resonance energy transfer (FRET) and contact-mediated quenching (Cardullo et al., 1988; Li et al., 2002; Morrison et al., 1989; Sando & Kool, 2002; Santangelo et al., 2004; Sixou et al., 1994; Tyagi & Kramer, 1996). Several different probes whose fluorescent properties change upon sequence-specific hybridization have been explored, including competitive hybridization probes (Li et al., 2002; Morrison et al., 1989; Sixou et al., 1994), side-by-side probes (Cardullo et al., 1988), quenched autoligation probes (Sando & Kool, 2002), molecular beacon probes (Tyagi & Kramer, 1996), and dual molecular FRET probes (Santangelo et al., 2004) (Figure 2). Probe-based imaging features several distinct advantages: probes detect mRNA in cells directly without the need to engineer target genes and GFP constructs. The approach can be multiplexed by using spectrally distinguishable fluorophores and cells can

9

be sorted based on gene expression levels but not only the ‘‘positive’’ or ‘‘negative’’ signals (Tyagi, 2009). The limitation of this method includes its lower sensitivity since these probes have only one fluorophore in each molecule resulting in lower overall signals compared to GFP tags, and the need for delivering probes into cells and the degradation of probes (Tyagi, 2009). In addition, the cell structure of bacterial cells may also be issues for introducing probes into the cell. Strategies such as lipofaction, pore-forming agents and cellpenetrating peptides (Tyagi, 2009), can be used to improve the probe delivery into cells. The third approach employs tagging of artificial RNA motifs (aptamers) with small nonfluorescent dyes to induce fluorescence when combined with specified aptamers (Babendure et al., 2003; Sando et al., 2007). The free dye molecules are nonfluorescent because of the strong dissipation of the excitation energy through vibrational (radiationless) relaxation. Once a selected aptamer binds to the dye molecule restricting its vibrational freedom, the dyes become fluorescent resulting in an increase of the fluorescence signal by more than, 2,000 fold (Babendure et al., 2003). Examples include Hoechst dye variants that are non-fluorescent in the unbound form but show strong fluorescence when bound to pre-selected effective RNA aptamers (Sando et al., 2007). The availability of many aptamer-dye combinations allows imaging of multiple mRNA simultaneously. However, the free radicals created by the irradiated dye can destroy the RNA motifs (Grate & Wilson, 1999). Taniguchi et al. (2010) used a DNA oligomer probe labeled with a single fluorophore, also known as RNA fluorescence in situ hybridization (RNA FISH) (Figure 2), to successfully hybridize it with the mRNA on a microfluidic device. RNA FISH is a high sensitive technique for visualizing mRNA of almost all the genes in single cells (Broude, 2011; Montero Llopis et al., 2010; Raj et al., 2008; Wijgerde et al., 1995). It is also a direct method for mRNA tagging without modifying RNAs, but the real-time measurements have not yet been achieved to date (Broude, 2011). Other recent methods, such as the silver nanocluster DNA probe for miRNA (Yang & Vosch, 2011) and RNA-based variants of green fluorescent protein (Paige et al., 2011), are also expected to contribute to quantitative imaging of multiple mRNAs and small RNAs simultaneously in single bacterial cells in the near future.

Conclusion The existence of substantial bacterial cell-to-cell diversity at the gene expression level is becoming accepted by the research community (Wang & Bodovitz, 2010). However, proper measurements and interpretation of the cell-to-cell heterogeneities are still challenging and may hold the ‘‘keys’’ for further insights into bacterial metabolism, stimulus responses and survival in various environments. In this review, we summarized the recent progress on developing new molecular methods and devices for gene expression measurements in single bacterial cells. These efforts, although still in their infancy, have already contributed greatly to a better understanding of bacterial metabolism (Heinemann & Zenobi, 2011).

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

10

X. Shi et al.

One major obstacle for general microbiologists to analyze gene expression at the single-cell level is the manipulation of individual cells. Although Section ‘‘Manipulation of single bacterial cells’’ lists the manipulation methods that can be applied to bacteria, the selection is mostly depending on the accessibility of instruments and compatibility with the downstream measurements. At this moment, there are no bacterial cell trapping devices commercially available. Microbiologists have to collaborate with researchers with engineering expertise to fabricate custom-made trapping devices. Meanwhile, some simple techniques such as dilution-to-extinction and micromanipulation can also be used to achieve satisfactory results in some cases, which may be more suitable for current application. To fully explore the gene expression heterogeneity in microbes, future technology developments are still needed in the following areas: (i) improved or novel molecular biology protocols with better sensitivity and specificity with focus on full-length mRNAs, non-coding RNAs and low-abundance transcripts, for example, the efficacy and amplification bias associated with current total transcriptome amplification protocols still need improvement; (ii) improved devices and detection technology with higher capacity so that large number of cells can be measured quickly with high sensitivity; (iii) low cost measurement, further development and optimization of microfluidic devices to make the technology available to the larger community of microbiologists at much lower cost; and (iv) integration of each individual analytic module into a fully functional system. The future systems will include all the major steps involved in single cell analysis, such as cell isolation, mRNA analysis and even data analysis, in one system so that the analysis can be accessed easily by microbiologists and the data reproducibility across the laboratories can be assured. In summary, single bacterial cell level analysis is expected to reveal novel insights in microbial research. With the help of new engineering technologies such as special designed microfluidic devices, the analysis tool will be more powerful.

Acknowledgements We thank ASU’s NEPTUNE fund to Prof. Deirdre Meldrum for the support of this research and Dr. Laimonas Kelbauskas of ASU for editing.

Declaration of interest The authors declare no conflict of interest. Dr. Weiwen Zhang is currently funded by a grant from the National Natural Science Foundation of China (project no. 31170043).

References Altindal T, Chattopadhyay S, Wu X-L. (2011). Bacterial chemotaxis in an optical trap. PLoS ONE, 6, e18231. Amann RI, Ludwig W, Schleifer KH. (1995). Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol Rev, 59, 143–69. An D, Parsek MR. (2007). The promise and peril of transcriptional profiling in biofilm communities. Curr Opin Microbiol, 10, 292–6. Anis Y, Holl M, Meldrum D. (2010). Automated selection and placement of single cells using vision-based feedback control. IEEE Trans Automat Sci Eng, 7, 598–606.

Crit Rev Biotechnol, Early Online: 1–13

Anis Y, Houkal J, Holl M, et al. (2011). Diaphragm pico-liter pump for single-cell manipulation. Biomed Microdevices, 13, 651–9. Aoi Y. (2002). In situ identification of microorganisms in biofilm communities. J Biosci Bioeng, 94, 552–6. Arumugam PU, Chen H, Cassell AM, Li J. (2007). Dielectrophoretic trapping of single bacteria at carbon nanofiber nanoelectrode arrays. J Phys Chem A, 111, 12772–7. Ashida N, Ishii S, Hayano S, et al. (2010). Isolation of functional single cells from environments using a micromanipulator: application to study denitrifying bacteria. Appl Microbiol Biotechnol, 85, 1211–17. Ashkin A, Dziedzic JM. (1987). Optical trapping and manipulation of viruses and bacteria. Science, 235, 1517–20. Ashkin A, Dziedzic JM, Yamane T. (1987). Optical trapping and manipulation of single cells using infrared laser beams. Nature, 330, 769–71. Babendure JR, Adams SR, Tsien RY. (2003). Aptamers switch on fluorescence of triphenylmethane dyes. J Am Chem Soc, 125, 14716–17. Banaeiyan AA, Ahmadpour D, Adiels CB, Gokso¨r M. (2013). Design and fabrication of high-throughput application-specific microfluidic devices for studying single-cell responses to extracellular perturbations. Proc SPIE, 8765, 87650K. Banerjee B, Balasubramanian S, Ananthakrishna G, et al. (2004). Tracking operator state fluctuations in gene expression in single cells. Biophys J, 86, 3052–9. Baptista M, Rodrigues P, Depardieu F, et al. (1999). Single-cell analysis of glycopeptide resistance gene expression in teicoplanin-resistant mutants of a VanB-type Enterococcus faecalis. Mol Microbiol, 32, 17–28. Bardy SL, Ng SY, Jarrell KF. (2003). Prokaryotic motility structures. Microbiology, 149, 295–304. Becskei A, Kaufmann BB, van Oudenaarden A. (2005). Contributions of low molecule number and chromosomal positioning to stochastic gene expression. Nat Genet, 37, 937–44. Beer NR, Hindson BJ, Wheeler EK, et al. (2007). On-chip, real-time, single-copy polymerase chain reaction in picoliter droplets. Anal Chem, 79, 8471–5. Benı´tez JJ, Topolancik J, Tian HC, et al. (2012). Microfluidic extraction, stretching and analysis of human chromosomal DNA from single cells. Lab Chip, 12, 4848–54. Bertrand E, Chartrand P, Schaefer M, et al. (1998). Localization of ASH1 mRNA particles in living yeast. Mol Cell, 2, 437–45. Beta C, Bodenschatz E. (2011). Microfluidic tools for quantitative studies of eukaryotic chemotaxis. Eur J Cell Biol, 90, 811–16. Block SM, Blair DF, Berg HC. (1989). Compliance of bacterial flagella measured with optical tweezers. Nature, 338, 514–18. Boedicker J, Li L, Kline T, Ismagilov R. (2008). Detecting bacteria and determining their susceptibility to antibiotics by stochastic confinement in nanoliter droplets using plug-based microfluidics. Lab chip, 8, 1265–72. Brady G, Iscove NN. (1993). Construction of cDNA libraries from single cells. Meth Enzymol, 225, 611–23. Brehm-Stecher BF, Johnson EA. (2004). Single-cell microbiology: tools, technologies, and applications. Microbiol Mol Biol Rev, 68, 538–59. Broude NE. (2011). Analysis of RNA localization and metabolism in single live bacterial cells: achievements and challenges. Mol Microbiol, 80, 1137–47. Button DK, Schut F, Quang P, et al. (1993). Viability and isolation of marine bacteria by dilution culture: theory, procedures, and initial results. Appl Environ Microbiol, 59, 881–91. Calapez A, Pereira HM, Calado A, et al. (2002). The intranuclear mobility of messenger RNA binding proteins is ATP dependent and temperature sensitive. J Cell Biol, 159, 795–805. Cardullo RA, Agrawal S, Flores C, et al. (1988). Detection of nucleic acid hybridization by nonradiative fluorescence resonance energy transfer. Proc Natl Acad Sci USA, 85, 8790–4. Carmon G, Feingold M. (2011). Rotation of single bacterial cells relative to the optical axis using optical tweezers. Opt Lett, 36, 40–2. Chalfie M, Tu Y, Euskirchen G, et al. (1994). Green fluorescent protein as a marker for gene expression. Science, 263, 802–5. Clausell-Tormos J, Lieber D, Baret J-C, et al. (2008). Droplet-based microfluidic platforms for the encapsulation and screening of mammalian cells and multicellular organisms. Chem Biol, 15, 427–37.

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

DOI: 10.3109/07388551.2014.899556

Colman-Lerner A, Gordon A, Serra E, et al. (2005). Regulated cell-tocell variation in a cell-fate decision system. Nature, 437, 699–706. Daigle N, Ellenberg J. (2007). N-GFP: an RNA reporter system for live-cell imaging. Nat Methods, 4, 633–6. Di Carlo D, Wu LY, Lee LP. (2006). Dynamic single cell culture array. Lab Chip, 6, 1445–9. Dupont CL, Rusch DB, Yooseph S, et al. (2012). Genomic insights to SAR86, an abundant and uncultivated marine bacterial lineage. ISME J, 6, 1186–99. Enger J, Gokso¨r M, Ramser K, et al. (2004). Optical tweezers applied to a microfluidic system. Lab Chip, 4, 196–200. Eun YJ, Utada AS, Copeland MF, et al. (2011). Encapsulating bacteria in agarose microparticles using microfluidics for high-throughput cell analysis and isolation. ACS Chem Biol, 6, 260–6. Fritzsch FS, Rosenthal K, Kampert A, et al. (2013). Picoliter nDEP traps enable time-resolved contactless single bacterial cell analysis in controlled microenvironments. Lab Chip, 13, 397–408. Fu AY, Spence C, Scherer A, et al. (1999). A microfabricated fluorescence-activated cell sorter. Nat. Biotechnol, 17, 1109–11. Fuchs P, Weichel W, Du¨bel S, et al. (1996). Separation of E. coli expressing functional cell-wall bound antibody fragments by FACS. Immunotechnology, 2, 97–102. Furutani S, Nagai H, Takamura Y, Kubo I. (2010). Compact disk (CD)-shaped device for single cell isolation and PCR of a specific gene in the isolated cell. Anal Bioanal Chem, 398, 2997–3004. Gao W, Navarroli D, Naimark J, et al. (2013). Microbe observation and cultivation array (MOCA) for cultivating and analyzing environmental microbiota. Microbiome, 1, 1–8. Gao W, Zhang W, Meldrum DR. (2011). RT-qPCR based quantitative analysis of gene expression in single bacterial cells. J Microbiol Methods, 85, 221–7. Goetz JJ, Trimarchi JM. (2012). Transcriptome sequencing of single cells with Smart-Seq. Nat Biotech, 30, 763–5. Golding I, Cox EC. (2004). RNA dynamics in live Escherichia coli cells. Proc Natl Acad Sci USA, 101, 11310–15. Golding I, Paulsson J, Zawilski SM, Cox EC. (2005). Real-time kinetics of gene activity in individual bacteria. Cell, 123, 1025–36. Grate D, Wilson C. (1999). Laser-mediated, site-specific inactivation of RNA transcripts. Proc Natl Acad Sci USA, 96, 6131–6. Gru¨nberger A, Paczia N, Probst C, et al. (2012). A disposable picolitre bioreactor for cultivation and investigation of industrially relevant bacteria on the single cell level. Lab Chip, 12, 2060–8. Guo MT, Rotem A, Heyman JA, Weitz DA. (2012). Droplet microfluidics for high-throughput biological assays. Lab Chip, 12, 2146–55. Heinemann M, Zenobi R. (2011). Single cell metabolomics. Curr Opin Biotechnol, 22, 26–31. Hodson RE, Dustman WA, Garg RP, Moran MA. (1995). In situ PCR for visualization of microscale distribution of specific genes and gene products in prokaryotic communities. Appl Environ Microbiol, 61, 4074–82. Huang B, Wu H, Bhaya D, et al. (2007). Counting low-copy number proteins in a single cell. Science, 315, 81–4. Ibrahim SF, van den Engh G. (2003). High-speed cell sorting: fundamentals and recent advances. Curr Opin Biotechnol, 14, 5–12. Kang Y, Norris MH, Zarzycki-Siek J, et al. (2011). Transcript amplification from single bacterium for transcriptome analysis. Genome Res, 21, 925–35. Kelly CD, Rahn O. (1932). The growth rate of individual bacterial cells. J. Bacteriol, 23, 147–53. Kerppola TK. (2006). Visualization of molecular interactions by fluorescence complementation. Nat Rev Mol Cell Biol, 7, 449–56. Kitano H. (2002). Computational systems biology. Nature, 420, 206–10. Kuang Y, Biran I, Walt DR. (2004). Simultaneously monitoring gene expression kinetics and genetic noise in single cells by optical well arrays. Anal Chem, 76, 6282–6. Kubista M, Andrade JM, Bengtsson M, et al. (2006). The real-time polymerase chain reaction. Mol Aspects Med, 27, 95–125. Ku¨hn S, Measor P, Lunt EJ, et al . (2009). Loss-based optical trap for on-chip particle analysis. Lab Chip, 9, 2212–16. Kurimoto K, Yabuta Y, Ohinata Y, Saitou M. (2007). Global single-cell cDNA amplification to provide a template for representative highdensity oligonucleotide microarray analysis. Nat Protoc, 2, 739–52. Lange M, Tolker-Nielsen T, Molin S, Ahring BK. (2000). In situ reverse transcription-PCR for monitoring gene expression in individual

Gene expression in single bacterial cells

11

Methanosarcina mazei S-6 cells. Appl Environ. Microbiol, 66, 1796–800. Le TT, Harlepp S, Guet CC, et al. (2005). Real-time RNA profiling within a single bacterium. Proc Natl Acad Sci USA, 102, 9160–4. Lenz AP, Williamson KS, Pitts B, et al. (2008). Localized gene expression in Pseudomonas aeruginosa biofilms. Appl Environ Microbiol, 74, 4463–71. Leung K, Zahn H, Leaver T, et al. (2012). A programmable dropletbased microfluidic device applied to multiparameter analysis of single microbes and microbial communities. Proc Natl Acad Sci USA, 109, 7665–70. Li Q, Luan G, Guo Q, Liang J. (2002). A new class of homogeneous nucleic acid probes based on specific displacement hybridization. Nucleic Acids Res, 30, E5. Lian M, Collier CP, Doktycz MJ, Retterer ST. (2012). Monodisperse alginate microgel formation in a three-dimensional microfluidic droplet generator. Biomicrofluidics, 6, 044108. Lidstrom ME, Meldrum DR. (2003). Life-on-a-chip. Nat Rev Microbiol, 1, 158–64. Lin L.-I., Chao S.-H., Meldrum DR. (2009). Practical, microfabricationfree device for single-cell isolation. PLoS ONE, 4, e6710. Lindqvist N, Vidal-Sanz M, Hallbo¨o¨k F. (2002). Single cell RT-PCR analysis of tyrosine kinase receptor expression in adult rat retinal ganglion cells isolated by retinal sandwiching. Brain Res Protoc, 10, 75–83. Lomas MW, Bronk DA, Van den Engh G. (2011). Use of flow cytometry to measure biogeochemical rates and processes in the ocean. Ann Rev Mar Sci, 3, 537–66. Lutz BR, Chen J, Schwartz DT. (2006a). Hydrodynamic tweezers: 1. Noncontact trapping of single cells using steady streaming microeddies. Anal Chem, 78, 5429–35. Lutz BR, Chen J, Schwartz DT. (2006b). Characterizing homogeneous chemistry using well-mixed microeddies. Anal Chem, 78, 1606–12. Macfarlane S, Dillon JF. (2007). Microbial biofilms in the human gastrointestinal tract. J Appl Microbiol, 102, 1187–96. Magliery TJ, Wilson CG, Pan W, et al. (2005). Detecting protein-protein interactions with a green fluorescent protein fragment reassembly trap: scope and mechanism. J Am Chem Soc, 127, 146–57. Maloney PC, Rotman B. (1973). Distribution of suboptimally induces b-D-galactosidase in Escherichia coli. The enzyme content of individual cells. J Mol Biol, 73, 77–91. McAdams HH, Arkin A. (1997). Stochastic mechanisms in gene expression. Proc Natl Acad Sci USA, 94, 814–9. Mckenna BK, Selim AA, Richard Bringhurst F, Ehrlich DJ. (2009). 384-Channel parallel microfluidic cytometer for rare-cell screening. Lab Chip, 9, 305–10. Meldrum DR, Holl MR. (2002). Microscale bioanalytical systems. Science, 297, 1197–8. Min TL, Mears PJ, Chubiz LM, et al. (2009). High-resolution, long-term characterization of bacterial motility using optical tweezers. Nat Methods, 6, 831–5. Mirsaidov U, Timp W, Timp K, et al. (2008). Optimal optical trap for bacterial viability. Phys Rev E, 78, 021910. Molter TW, Holl MR, Dragavon JM, et al. (2008). A new approach for measuring single-cell oxygen consumption rates. IEEE Trans Automat Sci Eng, 5, 32–42. Molter TW, McQuaide SC, Suchorolski MT, et al. (2009). A microwell array device capable of measuring single-cell oxygen consumption rates. Sensors Actuators B: chemical, 135, 678–86. Montero Llopis P, Jackson AF, Sliusarenko O, et al. (2010). Spatial organization of the flow of genetic information in bacteria. Nature, 466, 77–81. Morrison LE, Halder TC, Stols LM. (1989). Solution-phase detection of polynucleotides using interacting fluorescent labels and competitive hybridization. Anal Biochem, 183, 231–44. Munce NR, Li J, Herman PR, Lilge L. (2004). Microfabricated system for parallel single-cell capillary electrophoresis. Anal Chem, 76, 4983–9. Neuman KC, Chadd EH, Liou GF, et al. (1999). Characterization of photodamage to Escherichia coli in optical traps. Biophys J, 77, 2856–63. Nevo-Dinur K, Nussbaum-Shochat A, Ben-Yehuda S, Amster-Choder O. (2011). Translation-independent localization of mRNA in E. coli. Science, 331, 1081–4.

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

12

X. Shi et al.

Newman JR, Ghaemmaghami S, Ihmels J, et al. (2006). Single-cell proteomic analysis of S. cerevisiae reveals the architecture of biological noise. Nature, 441, 840–6. Nolan T, Hands RE, Bustin SA. (2006). Quantification of mRNA using real-time RT-PCR. Nat Protoc, 1, 1559–82. Nossal GJ, Szenberg A, Ada GL, Austin CM. (1964). Single cell studies on 19S antibody production. J Exp Med, 119, 485–502. Ottesen EA, Hong JW, Quake SR, Leadbetter JR. (2006). Microfluidic digital PCR enables multigene analysis of individual environmental bacteria. Science, 314, 1464–7. Paige JS, Wu KY, Jaffrey SR. (2011). RNA mimics of green fluorescent protein. Science, 333, 642–6. Pamp SJ, Harrington ED, Quake SR, et al. (2012). Single-cell sequencing provides clues about the host interactions of segmented filamentous bacteria (SFB). Genome Res, 22, 1107–19. Pedraza JM, van Oudenaarden A. (2005). Noise propagation in gene networks. Science, 307, 1965–9. Peitz I, van Leeuwen R. (2010). Single-cell bacteria growth monitoring by automated DEP-facilitated image analysis. Lab Chip, 10, 2944–51. Pijper A, Discombe G. (1946). Shape and motility of bacteria. J Pathol Bacteriol, 58, 325–42. Pohl HA, Kaler K, Pollock K. (1981). The continuous positive and negative dielectrophoresis of microorganisms. J Biol Phys, 9, 67–86. Rackham O, Brown CM. (2004). Visualization of RNA-protein interactions in living cells: FMRP and IMP1 interact on mRNAs. EMBO J, 23, 3346–55. Raj A, van den Bogaard P, Rifkin SA, et al. (2008). Imaging individual mRNA molecules using multiple singly labeled probes. Nat Meth, 5, 877–9. Rajilic´-Stojanovic´, M., Smidt H, De Vos WM. (2007). Diversity of the human gastrointestinal tract microbiota revisited. Environ Microbiol, 9, 2125–36. Ramsko¨ld D, Luo S, Wang Y-C, et al. (2012). Full-length mRNA-Seq from single-cell levels of RNA and individual circulating tumor cells. Nat Biotech, 30, 777–82. Rappe´ MS, Connon SA, Vergin KL, Giovannoni SJ. (2002). Cultivation of the ubiquitous SAR11 marine bacterioplankton clade. Nature, 418, 630–3. Rasmussen MB, Oddershede LB, Siegumfeldt H. (2008). Optical tweezers cause physiological damage to Escherichia coli and Listeria bacteria. Appl Environ Microbiol, 74, 2441–6. Roeder B, Wagner M, Rossmanith P. (2010). Autonomous growth of isolated single Listeria monocytogenes and Salmonella enterica serovar typhimurium cells in the absence of growth factors and intercellular contact. Appl Environ Microbiol, 76, 2600–6. Roman GT, Chen Y, Viberg P, et al. (2007). Single-cell manipulation and analysis using microfluidic devices. Anal Bioanal Chem, 387, 9–12. Rosenfeld N, Young JW, Alon U, et al. (2005). Gene regulation at the single-cell level. Science, 307, 1962–5. Sando S, Kool ET. (2002). Imaging of RNA in bacteria with self-ligating quenched probes. J Am Chem Soc, 124, 9686–7. Sando S, Narita A, Aoyama Y. (2007). Light-up Hoechst-DNA aptamer pair: generation of an aptamer-selective fluorophore from a conventional DNA-staining dye. Chembiochem, 8, 1795–803. Santangelo PJ, Nix B, Tsourkas A, Bao G. (2004). Dual FRET molecular beacons for mRNA detection in living cells. Nucleic Acids Res, 32, e57. Schmid A, Kortmann H, Dittrich PS, Blank LM. (2010). Chemical and biological single cell analysis. Curr Opin Biotechnol, 21, 12–20. Schneider J, Buness A, Huber W, et al. (2004). Systematic analysis of T7 RNA polymerase based in vitro linear RNA amplification for use in microarray experiments. BMC Genomics, 5, 29. Schut F, de Vries EJ, Gottschal JC, et al. (1993). Isolation of typical marine bacteria by dilution culture: growth, maintenance, and characteristics of isolates under laboratory conditions. Appl Environ Microbiol, 59, 2150–60. Sezonov G, Joseleau-Petit D, D’Ari R. (2007). Escherichia coli physiology in Luria-Bertani broth. J. Bacteriol, 189, 8746–9. Shi X, Gao W, Chao S, et al. (2013). Monitoring the single-cell stress response of the diatom Thalassiosira pseudonana by quantitative realtime reverse transcription-PCR. Applied and environmental microbiology, 79, 1850–8. Shi X, Lin L-I, Chen S-Y, et al. (2011). Real-time PCR of single bacterial cells on an array of adhering droplets. Lab Chip, 11, 2276–81.

Crit Rev Biotechnol, Early Online: 1–13

Shim J, Olguin LF, Whyte G, et al. (2009). Simultaneous determination of gene expression and enzymatic activity in individual bacterial cells in microdroplet compartments. J Am Chem Soc, 131, 15251–6. Siegele DA, Hu JC. (1997). Gene expression from plasmids containing the araBAD promoter at subsaturating inducer concentrations represents mixed populations. Proc Natl Acad Sci USA, 94, 8168–72. Singh R, Maganti RJ, Jabba SV, et al. (2005). Microarray-based comparison of three amplification methods for nanogram amounts of total RNA. Am J Physiol, Cell Physiol, 288, C1179–89. Sixou S, Szoka Jr FC, Green GA, et al. (1994). Intracellular oligonucleotide hybridization detected by fluorescence resonance energy transfer (FRET). Nucleic Acids Res, 22, 662–8. Sizova MV, Hohmann T, Hazen A, et al. (2012). New approaches for isolation of previously uncultivated oral bacteria. Appl Environ Microbiol, 78, 194–203. Spits C, Le Caignec C, De Rycke M, et al. (2006). Whole-genome multiple displacement amplification from single cells. Nat Protocols, 1, 1965–70. Spudich J, Koshland D. (1976). Non-genetic individuality: chance in single cell. Nature, 262, 467–71. Stewart PS, Franklin MJ. (2008). Physiological heterogeneity in biofilms. Nat Rev Microbiol, 6, 199–210. Strovas TJ, Lidstrom ME. (2009). Population heterogeneity in Methylobacterium extorquens AM1. Microbiology, 155, 2040–8. Strovas TJ, Sauter LM, Guo X, Lidstrom ME. (2007). Cell-to-cell heterogeneity in growth rate and gene expression in Methylobacterium extorquens AM1. J Bacteriol, 189, 7127–33. Tang F, Barbacioru C, Wang Y, et al. (2009). mRNA-Seq wholetranscriptome analysis of a single cell. Nat Methods, 6, 377–82. Tang F, Lao K, Surani MA. (2011). Development and applications of single-cell transcriptome analysis. Nat Methods, 8, S6–11. Taniguchi K, Kajiyama T, Kambara H. (2009). Quantitative analysis of gene expression in a single cell by qPCR. Nat Methods, 6, 503–6. Taniguchi Y, Choi PJ, Li G-W, et al. (2010). Quantifying E. coli proteome and transcriptome with single-molecule sensitivity in single cells. Science, 329, 533–8. Tanyeri M, Johnson-Chavarria EM, Schroeder CM. (2010). Hydrodynamic trap for single particles and cells. Appl Phys Lett, 96, 224101. Tanyeri M, Schroeder CM. (2013). Manipulation and confinement of single particles using fluid flow. Nano Lett, 13, 2357–64. Teramoto J, Yamanishi Y, Magdy E-SH, et al. (2010). Single livebacterial cell assay of promoter activity and regulation. Genes Cells, 15, 1111–22. Tischler J, Surani MA. (2013). Investigating transcriptional states at single-cell-resolution. Curr Opin Biotechnol, 24, 69–78. Tolker-Nielsen T, Holmstrøm K, Molin S. (1997). Visualization of specific gene expression in individual Salmonella typhimurium cells by in situ PCR. Appl Environ Microbiol, 63, 4196–203. Tsang PH, Li G, Brun YV, et al. (2006). Adhesion of single bacterial cells in the micronewton range. Proc Natl Acad Sci USA, 103, 5764–8. Tyagi S. (2009). Imaging intracellular RNA distribution and dynamics in living cells. Nat Methods, 6, 331–8. Tyagi S, Kramer FR. (1996). Molecular beacons: probes that fluoresce upon hybridization. Nat Biotechnol, 14, 303–8. Um E, Rha E, Choi S-L, et al. (2012). Mesh-integrated microdroplet array for simultaneous merging and storage of single-cell droplets. Lab Chip, 12, 1594–7. Valencia-Burton M, McCullough RM, Cantor CR, Broude NE. (2007). RNA visualization in live bacterial cells using fluorescent protein complementation. Nat Methods, 4, 421–7. Wacker MJ, Tehel MM, Gallagher PM. (2008). Technique for quantitative RT-PCR analysis directly from single muscle fibers. J Appl Physiol, 105, 308–15. Wang D, Bodovitz S. (2010). Single cell analysis: the new frontier in ‘‘omics’’. Trends Biotechnol, 28, 281–90. Wang E, Miller LD, Ohnmacht GA, et al. (2000). High-fidelity mRNA amplification for gene profiling. Nat Biotechnol, 18, 457–9. White AK, VanInsberghe M, Petriv OI, et al. (2011). High-throughput microfluidic single-cell RT-qPCR. Proc Natl Acad Sci USA, 108, 13999–4004.

DOI: 10.3109/07388551.2014.899556

Critical Reviews in Biotechnology Downloaded from informahealthcare.com by UMEA University Library on 08/26/14 For personal use only.

Wijgerde M, Grosveld F, Fraser P. (1995). Transcription complex stability and chromatin dynamics in vivo. Nature, 377, 209–13. Wood TH. (1967). Genetic recombination in Escherichia coli: clone heterogeneity and the kinetics of segregation. Science, 157, 319–21. Woyke T, Xie G, Copeland A, et al. (2009). Assembling the marine metagenome: one Cell at a Time. PLoS One, 4, E5299. Xie C, Mace J, Dinno MA, et al. (2005). Identification of single bacterial cells in aqueous solution using confocal laser tweezers Raman spectroscopy. Anal Chem, 77, 4390–7. Yang SW, Vosch T. (2011). Rapid detection of MicroRNA by a silver nanocluster DNA Probe. Anal Chem, 83, 6935–9.

Gene expression in single bacterial cells

13

Yi WC, Hsiao S, Liu JH, et al. (1998). Use of fluorescein labelled antibody and fluorescence activated cell sorter for rapid identification of Mycobacterium species. Biochem Biophys Res Commun, 250, 403–8. Zeng Y, Novak R, Shuga J, et al. (2010). High-performance single cell genetic analysis using microfluidic emulsion generator arrays. Anal Chem, 82, 3183–90. Zhang H, Liu K-K. (2008). Optical tweezers for single cells. J R Soc Interface, 5, 671–90. Zhu H, Tian Y, Bhushan S, et al. (2012). High throughput micropatterning of optical oxygen sensor for single cell analysis. IEEE Sensors J, 12, 1668–72.