Expanding Tn-Seq with Microfluidics: Droplet Tn-Seq
Tn-Seq (Transposon Sequencing) is a powerful tool used to determine an organism's fitness, particularly bacteria in high throughput by culturing the transposon library and analyzing the complex single-cell phenotypes. In a Tn-Seq, the bacteria is isolated and a transposon library (addition of transposon with a flanking Mmel digestion site) is created which is further cultured by transformation into the host. On growing in the medium, random gene silencing may lead to conditions where the host's survival is difficult due to their association with metabolism, resistance, virulence, etc. This is analyzed by isolating the DNA and sequencing by digestion of the Mmel and addition of adapters which is essential for PCR (Polymerase Chain Reaction) and sequencing. This analysis can also reveal essential genes, functional genes, pathogen virulent factors, gene-gene interaction, host-pathogen interaction, and how environmental conditions can influence the gene. Further, Tn-Seq can also provide insights into the biology, behavior, and activity of the transposon like insertion preference, Mobility, and transposon-host interaction. So this technique was considered as the gold standard for determining the quantitative contribution of a gene to fitness under a specific growth condition in high-throughput and genome-wide.
Complication in Tn-Seq
Even though Tn-Seq can screen thousands of mutants in a single experiment, the growing mass of mutants i.e. in a pool leads to masking. Masking is a phenomenon where the collective growth of the transposon mutant in the same environment leads to the masking of the fitness effect of an individual mutation. For example, enzymes that break down complex glycans into smaller units for energy utilization in one stain may help in mutants that do not produce these enzymes as they can “cheat” and reap the carbon-source benefits. This leads to inappropriate errors or background noise in analysis, making the result partially fallacious.
dTn-Seq (Droplet Transposon Sequencing)
Droplet Tn-Seq is a solution created by combining microfluidics and Tn-Seq which solves the limitation of conventional Tn-seq. dTn-Seq contains a droplet microfluidic device that sorts the host cells into micron-sized droplets. This encapsulation prevents the masking phenomena that are seen in traditional Tn-Seq. So in a dTn-Seq, the mutant library is passed through a droplet microfluidic device along with a fluorinated oil-surfactant which results in monodisperse droplets with outer fluorinated oil-surfactant layer and an inner filled with growth medium with one cell each. This is possible due to the diluting of the culture.
However, assuming a Poisson distribution, a cell concentration of ~2 × 106 cells/ml passed in a microfluidic device will generate a droplet population that consists of ~74% empty droplets, ~22% with single cells, and ~3% with two or more cells. The device yields droplets of ~65–67 µm diameter, ~144–157 pL volume, at a production rate of ~5 × 104 droplets/min. These after-growth are broken by 1H, 1H, 2H, 2H-perfluoro-1-octanol (PFO) which is further amplified by Whole-genome sequencing (mediated by phi29 DNA polymerase). Further digestion of Mmel and insertion of adapters are done for PCR by which the sequences are correlated with the initial library for the appropriate analysis. On the medium, 1% agarose is added for stable monodisperse agarose droplets that provide a matrix, which supports microcolony formation. Significantly. The amount of medium inside a droplet is essential for 5–8 generations of bacterial growth for Gram-negative and positive bacteria.
Additionally, comparing the cell's growth pattern in batch (8 ml liquid batch culture) and droplet states that each culture on droplet has growth three times higher than the batch. To assess if each droplet contains the correct number of cells (one cell per droplet), GFP (JWV500) and RFP (MK119) fluorescent Streptococcus pneumoniae strains were mixed in equal proportions. These strains were encapsulated and cultured in agarose droplets. Brightfield and fluorescence microscopy were then used to evaluate the encapsulation frequencies. GFP-only (green arrow), RFP-only (red arrow), and droplets containing both GFP and RFP fluorescing cells (yellow arrow) were shown in the magnified area. For concentrations 1.75×106 cells/ml and 2.5×106 cells/ml, 16.3% and 21.2% of the total droplets were filled, and 1.0-1.8% of droplets contained multiple cells which proves the Poisson distribution.
Significance growth of different microorganisms in Batch and Droplet culture. Liquid droplet and Agarose droplet size optimization
Fabrication of Droplet Microfluidic Device
The microfluidic channels are the tiny paths through which fluids (such as bacterial cultures or chemical solutions) will flow on the microfluidic chip. To create these channels, a design pattern of the channels is laid out on the chip. This is done using AutoCAD 2016, a software commonly used for precise engineering designs.
Once the design was finalized, it was transferred onto a photomask. In this case, the photomask is printed onto acetate transparency film, which is similar to a clear sheet with the design printed in black. The black areas block light, while the transparent areas allow light to pass through. The photomask will be used later in the UV exposure process to transfer the channel pattern onto the silicon wafer. This photomask is ordered from a company (CAD/Art Services), which specializes in producing high-quality photomasks.
A negative photoresist called SU-8 3025 is applied to the silicon wafer. A photoresist is a light-sensitive material that hardens when exposed to UV light. The SU-8 is spread over the silicon wafer using a technique called spin coating.
Spin coating is a method used to apply a uniform thin film of the photoresist on the surface of the wafer. The wafer is placed on a spinning platform, and the photoresist is applied. As the platform spins, the centrifugal force spreads the photoresist evenly across the surface, creating a thin layer. The thickness of the SU-8 layer (in this case, 40 μm) is controlled by adjusting the spin speed and the amount of SU-8 applied. After spin coating, the wafer is baked at 95°C. This baking step (called a soft bake) solidifies the photoresist and prepares it for UV exposure.
The photomask (created in step 1) was aligned with the silicon wafer, which is coated with the photoresist. The transparent and opaque areas on the photomask determine which parts of the photoresist get exposed to UV light.
Photolithography is the process of transferring the pattern from the photomask onto the photoresist using UV light. The silicon wafer (coated with photoresist) is exposed to UV light through the photomask. Wherever the UV light passes through the transparent areas of the photomask, it will harden the SU-8 photoresist underneath. The areas covered by the black regions of the photomask will not be exposed to UV light, so the SU-8 in those areas will remain soft.
After the UV exposure, the wafer undergoes a post-exposure bake. This step further hardens the areas of the SU-8 that were exposed to UV light. The temperature ramps up from 65°C to 95°C over a period of 4 minutes, which ensures that the exposed photoresist fully solidifies.
Now that the exposed areas of the photoresist are hardened, the unexposed areas (the soft SU-8) need to be removed. This is done using a chemical called SU-8 developer. The wafer is immersed in the developer solution, which dissolves the unexposed SU-8, leaving behind the patterned, hardened SU-8 where the UV light shone through.
This process creates raised features on the silicon wafer, corresponding to the design of the microfluidic channels on the photomask. The exposed parts of the wafer are now protected by the hardened SU-8, while the rest of the wafer is left bare. After development, the wafer is rinsed with isopropanol and deionized water (dH2O) to remove any remaining developer. A final heating step is performed, gradually increasing the temperature from 100°C to 200°C over 5 minutes. This final heat treatment strengthens the remaining SU-8, making the mold more durable.
The design of the droplet microfluidic device enables syringe pumps to deliver surfactant in fluorinated oil through tubing to the oil inlet, while the cell-containing culture medium is introduced via the aqueous inlet. Filters are in place to prevent debris from blocking downstream channels, and resistors help minimize fluctuations in liquid flow rates. At the flow-focus junction, the oil separates the continuous stream of cell culture into uniform droplets. These droplets then exit the device through the droplet outlet and are collected.
Fabrication of Droplet Microfluidics
Bacterial cell-cell and cell–host interaction models
Other than the ability to separate the individual bacterial cells in a population, the author hypothesized that they can also be used in applications of screening the phenotypes that are not influenced by the differences in growth rate. Interactions between bacterial cells, or between bacteria and host cells, can be studied, as these interactions can significantly influence bacterial population composition and survival. To understand this two studies were done stating
- Competence signaling between S. pneumoniae strains through competence-stimulating peptide (CSP)
- The interaction between Yersinia pseudotuberculosis (Yptb) and bone marrow-derived macrophages (BMDMs)
CSP signaling is a quorum-sensing mechanism in S. pneumoniae that activates competence processes, such as the uptake and recombination of extracellular DNA. A strain of sfCSPr with the fusion of CSP promoter (SPD_2065/comC) with the superfolder green-fluorescent protein (sfGFP) was created which produces a GFP signal in the presence of CSP-1(comC). sfCSPr was mixed with S. pneumoniae strain ADP112, a strain that produces CSP upon the addition of isopropyl-β-D-thiogalactopyranoside (IPTG) at a ratio of 1:40. After encapsulation and incubating for 3 hours in the absence of IPTG, no GFP signal could be observed in fluorescent microscopy or FACS (Fluorescence-activated cell sorting) whereas population isolated with IPTG showed fluorescent indicating that cell-cell interaction can be seen in the agarose. However, it was also identified that the removal of oil from agarose droplets aids in the encapsulation of a second droplet into the first. To explore this model, sfCSPr was encapsulated into an agarose droplet and incubated, then oil was removed with PFO and re-encapsulated by the device with slightly bigger channels (80 × 40 μm) and aqueous inlet with CSP, medium, and agarose. Within 2 h, CSP diffused from the second layer into the first layer and reached the microcolony, as demonstrated by the induction of a strong GFP signal. This indicates that any diffusible compound can easily pass through agarose droplets, allowing for the screening of signal-mediated interactions between bacteria, such as compounds that either inhibit or promote growth. When combined with FACS, this could be developed into a high-throughput screening method.
Another study was done where agarose droplets were combined with Yersinia pseudotuberculosis and bone marrow-derived macrophages (BMDMs). Y. pseudotuberculosis (Yptb) can form microcolonies in deep tissues of the liver and spleen which then break and start replication. However, extracellular clusters are surrounded by host innate immune cells, including macrophages and inflammatory monocytes. Even though the macrophage doesn't directly destroy the microcolonies, they secrete factors that aid in the destruction. To mimic this, Yptb was encapsulated into agarose/hydrogel droplets, and BMDMs from C57/BL6 mice were added to the agarose droplets after the growth of the Yptb microcolony. The BMDM attachment to the agarose droplet was observed indicating that non-contact-mediated interactions between Yersinia and macrophages can be studied through these models. A cell-host interaction can be studied through the droplets created by the droplet microfluidic device.
Conclusion
dTn-Seq applies to a wide variety of bacteria and can be used as an extension to any variation of Tn Seq to uncover (complex) single-cell phenotypes due to population effects, environment size, or interactions with the extracellular microbial and/or host environment.
Besides the separation application, these droplets have potential application as a cell-cell and cell–host interaction model which can be further expanded. Further, it should be implemented in other samples like viruses, fungi, and cells that need optimization and refabrication of the droplet microfluidic device.
Reference
Thibault, D., Jensen, P.A., Wood, S. et al. Droplet Tn-Seq combines microfluidics with Tn-Seq for identifying complex single-cell phenotypes. Nat Commun 10, 5729 (2019). https://doi.org/10.1038/s41467-019-13719-9
Image Credits
- PDB-101 - RCSB PDB, https://images.app.goo.gl/SkCqARVeVxNbXyeA9
- Nature, https://www.nature.com/articles/s41467-019-13719-9
Comments
Post a Comment