Directed Signaling Cascades in Monodisperse Artificial Eukaryotic Cells

Abstract

The bottom-up assembly of multicompartment artificial cells that are able to direct biochemical reactions along a specific spatial pathway remains a considerable engineering challenge. In this work, we address this with a microfluidic platform that is able to produce monodisperse multivesicular vesicles (MVVs) to serve as synthetic eukaryotic cells. Using a two-inlet polydimethylsiloxane channel design to co-encapsulate different populations of liposomes we are able to produce lipid-based MVVs in a high-throughput manner and with three separate inner compartments, each containing a different enzyme: α-glucosidase, glucose oxidase, and horseradish peroxidase. We demonstrate the ability of these MVVs to carry out directed chemical communication between the compartments via the reconstitution of size-selective membrane pores. Therefore, the signal transduction, which is triggered externally, follows a specific spatial pathway between the compartments. We use this platform to study the effects of enzyme cascade compartmentalization by direct analytical comparison between bulk, one-, two-, and three-compartment systems. This microfluidic strategy to construct complex hierarchical structures is not only suitable to study compartmentalization effects on biochemical reactions but is also applicable for developing advanced drug delivery systems as well as minimal cells in the field of bottom-up synthetic biology.

Keywords: artificial cells; bottom-up; directed signaling; enzyme cascade; microfluidics; multicompartmentalization; synthetic biology.

Figure 1

Microfluidic assembly of compartmentalized MVVs. (a) Schematic and bright-field image of the microfluidic platform. An aqueous phase containing LUVs forms water-in-oil droplets at the first cross junction, which are then sheared at the second cross junction to form double-emulsion templates. These undergo spontaneous dewetting of the oil to render MVVs. Scale bar: 100 μm. (b) Generation of MVVs in the absence of PEGylated lipids resulted in LUV bursting and aggregation at the inner leaflet of the GUVs (top). The addition of 1 mol % PEG-DSPE in the lipid mixture resulted in successful encapsulation of inner compartments (i.e., LUVs) in outer compartments (i.e., GUVs) without rupture (bottom). Inner LUVs consist of POPC:DOPG:Cholesterol:mPEG-DSPE with additional NBD-PE for labeling, and the outer POPC:DOPG:Cholesterol:mPEG-DSPE GUVs are labeled with Atto-633 DOPE. Scale bars: 50 μm. (c) Confocal fluorescence image of the monodisperse GUVs (purple) with inner LUVs (green). Scale bar: 100 μm. (d) Histograms showing the size distribution of MVVs and mean intensities of the inner LUVs (n = 123).

Figure 2

Chemical cascade reaction network in one- and two-compartment systems. (a) Scheme of a one-compartment system with HRP, GOx, and α-Glc in the lumen of GUVs functionalized with αHL pores. (b) Confocal fluorescence time series of multiple homogeneous one-compartment GUVs with a mean diameter of 75.3 ± 6.1 μm after being triggered externally with stachyose molecules via the αHL pores. (c) Average kinetic traces (left) and end point measurements (right) of the resorufin signal (P < 0.005, unpaired t test, N ≥ 2 for the one-compartment system, −stachyose and −GOx−α-Glc controls, respectively, n ≥ 50). (d) Schematic representation of the two-compartment system with HRP in the lumen of the outer GUV and with GOx and α-Glc further encapsulated within a population of LUVs (also embedded with αHL pores). (e) Confocal fluorescence time series of the MVVs with a mean diameter of 70.2 ± 4.9 μm after addition of stachyose. Note that the bright spots are the detached oil droplets. (f) Average kinetic traces (left) and end points (right) of the resorufin signal (P < 0.005, unpaired t test, N ≥ 2 for the two-compartment system, −stachyose and −GOx−α-Glc controls, respectively, n ≥ 50). Error bars in (c) and (f) are taken from the standard error of the mean. Scale bars: 100 μm.

Figure 3

Chemical cascade reaction network in three-compartment MVVs. (a) Scheme and bright-field image of the two-inlet platform encapsulating two distinct LUV populations in the double-emulsion templates and subsequently undergoing dewetting to render MVVs. (b) Scheme of three-compartment MVVs with an outer GUV enclosing GOx-LUVs, α-Glc-LUVs, and HRP. (c) Confocal fluorescence time series of the MVVs with a mean diameter of 73.1 ± 7.2 μm after input of the chemical trigger. Channel 1 shows GOx containing LUVs tagged with Atto-390-DOPE and reconstituted with OmpF pores. Channel 2 shows α-Glc containing LUVs tagged with NBD-PE and embedded with αHL pores. Channel 3 shows the outer microfluidic GUVs tagged with Atto-633-DOPE, and channel 4 shows the resulting resorufin fluorescence over time. (d) Average kinetic traces (left) and end points (right) of the fluorescent resorufin product formed as a final output of the successful initiation of the enzyme cascade (P < 0.005, unpaired t test, N ≥ 2 for the whole system, −stachyose, −OmpF, and −GOx−α-Glc controls, respectively, n ≥ 50). Error bars are taken from the standard error of the mean. Scale bars: 100 μm.

Figure 4

Comparison of the three-step enzyme reaction cascade network in systems with increasing complexity. (a) Schematic representation of the increasing levels of complexity from the open system to multicompartmentalized systems. (b) Average kinetic traces of bulk and one-, two-, and three-compartment systems. (c, d) Final resorufin intensities and apparent rate constants respectively (N ≥ 2). Error bars in (b)–(d) are taken from the standard error of the mean. (P < 0.05, unpaired t test, n = 6, 85, 88, and 116 for the bulk system and single-compartment, two-compartment, and three-compartment systems, respectively). (e) Violin plots of resorufin intensities across each of the different compartmentalized systems.