Redox sensitive protein droplets from recombinant oleosin

Abstract

Protein engineering enables the creation of materials with designer functionality and tailored responsiveness. Here, we design a protein with two control motifs for its phase separation into micron sized liquid droplets - one driven by a hydrophobic domain and the other by oxidation of a disulfide bond. Our work is based on the plant surfactant protein, oleosin, which has a hydrophobic domain but no cysteines. Oleosin phase separates to form liquid droplets below a critical temperature akin to many naturally occurring membrane-less organelles. Sequence mutations are made to introduce a cysteine residue into oleosin. The addition of a cysteine causes phase separation at a lower concentration and increases the phase transition temperature. Adding a reducing agent to phase-separated, cysteine-containing oleosin rapidly dissolves the droplets. The transition temperature is tuned by varying the location of the cysteine or by blending the parent cysteine-less molecule with the cysteine-containing mutant. This provides a novel way to control protein droplet formation and dissolution. We envision this work having applications as a system for the release of a protein or drug with engineered sensitivity to reducing conditions and as a mimic of membrane-less organelles in synthetic protocells.

Figure 1.

(A) Hydropathy of Oleo30G calculated using the Kyte-Doolitte scale shown in black, left axis. In this plot, positive values indicate hydrophilic residues and negative values indicate hydrophobic residues. Residues that were mutated to a cysteine are indicated with a blue X. Prediction of disorder of Oleo30G made using the PONDR VL3-BA algorithm shown in grey, right axis. Sections with a PONDR score above 0.5 are predicted to be disordered. The N and C terminal segments of Oleo30G are predicted to be both hydrophilic and disordered. (B) Schematics depicting wild type oleosin, Oleo30G, and Oleo30G-cys. The hydrophilic segments are shown in blue and the hydrophobic core is shown in red. Numbers indicate the number of amino acids in each segment. Schematics are simplifications of the protein and do not show the secondary structure. A family of Oleo30G-cys mutants was designed with the single cysteine residue located at various locations in the N-terminal hydrophilic arm.

Figure 2.

Phase diagrams for Oleo30G (A) and Oleo30G_S2C (B). Compositions at which droplets formed are shown as colored circles and compositions in which droplets did not form are shown as grey squares. As a guide to the eye, an approximate phase boundary is shown. Oleo30G_S2C forms droplets at lower protein concentrations than Oleo30G. Samples were cooled on ice for 10 minutes before imaging on chambered coverglass. The presence of droplets was determined visually. For comparison, Oleo30G and Oleo30G_S2C images are shown in (C) and (D) at the same protein and salt concentration, 80 μM protein and 140 mM salt. At this composition, Oleo30G_S2C droplets are noticeably larger and more numerous than Oleo30G droplets. Scale bars = 20 μm.

Figure 3.

(A) Brightfield DIC microscopy images of Oleo30G_S2C shows spherical droplets. The sample was at a concentration of 80 μM protein in DPBS with 1 mM DTT. The sample was chilled on ice for 10 minutes before transferring to chambered covergla0073s coated with pluronic F-127. Because the droplets are denser than water, they fell to the bottom of the chamber. Image was taken at the coverglass after droplets settled at the bottom. Scale bar = 20 μm. (B) Time-lapse images of Oleo30G_S2C shows the fusion of two droplets over the course of several minutes. Time starts after droplets settled onto the coverglass. Scale bar = 5 μm. (C) UV-vis spectroscopy traces of Oleo30G_S2C. The sample was at a concentration of 80 μM protein in DPBS with 1 mM DTT. Cooling trace was taken starting at 37 °C and cooling at a rate of 1 °C per minute. An increase in the absorbance indicates the formation of droplets. Heating trace was taken starting at 0 °C and heating at a rate of 1 °C per minute. Return of absorbance to near zero indicates return to a single phase. Measurements were taken in increments if 0.5 °C. (D) UV-vis spectroscopy traces of Oleo30G_S2C at various salt concentrations. The sample was at a concentration of 80 μM protein, 1mM DTT, and salt concentrations of 280 mM, 140 mM, and 70 mM NaCl. Measurements were taken starting at 50 °C and cooling at a rate of 1 °C per minute. Measurements were taken in increments if 0.5 °C.

Figure 4.

Brightfield DIC microscopy images of (A) Oleo30G_S2C, (B) Oleo30G_T3C, (C) Oleo30G_T4C, (D) Oleo30G_T5C, (E) Oleo30G_T12C, (F) Oleo30G_T24C, (G) Oleo30G_S39C, and (H) Oleo30G. Protein solutions were at a concentration of 80 μM protein in DPBS with 1 mM DTT. Samples were chilled on ice for 10 minutes before transferring to chambered coverglass coated with pluronic F-127. As the droplets are denser than water, they fell to the bottom of the chamber. Images were taken at the coverglass. Scale bar = 20 μm

Figure 5.

(A) UV-vis spectroscopy traces of Oleo30G and Oleo30G-cys mutants. Protein solutions were at a concentration of 80 μM protein in DPBS with 1 mM DTT. Measurements were taken starting at 37 °C and cooling at a rate of 1 °C per minute. Measurements were taken in increments if 0.5 °C. An increase in the absorbance indicates the formation of droplets. For clarity, data is shown only for 20 °C and below. (B) Temperature at which the absorbance measured by UV-vis spectroscopy while cooling at rate of 1 °C per minute reached 0.2 (a.u.) plotted against the position of the added cysteine. Scale bars show the standard deviation from three independent measurements. The transition temperature drops for variants with the cysteine further from the N-terminus of the protein. No values are shown for Oleo30G or Oleo30G_S39C because the absorbance did not reach 0.2.

Figure 6.

(A) UV-vis spectroscopy traces of Oleo30G_S2C, Oleo30G, and Oleo30G_S2C/Oleo30G blends. Protein solutions were at a concentration of 80 μM protein in DPBS with 1 mM DTT. Blends had a total protein concentration of 80 μM. Blend molar ratios used were 75%/25%, 50%/50%, and 25%/75% Oleo30G_S2C/Oleo30G. Measurements were taken starting at 37 °C and cooling at a rate of 1 °C per minute. Measurements were taken in increments if 0.5 °C. An increase in the absorbance indicates the formation of droplets. For clarity, data is shown only for 20 °C and below. Brightfield DIC microscopy images were taken of (B) Oleo30G_S2C, (C) 75%/25% Oleo30G_S2C/Oleo30G, (D) 50%/50% Oleo30G_S2C/Oleo30G, (E) 25%/75% Oleo30G_S2C/Oleo30G, (F) Oleo30G. Solutions were at a total protein concentration of 80 μM in DPBS with 1 mM DTT. Samples were chilled on ice for 10 minutes before transferring to chambered coverglass coated with pluronic F-127. As the droplets are denser than water, they fell to the bottom of the chamber. Images were taken at the coverglass. Scale bar = 20 μm.

Figure 7.

(A) Brightfield DIC microscopy image of Oleo30G_S2C at a concentration of 80 μM protein in DPBS with 1 mM DTT. (B) Brightfield DIC microscopy image of Oleo30G_S2C after the addition of βME to a final concentration of 80 mM. Samples were chilled on ice for 10 minutes before imaging. Scale bar = 20 μm. (C) UV-vis spectroscopy traces of Oleo30G_S2C. The protein solution was at a concentration of 80 μM protein in DPBS with 1 mM DTT (blue curve). Measurements were taken in increments if 0.5 °C starting at 37 °C and cooling at a rate of 1 °C per minute. Measurements were taken again after addition of βME to a final concentration of 80 mM (grey curve). For clarity, data is shown only for 25 °C and below. (D) UV-vis spectroscopy traces of Oleo30G_S2C/T5C. Addition of βME reduced the phase transition temperature of the protein. Unlike the Oleo30G variants containing only a single cysteine, the double cysteine mutant had a measurable increase in turbidity upon cooling after the addition of βME. Solutions were at a concentration of 80 μM protein in DPBS with 1 mM DTT. Measurements were taken starting at 37 °C and cooling at a rate of 1 °C per minute to a final temperature of 0 °C. Measurements were taken in increments if 0.5 °C. For clarity, data is shown only for 25 °C and below.