Micellar self-assembly of perfectly sequence-defined recombinant resilin-like/elastin-like block copolypeptides

Abstract

This paper reports the synthesis of perfectly sequence defined, monodisperse diblock copolypeptides of a hydrophilic elastin-like polypeptide block and a hydrophobic resilin- like polypeptide block and characterization of their self-assembly as a function of structural parameters by light scattering, cryo-TEM, and small-angle neutron scattering. A subset of these diblock copolypeptides exhibit LCST and UCST phase behavior and self-assemble into spherical or cylindrical micelles. Their morphology is dictated by their chain length, degree of hydrophilicity and hydrophilic weight fraction of the ELP block. We find that: (1) independent of the length of the corona forming ELP block there is a minimum threshold in the length of the RLP block below which self-assembly does not occur, but that once that threshold is crossed, (2) the RLP block length is a unique molecular parameter to independently tune self-assembly; and (3) increasing the hydrophobicity of the corona-forming ELP drives a transition from spherical to cylindrical morphology. Unlike the self-assembly of purely ELP based block copolymers, the self-assembly of RLP-ELPs can be understood by simple principles of polymer physics relating hydrophilic weight fraction, polymer-polymer and polymer-solvent interactions to micellar morphology, which is important as it provides a route for the de novo design of desired nanoscale morphologies from first principles.

Graphical Abstract

1. Introduction

In recent decades, block copolymers —polymers composed of two or more distinct blocks— have been studied extensively. ^1–2 When the blocks have a different solubility in a solvent of interest, block copolymers can self-assemble into a range of structures including micelles ³, vesicles (sometimes referred to as “polymersomes”) ⁴, and lamellae. ⁵ At the same time, polymers that display aqueous, temperature sensitive phase behavior, namely lower critical solution temperature (LCST) and the converse, upper critical solution temperature (UCST) transition have also been studied for decades. ^6–7 With advances in polymer synthesis methodologies, these polymers have been combined to create diblock copolymers with an LCST and UCST exhibiting block. ⁸ Some of these synthetic LCST-UCST diblock copolymers exhibit dual LCST and UCST behavior that are capable of temperature triggered self-assembly ⁸, and a subset of these also exhibit micelle inversion in response to temperature, as the two blocks switch their hydrophobicity in response to an increase in solution temperature. ^9–12

Despite the significant body of work on synthetic LCST-UCST diblock copolymers, these studies have also revealed some significant lacunae. First, these diblock copolymers have shown limited morphologies upon self-assembly. Most extant studies report spherical morphologies ^13–14, and few other morphologies such as vesicles ¹⁰, but in general, these observations are not backed by direct and unequivocal experimental conformation of the morphology of the putative self-assembled structures by scattering techniques or electron microscopy. Second, none of these studies have systematically explored structural parameters such as composition and weight fraction that critically impact the morphology of the self-assembled structures. Finally, none of these studies have elucidated the design rules that enable prediction of the morphologies of the self-assembled structures, even for systems with a narrowly defined composition. This paper addresses this important gap by the design of sequence controlled, monodisperse diblock copolymers of an LCST and UCST block and characterization of their self-assembled morphologies as a function of key structural parameters.

These studies were made possible by the fact that we recently identified a new set of recombinant peptide polymers that exhibit UCST behavior, that we have named resilin-like polypeptides (RLPs) ¹⁵, as they are inspired by the consensus sequence of native resilin. ^16–18 The recent availability of a set of UCST exhibiting polypeptides that can be combined with well- studied elastin-like polypeptides (ELPs) that exhibit LCST behavior, provided for the first time an opportunity for the systematic exploration of LCST-UCST diblock copolymers.

In this study, we systematically varied the length of the RLP block, which forms the core of the self-assembled micelles, to investigate the effect of hydrophilic weight fraction on self-assembly. We also independently varied the hydrophilicity of the corona-forming ELP block, to investigate the effect of corona chain hydrophilicity on self-assembly. Finally, we varied the overall length of the chains while keeping the hydrophilic weight fraction constant, to investigate the effect of molecular weight on self-assembly. To develop a complete picture of their self-assembly, we characterized the block copolypeptides by dynamic and static light scattering (DLS and SLS), cryogenic transmission electron microscopy (cryo-TEM), and small-angle neutron scattering (SANS).

We find that unlike most previously reported recombinant block copolypeptides ^19–21, the presence or absence of nanoscale self-assembly and the morphology of the self-assembled constructs of resilin-like/elastin-like block copolypeptides are all predictably governed by simple principles of polymer physics. Specifically, their self-assembly is consistent with the relationship reported by Eisenberg and Discher that the balance between the hydrophilic and hydrophobic blocks (i.e. the hydrophilic weight fraction) determines the morphology, with a higher hydrophilic weight fraction leading to spherical micelles and a lower hydrophilic weight fraction leading to cylindrical micelles.^22–24

2. Experimental section

2.1. Materials

Oligonucleotides encoding all polypeptide sequences were purchased from Integrated DNA Technologies (Coralville, IA). Competent E. coli EB5α cells, used for cloning, and BL21(DE3) cells, used for expression, were purchased from Edge BioSystems (Gaithersburg, MD). Terrific Broth (TB) Dry powder growth media was purchased from MO BIO Laboratories (Carlsbad, CA). Kanamycin sulfate was purchased from EMD Millipore (Billerica, MA) and isopropyl-beta-D-thiogalactoside (IPTG) was purchased from Gold Biotechnology (St. Louis, MO). Calbiochem phosphate buffered saline (PBS) tablets (10 mM phosphate buffer, 140 mM NaCl, 3 mM KCl, pH 7.4 at 25°C) were purchased from EMD Millipore (Billerica, MA). Low retention 0.02 μm and 0.1 μm Whatman Anotop 10 filters were purchased from GE Healthcare Life Sciences (Pittsburgh, PA).

2.2. Synthesis of block copolypeptides genes

All block copolypeptide genes were synthesized from synthetic oligomers using plasmid reconstruction recursive directional ligation (PRe-RDL).²⁵ The block copolypeptides have the form G - (QYPSDGRG)_n-(XGVPG)₈₀-Y where the resilin-like polypeptide (RLP) length n (20, 40, 60, 80, or 100) and the elastin-like polypeptide (ELP) guest residue (A/G, S, V) are systematically varied (Table 1). A guest residue of A/G indicates that the guest residue alternates between A and G from one pentapeptide to the next, resulting in a 50/50 ratio between A and G.

Table 1.

Block copolypeptide nomenclature, sequences, molecular weights, and hydrophilic weight percentages.

	Block copolypeptides	Sequence	MW (kDa)	Hphilic wt%
Set 1	RLP20 — ELPA/G,80	G-(QYPSDGRG)20- [(A/G)GVPG]80-Y	47.79	0.63
	RLP40 — ELPA/G,80	G-(QYPSDGRG)40- [(A/G)GVPG]80-Y	65.01	0.46
	RLP60 — ELPA/G,80	G-(QYPSDGRG)60- [(A/G)GVPG]80-Y	82.24	0.37
	RLP80 — ELPA/G,80	G-(QYPSDGRG)80- [(A/G)GVPG]80-Y	99.46	0.30
	RLP100 — ELPA/G,80	G-(QYPSDGRG)100- [(A/G)GVPG]80-Y	116.68	0.26
Set 2	RLP40 — ELPS,80	G-(QYPSDGRG)40- [SGVPG]80-Y	66.85	0.49
	RLP80 — ELPS,80	G-(QYPSDGRG)80- [SGVPG]80-Y	101.30	0.32
	RLP40 — ELPV,80	G-(QYPSDGRG)40- [VGVPG]80-Y	67.82	0.50
	RLP80 — ELPV,80	G-(QYPSDGRG)80- [VGVPG]80-Y	102.27	0.33
Set 3	RLP40 — ELPA/G,160	G-(QYPSDGRG)40- [(A/G)GVPG]160-Y	94.98	0.63
	RLP20 — ELPA/G,40	G-(QYPSDGRG)20- [(A/G)GVPG]40-Y	32.41	0.46
	RLP80 — ELPA/G,160	G-(QYPSDGRG)80- [(A/G)GVPG]160-Y	129.08	0.46
	RLP40 — ELPA/G,40	G-(QYPSDGRG)40- [(A/G)GVPG]40-Y	49.65	0.30

2.3. Expression and purification of block copolypeptides

Each block copolypeptide was expressed in BL21(DE3) E. coli using a previously published hyperexpression protocol, which relies on the leakiness of the T7 promoter.²⁶ 5 mL cultures were grown overnight and used to inoculate 1 L flasks of TBDry supplemented with 45 μg/mL kanamycin. The flasks were then incubated at 37°C for 24 h and 190 rpm. Each construct was purified using the previously described inverse transition cycling method.^27–28 Briefly, the cell suspension was centrifuged at 3,000 rpm for 10 min at 4°C, the cell pellet resuspended in PBS and then lysed by sonication on ice for 3 min (10 s on, 40 s off) (Masonix S-4000; Farmingdale, NY). Polyethyleneimine (PEI) 0.7% w/v was added to the lysate to precipitate nucleic acid contaminants. The supernatant was then subjected to multiple rounds of ITC as follows: the solution was kept on ice, and 3 M NaCl was added to isothermally trigger the phase transition of the ELP. The coacervate was then centrifuged for 20 min at 14,000 g and 20°C, the supernatant was decanted and discarded, and the pellet was resuspended in phosphate buffer. This suspension was cooled to 4°C, and then centrifuged for 10 min at 14,000 and 4°C to remove any insoluble contaminants. Purity of the block copolypeptides was assessed via SDS-PAGE gel and CuCl₂ staining (Figure S2).

2.4. Thermal turbidimetry

Turbidity profiles were obtained for each of the constructs by recording the optical density as a function of temperature (1°C/min ramp) on a temperature controlled UV-Vis spectrophotometer (Cary 300 Bio; Varian Instruments; Palo Alto, CA). The transition temperature (Tt) was defined as the inflection point of the turbidity profile. Samples were measured in PBS at 10 μM. Because some of the block copolypeptides which form larger micelles are slightly turbid when soluble, all measurements were taken after zeroing with PBS.

2.5. Static and dynamic light scattering

Static and dynamic light scattering measurements (SLS/DLS) were performed using an ALV/CGS-3 goniometer system (Langen, Germany). Samples for the ALV/CGS-3 goniometer system were prepared at a concentration of 10 μM in PBS and filtered through 0.45 μm Millex- GV filters into a 10 mm diameter disposable borosilicate glass tube (Fischer). Simultaneous SLS and DLS measurements were obtained at 15 °C of the ELP for angles between 30°−150° at 5° increments, with each angle consisting of 3 runs for 15 s. SLS experiments were only conducted for self-assembling block copolypeptides, since the molecular weight of a single block copolypeptide chain is already known, and the R_g of a single chain is likely near or below the detection limit of the SLS instrument. The differential refractive index (dn/dc) was determined by measuring the refractive index at different concentrations using an Abbemat 500 refractometer (Anton Paar, Graz, Austria). DLS data were analyzed by fitting the autocorrelation function with a cumulant fit, using the built-in ALV software. Hydrodynamic radius (Rh) was plotted against angle and extrapolated to zero. SLS data were analyzed by partial Zimin plots using ALVSTAT software to determine the R_g and molecular weight (MW).

2.6. Temperature-programmed dynamic light scattering

For one sample whose LCST phase transition was not visible via thermal turbidimetry (RLP₄₀ - ELP_A/G,80), temperature-programmed dynamic light scattering was used to identify the transition temperature (T_t). Temperature-programmed dynamic light scattering experiments were carried out using a Dynapro plate reader (Wyatt Technology; Santa Barbara, CA) with samples filtered through 0.45 μm Millex-GV filters. Data was collected at increments of 1 °C, and the cumulant fit hydrodynamic radius was taken as the radius. The T_t was defined as the temperature at which aggregates of size hundreds of nanometers were formed.

2.7. Cryogenic transmission electron microscopy

Cryo-TEM experiments were performed at Duke University’s Shared Materials Instrumentation Facility (Durham, NC). Lacey holey carbon grids (Ted Pella, Redding, CA) were glow discharged in a PELCO EasiGlow Cleaning System (Ted Pella, Redding, CA). A 3 μL drop (200 μM ELP concentration) was deposited onto the grid, blotted for 3 s with an offset of −3 mm, and vitrified in liquid ethane using the Vitrobot Mark III (FEI, Eindhoven, Netherlands). Prior to vitrification, the sample chamber was maintained at 15 °C and 100% relative humidity to prevent sample evaporation. Grids were transferred to a Gatan 626 cryoholder (Gatan, Pleasanton, CA) and imaged with an FEI Tecnai G2 Twin TEM (FEI, Eindhoven, Netherlands), operating at 80 keV. Feature sizes and spacing distances were measured in ImageJ by manual measurement of at least 25 distances.²⁹

2.8. Small-angle neutron scattering

SANS experiments were conducted at the NIST Center for Neutron Research (NCNR) on the NGB 30m SANS instrument. Samples were poured into quartz cuvettes and acquisitions were performed at 15 °C. All samples were measured in PBS with D2O as a solvent to have good contrast conditions (difference in scattering length densities, ΔSLD) and a lower incoherent background (mainly caused by hydrogen atoms). Three configurations were used with a fixed wavelength of 0.6 nm and sample-detector distances of 1.33 m, 4 m, and 13.17 m corresponding to q ranging from approximately 0.05 nm-1 to 5 nm-1. Data were reduced using the NCNR SANS reduction macros in Igor Pro ³⁰, which allows for the correction of intensities for the transmission, dead-time, detector background (with B4C as a neutron absorber at the sample position), and sample background (either the empty cuvette or the solvent). Absolute intensity was obtained from a tabulated value of a 1.5 mm sheet of Plexiglass.

3. Results and discussion

Recombinant DNA techniques were used to synthesize a family of block copolypeptides in which the RLP length n and the ELP guest residue X were systematically varied. (See Supplemental Information for gene sequences.) We refer to the RLP block with the notation RLP_n, where n denotes the number of octapeptide repeats of the sequence QYPSDGRG, and the ELP block with the notation ELP_x,80 where X denotes the guest residue, and 80 denotes the number of VPGXG pentapeptide repeats in the ELP segment. The first set of five diblock polypeptides consists of polymers in which the ELP block is ELP_A/G,80 that has an equal number of Ala (A) and Gly (G) residues at the guest residue position of the ELP, and the RLP length is varied between 20 and 100 octapeptides. This set allows us to determine how the molecular weight of the core-forming RLP block affects self-assembly. The second set consists of RLP₄₀ and RLP₈₀ fused to ELP_S,80 and ELP_V,80, with Ser (S), and Val (V) as the guest residues, respectively, allowing us to investigate the effect of replacing the ELP_A/G,80 with a more hydrophilic (ELP_S,80) or more hydrophobic ELP (ELP_V,80). The third set consists of RLP₄₀-ELP_A/G,160, RLP₂₀-ELP_A/G,40, RLP₈₀-ELP_A/G;160, and RLP₄₀-ELP_A/G,40, allowing us to investigate the effect of overall chain length, independent of the hydrophilic weight fraction.

The phase behavior of all block copolypeptides with 80 ELP repeats is shown in Figure 1; the turbidity profiles are shown in Figure S3. At the transition temperature (T_t) of the ELP block, the block copolypeptides phase transition into a polypeptide-rich coacervate. The only exception was RLP₂₀-ELP_A/G,80, which has the most hydrophilic RLP block, and which did not exhibit an LCST phase transition in the observable range (5^⁰C-90^⁰C). No UCST phase transition was observed as the solution temperature was raised above the LCST phase transition for any block copolypeptide. The T_t of the ELP block depends upon its hydrophilicity, with more hydrophilic ELP blocks exhibiting a higher T_t. Interestingly, the RLP block plays a role in modulating the T_t of the ELP block, as fusion to longer RLP blocks lowers the T_t.

Figure 1.

LCST phase behavior of block copolypeptides. (RLP₄₀ - ELP_S,80 undergoes a UCST phase transition from micelles to soluble chains at 35 °C before undergoing its LCST at 58 °C.)

A UCST phase transition was only observed for one diblock copolypeptide RLP₄₀-ELP_S,80, which transitions from micelles to soluble chains at 35 °C before undergoing its LCST phase transition from soluble chains to a coacervate at 58 °C (Fig. S2e). The absorbance of a second block copolypeptide - RLP₁₀₀-ELP_A/G,80 - also decreases before increasing at its LCST phase transition (Figure S3b). However, unlike RLP₄₀-ELP_S,80, RLP₁₀₀-ELP_A/G,80 is turbid even when soluble because it forms large micelles at low temperatures (Table 2). Thus, unlike RLP40- ELP_S,80, the turbidity decreases but the solution still remains turbid, indicating that there is not a UCST phase transition and the micelles do not disassemble.

Table 2.

Static and dynamic light scattering characterization of block copolypeptides. Aggregation number N_agg, radius of gyration R_g, hydrodynamic radius R_h and the shape factor ρ = R_g/R_h as derived from the light scattering experiments. (Full light scattering data in Figures S4 - S14 and Table S1.)

	Block copolypeptide	Nagg	Rg(nm)	Rh(nm)	ρ = Rg/Rh
Set 1	RLP20—ELPA/G,80	-	-	5.5	-
	RLP40—ELPA/G,80	68 ± 0.3	29.9 ± 0.6	33.3	0.89
	RLP60—ELPA/G,80	231 ± 0.7	23.8 ± 0.5	36.7	0.65
	RLP80—ELPA/G,80	2240 ± 84	145.4 ± 5.8	114.3	1.27
	RLP100—ELPA/G,80	4463 ±128	187.2 ± 2.5	139.0	1.35
Set 2	RLP40—ELPS,80	80 ± 0.5	27.2 ± 0.8	33.0	0.82
	RLP80—ELPS,80	392 ± 5.3	36.0 ± 1.4	49.4	0.73
	RLP40—ELPV,80	213 ± 0.5	33.1 ± 0.3	31.7	1.04
	RLP80—ELPV,80	1329±72	177.3 ± 4.9	176.8	1.00
Set 3	RLP40—ELPA/G,160	96 ± 0.4	17.1 ± 0.8	33.8	0.51
	RLP20—ELPA/G,40	-	-	5.4	-
	RLP80—ELPA/G,160	780 ± 18	65.9 ± 1.5	125.2	0.53
	RLP40—ELPA/G,40	83 ± 1.3	42.9 ± 1.3	31.6	1.36

All block copolypeptides were next characterized by static and dynamic light scattering (SLS/DLS) at 15 °C, to determine the number of polymer chains (N_agg) per nanostructure, and the radius of gyration (R_g) and hydrodynamic radius (R_h) of the nanostructures (Table 2). The shape factor ρ = R_g/R_h gives an indication of the morphology, with 1.505 corresponding to a Gaussian polymer chain, 1.0 to a hollow sphere or vesicle, and 0.775 to a solid sphere. For an elongated scatterer, the shape factor depends upon the aspect ratio. ³¹

Within the first set of RLP-ELPs wherein we systematically varied the length of the RLP block from 20 to 100 pentapeptides, while keeping the composition and length of the ELP block constant, the light scattering data reveal three distinct behaviors of self-assembly. First, RLP₂₀ – ELP_A/G,80 is fully soluble and does not self-assemble. Second, RLP₄₀-ELP_A/G,80 and RLP₆₀- ELP_A/G,80 self-assemble into structures with R_h of 30–40 nm, ρ < 1, and low N_agg, indicating that both diblocks likely self-assemble into spherical micelles. Third, RLP₈₀-ELP_A/G,80 and RLP₁₀₀- ELP_A/G,80 self-assemble into much larger structures with characteristic ratios above 1 and high aggregation numbers, indicating that they likely self-assemble into much larger, non-spherical structures. These results clearly show that increasing the length of a RLP block while holding the composition and length of the ELP block is enough to modulate self-assembly, indicating that the RLP block length is a unique molecular parameter to independently tune self-assembly of these diblock copolymers.

In the second set, we varied the hydrophilicity of the corona-forming elastin-like polypeptide (ELP) block, to investigate the effect of corona chain hydrophilicity on self-assembly. To do so, the ELP block length was held constant at 80 pentapeptides, but the composition was varied from S (hydrophilic) to V (hydrophobic) guest residues, to investigate the impact of the hydrophobicity of the ELP block on self-assembly. In this series, the ELP length was also toggled between 40 and 80 pentapeptides to examine how the RLP block length affects self-assembly. Equivalent length block copolypeptides — RLP₄₀-ELP_A/G,80 and RLP₈₀-ELP_A/G,80 — from the first set are also included in the analysis of this set, as their A and G guest residues have a hydrophobicity that is intermediate between S and V. Comparison of these three sets by varying hydrophobicity shows that when ELP_A/G,80 is replaced by the more hydrophilic ELP_S,80, the light scattering results indicate that the spherical morphology of ELP_A/G,80 is unaffected. However, when ELP_A/G,80 is replaced by the more hydrophobic ELP_V,80, larger, non-spherical structures are formed, as suggested by the p of 1.00, indicating that the morphology is affected above a threshold hydrophobicity of the ELP block. The trends in morphology for the RLP₄₀- ELP₈₀ copolypeptides in this set are consistent with this notion, as an increase in hydrophobicity from S to A/G in this series appear to have no impact on the spherical morphology, as the p remains close to 0.8, but the increase in ρ to 1.04 for the RLP₄₀ - ELP_V,80 is suggestive of a switch to a more elongated rod-like morphology. The most important observation from this set is that increasing the hydrophobicity of the corona-forming ELP decreases the repulsion between corona chains, causing a transition from spherical to cylindrical morphology.

The third set contains RLP-ELPs that when analyzed with corresponding RLP-ELPs from sets 1 and 2 allowed us to examine the effect of molecular weight on self-assembly, while keeping the hydrophilic weight fraction constant. In this set, RLP₄₀-ELP_A/G,160 self-assembled into spherical micelles, as suggested by its shape factor of approximately 0.5. This demonstrates that the hydrophilic weight fraction alone is not a sufficient condition for self-assembly, as RLP₂₀- ELP_A/G,80 from set 1, despite having the same hydrophilic weight fraction as RLP₄₀-ELP_A/G,160, did not self-assemble (Table 2, R_h = 5.4 nm). Consistent with this observation, RLP₂₀-ELP_A/G,40 also did not self-assemble while RLP₈₀-ELP_A/G,160, which has the same hydrophilic weight fraction, in contrast, self-assembles into spherical micelles (Table 2, shape factor = 0.84). The lack of self-assembly of the two RLP-ELPs with the shortest, 20 octapeptide blocks, is driven by the short length of the RLP block, suggesting that independent of the length of the corona forming ELP block there is a minimum threshold in the length of the RLP block, below which self-assembly does not occur. The formation of a spherical morphology is driven by the hydrophilic weight fraction: both RLP₄₀-ELP_A/G,80 (from Set 1) and RLP₈₀ – ELP_A/G,160 self- assemble into spherical micelles. The micelles formed by RLP₈₀-ELP_A/G,₁₆₀ are larger than those formed by RLP₄₀-ELP_A/G,80 that have the same hydrophilic: hydrophobic ratio, consistent with the higher molecular weight. Finally, RLP₄₀-ELP_A/G,40 has a characteristic ratio well above 1, indicating likely self-assembly into cylindrical micelles, consistent with RLP₈₀-ELP_A/G,80 (from set 1) which has the same hydrophilic weight fraction.

To further characterize the morphology of the self-assembled structures, all self-assembling block copolypeptides were characterized by cryo-TEM (Figures 2–4). In the first set (Figure 2), RLP₄₀-ELP_A/G,80 (Figure 2A) and RLP₆₀-ELP_A/G,80 (Figure 2B) both self-assemble into spherical micelles. Measurements of the core radii indicate that RLP₄₀-ELP_A/G,80 and RLP₆₀-ELP_A/G,80 cores are approximately 12.8 nm and 17.5 nm, consistent with a larger core-forming block leading to a larger micelle core (Table S2). RLP₄₀-ELP_A/G,80 also appears to form a few elongated, cylindrical micelles in the center of the film, where the ice layer is the thinnest. It is not clear from cryo-TEM whether this morphology exists normally in the dilute solution conditions that the light scattering was carried out in, or if it is an artifact of the sample preparation process, which concentrates the block copolypeptide into a thin vitreous ice layer and selectively partitions it to the center of the film. Both light scattering (Table 2) and small-angle neutron scattering (below, Figure 5) help resolve this question.

Figure 2.

Cryo-TEM micrographs of RLP-ELPs in set 1 (a, b) spherical micelles formed by RLP₄₀-ELP_A/G,80 and RLP₆₀-ELP_A/G,80, (c,d) cylindrical micelles formed by RLP₈₀-ELP_A/G,80, RLP₁₀₀-ELP_A/G,80. Scale bars 500 nm. All insets 250 nm by 250 nm. (Additional micrographs in Figures S14 - S17).

Figure 4.

Cryo-TEM micrographs of RLP-ELPs in set 3: (a) spherical micelles formed by RLP₄₀-ELP_A/G,160, (b) spherical micelles formed by RLP₈₀-ELP_A/G,160, and (c) cylindrical micelles formed by RLP₄₀-ELP_A/G,₄₀. Scale bars 500 nm. All insets 250 nm by 250 nm. (Additional micrographs in Figures S14 - S24).

Figure 5.

SANS spectra and analytical model fits for RLP₄₀-ELP_A/G,80 and (C) RLP₈₀-ELP_A/G,80. (The spectrum for RLP₈₀-ELP_A/G,80 is offset by a factor of 100 for clarity.) (Analytical model details in SI).

Cryo-TEM reveals that RLP₈₀-ELP_A/G,80 (Figure 2C) and RLP₁₀₀-ELP_A/G,80 (Figure 2D) form an entirely different nanostructure. Both block copolypeptides form long, overlapping cylindrical structures, with some spherical structures visible as well. Because these experiments were conducted at 10 μM, which is well below the overlap concentration, these cryo-TEM data indicate that during the vitrification process, these block copolypeptides preferentially partition into the thin vitreous ice layer, leading to a higher concentration than is observed in solution. Thus, it is unclear from cryo-TEM alone whether these block copolypeptides truly form micellar structures in dilute solution. Light scattering (Table 2) and small-angle neutron scattering (below, Figure 3) experiments were hence conducted to resolve this question.

Figure 3.

Cryo-TEM micrographs of RLP-ELPs in set 2: (a,b,c) spherical micelles formed by RLP₄₀ – ELP_S,80, RLP₈₀ – ELP_S,80, and RLP₄₀ – ELP_V,80, (d) cylindrical micelles formed by RLP₈₀ – ELP_V,80. Scale bars 500 nm (a-c), 200 nm (d). All insets 250 nm by 250 nm. (Additional micrographs in Figures S18 - S21).

Analyzing the cryo-TEM data, we find that the spacing between micelle cores increased considerably from 47.7 nm for RLP₈₀-ELP_A/G,80 to 70.1 nm for RLP_i00-ELP_A/G80 (Table S2). This is consistent with geometric arguments from polymer physics: as the cylindrical micelles increase in size and incorporate more polypeptide chains, the density of chains increases (consistent with light scattering, Table 2), leading to increased repulsion between corona chains, a more extended ELP chain conformation, and increased spacing of RLP₁₀₀-ELP_A/G80 relative to RLP80-ELP_A/G,80.

Cryo-TEM results for the second set (Figure 3) indicate that when ELP_A/G,80 is replaced by the more hydrophilic ELP_S,80, spherical micelles are formed for both RLP block lengths — RLP₄₀-ELP_S,80, (Figure 3A) and RLP₈₀-ELP_S,80 (Figure 3B). In contrast, when ELP_A/G,80 is replaced by the more hydrophobic ELP_V,80, cylindrical micelles are formed with both RLP blocks — RLP₄₀-ELP_V,80, (Figure 3C) and RLP₈₀-ELP_V,80, (Figure 3D).

In the third set (Figure 4), RLP₄₀-ELP_A/G,160 and RLP₈₀-ELP_A/G,160 both self-assemble into spherical micelles (Figure 4A and 4B), with RLP₈₀-ELP_A/G,160 having a larger micelle core than RLP₄₀-ELP_A/G,160 (33.0 nm vs 12.0 nm, Table S2). RLP₄₀-ELP_A/G,40 forms cylindrical micelles (Figure 4C) with cores smaller than those of RLP₈₀ - ELP_A/G,80 (13.2 nm vs 27.7 nm, Table S2).

To better understand the morphology of the self-assembled micelles, and specifically to determine whether the overlapping cylindrical micelle structures observed in Figure 2 truly occur in dilute solution, RLP₄₀-ELP_A/G,80 and RLP₈₀-ELP_A/G,80 were characterized by small-angle neutron scattering (SANS). These diblock polypeptides were selected because they represent the two main morphologies observed by light scattering and cryo-TEM. A SANS spectrum for each diblock polypeptide is shown in Figure 5.

Even in the absence of a model-dependent analysis, one can observe visually that RLP₈₀- ELP_A/G,80 has a larger size (due to its kink at a lower q of ~ 0.15 nm⁻¹) and a larger aggregation number due to its higher forward scattering value than RLP₄₀-ELP_A/G,₈₀ (30.9 cm⁻¹ vs. 5.9 cm⁻¹). Additionally, the RLP₈₀-ELP_A/G,80 curve continues to increase at low q, indicating that the long dimension of these micelles is beyond the range observable by SANS. To obtain a more complete picture of the self-assembly of these diblock polypeptides, we then fit their SANS spectra using analytical shape models (Figure 5, Supporting Information Table S3). The data were best fit by the spherical (RLP₄₀-ELP_A/G,80) and cylindrical (RLP₈₀-ELP_A/G,80) block copolymer micelle form factors developed by Gerstenberg and Pedersen. ³² (Full model details are provided in the Supporting Information.) The fit parameters (Table 3) are in good agreement with both the light scattering data (Table 2) and cryo-TEM images (Figures 2–4). (The model does not fit the data well at high q for RLP₈₀ - ELP_A/G,80 because the data is highly sensitive to the subtraction of the incoherent background.)

Table 3.

SANS analytical model fit parameters. ^[a] Rg of corona-forming ELP chains. ^[b] Volume fraction of water in the micelle core.

Block copolypeptide	Model	Rcore (nm)	Lcore (nm)	Rg,corona (nm) [a]	ϕH2O,core[b]
RLP40 — ELPA/G,80	spherical	16.0	N/A	6.3	0.67
RLP80 — ELPA/G,80	cylindrical	20.2	>250	6.0	0.73

The self-assembly behavior of the RLP-ELP diblock polypeptides as a function of structural parameters are summarized in Figure 6. For a hydrophilic block of ELP_A/G,80, increasing the size of hydrophobic block from RLP₂₀ to RLP₄₀ leads to increased interactions between the hydrophobic blocks and a transition from fully soluble unimers to self-assembly into spherical micelles (Figure 6A, 1). With the same ELP_A/G,80 hydrophilic block, further increasing the size of the hydrophobic block from RLP₆₀ to RLP₈₀ leads to a transition from spherical to cylindrical micelles (Figure 6A, 2). With a hydrophobic block of RLP₈₀, substituting the more hydrophilic ELP_S,80 for ELP_A/G,80 leads to increased repulsion between corona blocks and a transition from a cylindrical to spherical morphology (Figure 6A, 3). Conversely, substituting a more hydrophobic ELP_V,80 for ELP_A/G,80 decreases repulsion within the corona chains and leads to a transition from a spherical to cylindrical morphology (Figure 6A, 4). With ELP_A/G,80, further increasing the size of the hydrophobic block from RLP₈₀ to RLP₁₀₀ enhances the tendency towards a cylindrical morphology even further and leads to longer cylinders with a larger aggregation number.

Figure 6.

Quasi-phase diagrams illustrating trends in RLP-ELP self-assembly. (a) Trends in set 1 and set 2, block copolypeptides with the motif ELP_Y,80-RLP_x. ELP hydrophilicity increases from ELP_V to ELP_A/G and ELP_S. (b) Trends in set 1 and set 3, block copolypeptides with the motif ELP_A/G,_Y-RLP_X.

The self-assembly trends remain consistent when the overall chain length is varied. Similar to RLP₂₀-ELP_A/G,80, RLP₂₀-ELP_A/G,40 does not self-assemble because RLP₂₀ remains fully soluble (Figure 6B, 5). RLP₄₀-ELP_A/G,160 and RLP₈₀-ELP_A/G,160 (Figure 6B, 6 and 7) both self-assemble into spherical micelles because of their high hydrophilic weight fractions, similar to RLP₄₀ - ELP_A/G,80. RLP₄₀-ELP_A/G,40 (Figure 4B, 8) self-assembles into cylindrical micelles because of its lower hydrophilic weight fraction, similar to RLP₈₀-ELP_A/G,80.

Unlike purely ELP based block copolymers, whose behavior does not obey the canonical principles of self-assembly for diblock polymers predicted by polymer physics, ³³ the selfassembly of ELP-RLP diblock copolymers can be understood by simple principles of polymer physics that relate the hydrophilic weight fraction, polymer-polymer and polymer-solvent interactions to micellar morphology. Specifically, the major findings of this study are that: (1) independent of the length of the corona forming ELP block there is a minimum threshold in the length of the RLP block below which self-assembly does not occur, but that once that threshold is crossed, (2) the RLP block length is a unique molecular parameter to independently tune self-assembly; and (3) increasing the hydrophobicity of the corona forming ELP drives a transition from spherical to cylindrical morphology. These results are unique because they are the first instance of a recombinant diblock polypeptide system whose self-assembly is so clearly governed by simple principles from polymer physics. These findings are also important because they provide a rational approach for the de novo design of desired nanoscale morphologies from first principles, that we believe will lead to biomedical applications, such as the design of nanoscale drug delivery vehicles. ³⁴ We anticipate exploring these applications in future studies.