Fibrillins, fibulins, and matrix-associated glycoprotein modulate the kinetics and morphology of in vitro self-assembly of a recombinant elastin-like polypeptide

Abstract

Elastin is the polymeric protein responsible for the properties of extensibility and elastic recoil of the extracellular matrix in a variety of tissues. Although proper assembly of the elastic matrix is crucial for its durability, the process by which this assembly takes place is not well-understood. Recent data suggest the complex interaction of tropoelastin, the monomeric form of elastin, with a number of other elastic matrix-associated proteins, including fibrillins, fibulins, and matrix-associated glycoprotein (MAGP), is important to achieve the proper architecture of the elastic matrix. At the same time, it is becoming clear that self-assembly properties intrinsic to tropoelastin itself, reflected in a temperature-induced phase separation known as coacervation, are also important in this assembly process. In this study, using a well-characterized elastin-like polypeptide that mimics the self-assembly properties of full-length tropoelastin, the process of self-assembly is deconstructed into "coacervation" and "maturation" stages that can be distinguished kinetically by different parameters. Members of the fibrillin, fibulin, and MAGP families of proteins are shown to profoundly affect both the kinetics of self-assembly and the morphology of the maturing coacervate, restricting the growth of coacervate droplets and, in some cases, causing clustering of droplets into fibrillar structures.

Figure 1

Monitoring the kinetics of coacervate formation and maturation by absorbance at 440 nm. Open squares correspond to measured absorbance. The solid line is the curve mathematically fitted to the absorption and described by the equation abs = −a e^−k_ct + b e^−k_mt + c (see text). The dashed line shows the solution temperature as a function of time.

Figure 2

Effect of stirring rate on the kinetics of coacervate formation and maturation: (A) effect of various stirring rates on changes in absorbance with time; (B) effect of stirring rate on coacervation temperature (mean ± SD); (C) effect of stirring rate on velocity of coacervation, V_c (○), and velocity of maturation, V_m (□), parameters extracted from the equation fitted to the change in absorbance with time (mean ± SD).

Figure 3

Effect of elastin polypeptide concentration on temperature of coacervation and velocity of maturation in unstirred samples: (A) Temperature of coacervation (T_c) decreases nonlinearally with increased polypeptide concentration. (B) Decrease in turbidity with time after coacervation is more rapid for higher concentrations of elastin polypeptide (EP). In all cases, the temperature was increased stepwise and held at approximately 7–10 °C above the coacervation temperature, and samples were unstirred. (C) Velocity of maturation (V_m), calculated from these tubidity vs time curves as described in Figure 1, increases with increased concentration of elastin polypep-tide (mean ± SD).

Figure 4

Coacervate droplet size increases with maturation time. See text for details of methodology for visualization and measurement of coacervate droplet area (mean ± SEM).

Figure 5

Effect of fibulin-5 and fibulin-4 on the kinetics of elastin polypeptide (EP) coacervation and maturation. Coacervation conditions and methodology for mathematical analysis of the curves are described in the text. Addition of fibulin-5 resulted in a concentration-dependent inhibition of the decrease in absorbance with time (A). Velocity of maturation was significantly reduced, while the velocity of coacervation (V_c) was not affected (B). Addition of fibulin-4 also inhibited the fall in absorbance with time in a concentration-dependent manner (C), again the result of a decreased velocity of maturation (V_m) with no change in velocity of coacervation, V_c (D). In all cases, the concentration of the elastin polypeptide was 6.25 µM. Fibulin-5 and fibulin-4 were added at the molar ratios indicated, relative to the concentration of elastin polypeptide. Mean ± SD; *, p < 0.05; **, p < 0.01 vs elastin polypeptide alone.

Figure 6

Temperature of coacervation (T_c), of the elastin polypeptide is unaffected by coacervation in the presence of matrix-associated proteins. In all cases, the concentration of the elastin polypeptide is 6.25 µM and the concentration of the matrix-associated protein is 0.25 µM (molar ratio = 1.0/0.04). Mean ± SD.

Figure 7

Effect of deleting regions of fibulin-5 on kinetics of coacervation and maturation of the elastin polypeptide (EP). (A) Domain structure of wild-type fibulin-5 (WT) and the fibulin-5 deletions tested (adapted from ref 15). Shaded regions are EGF domains; hatched region is fibulin domain; EBD is elastin binding domain. (B) Effects of wild-type fibulin-5 (WT) and deletion mutations, as indicated, on change of absorption with time. (C) Effects of fibulin-5 deletions on coacervation velocity (V_c) and maturation velocity (V_m). In all cases, the concentration of elastin polypeptide was 6.25 µM and the molar ratio of elastin polypeptide/fibulin-5 was 1.0/0.04. Mean ± SD; *, p < 0.05; **, p < 0.01 vs elastin polypeptide alone; ††, p < 0.01 vs elastin polypeptide/WT.

Figure 8

Diagram mapping domains contained in the N- and C-terminal halves of fibrillin-1 and fibrillin-2 (adapted from ref 25).

Figure 9

Effect of N- and C-terminal halves of fibrillin-1 and fibrillin-2 on kinetics of elastin polypeptide (EP) coacervation and maturation. Coacervation conditions and methodology for mathematical analysis of the curves are described in the text. Addition of the N-terminal half of fibrillin-1 (Fibr-1N) resulted in a concentration-dependent acceleration of the decrease in absorbance with time, while the C-terminal half of fibrillin-1 (Fibr-1C) had little or no effect (A). The effect of Fibr-1N was the result of a significant increase in the velocity of maturation (V_m), with no effect on the velocity of coacervation (V_c) (B). In contrast, neither the N-terminal or C-terminal halves of fibrillin-2 (Fibr-2N and Fibr-2C, respectively) had any significant effect on either the change in absorbance with time (C) or the magnitudes of V_c and V_m (D). In all cases, the concentration of the elastin polypeptide was 6.25 µM. Fibr-1N, Fibr-1C, Fibr-2N, and Fibr-2C were added at the molar ratios indicated, relative to the concentration of elastin polypeptide. Mean ± SD; **, p < 0.01 vs elastin polypeptide alone.

Figure 10

Effect of MAGP-1 on the kinetics of elastin polypeptide (EP) coacervation and maturation. Coacervation conditions and methodology for mathematical analysis of the curves are described in the text. Addition of fibulin-5 resulted in a concentration-dependent acceleration of the decrease in absorbance with time (A). Velocity of maturation (V_m) was significantly increased (B). The apparent increase in velocity of coacervation was not statistically significant (p = 0.054). In all cases, the concentration of the elastin polypeptide was 6.25 µM. MAGP-1 was added at the molar ratios indicated, relative to the concentration of elastin polypeptide. Mean ± SD; *, p < 0.05 vs elastin polypeptide alone.

Figure 11

Effects of fibulin-5 (Fibu-5), N- and -C terminal halves of fibrillin-1 (Fibr-1N and Fibr-1C, respectively), MAGP-1, and a combination of Fibr-1N and Fibu-5 on growth of coacervation droplets of the elastin polypeptide (EP) with time. Images are taken at 5, 30, and 120 min after initiation of coacervation. With the exception of Fibr-1C, all of these matrix-associated proteins had the effect of inhibiting growth of the coacervation droplets. In all cases, the concentration of EP was 100 µM, and Fibu-5, Fibr-1N, Fibr-1C, and MAGP-1 were each added at a concentration of 2.0 µM (molar ratio of 1.0/0.02). In the case of Fibr-1N + Fibu-5, both proteins were present at a concentration of 2.0 µM. Scale bar refers to all panels.

Figure 12

Higher magnification images of coacervation droplets 120 min after initiation of coacervation. In all cases, the concentration of the elastin polypeptide was 100 µM and Fibu-5, Fibr-1N, and MAGP-1 were each added at a concentration of 2.0 µM (molar ratio of 1.0/0.02). In the case of Fibr-1N + Fibu-5, both proteins were present at a concentration of 2.0 µM. Particularly in the case of added Fibr-1N and Fibr-1N + Fibu-5, the effect is to cluster small coacervation droplets into chains or “strings”, forming an organized network.