A thermo-responsive protein treatment for dry eyes

Abstract

Millions of Americans suffer from dry eye disease, and there are few effective therapies capable of treating these patients. A decade ago, an abundant protein component of human tears was discovered and named lacritin (Lacrt). Lacrt has prosecretory activity in the lacrimal gland and mitogenic activity at the corneal epithelium. Similar to other proteins placed on the ocular surface, the durability of its effect is limited by rapid tear turnover. Motivated by the rationale that a thermo-responsive coacervate containing Lacrt would have better retention upon administration, we have constructed and tested the activity of a thermo-responsive Lacrt fused to an elastin-like polypeptide (ELP). Inspired from the human tropoelastin protein, ELP protein polymers reversibly phase separate into viscous coacervates above a tunable transition temperature. This fusion construct exhibited the prosecretory function of native Lacrt as illustrated by its ability to stimulate β-hexosaminidase secretion from primary rabbit lacrimal gland acinar cells. It also increased tear secretion from non-obese diabetic (NOD) mice, a model of autoimmune dacryoadenitis, when administered via intra-lacrimal injection. Lacrt ELP fusion proteins undergo temperature-mediated assembly to form a depot inside the lacrimal gland. We propose that these Lacrt ELP fusion proteins represent a potential therapy for dry eye disease and the strategy of ELP-mediated phase separation may have applicability to other diseases of the ocular surface.

Keywords: Elastin-like polypeptides (ELPs); Lacrimal gland; Lacritin; Prosecretory; Thermo-responsive; Uptake.

Figure 1. Purification and thermal characterization of LV96

A) Cartoon of the LV96 fusion protein showing Lacrt at the N-terminus and an ELP tag at the C-terminus, with a thrombin recognition site between the two moieties. B) SDS-PAGE of purified LV96, ELP alone (V96) and free Lacrt. Gels were stained using CuCl₂. C) Representative optical density profiles for LV96, V96 and Lacrt at 25 μM as a function of temperature, which indicate a phase separation at 26.8 (LV96) and 31.6°C (V96) respectively. Lacrt alone remains soluble, and does not increase optical density. D) A concentration temperature phase diagram was constructed. Best fit lines are indicated that follow Eq. 1 (Table 1).

Figure 2. Purified lacritin is susceptible to proteolysis of an unidentified origin

A) Western blot of purified Lacrt probed with an anti-Lacrt antibody (raised against Lacrt lacking 65 amino acids at the amino terminus) revealed a major band around 18 kDa, which is consistent with that observed previously for purified Lacrt. B) MALDI-TOF analysis of Lacrt revealed the appearance of major lower molecular weight fragments (Table 2) upon incubation at 37 °C in PBS. C) Time dependent disappearance of the purified Lacrt band by was tracked by SDS-PAGE stained with Coomassie blue. D) Lacrt disappearance was quantified and fitted to a single exponential decay model, which yielded a half-life 23.7 h (R²=0.99).

Figure 3. Lacrt-ELP fusion proteins are prosecretory in LGACs

A) A cartoon depicts structure for ex vivo clusters of LGACs obtained from rabbits. These primary cultures form an apical lumen (L) that is bounded by a thick network of actin filaments. SV: secretory vesicles. B) Rabbit secretory vesicles (SV) release β-hexosaminidase in a dose dependent manner in response to secretagogues. The percentage of cellular secretion for each treatment (Eq. 2) has been normalized to a positive control defined by CCh-stimulation, which is defined as 100 %. LGACs were treated with 0.1 to 20 μM of LV96, Lacrt, V96, or no treatment for 1 h at 37 °C. 10 and 20 μM LV96 significantly enhanced secretion compared to the V96 group (**p<0.01), and a similar effect was found with 20 μM Lacrt (*p<0.05). Data were shown as mean ± S.D. and analyzed by ANOVA followed by Bonferroni’s Post-Hoc Test. C) Live-cell confocal microscopy was performed using LGACs labeled for F-actin (red). Actin-RFP is enriched beneath the apical membrane surrounding the lumen. Shortly after 10 min, increased irregularity of apical actin filaments and actin-coated secretory vesicle (SV) formation beneath the apical membrane (white arrows) were observed in CCh treated cells (100 μM). Similarly LV96 (20 μM) induced time-dependent actin remodeling after 20 min, which increased the irregularity of apical actin filaments and formation of secretory vesicles (white arrows). No significant remodeling of actin filaments was observed in a V96 treated control group. Apical lumen are indicated (white asterisk *). Scale bar: 10 μm.

Figure 4. Fusion with V96 influences uptake of exogenous Lacrt into LGACs

A) Time-dependent uptake of rhodamine-labeled Lacrt into rabbit LGACs was assessed by live-cell confocal microscopy. After 30 min, significant numbers of fluorescent puncta were detected in the cytosol proximal to the basolateral membrane (white arrow). More diffuse staining was observed within the lumen encircled by the apical membrane (white *), which suggest possible transcytosis. After 2 h, basolateral binding became less uniformly distributed. B) Time-dependent uptake of LV96 revealed a less intense labeling pattern at the basolateral membranes; however, there were significant levels of intracellular puncta (white arrows). Diffuse accumulation was detected in the apical lumen by 2 h (white *), although this effect was less pronounced than for free Lacrt. C) A negative control V96 did not show significant uptake into LGACs. Scale bar: 10 μm. D) Lacrt, LV96 and V96 intensity in LGACs was quantified at three time points. Both Lacrt and LV96 exhibited significantly (****p<0.0001) higher uptake than V96. Lacrt entered LGACs to a greater extent than LV96, most obviously at 30 min (****p<0.0001). Data were analyzed by a two-way ANOVA followed by Tukey’s multiple comparisons test (n=9).

Figure 5. Lacrt-ELP fusion proteins stimulate tear secretion in NOD mice

A) Representative pictures showing tear secretion stimulated by 100 μM LV96 (5 μl) after an intra-lacrimal injection in a male NOD mouse. Blue arrow: collected tear volume after 30 min. B) Lacrimal glands injected with LV96 were collected after infusion and visualized using immunofluorescence to identify Lacrt by anti-C terminus Lacrt antibody. Green: anti-Lacrt antibody; Red: actin stained using Rho-phalloidin; Blue: nucleus stained by DAPI. Scale bar: 10 μm. C–D) Representative H&E staining images of NOD mice lacrimal gland. C) Severe lymphocytic infiltration was observed in male NOD mice LG. D) Female NOD mice exhibited normal morphology. E–F) Tear volume quantification showing significant enhancement of tear secretion by LV96 and Lacrt compared to negative V96 controls (**p<0.01, *p<0.05, n=9). Data are shown as mean ± S.D. and are compared by ANOVA followed by Tukey’s Post Hoc Test. E) Male NOD mice; F) Female NOD mice.

Figure 6. Intra-lacrimal injection of Lacrt-ELP fusion protein produces a depot

A,B) Representative confocal images showing exogenous rhodamine-labeled LV96 forms a depot in the LG of female C57BL/6 mice. LV96 was strongly retained over the 24 h time course, while free Lacrt was not observed after just 4 hours. A) Rhodamine signal alone. B) Merged combination of phase contrast and rhodamine signal. C) Quantification of average fluorescence intensity in the section of LG centered on the injection, which shows that LV96 is retained longer than free Lacrt (*p<0.05, **p<0.01). D&E) After injection, depots of LV96 maintained a lower concentration (****<0.0001) at a distance of 300 µm from the depot center; however, this fluorescence was greater than that detected either in surrounding or untreated acini. Scale bar: 100 μm.