Age related changes in hyaluronan expression leads to Meibomian gland dysfunction

Abstract

The prevalence of dry eye disease (DED) ranges from ∼5 to 50 % and its associated symptoms decrease productivity and reduce the quality of life. Approximately 85 % of all DED cases are caused by Meibomian gland dysfunction (MGD). As humans and mice age, their Meibomian glands (MGs) undergo age-related changes resulting in age related-MGD (ARMGD). The precise cause of ARMGD remains elusive, which makes developing therapies extremely challenging. We previously demonstrated that a hyaluronan (HA)-rich matrix exists surrounding the MG, regulating MG morphogenesis and homeostasis. Herein, we investigated whether changes to the HA matrix in the MG throughout life contributes towards ARMGD, and whether altering this HA matrix can prevent ARMGD. For such, HA synthase (Has) knockout mice were aged and compared to age matched wild type (wt) mice. MG morphology, lipid production, PPARγ expression, basal cell proliferation, stem cells, presence of atrophic glands and MG dropout were analyzed at 8 weeks, 6 months, 1 year and 2 years of age and correlated with the composition of the HA matrix. We found that as mice age, there is a loss of HA expression in and surrounding the MGs of wt mice, while, in contrast, Has1^-/-Has3^-/- mice present a significant increase in HA expression through Has2 upregulation. At 1 year, Has1^-/-Has3^-/- mice present significantly enlarged MGs, compared to age-matched wt mice and compared to all adult mice. Thus, Has1^-/-Has3^-/- mice continue to develop new glandular tissue as they age, instead of suffering MG atrophy. At 2 years, Has1^-/-Has3^-/- mice continue to present significantly larger MGs compared to age-matched wt mice. Has1^-/-Has3^-/- mice present increased lipid production, increased PPARγ expression and an increase in the number of proliferating cells when compared to wt mice at all-time points analyzed. Taken together, our data shows that a loss of the HA matrix surrounding the MG as mice age contributes towards ARMGD, and increasing Has2 expression, and consequently HA levels, prevents ARMGD in mice.

Keywords: Aging; Hyaluronan; Meibomian gland; Ocular surface; Stem cells.

Fig. 1.

Has1^−/−; Has3^−/− mice present an increase in MG area when compared to wt mice. Eyelids were collected from wt and Has1^−/−; Has3^−/− mice at (A) 8 weeks, (C) 1 year and (E) 2 years, and immediately mounted as flat mounts and imaged under a stereomicroscope (Discovery.V12, ZEISS, Oberkochen, Germany). The outer perimeter of the MG was demarcated with a white dashed line. The total MG area was calculated using ImageJ2 version 2.3.0/1.53q and Ilastik version 1.4.0 and data represented in a bar graph for MGs at (B) 8 weeks, (D) 1 year and (F) 2 years. * ≤ 0.05 and ** ≤ 0.01 when comparing wt and Has1^−/−; Has3^−/− mice.

Fig. 2.

Has1^−/−; Has3^−/− mice present an increase in ORO staining when compared to wt mice. (A) Eyelids were collected from wt and Has1^−/−; Has3^−/− mice at (A) 8 weeks, (C) 6 months, (E) 1 year and (G) 2 years, fixed and stained with ORO. Thereafter, eyelids were mounted and imaged under a stereomicroscope. The outer perimeter of the MG was demarcated with a white dashed line. (B) The mean pixel intensity was calculated using ImageJ2 version 2.3.0/1.53q and Ilastik version 1.4.0 and data represented in a bar graph for MGs at (B) 8 weeks, (D) 6 months, (F) 1 year and (H) 2 years. * ≤ 0.05 when comparing wt and Has1^−/−; Has3^−/− mice.

Fig. 3.

There is a change in the expression and distribution of HA surrounding the MG and in the tarsal plate as mice age. Eyelids were collected from wt and Has1^−/−; Has3^−/− mice at 8 weeks, 1 year and 2 years, and excess tarsal plate trimmed away and MGs processed for real time PCR to evaluate the expression of Has1 (A), Has2 (B), and Has3 (C). HA was isolated from the MGs of wt and Has1^−/−; Has3^−/− mice at 8 weeks (Adult) (D), 1 year (Aging) (G) and 2 years (Aged) (J) and subjected to molecular weight characterization via solid-state nanopore sensing. The overlaid graphs for molecular weight characterization via solid-state nanopore sensing of the wt mice at different ages are presented in (E), and for the Has1^−/−; Has3^−/− mice at different ages in (H). Box plots of the molecular weight characterization via solid-state nanopore sensing of wt and Has1^−/−; Has3^−/− mice at 8 weeks (F), 1 year (I) and 2 years (K) are also presented. Eyelids of aged wt (L) and Has1^−/−; Has3^−/− mice (M) were processed for whole mount staining of HA (red) and the distribution of HA analyzed under a confocal microscope using Aryscan. An image of the entire eyelid under a stereomicroscope is presented and areas imaged under the confocal microscope demarcated. For wt mice, the distribution of HA was analyze in healthy MGs (Li”), MGs undergoing atrophy (Lii”) and in areas after MG dropout with altered settings to reduce the intensity of HA staining (Liii”). For Has1^−/−; Has3^−/− mice, the distribution of HA was analyze in healthy MGs (Mi). The distribution of HA was also analyzed in the aged tarsal plate of wt and Has1^−/−; Has3^−/− mice (N). * ≤ 0.05 when comparing wt and Has1^−/−; Has3^−/− mice, and φ ≤ 0.05 when comparing adult to aging or aged mice of the same genotype.

Fig. 4.

Has1^−/−; Has3^−/− mice present an increase in PPARγ staining compared to wt mice. Eyelids were collected from wt and Has1^−/−; Has3^−/− mice at 8 weeks (adult), 1 year (aging) and 2 years (aged), fixed, processed for cryosectioning and stained for PPARγ (green) and HA (red). Nuclei were counter stained with DAPI (blue). Entire eyelids of adult (A) and aged (B) wt and Has1^−/−; Has3^−/− mice were scanned under the confocal microscope using the tiling mode. The outer perimeter of the MGs were demarcated with a white dashed line. Scale bar represents 100 μm. (C) Higher magnification images of MGs of adult and aging wt and Has1^−/−; Has3^−/−. Scale bar represents 100 μm. (D) Higher magnification images of MGs of adult and aged wt (atrophic) and Has1^−/−; Has3^−/− (healthy) to highlight the differences in PPARγ expression in atrophic glands. Scale bar represents 100 μm. (E) The expression of PPARg was analyzed in the MGs of wt and Has1^−/−; Has3^−/− mice at 8 weeks, 1 year and 2 years. * ≤ 0.05 when comparing wt and Has1^−/−; Has3^−/− mice, and ς ≤ 0.05 when comparing adult to aging or aged mice of the same genotype.

Fig. 5.

Has1^−/−; Has3^−/− mice present an increase in LRIG1 staining compared to wt mice. Eyelids were collected from wt and Has1^−/−; Has3^−/− mice at 8 weeks (Adult)and 2 years (Aged), fixed, processed for cryosectioning and stained for LRIG 1 (green) and HA (red). Nuclei were counter stained with DAPI (blue). (A) Entire eyelids of adult wt and Has1^−/−; Has3^−/− mice were scanned under the confocal microscope using the tiling mode. The outer perimeter of the MGs were demarcated with a white dashed line. Scale bar represents 100 μm. (B) Higher magnification images of MGs from adult (top panel) and aged (bottom panel) wt and Has1^−/−; Has3^−/− mice. White arrows indicate clusters of LRIG1⁺ cells. Scale bar represents 100 μm. (C) The expression of LRIG1 was analyzed in the MGs of wt and Has1^−/−; Has3^−/− mice at 8 weeks, 1 year and 2 years. * ≤ 0.05 when comparing wt and Has1^−/−; Has3^−/− mice, and ς ≤ 0.05 when comparing adult to aging or aged mice of the same genotype. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 6.

Has1^−/−; Has3^−/− mice present an increase in basal cell proliferation when compared to wt mice. (A) Eyelids were collected from wt and Has1^−/−; Has3^−/− mice at 8 weeks (adult), 1 year (aging), and 2 years (aged), fixed, processed for cryosectioning and stained for Ki67 (green) and nuclei were counter stained with DAPI (blue). Entire eyelids of adult wt and Has1^−/−; Has3^−/− mice were scanned under the confocal microscope using the tiling mode. The outer perimeter of the MGs were demarcated with a white dashed line. Scale bar represents 100 μm. (B) The number of Ki67⁺ cells were counted per acini and data represented as a graph. * ≤ 0.05 when comparing wt and Has1^−/−; Has3^−/− mice, and ς ≤ 0.05 when comparing adult to aging or aged mice of the same genotype. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 7.

Inhibition of HA synthesis with 4MU induces MG atrophy. (A) 4MU was injected into the eyelid of Has1^−/−; Has3^−/− mice every week from 10 months to 1 year and eyelids collected and compared to the eyelids of wt and Has1^−/−; Has3^−/− mice treated with PBS. Eyelids were immediately mounted as flat mounts and imaged under a stereomicroscope and thereafter fixed and processed for whole mount staining. The outer perimeter of the MG was demarcated with a white dashed line. (B) The total MG area was calculated using ImageJ2 version 2.3.0/1.53q and Ilastik version 1.4.0 and data represented in a bar graph. * ≤ 0.05 when comparing wt and Has1^−/−; Has3^−/− mice. (C) Whole-mounted eyelids were stained for HA (red) and counter stained with DAPI (blue) and imaged under a confocal microscope using Aryscan. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 8.

Cartoon representation of the distribution of HA within the tarsal plate and MG of healthy MGs (homeostasis), MGs undergoing atrophy and an area following MG dropout.