1. Characteristics of Exosomes Obtained from tear film
All participants had been examined to excluded other diseases and DM had been confirmed by HbAlc (Table1).
Table 1
the basic information of patients
Age
|
Gender (F/M)
|
HbAlc (%)
|
53.5 ± 10.5
|
13/12
|
9.05 ± 2.1
|
Transmission electron microscopy (TEM) detection showed the ultrastructurally morphologies of exosomes were cup-shaped (Fig. 1a); DLS shows the mean diameter was 120.75 ± 6.0 nm (Fig. 1b); Western blot detected the presence of known conventional exosomal markers CD81, CD63 (Fig. 1c).
Table 1The signature of diabetes patients
Figure 1 The identification of tear derived exosomes
a. The morphology of vesicles extracted from tears was determined by TEM (scale bar = 200 µm)
b. The diameter size of vesicles extracted from tears is consistent with the known size of exosomes.
c. The exosomal markers, CD6 and CD81 were detected by Western blot assay.
2. Dia-EVs deteriorate environment control induced dry eye-like ocular surface damages
Dry eye syndrome (DES) has been reported influence almost 15–33% DM patients over 65 years old cross the whole world. [20], the incidence and severity have positive related with the course of diabetes disease [21]. Tear film comprise molecules, lysozyme; metabolites play an important role in maintaining the stability of ocular surface, while tear film dysfunction, including hyperosmolarity, instability or compositions change will aggravate the feedback of corneal epithelium and surrounding tissues deterioration[22]. We asked whether Dia-EVs could be closely associated with desiccation stress-induced corneal epithelium damage, after environment humidity was controlled below 20% prior to subconjunctival injection for consecutive seven days, exosomes administration contributed to more puncate defect compared with control exosomes (Fig. 2a, c).
HE staining (Fig. 2b) showed the corneal epithelial cells were indistinctly stratified, with large cell volume, disordered arrangement and surface unsmooth. These observations make us conclude the function of our experimental exosome may be not enough for inducing diabetic dry eye, but still toxic for corneal epithelium already under stresses.
Figure 2 The different effect of tear derived exosomes on ocular surface
a. The fluorescein stain of tear derived exosomes induced corneal epithelium damage under desiccation stress.
b. HE staining of damaged corneal epithelium.
c. The statistical analysis of corneal staining (n = 8).
3. Dia-EVs delay corneal epithelial wound healing
A number of published studies have indicated the diabetic cornea exhibits a delayed wound-healing and the persistent impaired cornea epithelium defect may lead to sight-threatening complications, including ocular surface dysfunction, microbial keratitis and scarring. Various molecular, including insulin-like growth factor-binding protein 1, advanced glycation end products have been indicated to accumulation within tear [23]. Although previous studies characterize the altered corneal protein concentrations, nerve signaling have regulatory effects on wound healing process, there still no reports focus on Dia-EVs.
So we evaluate the effect of Dia-EVs on cornea wound healing in vivo by an corneal wound healing model with the method described previously. After abrasion of corneal epithelium, the cornea treated with Dia-EVs or Con-EVs, the epithelial wound was still defective at 24 hours, while completed healed in the mice after treating with Con-EVs (Fig. 3a, b), even the edema faded away (Fig. 3a, c).
Figure 3 The different recovery rate of wounded corneal epithelium under tear derived exosomes
a. The fluorescein stain in the wounded cornea.
b. The statistical analysis of wound healing rate (n = 6).
c. The statistical analysis of edema area (n = 6).
4. Dia-EVs contribute to deteriorate cornea epithelial cells proliferation and migration
Cornea epithelium defect is the most frequently exposed various damage, the rapid cornea healing not for maintain physical structure, but present as a barrier for further secondary injury, the normal process corneal healing comprises cell proliferation; migration and programmed death. While when people have already suffered diabetes, various growth factors or signal transduction have made a large alteration in tears thus delay whole cornea recovery[24].
To further clarify whether Dia-EVs contained in tears regulate corneal epithelial wound healing, HCECs were applicated separately with Dia-EVs or Con-EVs. After 24 h treatment, Dia-EVs administration resulted in a cytotoxic effect on HCEC viability, and Con-EVs has no obvious influence (Fig. 4b).
Similarly, consistent with this observation, scratch wound assay results showed exposure to Dia-EVs suppressed the migration of HCECs, the re-epithelialization of monolayers showed a time dependence (Fig. 4a, c).
Figure 4 Dia-EVs deteriorated HCEC cell proliferation and migration in vitro
a. The healing area of HCECs was quantitated by wound healing assay
b. CCK-8 assay was used to measure the proliferation of HCECs after treated with Dia-EVs or Con-EVs for 24 hours.
c. The statistical analysis of healing rate (n = 3)
5. Differential expression of miRNAs in Dia-EVs
Previously, numerous inflammatory factors, such as TNF-α, IL-4 and IL-8 were detected at higher levels in the tears of type 2 diabetic patients [25, 26], even microbiota in ocular surface get more complex in abundance and diversity distribution [13] make the different function of tear films.
To determine underlying mechanism of the different function of exosomes between health and diabetes, herein, microarray was used to determine the levels of Dia-EVs microRNAs.
Based on fold change greater than 1, a set of 12 small RNAs from Dia-EVs and Con-EVs was assessed and identified as differentially expressed. Among them, the concentrations of miR-4443 and significantly increased compared with control exosomes, whereas three miRNAs were suppressed (Table2). The heat map generated according 12 differentially expressed miRNAs illustrated the distinguishable miRNAs expression profile of the samples (Fig. 5a). The volcano plot of miRNAs those expression fold change > 2.0 was created to show the quick visual identification of miRNAs (Fig. 5b). The hierarchical cluster is constructed based on the relative abundance of miRNAs demonstrated that the DM and control groups were different (Fig. 5c).
Table 2
Total significantly differentially expressed miRNAs between Dia-EVs VS Con-EVs
Systematic Name
|
P values
|
foldchange
|
Regulation
|
hsa-miR-1254
hsa-miR-208a-5p
hsa-miR-3156-5p
hsa-miR-423-5p
hsa-miR-4443
hsa-miR-4687-3p
hsa-miR-4707-3p
hsa-miR-5585-3p
hsa-miR-6073
hsa-miR-6087
hsa-miR-671-5p
hsa-miR-6847-5p
|
0.011064416
0.006692563
0.012227109
0.037746593
0.001135519
0.045887284
0.035154498
0.036176931
0.029700396
0.020402541
0.040076376
0.011340659
|
0.319579453
0.445251528
2.040317703
0.41132038
0.335312367
0.486188256
0.489150825
0.270891312
0.432553573
0.489614159
9.933437101
0.379851043
|
up
up
down
up
up
up
up
up
up
up
up
up
|
Table 2 the different expressed genes compared between Dia-EVs and Con-EVs treatments
Figure 5 the signature of different expressed miRNAs
a. The heat map of miRNAs showed the different expression within different samples.
b. the volcano plot showed the different miRNAs change according the fold change.
c. the hierarchical cluster showed the distance contained within different expressed miRNAs.
6.The KEGG and GO analysis of different express miRNA
Numerous studies showed that miRNAs interact with the 3’ UTR of target mRNAs and induce their degradation[27], miRNAs are critical for physiological development and participate in a variety of biological processes and expression alteration are involved in many human diseases[28]
We performed bioinformatics analysis, including Gene ontology enrichment and pathway analyses with their identified target genes to explore potential function associated with differentially expressed miRNAs.
After mapping the differentially expressed miRNAs to their target genes, the top 20 most significant KEGG terms were found (Fig. 6a). glycosphingolipid biosynthesis is the highest enrichment one among all of these pathways. Functional annotation of these miRNAs was speculated by GO function analysis (Fig. 6b).
Figure 6 the bioinformatics analysis of target genes to the identified miRNAs.
a. The KEGG analysis showed the target gened participate in AMPK, Ras which related to apoptosis pathway.
b. All different miRNAs were annotated by GO analysis.