Sequential in-office vitreous aspirates demonstrate vitreous matrix metalloproteinase 9 levels correlate with the amount of subretinal fluid in eyes with wet age-related macular degeneration

Abstract

Purpose: To evaluate levels of 37 native pathway proteins of the vitreous proteome from a subset of wet age-related macular degeneration (AMD) patients with and without subretinal fluid (SRF).

Methods: A total of 62 consecutive samples were aspirated from 12 patients with AMD, six who had SRF at baseline, and six who did not have SRF at any point during the study. Vitreous levels of the 37 native pathway proteins were analyzed in these patients using reverse phase protein microarray technology. At each visit, at which the 62 samples were taken, SRF and central retinal thickness were measured. These values were then compared to the relative intensity level of the 37 proteins screened.

Results: In the subset of AMD patients with SRF, the average matrix metalloproteinase 9 (MMP-9), interleukin (IL)-12, Abelson murine leukemia viral oncogene homolog 1 (cABL) Thr735, heme oxygenase-1, Musashi, platelet-derived growth factor receptor beta Tyr751 (PDGFRβ), IL-8, and BCL-2 associated death promoter (BAD) Ser112 levels in the vitreous were found to be significantly different with a 21%-82% increase in expression compared to those without SRF (p<0.0001). Within the SRF group, there was a positive correlation between the vitreous MMP-9 levels and the SRF level. MMP-9 levels in the vitreous proteome varied with the level of SRF but not retinal edema. Compared to patients without SRF, the patients with initial SRF had persistent or progressive disease.

Conclusions: This is the first prospective case series sequentially monitoring the vitreous proteome in patients with wet AMD. The results suggest that MMP-9 is a proteomic biomarker of SRF accumulation, separate from macular edema.

Figure 1

Validation of matrix metalloproteinase 9 (MMP-9) antibody (Cell Signaling Cat# 3852, Lot#2) via western blot. The molecular weight of MMP-9 is 84, 92 kDa. Lanes 1–3 go Left to Right. Lane 1: is a Magic Mark Western Standard (Cat # LC5602, Life Technologies) molecular weights ranging from 20 to 220 kDa. Lane 2, shows the validation of the MMP-9 antibody, using human endothelial cell lysates (BD Cat# 611450, Lot: 86722). Lane 3 shows is a negative control loaded with BSA (BSA; Thermo Fisher Product # 23209).

Figure 2

Matrix metalloproteinase 9 (MMP-9) expression in the vitreous is predictive of subretinal fluid (SRF) accumulation . Here it can be seen that when MMP-9 expression levels are below 1.0 patients generally have lower levels or no SRF accumulation, but when MMP-9 levels go above 1.0, patients have higher levels of SRF accumulation.

Figure 3

Matrix metalloproteinase 9 (MMP-9) expression differences between patients with SRF versus patients with no subretinal fluid (SRF). The MMP-9 mean expression level (±SD) is shown for each patient in this study. The Group 1 (Patients with SRF) graph shows that most patients have a mean MMP-9 expression level greater than 1.0. Whereas the Group 2 (Patients with NO SRF) graph shows that all patients have a mean MMP-9 expression level much lower than 1.0.

Figure 4

Matrix metalloproteinase 9 (MMP-9) tracks patient subretinal fluid (SRF) levels. The MMP-9 and SRF levels for patients 1–6 are plotted for each patient treatment visit. These graphs demonstrate how MMP-9 expression levels in the vitreous track the rise and fall of SRF in each patient over time.

Figure 5

Sensitivity and specificity of matrix metalloproteinase 9 (MMP-9) association with subretinal fluid (SRF) accumulation. Receiver operator characteristic (ROC) curve analysis demonstrates MMP-9 to have a significant ability to distinguish patients who have SRF from patients who do not have SRF accumulation, with an area under curve (AUC) of 0.8492 and a p-value of <0.0001. This data could be used to determine a threshold level of MMP-9 that would be indicative of SRF accumulation, and possible worsening of AMD pathology.