Altered miRNA expression in canine retinas during normal development and in models of retinal degeneration

Abstract

Background: Although more than 246 loci/genes are associated with inherited retinal diseases, the mechanistic events that link genetic mutations to photoreceptor cell death are poorly understood. miRNAs play a relevant role during retinal development and disease. Thus, as a first step in characterizing miRNA involvement during disease expression and progression, we examined miRNAs expression changes in normal retinal development and in four canine models of retinal degenerative disease.

Results: The initial microarray analysis showed that 50 miRNAs were differentially expressed (DE) early (3 vs. 7 wks) in normal retina development, while only 2 were DE between 7 and 16 wks, when the dog retina is fully mature. miRNA expression profiles were similar between dogs affected with xlpra2, an early-onset retinal disease caused by a microdeletion in RPGRORF15, and normal dogs early in development (3 wks) and at the peak of photoreceptor death (7 wks), when only 2 miRNAs were DE. However, the expression varied much more markedly during the chronic cell death stage at 16 wks (118 up-/55 down-regulated miRNAs). Functional analyses indicated that these DE miRNAs are associated with an increased inflammatory response, as well as cell death/survival. qRT-PCR of selected apoptosis-related miRNAs ("apoptomirs") confirmed the microarray results in xlpra2, and extended the analysis to the early-onset retinal diseases rcd1 (PDE6B-mutation) and erd (STK38L-mutation), as well as the slowly progressing prcd (PRCD-mutation). The results showed up-regulation of anti-apoptotic (miR-9, -19a, -20, -21, -29b, -146a, -155, -221) and down-regulation of pro-apoptotic (miR-122, -129) apoptomirs in the early-onset diseases and, with few exceptions, also in the prcd-mutants.

Conclusions: Our results suggest that apoptomirs might be expressed by diseased retinas in an attempt to counteract the degenerative process. The pattern of expression in diseased retinas mirrored the morphology and cell death kinetics previously described for these diseases. This study suggests that common miRNA regulatory mechanisms may be involved in retinal degeneration processes and provides attractive opportunities for the development of novel miRNA-based therapies to delay the progression of the degenerative process.

Figure 1

Heat map representation of all miRNAs present on the microarray. The heat map illustrates the expression differences of all miRNAs on the microarray between xlpra2-mutants vs. normals at the 3 tested ages (3, 7, and 16 wks). miRNAs are listed from the highest to the lowest fold change difference at 16 wks. The x-axis shows the ages, while the y-axis displays the different miRNAs. The map contains log2 intensity/fold change ratios that are color coded with red corresponding to up-regulation and blue to down-regulation.

Figure 2

Heat map representation over time of the DE miRNAs at 16 wks. The heat map illustrates the fold change differences identified by microarray analysis between xlpra2-mutant and normal retinas at all three ages (3, 7, and 16 wks) for the up-regulated (A) and down-regulated (B) miRNAs at 16 wks. miR-183 and miR-122 were also tested by qRT-PCR, therefore they were also included in spite of their expression differences being not statistically significant. miRNAs are listed from the highest to the lowest fold change difference at 16 wks. The x-axis shows the time points, while the y-axis displays the DE miRNAs at 16 wks. The map contains log2 intensity/fold change ratios that are color coded with red corresponding to up-regulation and blue to down-regulation. Apoptomirs that were selected for qRT-PCR analysis are boxed and marked in bold.

Figure 3

Expression changes of DE apoptomirs during development. Significant FC differences as measured by qRT-PCR are reported at 7 vs. 3, 16 vs. 7, and 16 vs. 3 wks in normal, xlpra2, and rcd1; at 8.3/9.9 vs. 6.4, 11.9/14.1 vs. 8.3/9.9, and 11.9/14.1 vs. 6.4 wks in erd; and at 24 vs. 10 wks in inferior and superior retinas of prcd-mutants. Bars show SD of biological triplicates. Only DE apoptomirs were displayed.

Figure 4

Expression changes of apoptomirs between mutant and normal retinas at different ages. FC differences of apoptomirs as measured by qRT-PCR are shown at 3, 5, 7, 16 wks in xlpra2 and rcd1, as well as 6.4, 8.3-9.9, 11.9-14.1 wks in erd compared to normals. Selected apoptomirs belong to different functional groups: A) anti-apoptotic that were DE at later ages: miR-9, -19a, -20; B) anti-apoptotic that were DE during the course of disease: miR-21, -155, -221; C) pro-apoptotic: miR-122, -129; D) dual properties, anti- and pro-apoptotic: miR-29b, -146a. An asterisk indicates statistical significance (p < 0.05; FC > +/-2); bars show SD of biological triplicates. Results for miR-183 are not illustrated because they were not significant in any of the diseases and ages tested.

Figure 5

Expression changes of apoptomirs in prcd vs. normal retinas. Values that significantly differ as measured by qRT-PCR are indicated with an asterisk (*: p < 0.05; FC > +/-2). A) FC differences between either superior or inferior retinas in 10 wks old prcd vs. age and retinal location matched normals. B) FC differences between either superior or inferior retinas in 24 wks old prcd vs. 16 wks old entire normal retinas.

Figure 6

Expression changes of apoptomirs between mutant and normal RPE/choroids. FC differences of apoptomirs as measured by qRT-PCR are shown between xlpra2, rcd1, and erd-mutants compared to normals at 7 wks of age. Bars show SD of biological triplicates and values that significantly differ are indicated with an asterisk (*: p < 0.05; FC > +/-2).