Application of small RNA sequencing to identify microRNAs in acute kidney injury and fibrosis

Abstract

Establishing a microRNA (miRNA) expression profile in affected tissues provides an important foundation for the discovery of miRNAs involved in the development or progression of pathologic conditions. We conducted small RNA sequencing to generate a temporal profile of miRNA expression in the kidneys using a mouse model of folic acid-induced (250mg/kgi.p.) kidney injury and fibrosis. From the 103 miRNAs that were differentially expressed over the time course (>2-fold, p<0.05), we chose to further investigate miR-18a-5p, which is expressed during the acute stage of the injury; miR-132-3p, which is upregulated during transition between acute and fibrotic injury; and miR-146b-5p, which is highly expressed at the peak of fibrosis. Using qRT-PCR, we confirmed the increased expression of these candidate miRNAs in the folic acid model as well as in other established mouse models of acute injury (ischemia/reperfusion injury) and fibrosis (unilateral ureteral obstruction). In situ hybridization confirmed high expression of miR-18a-5p, miR-132-3p and miR-146b-5p throughout the kidney cortex in mice and humans with severe kidney injury or fibrosis. When primary human proximal tubular epithelial cells were treated with model nephrotoxicants such as cadmium chloride (CdCl₂), arsenic trioxide, aristolochic acid (AA), potassium dichromate (K₂Cr₂O₇) and cisplatin, miRNA-132-3p was upregulated 4.3-fold after AA treatment and 1.5-fold after K₂Cr₂O₇ and CdCl₂ treatment. These results demonstrate the application of temporal small RNA sequencing to identify miR-18a, miR-132 and miR-146b as differentially expressed miRNAs during distinct phases of kidney injury and fibrosis progression.

Keywords: Folic Acid; Kidney Fibrosis; Kidney Toxicity; Small RNA Sequencing; miRNAs.

Figure 1

Kidney phenotype of mice injected with folic acid. The levels of blood urea nitrogen (BUN; A) were measured at 0, 1, 2, 3, 7, 14, and 28 days following the administration of folic acid. Kim-1, collagen, and fibronectin mRNA from kidney lysates was assessed by qRT-PCR, normalized to Gapdh, and are shown as fold change relative to the untreated group (B, C, and D, respectively). Data is represented as mean ± SEM and *p < 0.05 in comparison to the untreated group (n = 5 mice/group). Histological features of the kidneys were assessed in Periodic acid-Schiff (PAS) or Masson’s Trichrome (MTS) stained sections, and representative images from each time point are shown (E; PAS scale bar = 50 μm; MTS scale bar = 100 μm).

Figure 2

Small RNA sequencing of folic acid treated kidneys. Using miRDeep2, reads were trimmed, mapped to the genome (mm9), and annotated. The number of reads passing through each stage of the analysis is shown (A; the range for all samples shown in parentheses). The data for each sample was normalized to the total number of reads for that sample and expressed as counts/million. Each miRNA was assigned to one of the indicated count ranges based on the average counts/million from all samples, and the number of miRNAs within each count range (per million reads) is shown (B). Using the data from the untreated mice, the percentage of reads attributable to each miRNA was determined, and the miRNAs that comprised more than 5% of the total are represented (C). A two dimensional scaling of the 21 samples based on the 100 most heterogenous miRNAs with expression of at least 5 counts/million was prepared using the edgeR function ‘plotMDS’ (D).

Figure 3

Differentially expressed miRNAs during folic acid injury and fibrosis. Using edgeR, the small RNA sequencing data was analyzed to identify miRNAs that demonstrated a significant change in expression during at least one time point following folic acid administration. The expression of these miRNAs is represented as the average fold change in comparison to the untreated group (n = 3 mice/group).

Figure 4

Expression of miR-18a-5p, miR-132-3p, and miR-146b-5p in various models of kidney injury. Three miRNAs were chosen from the small RNA sequencing data to represent the acute (18a-5p), intermediate (132-3p), and fibrotic (146b-5p) stages of injury. The expression of these miRNAs was assessed in the folic acid model (FA; A–C), as well as unilateral ureteral obstruction (UUO; D–F), which causes a severe fibrotic injury, and bilateral ischemia reperfusion injury (IRI; G–I), which induces acute tubular injury. Data was normalized to mmu-let-7f-5p and is represented as mean ± SEM. *p < 0.05 in comparison to the untreated control group for each experiment (n = 5–6 mice/group). The data for the CoK group was derived from the contralateral kidney, which was left untouched during the UUO surgery.

Figure 5

In situ hybridization analysis of miR-18a-5p, miR-132-3p and miR-146b-5p in the kidneys of folic acid-treated mice. Expression of miR-18a-5p (A–C), miR-132-3p (D–F) and miR-146b-5p (G–I) was investigated in mice at day 0 (Normal), as well as 2 and 14 days following the administration of 250 mg/kg folic acid (FA). A scrambled (Scr; J–L) and U6 probe (M–O) were used as negative and positive controls, respectively. The development of purple staining indicates detection of the ISH probe. Black arrows represent injured tubules and blue arrows represent normal tubules and glomeruli. Representative images from each group are shown (scale bar = 50 μm; n = 3 mice/group).

Figure 6

In situ hybridization analysis of miR-18a-5p, miR-132-3p and miR-146b-5p in human kidney disease. The expression of miR-18a-5p (A–C), miR-132-3p (D–F) and miR-146b-5p expression (G–I) was investigated in the kidneys of patients with no evidence of kidney disease, patients with AKI due to renal allograft rejection, or patients with severe kidney fibrosis due to CKD. A scrambled (J–L) and U6 probe (M–O) were used as negative and positive controls, respectively. The development of purple staining indicates detection of the ISH probe. Asterisks represent tubular distension or hyaline casts. Representative images from each group are shown (scale bar = 50 μm; n = 5 samples/group).

Figure 7

Expression of miR-18a-5p, miR-132-3p, and miR-146b-5p in vitro following injury with various nephrotoxicants. The expression of these miRNAs was measured in human primary proximal tubular epithelial cells at 6h (A) and 24h (B) following treatment with 100μM aristolochic acid (AA), 50μM arsenic trioxide (As₂O₃), 10μM potassium dichromate (K₂Cr₂O₇), 25μM cadmium chloride (CdCl₂) and 40μM cisplatin (CDDP). Data was normalized to let-7f-5p and is represented as log₂ of mean ± SEM. *p < 0.05 in comparison to DMSO-treated cells for each experiment (n = 3 replicates/treatment).