Deep Sequencing in Microdissected Renal Tubules Identifies Nephron Segment-Specific Transcriptomes

Abstract

The function of each renal tubule segment depends on the genes expressed therein. High-throughput methods used for global profiling of gene expression in unique cell types have shown low sensitivity and high false positivity, thereby limiting the usefulness of these methods in transcriptomic research. However, deep sequencing of RNA species (RNA-seq) achieves highly sensitive and quantitative transcriptomic profiling by sequencing RNAs in a massive, parallel manner. Here, we used RNA-seq coupled with classic renal tubule microdissection to comprehensively profile gene expression in each of 14 renal tubule segments from the proximal tubule through the inner medullary collecting duct of rat kidneys. Polyadenylated mRNAs were captured by oligo-dT primers and processed into adapter-ligated cDNA libraries that were sequenced using an Illumina platform. Transcriptomes were identified to a median depth of 8261 genes in microdissected renal tubule samples (105 replicates in total) and glomeruli (5 replicates). Manual microdissection allowed a high degree of sample purity, which was evidenced by the observed distributions of well established cell-specific markers. The main product of this work is an extensive database of gene expression along the nephron provided as a publicly accessible webpage (https://helixweb.nih.gov/ESBL/Database/NephronRNAseq/index.html). The data also provide genome-wide maps of alternative exon usage and polyadenylation sites in the kidney. We illustrate the use of the data by profiling transcription factor expression along the renal tubule and mapping metabolic pathways.

Keywords: collecting ducts; nephron; transcription factors; transcriptional profiling.

Figure 1.

The distribution of RPKM values for several marker genes match prior knowledge, documenting the precision of the technique and the accuracy of segment identification. (A) Nomenclature for renal tubule segments. (B) Median RPKMs for segment-specific markers. Aqp1, aquaporin-1; Aqp2, aquaporin-2; Slc4a1 (AE1), anion exchanger 1; Slc12a1 (NKCC2), bumetanide–sensitive Na⁺-K⁺-2Cl⁻ cotransporter; Slc12a3 (NCC), thiazide–sensitive Na⁺-Cl⁻ cotransporter; Slc14a2 (UT-A), urea transporter A; Slc22a6 (OAT1), organic anion transporter 1; Slc34a1 (NaPi-IIa), sodium-phosphate cotransporter IIa.

Figure 2.

A dendrogram created by hierarchical clustering of 105 replicates of microdissected renal tubule segments demonstrates reproducibility among replicates from the same segment. (Supplemental Figure 3 shows the full heat map representation). Replicates from 14 different renal tubule segments were found to form six distinct clusters according to their anatomic and functional organization. Terminology is the same as in Figure 1A.

Figure 3.

Mapping of deep sequencing reads to gene structure reveals alternative exon usage in different renal tubule segments for some genes. (A) Alternative initial (5′ end) exon usage in the γ-subunit of Na⁺-K⁺ ATPase (Fxyd2). (B) Alternative terminal (3′ end) exon usage in glutaminase (Gls).

Figure 4.

Transcription factors show distinct patterns of expression along the renal tubule. The distributions can be mapped to general regions (proximal region [S1, S2, and S3], thin-limb region, thick-limb/DCT region [mTAL, cTAL, and DCT], and collecting duct region [CNT, CCD, OMCD, and IMCD]). (A) Transcription factors specific to a renal tubule region. (B) Transcription factors expressed in two contiguous regions. (C) Transcription factors with bimodal pattern of expression. (D) Transcription factors with bimodal pattern but expression only in the inner medulla.

Figure 5.

Genes coding for metabolic enzymes are expressed in patterns along the renal tubule that correlate with function. Expression of metabolic enzymes critical for specific metabolic pathways are presented as a heat map showing specific metabolic pathways.

Figure 6.

Some transcripts are selectively expressed in developmentally or anatomically defined renal elements. (A) Transcripts selectively expressed in nephron- and collecting duct–derived segments. (B) Transcripts selectively expressed in the cortex and the medulla.

Figure 7.

Forty-five medulla-enriched genes are associated with the GO cellular component term ‘extracellular region’.