Analysis of the human tissue-specific expression by genome-wide integration of transcriptomics and antibody-based proteomics

Abstract

Global classification of the human proteins with regards to spatial expression patterns across organs and tissues is important for studies of human biology and disease. Here, we used a quantitative transcriptomics analysis (RNA-Seq) to classify the tissue-specific expression of genes across a representative set of all major human organs and tissues and combined this analysis with antibody-based profiling of the same tissues. To present the data, we launch a new version of the Human Protein Atlas that integrates RNA and protein expression data corresponding to ∼80% of the human protein-coding genes with access to the primary data for both the RNA and the protein analysis on an individual gene level. We present a classification of all human protein-coding genes with regards to tissue-specificity and spatial expression pattern. The integrative human expression map can be used as a starting point to explore the molecular constituents of the human body.

Fig. 1.

The human tissues and organs analyzed by the transcriptomics analysis. A, The location of all 27 analyzed tissues and organs in the human body. B, The relationship between the tissues. Hierarchical clustering results showing the relationships between the 27 different tissues and organs based and a heat map showing the pairwise Spearman correlation. The numbers in parentheses show the number of replicate samples for each tissue.

Fig. 2.

The classification of all human protein-coding genes with regards to transcriptional levels in 27 tissues. A, The total number of genes with detected transcripts in each cell type using five different abundance levels for FPKM values; 1–5 FPKM (yellow), 5–20 FPKM (orange), 20–100 FPKM (light red), 100–500 FPKM (red), >500 FPKM (dark red). B, Venn-diagram showing the overlap between a total of 20,083 human Ensembl genes based on three different datasets: (1) genes detected in one or more tissues used in this study (RNA-seq), (2) the core set of genes defined by the Consensus CDS project (CCDS), (3) genes with transcript or protein evidence in UniProt. C, Piechart showing the distribution of all 20,050 genes into eight different categories based on the transcript abundance as well as the number of detected tissues.

Fig. 3.

Network plot showing the relationship of tissue enriched and group enriched genes in the various tissues and organs analyzed here. Blue circle nodes represent a group of expressed genes and are connected to tissues represented by gray circles. Dark blue nodes show the number of genes that are group enriched in up to five different tissue types (gray circles), with a minimum of three genes. The light blue nodes show the total number of highly and moderately tissue enriched genes in each tissue. The size of each blue node is related to the square root of the number of genes enriched in a particular combination of tissues.

Fig. 4.

Visualization of the RNA and protein expression levels across the 27 tissues. Examples of the presentation of expression data from the Protein Atlas, showing (A) a ubiquitously expressed gene (ribosomal protein L24, RPL24) and (B) a protein expressed specifically in one tissue (cytochrome P450, 2A13, CYP2A13).

Fig. 5.

Examples of genes with selective expression pattern in kidney, liver and testis. All examples show immunohistochemistry images from the Human Protein Atlas (www.proteinatlas.org). A, Examples of kidney-specific proteins localized to different parts of the human kidney (glomerus, proximal and distal tubuli). B, Examples of liver-specific enzymes with a gradient-like expression in hepatocytes from the central vein to the portal zone. C, Examples of testis specific proteins localized to the various cell types representing the different phases of the spermatogenesis.

Fig. 6.

Example of differential splicing in lung and stomach for the gene CLDN18. A, The exon structure (with introns scaled down 20-fold) of two different splice variants of CLDN18 are shown on top. Read coverage plots for lung (yellow) and stomach (blue) highlight the use of different initial exons in the two tissues. B, Transcript abundance (FPKM) plotted for all 95 samples, showing that each splice variant is specific for either of the two tissues. The yellow bars show expression levels for the five individual lung samples and the blue bars for the three stomach samples. Also some slight expression for one of the transcripts is detected in a single sample of heart muscle, shown by the gray bar.