Patterns of Unwanted Biological and Technical Expression Variation Among 49 Human Tissues.

Tissue gene expression studies are impacted by biological and technical sources of variation, which can be broadly classified into wanted and unwanted variation. The latter, if not addressed, results in misleading biological conclusions. Methods have been proposed to reduce unwanted variation, such as normalization and batch correction. A more accurate understanding of all causes of variation could significantly improve the ability of these methods to remove unwanted variation while retaining variation corresponding to the biological question of interest. We used 17,282 samples from 49 human tissues in the Genotype-Tissue Expression data set (v8) to investigate patterns and causes of expression variation. Transcript expression was transformed to z-scores, and only the most variable 2% of transcripts were evaluated and clustered based on coexpression patterns. Clustered gene sets were assigned to different biological or technical causes based on histologic appearances and metadata elements. We identified 522 variable transcript clusters (median: 11 per tissue) among the samples. Of these, 63% were confidently explained, 16% were likely explained, 7% were low confidence explanations, and 14% had no clear cause. Histologic analysis annotated 46 clusters. Other common causes of variability included sex, sequencing contamination, immunoglobulin diversity, and compositional tissue differences. Less common biological causes included death interval (Hardy score), disease status, and age. Technical causes included blood draw timing and harvesting differences. Many of the causes of variation in bulk tissue expression were identifiable in the Tabula Sapiens data set of single-cell expression. This is among the largest explorations of the underlying sources of tissue expression variation. It uncovered expected and unexpected causes of variable gene expression and demonstrated the utility of matched histologic specimens. It further demonstrated the value of acquiring meaningful tissue harvesting metadata elements to use for improved normalization, batch correction, and analysis of both bulk and single-cell RNA-seq data.

Beta blockers reduce uterine fibroid incidence in hypertensive women.

OBJECTIVE: Is prior beta blocker (BB) use associated with reduced odds of the clinical incidence of leiomyomas? WHAT IS KNOWN ALREADY: In-vitro and in-vivo evidence has supported the role of beta receptor blockade in reducing leiomyoma cell proliferation and growth. However, no population-based study to date has investigated this potential association. STUDY DESIGN, SIZE, DURATION: A nested case-control study was conducted in a population of women aged 18-65 with arterial hypertension (n = 699,966). Cases (n = 18,918) with a leiomyoma diagnosis were matched to controls (n = 681,048) with no such diagnosis at a 1:36 ratio by age and region of origin within the United States. PARTICIPANTS/MATERIALS, SETTING, METHODS: This population was assembled from the Truven Health MarketScan® Research Database, which includes health insurance claims from January 1st, 2012 to December 31st, 2017. Prior use of BB wasdetermined fromoutpatient drug claims and leiomyoma development was indicated by a first-time diagnosis code. We conducted a conditional logistic regression to determine the odds of uterine fibroid development in women with prior use of BB compared to women with no such history. We then conducted subset analyses, stratifying the women by age group and by type of BB. RESULTS: Women on a BB experienced 15% reduced odds of developing clinically recognized leiomyoma compared to non-users (OR 0.85, 95% CI 0.76-0.94). This association was significant for the 30-39 age group (OR 0.61, 95% CI 0.40-0.93) but no other age group. Of the BBs, propranolol (OR 0.58, 95% CI 0.36-95) demonstrated a significant association with reduced leiomyoma incidence and metoprolol (OR 0.82, 95% CI 0.70-0.97) was associated with lower uterine fibroid incidence after adjustment for comorbidities. CONCLUSIONS: Hypertensive women with prior BB use experienced reduced odds of developing clinically recognized leiomyoma compared to non-users. A key predisposing risk factor for uterine leiomyoma is elevated blood pressure. Thus, the results of this analysis may have clinical relevance to women with hypertension, as the use of this drug may introduce a dual benefit of managing hypertension as well as curbing an increased risk of leiomyomas.

Patterns of unwanted biological and technical expression variation across 49 human tissues.

All tissue-based gene expression studies are impacted by biological and technical sources of variation. Numerous methods are used to normalize and batch correct these datasets. A more accurate understanding of all causes of variation could further optimize these approaches. We used 17,282 samples from 49 tissues in the Genotype Tissue Expression (GTEx) dataset (v8) to investigate patterns and causes of expression variation. Transcript expression was normalized to Z-scores and only the most variable 2% of transcripts were evaluated and clustered based on co-expression patterns. Clustered gene sets were solved to different biological or technical causes related to metadata elements and histologic images. We identified 522 variable transcript clusters (median 11 per tissue) across the samples. Of these, 64% were confidently explained, 15% were likely explained, 7% were low confidence explanations and 14% had no clear cause. Common causes included sex, sequencing contamination, immunoglobulin diversity, and compositional tissue differences. Less common biological causes included death interval (Hardy score), muscle atrophy, diabetes status, and menopause. Technical causes included brain pH and harvesting differences. Many of the causes of variation in bulk tissue expression were identifiable in the Tabula Sapiens dataset of single cell expression. This is the largest exploration of the underlying sources of tissue expression variation. It uncovered expected and unexpected causes of variable gene expression. These identified sources of variation will inform which metadata to acquire with tissue harvesting and can be used to improve normalization, batch correction, and analysis of both bulk and single cell RNA-seq data.

Tissue, age, sex, and disease patterns of matrisome expression in GTEx transcriptome data.

The extracellular matrix (ECM) has historically been explored through proteomic methods. Whether or not global transcriptomics can yield meaningful information on the human matrisome is unknown. Gene expression data from 17,382 samples across 52 tissues, were obtained from the Genotype-Tissue Expression (GTEx) project. Additional datasets were obtained from The Cancer Genome Atlas (TCGA) program and the Gene Expression Omnibus for comparisons. Gene expression levels generally matched proteome-derived matrisome expression patterns. Further, matrisome gene expression properly clustered tissue types, with some matrisome genes including SERPIN family members having tissue-restricted expression patterns. Deeper analyses revealed 382 gene transcripts varied by age and 315 varied by sex in at least one tissue, with expression correlating with digitally imaged histologic tissue features. A comparison of TCGA tumor, TCGA adjacent normal and GTEx normal tissues demonstrated robustness of the GTEx samples as a generalized matrix control, while also determining a common primary tumor matrisome. Additionally, GTEx tissues served as a useful non-diseased control in a separate study of idiopathic pulmonary fibrosis (IPF) matrix changes, while identifying 22 matrix genes upregulated in IPF. Altogether, these findings indicate that the transcriptome, in general, and GTEx in particular, has value in understanding the state of organ ECM.

Single cell RNA-seq analysis of the flexor digitorum brevis mouse myofibers.

BACKGROUND: Skeletal muscle myofibers can be separated into functionally distinct cell types that differ in gene and protein expression. Current single cell expression data is generally based upon single nucleus RNA, rather than whole myofiber material. We examined if a whole-cell flow sorting approach could be applied to perform single cell RNA-seq (scRNA-seq) in a single muscle type. METHODS: We performed deep, whole cell, scRNA-seq on intact and fragmented skeletal myofibers from the mouse fast-twitch flexor digitorum brevis muscle utilizing a flow-gated method of large cell isolation. We performed deep sequencing of 763 intact and fragmented myofibers. RESULTS: Quality control metrics across the different gates indicated only 171 of these cells were optimal, with a median read count of 239,252 and an average of 12,098 transcripts per cell. scRNA-seq identified three clusters of myofibers (a slow/fast 2A cluster and two fast 2X clusters). Comparison to a public skeletal nuclear RNA-seq dataset demonstrated a diversity in transcript abundance by method. RISH validated multiple genes across fast and slow twitch skeletal muscle types. CONCLUSION: This study introduces and validates a method to isolate intact skeletal muscle myofibers to generate deep expression patterns and expands the known repertoire of fiber-type-specific genes.

Spatial Proteomic Approach to Characterize Skeletal Muscle Myofibers.

Skeletal muscle myofibers have differential protein expression resulting in functionally distinct slow- and fast-twitch types. While certain protein classes are well-characterized, the depth of all proteins involved in this process is unknown. We utilized the Human Protein Atlas (HPA) and the HPASubC tool to classify mosaic expression patterns of staining across 49,600 unique tissue microarray (TMA) images using a visual proteomic approach. We identified 2164 proteins with potential mosaic expression, of which 1605 were categorized as "likely" or "real." This list included both well-known fiber-type-specific and novel proteins. A comparison of the 1605 mosaic proteins with a mass spectrometry (MS)-derived proteomic dataset of single human muscle fibers led to the assignment of 111 proteins to fiber types. We additionally used a multiplexed immunohistochemistry approach, a multiplexed RNA-ISH approach, and STRING v11 to further assign or suggest fiber types of newly characterized mosaic proteins. This visual proteomic analysis of mature skeletal muscle myofibers greatly expands the known repertoire of twitch-type-specific proteins.

Angiotensin-Converting Enzyme Inhibitors Reduce Uterine Fibroid Incidence in Hypertensive Women.

CONTEXT: In vitro and in vivo evidence has supported the role of angiotensin II blockade in reducing leiomyoma cell proliferation and growth. However, no population-based study to date has investigated this potential association. OBJECTIVE: This work aims to determine whether prior angiotensin-converting enzyme inhibitor (ACEi) use is associated with a reduced odds of leiomyoma development. DESIGN: A nested case-control study was conducted. SETTING: The population was assembled from the Truven Health MarketScan Research Database, which includes private health insurance claims from January 1, 2012 to December 31, 2017. PATIENTS OR OTHER PARTICIPANTS: We included (n = 353 917) women age 18 to 65 with hypertension. Cases (n = 13 108) with a leiomyoma diagnosis were matched to controls (n = 340 808) with no such diagnosis at a 1:26 ratio by age and region of origin within the United States. INTERVENTION: Prior ACEi use was determined from outpatient drug claims. MAIN OUTCOME MEASURE: Leiomyoma development was indicated by a first-time diagnosis code. RESULTS: Women on an ACEi experienced a 31.8% reduced odds of developing clinically recognized leiomyoma compared to nonusers (odds ratio [OR] 0.68; 95% CI, 0.65-0.72). This association was significant for each age group: 30 to 39 years (OR 0.86; 95% CI, 0.74-0.99), 40 to 49 years (OR 0.71; 95% CI, 0.66-0.76), 50 to 59 years (OR 0.63; 95% CI, 0.58-0.69), and 60 to 65 years (OR 0.58; 95% CI, 0.50-0.69). Of the ACEis, lisinopril (OR 0.67; 95% CI, 0.64-0.71), quinapril (OR 0.62; 95% CI, 0.41-0.92), and ramipril (OR 0.35; 95% CI, 0.23-0.50) demonstrated a significant association with reduced leiomyoma incidence. CONCLUSIONS: ACEi use was associated with a reduced odds of developing clinically recognized leiomyoma in adult hypertensive women.

Consistent RNA sequencing contamination in GTEx and other data sets.

A challenge of next generation sequencing is read contamination. We use Genotype-Tissue Expression (GTEx) datasets and technical metadata along with RNA-seq datasets from other studies to understand factors that contribute to contamination. Here we report, of 48 analyzed tissues in GTEx, 26 have variant co-expression clusters of four highly expressed and pancreas-enriched genes (PRSS1, PNLIP, CLPS, and/or CELA3A). Fourteen additional highly expressed genes from other tissues also indicate contamination. Sample contamination is strongly associated with a sample being sequenced on the same day as a tissue that natively expresses those genes. Discrepant SNPs across four contaminating genes validate the contamination. Low-level contamination affects ~40% of samples and leads to numerous eQTL assignments in inappropriate tissues among these 18 genes. This type of contamination occurs widely, impacting bulk and single cell (scRNA-seq) data set analysis. In conclusion, highly expressed, tissue-enriched genes basally contaminate GTEx and other datasets impacting analyses.

HPAStainR: a Bioconductor and Shiny app to query protein expression patterns in the Human Protein Atlas.

The Human Protein Atlas is a website of protein expression in human tissues. It is an excellent resource of tissue and cell type protein localization, but only allows the query of a single protein at a time. We introduce HPAStainR as a new Shiny app and Bioconductor/R package used to query the scored staining patterns in the Human Protein Atlas with multiple proteins/genes of interest. This allows the user to determine if an experimentally-generated protein/gene list associates with a particular cell type. We validated the tool using the Panglao Database cell type specific marker genes and a Genotype Expression (GTEx) tissue deconvolution dataset.  HPAStainR identified 92% of the Panglao cell types in the top quartile of confidence scores limited to tissue type of origin results. It also appropriately identified the correct cell types from the GTEx dataset. HPAStainR fills a gap in available bioinformatics tools to identify cell type protein expression patterns and can assist in establishing ground truths and exploratory analysis. HPAStainR is available from: https://32tim32.shinyapps.io/HPAStainR/.

Human cardiac myosin light chain 4 (MYL4) mosaic expression patterns vary by sex.

Sex disparities modulate cardiac function, although the proteins and mechanisms remain to be elucidated. We recently demonstrated a mosaic pattern of protein expression in the heart for over 100 proteins. Here we investigate one of these proteins, myosin light chain 4 (MYL4), which is important for contractile functions by increasing force production. We assayed the expression pattern of MYL4 across 756 ventricular myocardial samples from 668 individuals utilizing a semi-automated Cell Profiler method on five tissue microarrays (TMAs) of cardiac tissues across a diverse set of diseases. The percentage of MYL4 positive cells was significantly higher in male subjects independently across all five TMAs, regardless of disease state (p = 8.66e-15). Higher MYL4 expression was also modestly associated with hypertrophic cardiomyopathy (p = 6.3e-04). MYL4 expression did not associate with sudden cardiac death or other cardiomyopathies. This study demonstrates a new mosaic pattern of protein expression that underlies sex disparities in the human heart.

An expanded proteome of cardiac t-tubules.

BACKGROUND: Transverse tubules (t-tubules) are important structural elements, derived from sarcolemma, found on all striated myocytes. These specialized organelles create a scaffold for many proteins crucial to the effective propagation of signal in cardiac excitation-contraction coupling. The full protein composition of this region is unknown. METHODS: We characterized the t-tubule subproteome using 52,033 immunohistochemical images covering 13,203 proteins from the Human Protein Atlas (HPA) cardiac tissue microarrays. We used HPASubC, a suite of Python tools, to rapidly review and classify each image for a specific t-tubule staining pattern. The tools Gene Cards, String 11, and Gene Ontology Consortium as well as literature searches were used to understand pathways and relationships between the proteins. RESULTS: There were 96 likely t-tubule proteins identified by HPASubC. Of these, 12 were matrisome proteins and 3 were mitochondrial proteins. A separate literature search identified 50 known t-tubule proteins. A comparison of the 2 lists revealed only 17 proteins in common, including 8 of the matrisome proteins. String11 revealed that 94 of 127 combined t-tubule proteins generated a single interconnected network. CONCLUSION: Using HPASubC and the HPA, we identified 78 novel, putative t-tubule proteins and validated 17 within the literature. This expands and improves our knowledge of this important subcellular structure of the cardiac myocyte. This information can be used to identify new structural targets involved in excitation-contraction coupling that may be altered in disease.