Enrichment of ovine gonadotropes via adenovirus gene targeting enhances assessment of transcriptional changes in response to estradiol-17 beta†

Abstract

Gonadotropes represent approximately 5-15% of the total endocrine cell population in the mammalian anterior pituitary. Therefore, assessing the effects of experimental manipulation on virtually any parameter of gonadotrope biology is difficult to detect and parse from background noise. In non-rodent species, applying techniques such as high-throughput ribonucleic acid (RNA) sequencing is problematic due to difficulty in isolating and analyzing individual endocrine cell populations. Herein, we exploited cell-specific properties inherent to the proximal promoter of the human glycoprotein hormone alpha subunit gene (CGA) to genetically target the expression of a fluorescent reporter (green fluorescent protein [GFP]) selectively to ovine gonadotropes. Dissociated ovine pituitary cells were cultured and infected with an adenoviral reporter vector (Ad-hαCGA-eGFP). We established efficient gene targeting by successfully enriching dispersed GFP-positive cells with flow cytometry. Confirming enrichment of gonadotropes specifically, we detected elevated levels of luteinizing hormone (LH) but not thyrotropin-stimulating hormone (TSH) in GFP-positive cell populations compared to GFP-negative populations. Subsequently, we used next-generation sequencing to obtain the transcriptional profile of GFP-positive ovine gonadotropes in the presence or absence of estradiol 17-beta (E2), a key modulator of gonadotrope function. Compared to non-sorted cells, enriched GFP-positive cells revealed a distinct transcriptional profile consistent with established patterns of gonadotrope gene expression. Importantly, we also detected nearly 200 E2-responsive genes in enriched gonadotropes, which were not apparent in parallel experiments on non-enriched cell populations. From these data, we conclude that CGA-targeted adenoviral gene transfer is an effective means for selectively labeling and enriching ovine gonadotropes suitable for investigation by numerous experimental approaches.

Keywords: anterior pituitary; gonadotropin-releasing hormone (GnRH/GnRH receptor): ovine/sheep; transcription.

Figure 1

Workflow for enrichment and analysis of sorted GFP⁺ gonadotropes from primary ovine pituitary cell cultures infected with Ad-hαCGA-eGFP.

Figure 2

Sorting of GFP-labeled ovine pituitary cells. (A) Gating was performed to remove dead cells, debris, and aggregates before sorting (cells located outside the black box R1). (B) Histogram of uninfected cells used to establish GFP⁺ and GFP⁻ gate parameters. (C) GFP⁺ cells isolated using wide and strict gating.

Figure 3

(A) LH and TSH content (means ± SEM) in GFP⁺ and GFP⁻ sorted cells. (B) LH and TSH enrichment (means ± SEM) in GFP⁺ cells sorted using wide or strict gating. Significant differences are indicated (^**P < 0.05, ^***—P < 0.001).

Figure 4

E2 regulation of GNRHR expression in sorted and unsorted ovine pituitary cells. Data expressed as the percent change from GNRHR expression in vehicle-treated GFP⁺ cells (defined as 100%). Significant differences are indicated (^****—P < 0.0001).

Figure 5

PCA plot of sorted GFP⁺ (red triangles) and infected unsorted (blue circles) cells in six isolated replicates.

Figure 6

(A) Heatmap of 372 overrepresented and 1596 underrepresented genes (FDR ≤ 0.05) in sorted (GFP⁺) and unsorted ovine pituitary cells. Each row represents a single gene; each column represents a single isolated replicate. The horizontal axis shows clustering within the two groups (sorted and unsorted). Color key represents the FCs. (B) Volcano plot of 17 082 genes. log₂ FCs are on the X-axis and log₁₀-transformed adjusted FDR values are on the Y-axis. Red points represent overrepresented genes (FDR ≤ 0.05 and absolute FC ≥ 3; right dashed vertical line) in the sorted (GFP⁺) population compared to the unsorted cells; the green points represent underrepresented genes (FDR ≤ 0.05 and absolute FC ≥ 3; left dashed vertical line). Gray points represent genes excluded from the analysis (FDR > 0.05 and absolute FC ≤ 3; horizontal dashed line).

Figure 7

(A) Fold enrichment of selected genes in GFP⁺ cells compared to the unsorted population as determined by qRT-PCR. Results are expressed as means ± SEM. (B) Fold enrichments obtained from RNA-seq data. White bars represent genes that are key gonadotrope markers; black bars represent genes that are markers for non-gonadotrope endocrine cells of the anterior pituitary gland. The dotted line represents the FC cutoff (FC ≥ 3) used in RNA-Seq for the overrepresented genes.

Figure 8

PCA plot of sorted control GFP⁺ cells (red triangles) versus E2-treated sorted GFP⁺ cells (blue circles) in six isolated replicates.

Figure 9

(A) Upper panel: Heatmap of 232 up-regulated and 31 down-regulated genes (FDR ≤ 0.1) in sorted GFP⁺ cells either untreated or treated with E2. Each row represents a single gene; each column represents a single isolated replicate. The horizontal axis shows clustering within the two groups (vehicle and E2 treated). Color key represents the FCs. Lower panel: log₂ FCs are on the X-axis and log₁₀-transformed adjusted FDR values are on the Y-axis. Red points represent up-regulated differentially expressed genes (FDR ≤ 0.1 and absolute FC ≥ 1.5; right dashed vertical line) in untreated cells compared with E2-treated samples. The green points represent down-regulated differentially expressed genes (FDR ≤ 0.1 and absolute FC ≥ 1.5; left dashed vertical line). Gray points represent genes excluded from the analysis (FDR > 0.1 and absolute FC ≤ 1.5; horizontal dashed line). (B) Upper panel: Heatmap of 93 up-regulated and 17 down-regulated genes (FDR ≤ 0.1) in unsorted cells either untreated or treated with E2. Each row represents a single gene; each column represents a single isolated sample. The horizontal axis shows clustering within the two groups (vehicle and E2 treated). Color key represents the FCs. Lower panel: log₂ FCs are on the X-axis and log₁₀-transformed adjusted FDR values are on the Y-axis. Red points represent up-regulated differentially expressed genes (FDR ≤ 0.1 and absolute FC ≥ 1.5; right dashed vertical line) in infected unsorted compared with E2-treated infected unsorted samples. Green points represent down-regulated differentially expressed genes (FDR ≤ 0.1 and absolute FC ≥ 1.5; left dashed vertical line). Gray points represent genes excluded from the analysis (FDR > 0.1 and absolute FC ≤ 1.5; horizontal dashed line). (C) Venn diagram showing the number of genes differentially regulated by E2 that are unique to either the sorted or unsorted populations or common to both populations.

Figure 10

qRT-PCR analysis of PGR, VEGFA, and LHB expression in sorted GFP⁺ and unsorted cells with and without E2 treatment. Data are expressed as an FC compared with GFP⁺ ovine cells, which was arbitrarily set as 1. Means ± SEM. Significant differences are indicated (^**P < 0.05, ^****P < 0.0001).