Dissecting Mammalian Spermatogenesis Using Spatial Transcriptomics

Background:

One of the main testicular functions is to produce male gametes (sperm cells), which are crucial for sexual reproduction. Two main compartments are distinguishable in the testis, the seminiferous tubules, fine coiled tubular structures where spermatogenesis takes place, and the peritubular space. The tubules contain layers of germ cells, that develop into sperm cells (also known as spermatozoa), in close association with Sertoli cells. Steroidogenic Leydig cells, macrophages and other interstitial cells are located in the space between the tubules.

Spermatogenesis is a controlled process that begins postnatally and culminates with the production of the first sperm at the onset of sexual maturity. In mice, the seminiferous tubules periodically cycle through 12 stages (I-XII) that are defined by the combination of germ cells present. Recently, testicular single-cell RNA sequencing (scRNA-seq) studies showed that not only the germ cells are stage specific but other cells like Sertoli cells also vary from stage to stage.

scRNA-seq introduced a fundamental change in the focus of the research and gene expression. Instead of using the whole organ as a research unit, it is partitioned to the smallest structural and functional unit, the single cell. One of the main issues concerning single-cell technologies is that spatial and tissue organization information is lost when generating a single-cell suspension. Spatial transcriptomics was developed as a new methodology to overcome this problem. Briefly, frozen tissue sections are placed on a slide containing barcoded spots (bead array) that provide a specific spatial orientation. The cells in the tissue sections are permeabilized, allowing mRNA to be bound to the most proximal barcode, followed by cDNA synthesis and sequencing. The barcode sequence associates each transcript with a spatial position, assigning mRNA expression to different tissue regions (Fig. 1).

The authors used spatial transcriptomics in mouse and human testis to develop a spatial transcriptome atlas for mammalian spermatogenesis, being able to assign information of cell type, seminiferous tubule, and stage to each bead with high accuracy. Furthermore, the authors provided evidence for the diagnostic value of this technology by characterizing the testicular disorganization in a diabetic mouse model.

Fig 1. Spatial transcriptomics workflow. (Preprint Fig. 1A and B).

Key findings

 1) Spatial transcriptome atlas for mouse spermatogenesis

Mouse 10 um frozen testicular sections were captured into a slide-seq array and subjected to spatial transcriptomics, detecting around 30,000 beads (“cells”) per array. The different beads were assigned with a cell type and were mapped into the array, which reflected the structure of the testicular seminiferous tubules (Fig. 2A). After the cell type assignment, the array was partitioned into the distinctive seminiferous tubules (Fig. 2B) and classified into one of the four major stages of seminiferous tubules using known stage-dependent markers: stages I-III, IV-VI, VII-VIII or IX-XII (Fig. 2C).

Stage-specific Sertoli and germ cells are well characterized in the literature, but the cells in the testicular interstitium and the peritubular space remain largely unexplored. The authors were able to identify transcriptomic signatures of stage specific Leydig (steroidogenic) cells (Fig. 2D) and macrophages (Fig. 2E) never reported before.

Fig 2. Mouse testicular spatial transcriptomics. (A) Spatial mapping of testicular cell types in mouse section. ES: elongated/elongating spermatid, RS: round spermatid, SPC: spermatocyte, SPG: spermatogonium. (B) Digital segmentation of the seminiferous tubules. (C) Spatial mapping of the four stage clusters. (D) Schematic representation of the spatial localization and marker genes for each stage dependent Leydig cell. (E) Schematic representation of the spatial localization and marker genes for each stage dependent macrophage. (Preprint Fig. 1C, 1F, 1H, 3A, 3C and 3D).

2) Spatial transcriptome atlas for human spermatogenesis

The next challenge was to apply this pipeline to human testicular samples, spatially mapping human testicular cell types (Fig 3A). In order to validate these results, the authors used the Human Protein Atlas database to corroborate the expression pattern of different spermatogonia genes into the seminiferous tubules (Fig. 3B). This resulted in around 74% of match, with the remaining 26% not being present in the protein database or showing non-specific immunostaining. An example of four genes with differential expression from the lumen (left) to the basement membrane (right) can be found in Fig. 3B.

Recently, human testicular single-cell RNA sequencing revealed 5 transcriptional states of human spermatogonia (0-4). Using the described expression markers, two different microenvironments of spermatogonia coexist in the same seminiferous tubule, one containing mostly state 0 spermatogonia (naïve/stem cell) and the other containing states 1-4 and the remaining state 0 (Fig. 3C).

Fig 3. Human testicular spatial transcriptomics. (A) Spatial mapping of testicular cell types in human section. ES: elongated/elongating spermatid, RS: round spermatid, SPC: spermatocyte, SPG: spermatogonium. (B) Spatial expression patterns of the genes ACTRT3, ACRBP, CLDN11 and GSTA1 from the spatial transcriptomic atlas and the human protein atlas. (C) Spatial mapping of the five human spermatogonia germ cells. (Preprint Fig. 4 A, B and D).

3) Testicular spatial transcriptomics in diabetes-induced testicular injuries

Several reports indicate that diabetes mellitus results in disturbances in the male reproductive system, but the exact mechanism is still unknown. To evaluate if the slide-seq workflow could be applied to study testicular diseases, leptin-deficient diabetic mice (ob/ob) testis were compared with wild type testis (WT), using spatial transcriptomics (Fig. 4A). Differential expression analysis identified a downregulation of elongating/elongated spermatids genes such as Smcp and Odf1 (Fig. 4B). This is consistent with the loss of elongated spermatids in ob/ob testes. Furthermore, mitochondrial genes were elevated, a condition associated with mitochondrial dysfunction and mtDNA damage.

A disorganization in the seminiferous tubules structure was also detected in the ob/ob mice testes. To quantify these changes, a purity score was generated, taking into account the spatial mixing of the elongated spermatids beads with beads of other cell types (Fig. 4C). In the wild type mice, the elongated spermatids beads clustered in the centre of the seminiferous tubule, showing a high purity score (0.63). In contrast, in ob/ob seminiferous tubules, the elongated spermatids beads were likely to make contacts with beads of other cell types, reflecting a lower purity score (0.44) and a structural disorganization.

Fig 4. Diabetes causes disruption in the testicular seminiferous tubules. (A) Spatial mapping of the testicular cell types in wild type (WT) and leptin-deficient diabetic (ob/ob) samples. ES: elongated/elongating spermatid, RS: round spermatid, SPC: spermatocyte, SPG: spermatogonium. (B) Spatial expression pattern of Smcp and Odf1 genes in a representative WT and ob/ob seminiferous tubule. (C) Elongated spermatid purity score for each Slide-seq bead in a representative WT and ob/ob seminiferous tubule. (Preprint Fig. 5A, C and S11).

Why I choose this paper:

One of the main issues around single-cell technologies is that spatial and tissue organization information is lost when generating a single-cell suspension. Spatial transcriptomics was developed as a new methodology to overcome this problem. This study demonstrated the utility of this emerging technique by characterizing testicular spermatogenesis in mouse and human tissue with high accuracy. Furthermore, the authors provided evidence for the diagnostic value of this technology by characterizing the testicular disorganization in a diabetic mouse model. The spatial atlases generated can be useful to identify cellular and molecular changes associated with different diseases.

The used protocol can easily be adapted to other tissues, animal models or human samples and provides the first workflow that could be used in the clinic. Tissue biopsies from testicular cancer, infertility or differences of sex development (DSDs) could potentially be subjected to spatial transcriptomics in the near future.

Future directions / questions for the authors:

• Sertoli and germ cells grow in really close association, how hard was it for you to differentiate both cell types? How often did you find Sertoli and sperm cells transcripts in the same bead?
• In your opinion, what would be the next step required to bring spatial transcriptomics into a clinical setting?

Posted on: 7th January 2021

doi: https://doi.org/10.1242/prelights.26837

(No Ratings Yet)

Report Content

• Issue: *
  Copyright infringement Spam Inappropriate content Broken links
• Your Name: *
  
• Your Email: *
  

• Details: *
  

Submit Report