Dynamic changes in EPCAM expression during spermatogonial stem cell differentiation in the mouse testis

Abstract

Background: Spermatogonial stem cells (SSCs) have the unique ability to undergo self-renewal division. However, these cells are morphologically indistinguishable from committed spermatogonia, which have limited mitotic activity. To establish a system for SSC purification, we analyzed the expression of SSC markers CD9 and epithelial cell adhesion molecule (EPCAM), both of which are also expressed on embryonic stem (ES) cells. We examined the correlation between their expression patterns and SSC activities.

Methodology and principal findings: By magnetic cell sorting, we found that EPCAM-selected mouse germ cells have limited clonogenic potential in vitro. Moreover, these cells showed stronger expression of progenitor markers than CD9-selected cells, which are significantly more enriched in SSCs. Fluorescence-activated cell sorting of CD9-selected cells indicated a significantly higher frequency of SSCs among the CD9(+)EPCAM(low/-) population than among the CD9(+)EPCAM(+) population. Overexpression of the active form of EPCAM in germline stem (GS) cell cultures did not significantly influence SSC activity, whereas EPCAM suppression by short hairpin RNA compromised GS cell proliferation and increased the concentration of SSCs, as revealed by germ cell transplantation.

Conclusions/significance: These results show that SSCs are the most concentrated in CD9(+)EPCAM(low/-) population and also suggest that EPCAM plays an important role in progenitor cell amplification in the mouse spermatogenic system. The establishment of a method to distinguish progenitor spermatogonia from SSCs will be useful for developing an improved purification strategy for SSCs from testis cells.

Figure 1. Characteristics of germ cells after CD9 or EPCAM selection.

(A) Appearance of CD9-selected (Left) or EPCAM-selected (Right) germ cells on laminin-coated dishes, after 7 days in culture. Testis cells were collected from 10-day-old pups and used to initiate GS cell cultures after MACS. No significant colony formation is seen in EPCAM-selected cells. CD9-selected cells started to proliferate to form spermatogonia chains under the same culture condition, and spermatogonial chains are seen. Note the contaminating testicular fibroblasts in cultures of CD9-selected cells. Arrows indicate magnetic beads used for cell separation. (B) Apoptosis of CD9-selected (Left) and EPCAM-selected cells (Right), after 8 days in culture. TUNEL-positive cells are stained red. Counterstained by Hoechst 33342 (blue). (C) Flow cytometric analyses of CD9-selected (Top) or EPCAM-selected (Bottom) cells collected from adult testes. Green lines indicate control staining. Note the increased KIT staining in EPCAM-selected cells. (D) Real-time PCR analyses of spermatogonial marker genes or cyclins in EPCAM- or CD9-selected cells. CD9-selected cells show increased expression of Nanos3, Bcl6b, Etv5, Ccnd1, and Ccnd2. Transcript levels were normalized to Hprt expression, with expression levels in EPCAM-selected cells. (E) Macroscopic appearance of recipient testes transplanted with EPCAM- or CD9-selected cells. Green tubules indicate germ cell colonies developed from donor SSCs. The same numbers of cells were transplanted at the same time. (F) Quantification of colonies. Both EPCAM-selected (Left) and CD9-selected (Right) cells produced significantly more germ cell colonies than control unselected testis cells, but CD9-selected cells contained a higher concentration of SSCs. Bars  = 20 µm (A); 100 µm (B); 1 mm (E).

Figure 2. Flow cytometric analyses of CD9- or EPCAM-selected cells after MACS.

(A) Light-scattering properties and double immunostaining of total testis cells (Top), CD9-selected cells (Middle) or EPCAM-selected cells (Bottom) stained with APC-conjugated anti-CD9 and PE-conjugated anti-EPCAM antibodies. Cells were gated according to forward scatter (size) and side scatter (cell complexity) values (Left). Gated cells were analyzed for CD9 and EPCAM (Right). Note the simpler light-scattering properties of EPCAM-selected cells. (B) Three subpopulations of CD9-selected cells. Fraction I shows high CD9 and low or no EPCAM immunostaining; fraction II shows low CD9 and high EPCAM immunostaining; and fraction III is low or no CD9 and no EPCAM immunostaining. (C) The three CD9-selected cell subpopulations, immunostained with APC-conjugated anti-CD9, PE-conjugated anti-EPCAM, and PE-Cy7-conjugated anti-KIT antibodies. KIT is strongly expressed in fraction II. Areas shaded in black indicate control staining. (D) Changes in immunostaining of total testis cells during postnatal testicular development. Total testis cells were stained with APC-conjugated anti-CD9 and PE-conjugated anti-EPCAM antibodies. Stronger CD9 and EPCAM immunostaining is seen in 10-day-old mouse testis cells compared with 1-day-old mouse testis cells. Only testes from 35-day-old mice show all three fractions.

Figure 3. Functional analyses of SSC activity by germ cell transplantation of each CD9-selected cell fraction.

(A) Macroscopic appearance of recipient testes. Approximately 4.9×10³, 2.3×10⁴, 3.5×10², and 1.6×10⁵ cells were transplanted for fractions I, II, III, and control cells, respectively. Recipient testes were stained with X-Gal 2 months after transplantation. Blue tubules indicate germ cell colonies developed from donor SSCs. Cells in fraction I have a uniform appearance (insert). (B) Quantification of colonies. The number of cells that could be recovered in each experiment varied, and thus the colony number was normalized to reflect donor cells at a concentration of 10⁵ cells injected/testis. Cells in fraction I are significantly enriched for SSCs. (C) Histological sections of recipient testes. Note the normal appearance of spermatogenesis of donor-derived cells in recipients of control cells (Top) and fraction I cells (Bottom). Stain: X-Gal (A); X-Gal, Hematoxylin and eosin (C). Bars  = 1 mm (A), 50 µm (C).

Figure 4. Overexpression of EpICD in ROSA GS cells.

(A) Upregulation of EPCAM by GDNF stimulation. ROSA GS cells were cultured on laminin with 1% FBS and without cytokines for 3 days and then stimulated with the indicated cytokines. The cells were recovered 3 days after cytokine stimulation, and stained with anti-EPCAM antibody. (B) Sorting of EpICD-transfected GS cells. GS cells (3×10⁵) on MEFs in 6-well plates were infected, expanded in vitro, and sorted. (C) Morphology of the sorted cells. (D, E) Flow cytometric (D) and RT-PCR (E) analyses of EpICD-transfected ROSA GS cells. No significant changes are seen. Green lines indicate controls. (F) Effects of cytokines on proliferation of EpICD-transfected ROSA GS cells. GS cells (3×10⁵) on MEFs were cultured with the indicated cytokines and recovered by trypsinization 6 days after initiation of culture. No significant differences are seen between the control and EpICD-transfected cells. (G) Quantification of colonies. No significant differences are seen between the control and EpICD-transfected cells. (H) Macroscopic appearance of recipient testes. Recipient testes were stained with X-Gal 2 months after transplantation. Blue tubules indicate germ cell colonies developed from donor SSCs. (I) Histological sections of recipient testes. Cells show apparently normal spermatogenesis. Stain: X-Gal (H); X-Gal, Hematoxylin and eosin (I). Bars  = 100 µm (C), 1 mm (H), 20 µm (I).

Figure 5. Epcam KD in ROSA GS cells by shRNA.

(A) Flow cytometric profile of ROSA GS cells 2 days after transduction with Epcam KD vector. Green lines indicate controls. (B) Expression of EPCAM represented by the mean fluorescence intensity. Data are expressed as mean fluorescence intensity minus background autofluorescence of cells stained with the control secondary antibody. (C) Reduced recovery of ROSA GS cells after transduction with shRNA against Epcam. GS cells (8×10⁵) on MEFs were infected with the lentivirus and recovered 2 days later. Results of three experiments are shown. (D) Macroscopic appearance of the recipient testes transplanted with Epcam KD GS cells. The same number of cells was transplanted at the same time. Blue tubules indicate germ cell colonies developed from donor SSCs. (E) Assessment of SSC activity by germ cell transplantation. ROSA GS cells were transplanted into the seminiferous tubules of W mice 3 days after transduction with shRNA against Epcam. (F) A model for EPCAM function during SSC differentiation. GDNF upregulates EPCAM expression on SSCs/progenitors. Proliferation/survival of progenitors may be stimulated by EPCAM expression on neighboring cells as well as by KITL on Sertoli cells. Stain: X-Gal (D). Bar  = 1 mm (D).