ErbB4, a Receptor Tyrosine Kinase, Coordinates Organization of the Seminiferous Tubules in the Developing Testis

Abstract

Although close to every fifth couple nowadays has difficulty conceiving, the molecular mechanisms behind the decline in human reproduction remain poorly understood. We report here that the receptor tyrosine kinase Erbb4 is a candidate causal gene, because it is expressed in a sexually dimorphic manner and is abundant in the developing and adult testes in the mouse. Sertoli cell–specific Erbb4-knockout mice have a compromised 3-dimensional organization of the testicular seminiferous tubules that affects their fertility. More specifically, adhesion defects are observed in the absence of Erbb4, which are characterized by changes in the expression of laminin-1, N-cadherin, claudin-3, and certain cell-cell junction components between the Sertoli and germ cells. Interestingly, Erbb4 knockout also had an effect on the Leydig cells, which suggests a paracrine influence of Sertoli cells expressing ErbB4. Many of the defects observed in Erbb4-knockout mice are rescued in targeted ERBB4 gain-of-function mice, pointing to a coordination role for ErbB4 in the developing testis. Thus, the ErbB4 receptor tyrosine kinase promotes seminiferous tubule development by controlling Sertoli cell and germ cell adhesion.

ErbB4 is a member of the epidermal growth factor receptor family, which includes ErbB1, ErbB2, ErbB3, and ErbB4. ErbB4 is a unique among the ErbB receptors, because it can generate 4 alternatively spliced isoforms with distinct signaling properties (1, 2). Its function is critical for the development of several organs, such as the heart (3), kidney (4), nervous system (5), and mammary glands (6). Deregulated ErbB4 signaling has been associated with diseases such as breast (7), kidney (8), and ovarian carcinogenesis (9). Interestingly, although ErbB4 expression is induced in yolk sac tumors (10), it is reduced in certain testicular cancers (11). In fact, the role of ErbB4 in testis cancers suggests that it may be involved in testis development, although this has not previously been investigated.

Male sexual characteristics are established and maintained by the testis, especially by its 2 key cell types: Sertoli and Leydig cells. The Sertoli cells are integrated within the seminiferous tubules and promote differentiation of the germ cells during spermatogenesis (12). They also establish the blood-testis barrier (BTB) through tight intercellular junctions. This barrier provides the isolated environment that is necessary for proper development of the spermatocytes into spermatids (13). The Leydig cells, which are external to the tubules, produce androgens such as testosterone in response to the pituitary gland-derived LH. Testosterone then signals from the Leydig cells to the Sertoli cells via the androgen receptor (14), promoting maturation of the testis. Androgens also control formation of the BTB (15), support the progression of germ cells through meiosis, and regulate differentiation of the spermatids into mature sperm (16).

At present, >12% of couples in the Western world are having difficulties conceiving, and in >50% of cases the man is the sole cause of this infertility or at least a contributing factor (17). This severe decline in fertility potential may involve azoospermia, oligospermia, cryptorchidism, and testicular cancers, but other contributing factors remain poorly understood (18). Among the latter, changes in the seminiferous tubules, cell adhesion, junctions between the Sertoli cells, and cell polarity have also been considered as possible causes of male infertility (19, 20).

We adopted a conditional knockout approach to examine whether ErbB4 function is involved in male sexual development. We show here that the Sertoli, Leydig, and germ cells in the developing and adult testis express Erbb4, the critical role of which was revealed by conditionally inactivating the Erbb4 gene in the Sertoli cells with MisCre. The mutant mice demonstrated reduced testis size, decreased sperm production, compromised fertility, and defective Sertoli cell and germ cell adhesion. Strikingly, some of these defects were due to secondary and reciprocal influences on the adjacent Leydig cells and germ cells. In addition, the Erbb4-knockout mice had altered LH, FSH, and testosterone levels, thus also indicating systemic consequences. ERBB4 gain-of-function in the Sertoli cells restored many of the anomalies caused by the Erbb4 knockout. The findings indicate an important role for ErbB4 in testis organogenesis, specifically in the coordination of seminiferous tubule development by influencing the interactions between Sertoli cells, germ cells, and Leydig cells.

Materials and Methods

Animals

Erbb4^Flox/Flox (21), R26E4⁺ (4), Rosa26LacZ (22), and MisCre (23) mice were used to generate Erbb4^Flox/Flox;MisCre⁺, R26E4^+/−ErbB4^Flox/Flox;MisCre⁺, and R26LacZ;MisCre⁺ indicator mice. The gonads of the wild-type (WT) littermate embryos were used as controls. All the experiments involving mice were performed in accordance with Finnish national legislation, the European Convention (ETS 123) and EU Directive 86/609/EEC. Samples were collected and processed as described previously (4).

Fertility assay and determination of hormone levels

To test fertility, single-pair matings were set up using 2-month-old Erbb4^Flox/Flox;MisCre⁺ males or WT siblings and 3-month-old Erbb4^Flox/Flox females. Ten separate mating assays were performed for each genotype. The numbers of plugs and the sizes of the litters generated were monitored. Blood from Erbb4^Flox/Flox;MisCre⁺ and WT sibling males was collected at 6 weeks and 3 months of age, and serum was purified by centrifugation and stored at −20°C until used. The amounts of FSH, LH, and testosterone were measured by immunofluorometric assays as described previously (24).

Quantitative RT-PCR (qRT-PCR) and in situ hybridization

The primers and probes used for the qRT-PCR are described in Supplemental Table 1 and by Veikkolainen et al (4). In situ hybridization with Erbb4, Mis, Hsdb17b1, and Hsd3b1 probes was performed as described previously (25).

Immunohistochemistry

The expression of ErbB4 was analyzed by immunohistochemistry in paraffin-embedded sections as described previously (4), using antibodies against Gata-1 (Santa Cruz Biotechnology), caspase-3 (Millipore), laminin-1 (Abcam), claudin-3 (Abcam), inhibin β B (LSBio), collagen IV (Millipore), N-cadherin (Abcam), and connexin-43 (Life Technologies). Immunofluorescence staining was performed essentially according to the method of Naillat et al (25). A TSA kit (PerkinElmer) was used for analyzing albumin expression according to the manufacturer's protocol. Alexa Fluor 488- or 546-conjugated antibodies were used as the secondary antibodies (Invitrogen). The nuclei were depicted with 4′,6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich).

Western blotting

Testes were prepared from embryonic and adult mice, and the proteins were purified immediately and subjected to Western blotting using anti-ErbB4 (E200; Abcam), anti-Akt (sc-1618; Santa Cruz Biotechnology), anti-phospho-Akt (9271; Cell Signaling Technology), anti-Erk (9102; Cell Signaling Technology), anti-phospho-Erk (9101; Cell Signaling Technology), anti-MIS (sc-28912; Santa Cruz Biotechnology), or anti-Actin (sc-1616; Santa Cruz Biotechnology) antibodies as described previously (26).

Sperm motility analysis

To analyze possible changes in sperm motility due to Erbb4 knockout, cauda epididymides were dissected from 6-month-old Erbb4^Flox/Flox;MIsCre⁺ and WT mice (n = 5 for each genotype) and transferred to 500 μL of KSOM Embryo Culture Medium (Speciality Media; Millipore). The sperm were allowed to disperse for 20 minutes at 37°C, triplicates of a well-mixed sperm suspension (10 μL) on preheated glass slides were placed on a heating plate (37°C) that was attached to a microscope, and the mobility of the sperm cells was examined immediately at ×400 magnification. The spermatozoa were classified as (1) progressively motile, (2) nonmotile, and (3) immotile, as judged by visual inspection.

Transmission electron microscopy

Testes dissected from 3-month-old mice (n = 3 for each genotype) were fixed in 1% glutaraldehyde and 4% formaldehyde in 0.1 M phosphate buffer (pH 7.4) and then 1% osmium tetroxide. The samples were dehydrated in acetone and embedded in Epon LX 112 (Ladd Research Industries). Semithin sections (1 μm) were stained with toluidine blue for light microscopy. Thin sections (80 nm) were cut with a Leica Ultracut UCT ultramicrotome, stained in uranyl acetate and lead citrate, and examined in a Tecnai Spirit G2 transmission electron microscope. The images were captured with a Veleta CCD camera (Olympus Soft Imaging Solutions GmbH).

Optical projection tomography (OPT)

To analyze the testes in 3 dimensions, a modification of the protocol of Short et al (27) was used with the Vasa antibody (Abcam). The images were taken with a Bioptonics 3001 OPT scanner (Bioptonics) at 480 nm. Imaris 64 7.5.2 software was used to reconstruct the OPT data and measure the testis volume. Drishti V2.0 software was used for rendering the testis views.

Statistics

Student's t test was used to analyze the morphometric, immunohistochemical, and hormone data. One-way ANOVA with the Tukey honestly significant difference or Games-Howell post hoc t test was used to estimate the statistical significance of the changes noted in the Leydig cells, the body, and the weight of the external genitalia.

Results

Erbb4 is expressed by Leydig, Sertoli, and germ cells

qRT-PCR was used to determine whether ErbB receptors would be expressed during gonad development. Whereas Early growth factor or Erbb1 Egfr and Erbb2 are expressed similarly in embryonic male and female gonads, Erbb3 and Erbb4 are more abundant in the testes than in the ovaries. Erbb4 in particular demonstrates a sexually dimorphic expression pattern (Supplemental Figure 1, A–D). Of the 4 alternatively spliced ErbB4 isoforms, only JM-a CYT-1 and JM-a CYT-2 (1, 2) are expressed in developing testes (data not shown). The JM-a isoforms also produce a soluble 80-kDa intracellular domain (ICD) that can translocate into the nucleus and participate in transcription regulation (28). As expected, ICD production and nuclear ErbB4 staining were also observed in the testis (Supplemental Figures 1, E, arrow), in line with its putative signaling roles.

Whole-mount in situ hybridization and immunohistochemical studies indicated that although ErbB4 is not expressed in the ovary (Supplemental Figure 1F), it is confined to the testis cords, more specifically to the Sertoli cells, Leydig cells, and germ cells (Supplemental Figure 1, G and H, arrows). Collectively, these findings demonstrate differential expression patterns for the ErbB receptors during gonad development and suggest that ErbB4 may have a role in testis organogenesis.

Sertoli cell–specific Erbb4-knockout mice reveal a role for ErbB4 in controlling testis development

The Sertoli cells are critical for male reproduction, because they promote elongation of the testis cords and germ line maturation and lay the foundation for sperm production (16). We examined whether the ErbB4 in Sertoli cells would be functional by inactivating the gene specifically in these cells using Müllerian inhibitory substance (Mis; also known as anti-Müllerian hormone [Amh]) MisCre-mediated gene recombination (23).

Successful targeting of Cre to the Sertoli cells by the Mis promoter was achieved by crossing MisCre⁺ mice with floxed R26LacZ indicator mice. As expected, this revealed LacZ staining in the seminiferous tubules and their Sertoli cells at embryonic days 12.5 and 16.5 (23) (Supplemental Figure 2A). To create Erbb4^Flox/Flox;MisCre⁺ mice in which ErbB4 would be inactivated in the Sertoli cells (Supplemental Figure 2B), mice that were homozygous for the Erbb4 floxed allele (Erbb4^Flox/Flox) (21) were first mated with MisCre⁺ mice, and then the Erbb4^Flox/+;MisCre⁺ progeny were further crossed with Erbb4^Flox/Flox progeny. The littermate Erbb4^Flox/Flox;MisCre⁻ mice served as controls and will be referred to as WT from here on.

Western blot analysis with ErbB4 antibody indicated a notable reduction in ErbB4 and its cleavable ICD in the testes of the Erbb4^Flox/Flox;MisCre⁺ mice (Figure 1A). In line with this, expression of ErbB4 in the Sertoli cells was weaker in the tubules (Figure 1C compare with B, asterisks), and the staining extended further toward the tubular lumen than normal (Figure 1, compare C with B, and Supplemental Figure 1, E and H). Unexpectedly, Erbb4 knockout in the Sertoli cells also led to a reduction in ErbB4 expression in the Leydig cells (Figure 1, compare C with B; arrow), and consistent with this, phosphorylation of the Erk and Akt factors, which represent 2 downstream mediators of ErbB4 signaling, was also decreased in the Erbb4^Flox/Flox;MisCre⁺ knockout mice (Supplemental Figure 2C).

Figure 1.

Conditional knockout of Erbb4 in the Sertoli cells alters the 3-dimensional organization of the seminiferous tubules. A, Western blot analysis of ErbB4 expression in the testis. ErbB4 expression is lowered in the testes of Erbb4^Flox/Flox;MisCre⁺ mice. NS, nonspecific. B and C, ErbB4 immunohistochemical analysis. The asterisks denote a reduction in ErbB4 expression in the Sertoli cells, and the arrow denotes a reduction in Erbb4 expression in the Leydig cells. D, D′, E, and E′, OPT of WT and Erbb4^Flox/Flox;MisCre⁺ newborn testes. The red arrows indicate disrupted testis cords in the mutant mice. F, The ratio of total testis cord volume to the whole testis demonstrates decreased testis cord formation in the mutant mice. Scale bars correspond to 100 μm. *, P < .05.

We then started to investigate whether the inactivation of Erbb4 in the Sertoli cells would indeed have consequences for testis development and set out first to examine possible changes in the 3-dimensional organization by means of OPT and morphometric inspection of the OPT data (29). Whereas the cords of the newborn WT testis are elongated and convoluted and have invaded the testis itself (Figure 1, D and D′), the Erbb4^Flox/Flox;MisCre⁺ mutant testis cords were more randomly arranged and further apart, as judged by visual inspection (Figure 1, D and D′ vs E and E′). Some of the mutant testis cords became discontinuous due to splitting (Figure 1E′, red arrows). In addition, the overall volume of the Erbb4^Flox/Flox;MisCre⁺ testis cords relative to the total volume of the testis was decreased by around 50% (Figure 1F). Consistent with these findings, the expression of MIS, a marker of immature Sertoli cells, was reduced in the Erbb4^Flox/Flox;MisCre⁺ newborn testis when analyzed by in situ hybridization (Supplemental Figure 2, D and E). These findings raised the possibility that ErbB4 may have some role in testis cord development.

The Erbb4^Flox/Flox mice mated with Erbb4^Flox/+;MisCre⁺ mice produced pups with the expected Mendelian and female/male ratios (n = 360, data not shown), pointing to normal viability and sex determination. We then analyzed the capacity of the Erbb4^Flox/Flox;MisCre⁺ mice to reproduce by counting the pups that were born from the crosses. Whereas the WT mice produced an average of 10 pups per litter, the Erbb4^Flox/Flox;MisCre⁺ males mated with WT females produced only 6 pups per litter on average (n = 12, P < .002) (Figure 2A); ie, the Erbb4^Flox/Flox;MisCre⁺ males had a reduced capacity to generate litters (n = 12, P < .001) (Figure 2A).

Figure 2.

Erbb4 deficiency in the Sertoli cells influences testis weight and morphology and also fertility. A, Numbers of progeny and pups per litter from crosses with WT females (n = 10 for each genotype). B–E, Weights of WT and Erbb4^Flox/Flox;MisCre⁺ testes: testes sizes of 6-week-old and 6-month-old WT and Erbb4^Flox/Flox;MisCre⁺ mice. Scale bars correspond to 100 μm. ***, P < .001.

Body weight (Supplemental Figure 3A), external genitalia, and testicular descent (data not shown) remained unchanged between the Erbb4^Flox/Flox;MisCre⁺ and WT mice, but the testes of the Erbb4^Flox/+;MisCre⁺ mice were greatly reduced in size (Figure 2, B–E) and weighed approximately 25% less than those of their WT counterparts around puberty (postnatal day [P] 25; n = 9, P < .001), 40% less at 6 weeks (n = 12, P < .001), and 50% less at 6 months (n = 12, P < .001) (Figure 2B). We conclude from these data that the phenotypes caused by Erbb4 deficiency point to a role for this receptor in testis development.

Erbb4 deficiency in Sertoli cells leads to seminiferous tubule defects

A detailed histological inspection of the Erbb4^Flox/Flox;MisCre⁺ knockout model revealed focal dysgenic tubules and a reduction in, or even a complete failure of, spermatogenesis (Figure 3, compare C–G with A and B). The seminiferous tubules of the Erbb4^Flox/Flox;MisCre⁺ testis had notably elongated spermatids, as expected (Figure 3D, black arrowheads), but also apoptotic germ cells and symplasts of elongated spermatids on the luminal side (Figure 3, E, F, and G, arrows). In addition, they contained elongated and misaligned spermatids embedded in the epithelium, as depicted in Figure 3, H and J, where plates H′ to J′ show the misaligned spermatids (yellow arrowheads) and those embedded in the epithelium (white arrowheads), and the DAPI staining illustrates symplasts and disturbed organization of the spermatids due to Erbb4 knockout (Figure 3, K–M, arrowheads). The data suggest that spermiation is delayed at stage VIII in the Erbb4^Flox/Flox;MisCre⁺ mutants. Taken together, the findings point to a role for ErbB4 in seminiferous tubule development and perhaps in the spermiation process.

Figure 3.

Lack of Erbb4 disrupts cell adhesion and delays spermiation. A–K, Histological sections of 3-month-old WT and Erbb4^Flox/Flox;MisCre⁺ testes. B, Higher magnification images of a stage VI of a WT seminiferous tubule. C, Asterisk denotes focal dysgenic tubules in an Erbb4^Flox/Flox;MisCre⁺ testis. D, Stage V of the mutant semiferous tubule. E–H, An Erbb4^Flox/Flox;MisCre⁺ testis shows apoptotic cells (E, arrow) and symplasts (F, G, arrow) in its seminiferous tubules. H–K, In addition, the mutant mice demonstrated defects in spermiation. Whereas elongated spermatids were released into the testis cord lumen of the mutant mice (D, black arrowheads), elongated spermatids were embedded in the epithelium (white arrow). The boxes I′–J′ are higher magnifications showing misaligned spermatids that are oriented away from the basement membrane (I′, J′, and K′, yellow arrowheads). L–M, Erbb4^Flox/Flox;MisCre⁺ testis cords stained with DAPI also reveal some elongated spermatids that are still embedded in the seminiferous epithelium (white arrowhead), and some of these are misaligned (yellow arrowhead) in stage VIII tubules of mutant mice. Round spermatids were occasionally observed in the middle of the tubule lumen (red arrow). Scale bars correspond to 100 μm (A–J, K–N); 50 μm (H′–J′).

ErbB4 function is connected to maturation of the Sertoli cells

The Sertoli cells play an important role in puberty by providing structural and nutritional support for the germ cells during spermatogenesis. They proliferate until P15 in mice (16). We speculated that the reduction in the volume of the testis cords in the Erbb4^Flox/Flox;MisCre⁺ mice may have been caused by deregulation of the organization of the Sertoli cells, which would also have influenced testis maturation. We tested this hypothesis by analyzing possible changes in the expression of MIS and a transcription factor, Gata-1. When the Sertoli cells become fully mature, the expression of MIS is down-regulated, whereas Gata-1 expression is up-regulated (16). As expected, Western blot analysis indicated that the testes of 3-month-old Erbb4^Flox/Flox;MisCre⁺ mice expressed considerably more MIS than those of WT mice (Figure 4A). In contrast, Gata-1 expression, which depicts more mature Sertoli cells, was decreased in the Sertoli cells of the Erbb4-deficient testes at P35 and at 3 months (Figure 4, B–E, P = .05). We interpret this as an indication that ErbB4 function is involved in the maturation of Sertoli cells.

Figure 4.

Maturation of the Sertoli cells is delayed in the Erbb4Flox^/Flox;MisCre⁺ mice. A, Western blot analysis of MIS expression in the testis. MIS expression is enhanced in 3-month-old Erbb4^Flox/Flox;MisCre⁺ mice. B–D, Gata-1, a mature Sertoli cell marker, is reduced in P35 and 3-month-old Erbb4^Flox/Flox;MisCre⁺ testis cords. E, Quantification of Sertoli cells per testis cord cross-section demonstrates significantly reduced numbers of Sertoli cells in the mutant mice. Scale bars correspond to 100 μm. *, P < .05.

Impaired ErbB4 function in the Sertoli cells has a reciprocal influence on the Leydig cells

Examination of the Erbb4^Flox/Flox;MisCre⁺ testes revealed that the absence of Erbb4 in the Sertoli cells influenced the Leydig cells in a reciprocal manner. The Leydig cells in the Sertoli cell–specific Erbb4 knockout mice typically contained more transparent cytoplasm than did the corresponding WT cells (Figure 5, A and B, arrow). Moreover, electron microscopy pointed to apoptotic features in the Leydig cells, seen as a condensation of the nuclear material or complete absence of the nucleus (Figure 5, D and E). The number of Leydig cells, however, was not altered in the mutant mice (P = .37) (Figure 5C).

Figure 5.

Sertoli cell–specific Erbb4 knockout affects Leydig cell maturation indirectly. A and B, Histological sections of 3-month-old WT and Erbb4^Flox/Flox;MisCre⁺ testes. H & E, hematoxylin and eosin. C, Quantification of Leydig cell nuclei. D and E, Electron microscopy images of WT and Erbb4^Flox/Flox;MisCre⁺ testes. BV, blood vessel. F, qRT-PCR analysis of expression levels of steroidogenesis enzymes. G and H, Hsd3b1 in situ hybridization of 3-month-old WT and Erbb4^Flox/Flox;MisCre⁺ testes demonstrates enhanced expression of Hsd3b1 in mutant mice. I–L, Hsd17b3 in situ hybridization of newborn and 6-month-old WT and Erbb4^Flox/Flox;MisCre⁺ testes shows a drastic reduction in mutant mice at birth but induction in adult mice. M and N, Inhibin βB immunofluorescence staining of 3-month-old WT and Erbb4^Flox/Flox;MisCre⁺ testes. The data indicate a reduction in inhibin βB expression in the mutant mice. Scale bars correspond to 10 μm (A); 5 μm (D and E); 100 μm (G–N). *, P < .05.

We addressed the changes to the Leydig cells caused by Erbb4 knockout in the Sertoli cells in further detail by studying a panel of genes involved in the steroid synthesis pathway by qRT-PCR. Whereas no significant changes were found in the expression of the Cyp19 (P = .26) or Star (P = .83) genes, expression of the Leydig cell marker gene Hsd3b1 (P = .02) (30) was enhanced in response to the Erbb4 knockout compared with that for the WT situation (Figure 5F), in line with the in situ hybridization data (Figure 5, G and H). Thus, the deficiency in ErbB4 function in the Sertoli cells has a reciprocal influence on the structure of the Leydig cells and on their steroidogenesis.

ErbB4 deficiency in the Sertoli cells leads to testicular hypogonadism associated with increased FSH and LH but reduced amounts of testosterone

Given the severe testicular atrophy of the Erbb4^Flox/Flox;MisCre⁺ mice, it was considered that this would involve changes in the gonad-pituitary axis. We tested this possibility by measuring the amounts of FSH, LH, and testosterone in the serum of these mice (Table 1).

Table 1.

Serology of FSH, LH, and Testosterone of Erbb4^Flox/Flox;MisCre⁺ and WT Male Mice

Group	FSH, ng/mL	LH, ng/mL	Intratesticular Testosterone, nmol/Testis	Serum Testosterone, nmol/L	Seminal Vesicles, mg
6 weeks
WT	48.52 ± 6.51 (n = 6)	0.22 ± 0.15 (n = 6)			63.2 ± 3.45 (n = 7)
Erbb4Flox/Flox; MisCre+	61.51 ± 1.33 (n = 6)	2.86 ± 0.46 (n = 6)a			44.13 ± 6.21 (n = 8)a
3 month
WT	43.27 ± 7.99 (n = 6)	0.39 ± 0.27 (n = 6)	7.24 ± 0.40 (n = 4)	2.48 ± 0,94 (n = 3)
Erbb4Flox/Flox; MisCre+	65.47 ± 4.22 (n = 6)b	4.4 ± 1.32 (n = 6)b	5.22 ± 0.48 (n = 5)	0.69 ± 0.23 (n = 6)b

Values are means ± SEM for the number of mice indicated (n).

^aP < .01 vs control.

^bP < .05 vs control.

The 6-week-old (P = .06) and 3-month-old (P = .02) Erbb4^Flox/Flox;MisCre⁺ males did have approximately 30% more FSH than their WT counterparts (Table 1), although their pituitary gland appeared to be histologically normal (Supplemental Figure 4, A–C). Similarly, the amount of LH in the serum of the Erbb4^Flox/Flox;MisCre⁺ mice was also several times higher at the ages of 6 weeks (P = .02) and 3 months (P = .005) (Table 1). In line with these data, expression of the LH receptor (LhcgR) gene was up-regulated in the Leydig cells of the mutant males (P = .04) (Figure 5F).

We then measured the amounts of serum testosterone and intratesticular testosterone in the 3-month-old Erbb4^Flox/Flox;MisCre⁺ mice. Although the data showed a significant reduction in systemic testosterone (P = .05), no changes were seen in the intratesticular testosterone levels in the absence of Erbb4 (Table 1). Interestingly, we also found that the final enzyme in the testosterone biosynthesis pathway, Hsd3b17, which shifts from the Sertoli cells to the Leydig cells during maturation in the WT testis, was initially low in the Erbb4^Flox/Flox;MisCre⁺ seminiferous tubules (Figure 5, I and K) and remained expressed in both the Leydig cells and the Sertoli cells later (Figure 5, compare J and L with I and K). Similarly, inhibin βB, which controls FSH secretion, was reduced in the absence of Erbb4 (Figure 5, compare N with M, in green). Thus, Sertoli cell–specific Erbb4 knockout leads to compromised production of systemic testosterone. This finding provides a basic explanation for the Erbb4-dependent reduction in seminal vesicle size (P = 4.5 × 10⁻⁵) (Table 1) and the occurrence of testicular atrophy.

ErbB4 signaling promotes between the Sertoli and germ cells

Because symplasts characterized by premature of the cellular bridges between the round spermatids (31) were noted in the Erbb4^Flox/Flox;MisCre⁺ mice, we speculated that the function of ErbB4 in the testis may be connected with the control of adhesion in the Sertoli cells.

We examined the association of Erbb4 deficiency with possible changes in cell adhesion first by analyzing the expression of a panel of critical cell-cell junction genes at P10 and P42 by qRT-PCR. Cadherin/Cdh2 (N-cadherin), as an adherens junction marker, and connexin-43/Cnx43 as a gap junction protein, for example, couple Sertoli cells and germ cells together (32), whereas type IV collagen and laminin-1 and -3 are present in the basement membrane around the seminiferous tubule (33).

While Cnx43 increased, laminin-1a/3 (Lama1a/3) and type IVa1 collagen gene expression decreased in the testes of adult Erbb4^Flox/Flox;MisCre⁺ mice relative to that in WT mice, especially at P42, being in part a reversal of the P10 situation (P = .0003 and P = .002) (Figure 6E). In line with this observation, immunohistochemical studies indicated a reduction in laminin-1 in particular, but also to some degree in collagen IV staining in the testis cord in the presence of Erbb4-deficient Sertoli cells (Figure 6, compare B with A, D with C, insets and arrows). Thus, it appears that ErbB4 signaling is critically involved in the maintenance of proper adhesion between the Sertoli cells and germ cells and also adhesion to the basement membrane in the developing and maturing seminiferous tubules.

Figure 6.

Erbb4 deficiency in the Sertoli cells induces cell junction defects. A–D, Laminin-1 and type IV collagen expression in 3-month-old WT and Erbb4^Flox/Flox;MisCre⁺ testes. E, qRT-PCR analysis of cell adhesion component gene expression at P10 and P42. Scale bars correspond to 100 μm. *, P < .05; **, P < .01; ***, P < .001. In C and D, the insets show in higher magnification the basement membrane.

Deregulated ErbB4 signaling affects BTB formation and reduces male fertility

The tight cell-cell junctions within the Sertoli cells contribute to the formation of the BTB, the components of which are the tight, cell adherens, and desmosomal junctions. The BTB represents a semipermeable barrier and establishes an immune-privileged microenvironment for the germ cells (34).

To address the putative roles of ErbB4 in establishing the BTB as a further indication of its roles in cell adhesion control, we examined possible changes in the tight junction (claudin-3), gap junction (connexin-43), and cell-cell adhesion (N-cadherin) by immunohistochemical analysis (15). Claudin-3, which is present in the spermatocytes at stage VIII (15) (Figure 7, A and B), was ectopically expressed in the seminiferous tubules of the Erbb4^Flox/Flox;MisCre⁺ mice at stage X (Figure 7, compare D with C, arrows). Connexin-43, in turn, is expressed between Sertoli cells and germ cells but was enhanced in the Sertoli cells of the WT mice, especially on the basal side of the seminiferous tubules (Figure 7E). In the Erbb4^Flox/Flox;MisCre⁺ mice, however, connexin-43 expression was decreased in the basal third component of the lumen of the seminiferous tubules compared with that in the WT mice (Figure 7, compare F and G with E, arrows). In addition, some residual expression is observed in the adluminal compartment of the Erbb4^Flox/Flox;MisCre⁺ mice (Figure 7, F and G). Finally whereas N-cadherin/Cdh2 was evenly distributed in the basal compartment of the WT seminiferous tubules, its expression was more irregular in the Erbb4^Flox/Flox;MisCre⁺ testes (Figure 7, compare I with H, arrow).

Figure 7.

Changes in tight junction and cell adhesion markers in Erbb4^Flox/Flox;MisCre⁺ testes. A–D, Claudin-3 immunofluorescence staining (arrow) to examine newly formed tight junctions (A and B, stage VIII; C and D, stage X) in 6-week-old WT and Erbb4^Flox/Flox;MisCre⁺ testes (arrows). E–G, Connexin-43 immunofluorescence staining reveals gap junctions in the testes of 6-week-old (WT and mutant) and 3-month-old (mutant) mice. Insets, connexin-43 staining. H–I, N-cadherin immunofluorescence staining to examine cell adhesion in 6-week-old WT and Erbb4^Flox/Flox;MisCre⁺ testes (arrows). Insets, N-cadherin staining in the basement membrane region. Scale bars correspond to 100 μm.

We noted changes in the expression of the Sertoli cell adhesion machinery, which raised the possibility that the BTB might be compromised in the absence of ErbB4 as well. To test this hypothesis, we used albumin accumulation within the testis as a marker of BTB function. The BTB did indeed prevent the entry of albumin into WT seminiferous tubules (Supplemental Figure 5, A and C), whereas albumin had penetrated the seminiferous tubules of the Erbb4^Flox/Flox;MisCre⁺ testes (Supplemental Figure 5, compare B, D, and E with A and C, arrows). Collectively, the changes in the cell adhesion machinery during sperm maturation had led to a compromised capacity of the testes of Erbb4^Flox/Flox;MisCre⁺ males for producing motile sperm (Figure 8).

Figure 8.

Sperm motility is lowered in the Erbb4^Flox/Flox;MisCre⁺ mice. Sperm motility assay showing the percentages of sperm cells with deficiencies in progressive motility (PR), in nonprogressive motility (NP), and in immobility (IM).

Human ERBB4 gain-of-function rescues many of the phenotypes caused by the absence of Erbb4 in Sertoli cells

Given that MisCre becomes active in the Sertoli cells during embryogenesis, the anomalies noted in the testes of the Erbb4^Flox/Flox;MisCre⁺ knockout mice may be not causally linked to primary Erbb4 function but may rather be secondary to this. Because of this possibility, we targeted human ERBB4 to the Erbb4-deficient Sertoli cells. To achieve this, the Erbb4^Flox/+;MisCre⁺ mice were crossed with mice that were heterozygous for the floxed Rosa26 promoter-driven ERBB4 JM-a CYT-2 (4).

As expected, the R26E4^+/−ErbB4^Flox/Flox;MisCre⁺ mice expressed ERBB4, both the full-length receptor and the cleavable ICD, robustly (Figure 9A). The ectopic ERBB4 also provided a striking rescue of many of the phenotypes noted in the Erbb4^Flox/+;MisCre⁺ mice when crossed with this genetic background. Importantly, the unchanged body and prostate weights (Table 2) and normalized testis weight (P = .001, Figure 9B) observed in 6-week-old R26E4^+/−ErbB4^Flox/Flox;MisCre⁺ mice indicated both tissue-specific expression of ERBB4 and a resulting functional rescue.

Figure 9.

Rescue of Erbb4^Flox/Flox;MisCre⁺ phenotypes with ERBB4 expression indicates direct roles for ErbB4 in the testis. A, Western blot analysis of ErbB4 expression in WT and R26E4^+/−Erbb4^Flox/Flox;MisCre⁺ rescue mice. NS, nonspecific, B, Testis weight. C and D, Histological sections of 6-week-old WT and R26E4^+/−Erbb4^Flox/Flox;MisCre⁺ testes. E–G, Type IV collagen immunofluorescence staining demonstrates restored expression in R26E4^+/−Erbb4^Flox/Flox;MisCre⁺ mice. Scale bars correspond to 100 μm. ***, P < .001.

Table 2.

External Genital Dry and Body Weight in WT, Erbb4^Flox/Flox;MisCre⁺ and E4A2Erbb4^Flox/Flox;MisCre⁺ Mice at 6 Weeks

Group	Prostate, mg	Body Weight, g
WT	12.02 ± 2.77 (n = 7)	21.50 ± 1.42 (n = 7)
Erbb4Flox/Flox;MisCre+	9.56 ± 2.0 (n = 8)	20.61 ± 2.08 (n = 8)
R26E4+/−Erbb4Flox/Flox; MisCre+	10.03 ± 1.54 (n = 6)	19.46 ± 0.40 (n = 6)

Values are means ± SEM for the number of mice indicated (n). Analysis was performed by one-way ANOVA followed by a post hoc Tukey honestly significant difference for the prostate and a post hoc Games-Howell test for body weight.

Inspection of histological sections from WT and R26E4^+/−ErbB4^Flox/Flox;MisCre⁺ testes also indicated rescue of many of the features that had been caused by Erbb4 deficiency, and the histological features appeared to be similar (Figure 9, C and D). The Leydig cells were of normal appearance in the R26E4^+/−ErbB4^Flox/Flox;MisCre⁺ mice (Figure 9, compare D with C, insets). Moreover, the ectopic expression of type IV collagen in the testis of the Erbb4^Flox/Flox;MisCre⁺ mice was restored by ERBB4 gain of function (Figure 9, E and G). Taken together, the capacity of ERBB4 to reverse many of the Sertoli cell–specific ErbB4 knockout features supports the conclusion that Erbb4 function is critical for the testis and has a role in the Sertoli cells as a coordinator of seminiferous tubule organization in the developing testis.

Discussion

A better understanding of the male reproductive system is needed to answer questions such as why sperm counts and fertility have declined in the industrialized countries (17). Today the diagnosis in around one third of all infertile men is an idiopathic infertility condition that may have resulted from genetic or endocrine causes (35, 36). Biopsy samples taken from the testes of infertile men have revealed that some of the seminiferous tubules still support relatively normal spermatogenesis, whereas others have undergone an arrest of spermatogenesis (37). There is thus a great need to gain more knowledge of the fundamental details of male reproductive development and health.

We found here that a receptor tyrosine kinase, ErbB4, is a functional component of testis organogenesis. Localization studies indicated that Erbb4 is expressed at a relatively high level not only in the Sertoli cells, but also in the germ cells and Leydig cells, which together represent the key cell types involved in male reproduction.

Targeted inactivation of ErbB4 function in the Sertoli cells by crossing MisCre and floxed Erbb4 mice led to a considerable reduction in testis size. Moreover, optical projection tomography revealed that the absence of Erbb4 in the Sertoli cells compromised the 3-dimensional organization of the seminiferous tubules. These data suggest that ErbB4 signaling has an important role in testis development.

Given the roles of FSF and LH, the synthesis of which is controlled by the pituitary gland, as testicular feedback signals, the reduction in testosterone in the Erbb4 Sertoli cell knockout mice led to an increase in the amounts of systemic FSH and LH. In humans, men who present with low testosterone accompanied by high levels of FSH and LH have hypogonadism due to defective function of the pituitary, hypothalamus, or testis (38). The likely reason for these alterations in the knockout mouse is reduced testicular androgen production, which in turn reflects changes on the testis-pituitary gland axis toward enhanced compensatory synthesis of FSF and LH. The notion of a reduction in androgen levels is consistent with the Erbb4 deficiency-induced reduction in seminal vesicle weight and spermatogenesis delay, the development of which are dependent on testosterone action.

O'Shaughnessy et al (39) demonstrated that 5-α-androstanediol and later testosterone, synthesized in the Leydig cells by the Hsd17b3 enzyme, are the major androgens in the mouse. During fetal development, this enzyme is expressed by the Sertoli cells and consistent with this, the Sertoli cell–specific Erbb4 knockout male mice had a defect in testosterone synthesis. This may in part have been due to the observed delay in the maturation of Sertoli cells and the associated delay in the timing of the shift of testosterone production from the Sertoli cells to the Leydig cells.

FSH influences the division of the Leydig cells, which is mediated through the action of androgen (40). Moreover, the Leydig cells produced androgens to promote their own differentiation via an androgen receptor–mediated positive feedback loop (40, 41). Similarly, pituitary FSH promotes this process in the Sertoli cells through the FSH receptor (42). The Sertoli cells are considered to influence the Leydig cells secondarily to promote their proliferation and apparently also their differentiation from precursor cells (40). Because we observed defects in cell adhesion in the Sertoli cells, germ cells, and Leydig cells, we can hypothesized that such defects influence the production of FSH, LH, and testosterone, which in turn affects the feedback loop, the hypothalamic-pituitary testis axis. Hence, based on the interplay described between the Sertoli cells, the Leydig cells, and the pituitary gland, the reduction observed in systemic testosterone levels in ErbB4^Flox/Flox;MisCre⁺ mice may lead to diminished autocrine and paracrine signaling in this complex and collectively contribute to a disturbance in the feedback loop and at the end lead to the atrophy of the testis. Taken together, the data suggest that the normal function of the ErbB4 in Sertoli cells is to coordinate communication between the Sertoli, Leydig, and germ cells to advance maturation of the testis in situ via control over testosterone production and also indirectly via pituitary FSH.

MIS expression was increased in the Erbb4^Flox/Flox;MisCre⁺ mice; thus, we consider it to be one factor contributing to ErbB4 function in the testis. A reduction in MIS expression is normally associated with entry into puberty and is therefore often used to score the degree of testis maturation during puberty (43). Given that we found elevated amounts of MIS in the Sertoli cells of our Erbb4^Flox/Flox;MisCre⁺ mice, we propose that this explains the compromised maturation observed in the Sertoli cells and the differentiation of the Leydig cells.

In the light of our results, we speculate that the primary function of the ErbB4 receptor in the testis is to contribute to the control of cell adhesion. The Sertoli cells need to be close to the germ cells by virtue of tight junctions, as they are nursing them (33). Similarly the sequential differentiation of germ cells during spermatogenesis needs to be orchestrated from the basal compartment to the adlumenal part of the lumen of the seminiferous tubule. This depends on establishing the proper BTB and proper adhesion between the Sertoli cells and germ cells.

We have demonstrated recently in an embryonic kidney model that ErbB4 signaling is connected with the coordination of cell division, which is important for normal tubulogenesis (4). As in the kidney, ErbB4 deficiency in the Sertoli cells deregulated a panel of cell adhesion parameters. The present immunohistochemical data, for example, depicted changes in the tight junctions and in associated proteins such as connexin-43, claudin-3, and N-cadherin between the Sertoli cells and defects in the BTB, with a misalignment of the spermatids shown by hematoxylin and eosin staining. These changes were also accompanied by a reduction in the components of the basement membrane, such as laminin-1 and type IV collagen. These data suggest that ErbB4 signaling within the Sertoli cells might provide a positional cue for maintaining the polarized epithelial status of these cells. When this process fails, the changes secondarily affect the Leydig cells and their testosterone synthesis.

In summary, this work has demonstrated involvement of signaling by the ErbB4 receptor tyrosine kinase as a novel functional player in the maturation of Sertoli cells. These findings should help us to gain a better understanding of the reasons behind the current reduction in human male fertility.

Funding Statement

This work was supported by the Academy of Finland (Grants 206056, 250900, and FiDiPro 263246 and Centre of Excellence Grant 2012–2017 251314), the Sigrid Jusélius Foundation, the Novo Nordisk Foundation, and the European Union, from the European Community's Seventh Framework Programme (FP7/2007-2013 under grant agreement FP7-HEALTH-F5-2012-INNOVATION-1 EURenOmics 305608). F.N. was also supported by an Academy of Finland Fellowship (243014583).