R

EPRODUCTIONRESEARCH

Plasticity of basal cells during postnatal developmentin the rat epididymis

Winnie W C Shum, Eric Hill, Dennis Brown and Sylvie Breton

Program in Membrane Biology and Nephrology Division, Harvard Medical School, Center for Systems Biology,Massachusetts General Hospital, 185 Cambridge Street Simches Research Center, Boston, Massachusetts 02114, USA

Correspondence should be addressed to W W C Shum; Email: shum.winnie@mgh.harvard.edu

Abstract

Our previous study has shown that basal cells sense luminal factors by forming a narrow body projection that can cross epithelial

tight junctions. As a first step toward characterizing the structural plasticity of basal cells, in this study, we followed their appearance

and morphology in the rat epididymis and vas deferens (VD) during postnatal development and examined their modulation by

androgens in adulthood. Immunofluorescence labeling for cytokeratin 5 showed that basal cells are absent at birth. They

progressively appear in a retrograde manner from the VD and cauda epididymis to the initial segments during the postnatal weeks

PNW1–3. At the onset of differentiation, basal cells are in contact with the lumen and their nucleus is located at the same level as

that of adjacent epithelial cells. Basal cells then position their nucleus to the base of the epithelium, and while some are still in

contact with the lumen, others have a ‘dome-shaped’ appearance. At PNW5–6, basal cells form a loose network at the base of the

epithelium, and luminal-reaching basal cells are rarely detected. The arrival of spermatozoa during PNW7–8 did not trigger the

development of projections in basal cells. However, cells with a narrow luminal-reaching projection began to reappear between

PNW8 and PNW12 in the corpus and the cauda. Treatment with flutamide from PNW10 to PNW12 significantly reduced the number

of luminal-reaching basal cell projections. In summary, basal cells exhibit significant structural plasticity during differentiation.

Fewer apical-reaching projections were detected after flutamide treatment in adulthood, indicating the role of androgens in the

luminal-sensing function of basal cells.

Reproduction (2013) 146 455–469

Introduction

The epididymis is an important organ in the malereproductive tract that performs a variety of functionsincluding sperm concentration, maturation, protection,and storage. Passage through this organ is, therefore,necessary for sperm to acquire their mobility andfertilizing ability (Orgebin-Crist et al. 1975, Robaire &Hermo 1988, Turner 1995, Cornwall 2009). Thesefunctions are carried out by the pseudostratifiedepithelium lining the highly convoluted tubule thatforms the epididymis. This epithelium comprises severalcell types that establish a changing luminal environmentalong the length of the epididymal tubule (Robaire &Hermo 1988, Turner 1991, 2002, Wong et al. 2002,Shum et al. 2011). At least four cell types have beendescribed in the epididymal epithelium: basal, clear,narrow, and principal cells (Sun & Flickinger 1979,Hermo & Robaire 2002). Principal cells are mainlyresponsible for fluid transport and nutrient secretion(Robaire & Hermo 1988, Hermo & Robaire 2002, Wonget al. 2002). Our laboratory has shown that narrow and

q 2013 Society for Reproduction and Fertility

ISSN 1470–1626 (paper) 1741–7899 (online)

clear cells secrete protons via the vacuolar HC-ATPase(V-ATPase) and contribute to the acidification of thelumen, a process that is critical for sperm maturation andviability (Breton et al. 1996, Brown & Breton 2000,Pastor-Soler et al. 2005, Breton & Brown 2007, Shumet al. 2009). The function of epididymal basal cells is lesswell documented, although several roles have beenproposed, including protection of the epithelium frompotentially harmful electrophiles (Veri et al. 1993,Hermo et al. 1994) or from elevated temperatures(Legare et al. 2004), transepithelial fluid transport viaaquaporin 3 (Hermo et al. 2004), immune defense(Yeung et al. 1994, Poulton et al. 1996, Li et al. 2010),and paracrine regulation of principal cell secretion viaPGE2 signaling (Leung et al. 2004, Cheung et al. 2005).The different morphological characteristics of the basalcells indicate that they are highly plastic, varying from adome-shaped cell that nestles at the base of epithelialcells to a cell that extends a long and narrow bodyprojection between adjacent epithelial cells in thedirection of the lumen (Veri et al. 1993, Robaire &Viger 1995, Shum et al. 2008). In addition, we have

DOI: 10.1530/REP-12-0510

Online version via www.reproduction-online.org

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

456 W W C Shum and others

recently shown that these ‘luminal-reaching’ basal cellextensions can cross the tight junctions (TJs) to scan theluminal environment and that basal cells then commu-nicate their findings to neighboring proton-secretingclear cells (Shum et al. 2008). These results providedevidence for the presence of a novel crosstalk betweenbasal cells and clear cells to control acidification of thelumen in the epididymis.

Currently, very little is known about the factors thatcontrol the morphological plasticity of basal cells. Theepididymis of several species including humans androdents is immature at birth and epithelial cells acquiretheir differentiated phenotypes over an extendedpostnatal period (Nilnophakoon 1978, Sun & Flickinger1979, Zondek & Zondek 1980, Francavilla et al. 1987,De Miguel et al. 1998, Rodriguez et al. 2002, Marty et al.2003). Based on morphological studies, the postnataldevelopment of the rat epididymis has been dividedinto three phases, namely an undifferentiated period(days 1–15), a differentiation period (days 16–44), and aperiod of expansion (days beyond 44) (Sun & Flickinger1979). We previously reported that markers specificfor principal cells (AQP9) and narrow and clear cells(V-ATPase) in rats are undetectable at birth and begin tobe expressed during the second postnatal week (PNW2;Breton et al. 1999, Pastor-Soler et al. 2001, Da Silva et al.2006), consistent with the notion that the epididymalepithelium is undifferentiated at birth. Based onimmunoreactivity studies carried out for the Yf subunitof glutathione S-transferase P (Yf-GST), a few basal cellshave been reported to first appear at postnatal day 21 inthe corpus region (Hermo et al. 1994). The transcriptionfactor p63 (also known as TP63 or TPR63), which isessential for the differentiation and maintenance of basalcells in several tissues including the epididymis (Mura-shima et al. 2011), has been detected as early aspostnatal day 10 in the vas deferens (VD; Atanassovaet al. 2005), at postnatal days 14–18 in the cauda region,and at postnatal day 18 in the caput of rats (Hayashi et al.2004, Atanassova et al. 2005). Analysis of conditionalknockout mice in which the androgen receptor wasdeleted in the Wolffian duct revealed a reducedexpression of p63 in the epididymal epithelium whenmeasured at postnatal day 35, indicating an early role ofandrogen in the establishment of basal cells (Murashimaet al. 2011).

As a first step toward understanding the factors thatregulate the structural plasticity of basal cells, in thisstudy, we followed their appearance and morphologyduring postnatal development in the rat epididymis.In adult rats, the integrity and function of the differ-entiated epididymal epithelium are tightly regulatedby androgens (reviewed in Robaire & Hamzeh (2011)).We, therefore, examined the role of androgens in theregulation of the novel luminal-sensing property of basalcells in sexually mature rats after treatment with theandrogen receptor antagonist flutamide.

Reproduction (2013) 146 455–469

Materials and methods

Animals

All animal studies were in compliance with the InstitutionalAnimal Care and Use Committees of the MassachusettsGeneral Hospital and in accordance with National Institutesof Health and Accreditation of Laboratory Animal Carerequirements. Sprague Dawley rats were acquired from CharlesRiver Laboratories (Wilmington, MA, USA).

Flutamide treatment

Ten rats at PNW10 (74 days of age; weight w400 g) weredivided into a control group and a flutamide-treated group. Therats were treated at a dose of 50 mg/kg per day for 2 weeks, asreported previously (Imperato-McGinley et al. 1992, O’Connoret al. 2002, Pastor-Soler et al. 2002, 2010, Atanassova et al.2005). Two flutamide tablets (200 mg/pellet), designed torelease flutamide continuously for 21 days (InnovativeResearch of America, Sarasota, FL, USA), were implantedsubcutaneously into the lateral side of the neck of each rat.Since the weight varied slightly among the rats, the calculateddose of flutamide ranged between 44 and 48 mg/kg per day.A sham procedure was carried out in control rats. Two weeksafter implantation, epididymis, testis, and sex accessory glandswere harvested, as described below.

Tissue fixation and preparation

Adult male rats and rat pups older than 3 weeks wereanesthetized with sodium pentobarbital (80 mg/kg body weight,i.p.). The rats were perfused via the left ventricle with PBS andthen with a paraformaldehyde-lysine-periodate (PLP) fixativecontaining 4% paraformaldehyde, 10 mM sodium periodate,75 mM lysine, and 5% sucrose in 0.1 M sodium phosphatebuffer, as described previously (Breton et al. 1999, Shum et al.2008). The male reproductive organs were harvested and furtherfixed by immersion in PLPovernight. Pups younger than 2 weekswere anesthetized with an overdose of isoflurane, and malereproductive organs were harvested and fixed by immersion inPLP overnight. Tissues were then washed three times with PBSand stored at 4 8C in PBS containing 0.02% Na azide.Epididymides were cryoprotected in PBS with 30% sucrose forat least 4 h at room temperature or overnight at 4 8C, embeddedin Tissue-Tek OCT compound (Sakura Fintek, Torrance, CA,USA), and mounted and frozen on a cutting block. Epididymalsections were cut in a microtome (Leica CM3050-S, LeicaMicrosystems, Bannockburn, IL, USA) at 5–25-mm thicknessand placed on Superfrost Plus microscope slides (FisherScientific, Pittsburgh, PA, USA) and stored at K4 8C until use.

Immunofluorescence staining procedure and antibodies

Single- and double-immunofluorescence labeling wascarried out as described previously (Breton et al. 1999, Shumet al. 2008). For routine staining, epididymal cryo-sectionswere hydrated in PBS for 5 min and incubated in PBScontaining 1% SDS for 4 min, followed by three washes with

www.reproduction-online.org

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

Postnatal regulation of basal cell plasticity 457

PBS (Brown et al. 1996). Non-specific binding was blocked byincubation with 1% BSA in PBS for at least 10 min at roomtemperature. The samples were then incubated with primaryantibodies for at least 60 min at room temperature orovernight at 4 8C. After two washes with high-salt PBS(PBS containing 2.7% NaCl) and one wash with PBS, thesections were incubated with secondary antibodies for atleast 60 min at room temperature. For sections thicker than10 mm, the incubation time for both primary and secondaryantibodies was extended up to 3 h at room temperatureor overnight at 4 8C. The slides were washed as had beendone for the primary antibodies and mounted in Vectashieldmedium (Vector Labs, Burlingame, CA USA) with or without4 0,6-diamidino-2-phenylindole (DAPI) to label the nucleus.

For basal cell identification, a rabbit MAB against theC-terminus of KRT5 (1:200) (Thermo Scientific, Rockford, IL,USA) and a mouse MAB against human cytokeratin 5/6 (DAKOCytomation, Carpinteria, CA, USA) were applied. An antigenretrieval technique using three to four microwave treatments of1–2 min with 5-min intervals, in a buffer containing 10 mM Trisand 1 mM EDTA (pH 9.0), was used to reveal KRT5 staining.For sections of 12-mm thickness or more, 1% saponin wasincluded in all the incubation steps. A rat monoclonal anti-ZO1 antibody (R40/76) (1:10 dilution), provided by EvelineSchneeberger (Department of Pathology, MassachusettsGeneral Hospital), and a rabbit polyclonal antibody againstclaudin-1 (Zymed, Life Technologies, Grand Island, NY,USA; 1:200 dilution) were also used as published previously(Shum et al. 2008). For clear cell identification, affinity-purifiedrabbit or chicken polyclonal antibodies against the B1 subunitof V-ATPase (B1-V-ATPase) (diluted 1:1000) were used aspublished previously (Breton et al. 1999, Shum et al. 2008).All secondary antibodies used in this study were affinitypurified and were purchased from Jackson ImmunoResearchLaboratories (West Grove, PA, USA). FITC- and CY3-conjugated donkey or goat anti-rabbit IgG; FITC-, CY3-, andCY5-conjugated donkey anti-chicken; and FITC- and CY3-conjugated goat anti-mouse IgG. All the antibodies werediluted in Dako antibody diluent (Dako).

Microscopy and image analysis

Conventional immunofluorescence images were obtainedusing a Nikon Eclipse E90i advanced automated microscopesystem, equipped with an Orca-100 CCD camera (Hamamatsu,Bridgewater, NJ, USA), and the NIS-Elements software. Mosaicimages were obtained using the automated module of the NikonEclipse 90i microscope. For widefield three-dimensional (3D)reconstruction, Z-series images were taken with a Nikon 90imicroscope. For confocal 3D reconstruction, Z-series imageswere taken with a Nikon A1R microscope using the NIS-Elementssoftware. The Z-stack images were then imported into theVolocity software (version 5.1) for final animations and exportedas TIFF projection pictures or as QuickTime movies. The finalpresentation of the projection pictures was made using the AdobePhotoshop software, and the levels command was used to adjustthe entire images to represent the raw data visualized underthe microscope.

www.reproduction-online.org

For basal cell quantification, 12-mm epididymal sectionsfrom the control and flutamide-treated groups were used, anda 10-mm stack of Z-series of widefield immunofluorescenceimages was acquired at 2-mm intervals using the 90i Nikonmicroscope with a 20! objective. Images were processedand 3D-reconstructed using the Volocity 5.1 software.Quantification was carried out on the projected stack ofimages. Each data set represented four images per rat, withthree rats per group.

Statistical analysis

The epididymides of at least three rats were examined per timepoint and three cryostat sections were analyzed for eachepididymis, unless otherwise stated in the text. Comparisonsbetween multi-groups were made by Student’s t-tests (forstudies including only two groups: e.g., flutamide treatment) orby two-way ANOVA followed by Bonferroni’s post hoc testusing the Prism 5 software (La Jolla, CA, USA). All tests weretwo-tailed unless otherwise stated, and the limit of statisticalsignificance was set at PZ0.05.

Results

Cytokeratin 5 is a marker of epididymal basal cells

Cytokeratin 5 is a putative basal cell marker inpseudostratified epithelia (Gusterson et al. 2005, Rocket al. 2009), and we confirmed in this study that it labelsbasal cells in the epididymis. Confocal immunofluores-cence pictures (Fig. 1A, B, and C) showed that basal cellsin the epididymis of adult rats, identified by theirexpression of claudin-1, a known marker of basal cellsin the epididymis (Gregory et al. 2001, Shum et al. 2008),also express high levels of intracellular KRT5. A highermagnification 3D reconstruction picture showed thatclaudin-1 is highly expressed in the membrane of basalcells and is also present at lower levels in the TJs, whereasKRT5 is exclusively expressed in the cytoplasm of basalcells (Fig. 1D). Consistent with previous observations(Veri et al. 1993, Robaire & Viger 1995, Shum et al.2008), numerous basal cells in the corpus epididymidishave a narrow body projection that extends toward thelumen and some of these projections reach the TJs (arrowsin Fig. 1C and D). No labeling was obtained when theKRT5 primary antibody was omitted (data not shown).Having determined that KRT5 is a specific marker ofbasal cells in the epididymis, we then followed theirappearance and morphology during postnatal develop-ment in epididymal sections labeled for KRT5.

PNW1–3

Mosaic images representing the proximal VD and entireepididymis of rats during early postnatal developmentshowed a progressive ‘ascending’ appearance of basalcells, identified by their positive labeling for KRT5, fromthe VD at the end of PNW1 (Fig. 2A; arrows) to the distal

Reproduction (2013) 146 455–469

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

PNW1

VD

CD

CPT

dCD

CPSCPS

CPT

A C

B

PN

Figure 2 Progressive appearance of basal cells in the VD and then in the caMosaic images showing that basal cells (labeled for KRT5) progressively apepididymidis at PNW2 (C), and eventually in all the epididymal regions at Pand only occasional basal cells can be detected in the proximal cauda, corcauda; CPS, corpus; CPT, caput. BarsZ500 mm.

Lu Lu

KRT5

KRT5

A B

C D

Claudin-1

Claudin-1

Lu Lu

LuLu

Lu Lu

Lu

Lu

Figure 1 Expression of KRT5 in epididymal basal cells. (A, B, and C)Projection pictures of a stack of 13 1-mm-interval confocal imagesshowing a rat corpus epididymal section doubled-labeled for KRT5(red in panels A and C) and the basal cell marker claudin-1 (green inpanels B and C). KRT5 is expressed intracellularly in basal cells,whereas claudin-1 is present in the basal cell membrane. Claudin-1 isalso present at lower levels in TJs. (D) Higher magnification 3D viewshowing three basal cells, double-labeled for KRT5 (red) and claudin-1(green), that extend a narrow body projection toward the lumen. DNAin spermatozoa and cell nuclei is visualized with DAPI in blue. Arrowsin panel D indicate the 3D X (green), Y (red), and Z (blue) axes, and 3Dspatial representation is illustrated by the cross-hatched sphere.Lu, lumen. Scale bars: 20 mm in A, B, and C and 5 mm in D.

458 W W C Shum and others

Reproduction (2013) 146 455–469

cauda epididymis at the end of PNW2 (Fig. 2C; arrows)and in most regions of the epididymis at the end ofPNW3 (Fig. 2D and E). Only occasional basal cells weredetected in the proximal cauda, corpus, and caput atPNW2 (Fig. 2B and C). Labeling for COX-1 or claudin-1,two other markers of basal cells (Gregory et al. 2001,Cheung et al. 2005, Shum et al. 2008), showed similarlabeling patterns (data not shown).

Higher magnification images showed that, atPNW1 (day 7), basal cells formed a continuous layerat the base of the epithelium in the VD (Fig. 3A).In agreement with our previous studies (Breton et al.1999), the VD also contained numerous clear cellslabeled for V-ATPase in their apical pole. At this earlytime point, basal cells in the VD showed variablemorphological patterns (Fig. 3A, inset): while most cellshad a dome-shaped appearance and were entirelylocated at the base of the epithelium, others had a longand narrow body projection that extended betweenadjacent epithelial cells toward the luminal border ofthe epithelium (arrows). In the distal cauda epididymi-dis, basal cells (arrow) and clear cells (arrowheads)were only occasionally detected (Fig. 3A 0), and bothcell types were still absent in the proximal caudaepididymidis (Fig. 3A 00) and more proximal regions ofthe epididymis (data not shown). On day 10, the distalcauda was populated with many basal cells (Fig. 3Band B 0). Similarly to that observed at earlier timepoints in the VD, basal cells exhibited severalmorphological patterns: some basal cells had amorphological appearance resembling that of adjacentcolumnar epithelial cells (arrows), some had theirnucleus located in the basal pole of the epithelium and

CPT

CD

CPSpCD

E

D

CPS

CPS

W2 PNW3

uda, corpus, and caput epididymidis during the first 3 postnatal weeks.pear first in the VD at PNW1 (A; arrows), then in the distal caudaNW3 (D and E). No basal cells are seen in the epididymis at PNW1 (A),pus, and caput at PNW2 (B and C). dCD, distal cauda; pCD, proximal

www.reproduction-online.org

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

A A′

B′B

C′C

A′′

Day

7D

ay 1

0D

ay 1

4

Vas deferens Distal CD Proximal CD

Proximal CD

Distal CPS

Distal CD

Distal CD

Proximal CDDistal CD

Caput

Figure 3 PNW1–2. (A, A 0, and A 00) Higher magnification pictures of theVD and epididymis of a 1-week-old (day 7) rat double-labeled for KRT5(red) and V-ATPase (green) showing that in the VD (A), a continuouslayer of basal cells covers the basal side of the epithelium, and numerousclear cells (positive for V-ATPase) are present. At this early time point,some basal cells have a long and narrow body in contact with theluminal surface of the epithelium (A, inset; arrows). Occasional basalcells (arrow) and clear cells (arrowheads) are visible in the distal cauda(A0), but they are still absent in the proximal cauda (A 00), corpus, andcaput (data not shown). (B and B 0) At day 10, numerous basal cells andclear cells are present in the distal cauda. Some basal cells resembleadjacent epithelial cells and have their nucleus located at the sameheight (arrows), some basal cells have their nucleus located at the baseof the epithelium and have a narrow body extension that still appears tobe in contact with the luminal side of the epithelium (arrowheads), andsome basal cells have their characteristic dome-shaped morphology(double arrows). (C and C 0) At day 14 (PNW2), several basal cells can bedetected in the proximal cauda and the caput. At this time point, theyresemble adjacent epithelial cells and have nuclei that are located at thesame level as those of adjacent epithelial cells (insets in C and C 0).Images shown in all the insets and panel B 0 are from 10-mm projectedstacks of Z-series optical sections. Nuclei are labeled in blue with DAPI.Scale bars: 100 mm in A and 20 mm in A 0, A 00, B, B 0, C, and C 0.

Postnatal regulation of basal cell plasticity 459

extended long and narrow body projections toward thelumen (arrowheads), and some were dome shapedentirely located underneath other epithelial cells(double arrows). At PNW2 (day 14), basal cells formeda continuous network covering the base of the epithelium

www.reproduction-online.org

in the distal cauda epididymidis (Fig. 3C). Interestingly, inthe proximal cauda (Fig. 3C; inset) and the caput (Fig. 3C 0

and inset) where only rare differentiated cells started toappear, all basal cells had a morphological appearancesimilar to that of adjacent columnar epithelial cells(inset). At this time point, V-ATPase-positive clearcells were highly abundant in the distal cauda and lessabundant in the proximal cauda and the caput.

Epididymis at PNW3–6

As shown in Fig. 4, basal cells continued to appearprogressively in an ascending pattern from the distalregions (cauda) to the more proximal regions of theepididymis (corpus, caput, and initial segments) fromPNW3 to PNW6. By PNW5, numerous basal cellsformed a continuous layer at the base of the epitheliumin all the regions, except in the initial segments, wheresome tubule regions devoid of basal cells were stilldetected (Fig. 4O; arrowheads). Interestingly, themorphology of basal cells became more uniform as theepididymis matured, and by PNW5 and PNW6, onlyrare basal cells with a luminal-reaching projection werestill detected (arrows in Fig. 4O, R, and S). Some cellswere detected in the lumen, and several of them wereweakly labeled for V-ATPase, indicating shedding ofclear cells (Fig. 4M, N, and T, asterisks).

In agreement with previous studies (Veri et al. 1993,Robaire & Viger 1995), basal cells extended lateral bodyprojections underneath adjacent epithelial cells, and atPNW5, they formed a remarkable network covering alarge area of the base of the epithelium (Fig. 5A and B;see Supplementary Movies S1 and S2, see section onsupplementary data given at the end of this article).A similar ‘loose’ basal cell network was also detected inadult rats (Fig. 5C; also see Supplementary Movie S3).

Epididymis at PNW7–12

At the peripubertal PNW7, spermatozoa were detectedin the caput (Fig. 6A) and proximal corpus (Fig. 6B),but they were still absent in the distal corpus (Fig. 6C)and the entire cauda (Fig. 6D). This confirmed thatPNW7 corresponds to the arrival of the first waveof sperm in the epididymis, as shown previously(Nilnophakoon 1978, Sun & Flickinger 1979, Scheer &Robaire 1980). In addition, a large amount of cell debriswas detected in the corpus and cauda (Fig. 6B, C, and D),indicating significant shedding of epithelial cells. At thistime point, most basal cells had a dome-shapedappearance and were located underneath adjacentepithelial cells. Only a few basal cells were observedwith a body extension that reached the apical pole of theepithelium (arrows in panels B and C). The presence ofthese luminal-reaching cells did not coincide with thearrival of spermatozoa in the lumen.

Reproduction (2013) 146 455–469

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

Week 3

A F K P

B G L Q

C H M R

D I N S

E J O T*

*

*

*

Dis

tal C

DP

roxi

mal

CD

CP

SC

PT

IS

Week 4 Week 5 Week 6

Figure 4 Epididymis at PNW3–6. At the end of PNW3 (A, B, C, D, and E), basal cells (labeled for KRT5 in red) form an almost continuous layer at thebase of the epithelium in the distal cauda (A). At this time point, a large number of them have a long and narrow body that extends toward thelumen. In the proximal cauda (B), corpus (C), caput (D), and initial segments (E), basal cells appear as individual cells and most of them have atriangular shape and their nucleus is located at the same level as that of the adjacent cells. Almost all of them are in contact with the luminal side ofthe epithelium. At this time point, numerous V-ATPase-positive clear cells (green) are present in all the segments. At PNW4 (F, G, H, I, and J),most basal cells have a hemispheric shape and are located underneath other epithelial cells. Only a few of them have a luminal-reachingbody extension. In the distal cauda, basal cells form a continuous layer at the base of the epithelium (F), while in the initial segments sometubule regions still do not contain basal cells. At PNW5 (K, L, M, N, and O), basal cells form a continuous layer at the base of the epithelium, exceptin the initial segments, where discrete tubule regions can be observed without basal cells (arrowheads in O). Very few basal cells with aluminal-reaching body extension can be detected (arrows in K and O). Some epithelial cells are present in the lumen (asterisks in M, N, and T). Theseluminal cells are weakly labeled for V-ATPase. At PNW6 (P, Q, R, S, and T), in the distal and proximal cauda, a few clusters of basal cells havetheir body extensions reaching the apical surface of the epithelium (curved arrows in Pand Q), while other basal cells have a triangular shape and arelocated beneath other epithelial cells. Some apical-reaching basal cells can also be seen in the corpus and caput regions, but their occurrenceis very low (arrow in R and S). Nuclei are labeled with DAPI in blue. CD, cauda; CPS, corpus; CPT, caput; IS, initial segment. Scale bar: 20 mm.

460 W W C Shum and others

Reproduction (2013) 146 455–469 www.reproduction-online.org

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

Week 5 Distal CD Week 5 Caput Week 12

A B C

Cauda

Figure 5 Formation of a cellular network at the base of the epididymal epithelium by basal cells. 3D view from a series of confocal images ofthe distal cauda (A) and caput (B) at PNW5 showing that basal cells form a network that covers a large area of the base of the epithelium. (C) A similarpattern is observed in adult rats (PNW12). Numerous clear cells (green) are also present. Arrows in panels A and B indicate the 3D X (green),Y (red), and Z (blue) axes, and 3D spatial representation is illustrated by the cross-hatched sphere (also see Supplementary Movies S1, S2, and S3).Images are from projected stacks of Z-series optical sections at different thicknesses (A: 21 mm; B: 25 mm; and C: 23 mm). Scale bars: 25 mm inA and 20 mm in B and C.

A B

**

*

CPT pCPS

**

*DC

dCDdCPS

Figure 6 Epididymis at PNW7. At PNW7, spermatozoa with normalmorphology are present in the lumen of the caput (A) and in some of theproximal corpus regions (B), but they are still absent in the distal corpus(C) and cauda epididymidis (D). Double labeling for KRT5 (red) andV-ATPase (green) showed that the vast majority of basal cells have ahemispherical shape and that only rare basal cells have a luminal-reaching body projection (arrows in B). No relationship was observedbetween the presence of sperm in the lumen and the formation of basalcell luminal-reaching body projections. In addition, some cells can bedetected in the lumen prior to the arrival of spermatozoa in theproximal and distal corpus and in the cauda epididymidis (asterisks inB and D). Some, but not all, are weakly labeled for V-ATPase. Nucleiare labeled with DAPI in blue. CPT, caput; pCPS, proximal corpus;dCPS, distal corpus; dCD, distal cauda. Scale bars: 20 mm.

Postnatal regulation of basal cell plasticity 461

At the end of PNW8, spermatozoa with normalmorphology were present throughout the entire lengthof the epididymis (Fig. 7), whereas the VD was still free ofspermatozoa (not shown). At this time point, basal cellswith a luminal-reaching extension started to reappear inthe distal corpus (arrow in Fig. 7A) and were moreprominent in the cauda epididymidis (arrows in Fig. 7B).In the other epididymal regions, most basal cellsremained dome shaped in appearance and were locatedunderneath adjacent epithelial cells (data not shown).No cell debris was detected in the lumen.

At PNW9, similarly to PNW8, in the initial segments andcaput, most basal cells had a dome-shaped morphologywith no distinct luminal-reaching extensions (not shown),and in the distal corpus, a few basal cells exhibited a bodyextension that reached towards the lumen (Fig. 8A, arrow).At this time point, fewer luminal-reaching basal cells weredetected in the distal cauda compared with PNW8 (notshown). The number of luminal-reaching basal cellscontinued to increase at PNW10 and PNW12 in thecorpus (Fig. 8B and C; arrows and arrowheads). Quan-titative analysis confirmed that the percentage of basalcells with body projections that reached the apical pole ofthe epithelium (defined as the body projections reachingthe apical region located above the nuclei of adjacentprincipal cells; arrowheads in panels B and C) becameprogressively higher in the proximal, middle, and distalcorpus at PNW10 and PNW12 compared with PNW9(Fig. 8D, left panel). The percentage of basal cells thatextended all the way to the apical border (arrows in panelsA, B, and C) was also significantly higher at PNW10 and12 compared with PNW9 (Fig. 8D, right panel).

In agreement with our previous report (Shum et al.2008), we found that basal cells extend their narrow bodyprojections toward the junctions between three epithelialcells (tricellular corners; Fig. 9). Several patterns ofinteraction with the TJ barrier were detected. Somebasal cells had body projections that reached the tripartitejunction, but did not cross it (Fig. 9A; see SupplementaryMovie S4, see section on supplementary data given at theend of this article), and some basal cells had bodyprojections that crossed the TJs to be in contact with the

www.reproduction-online.org

luminal environment of the epididymal tubule (Fig. 9B;see Supplementary Movie S5). Luminal-reaching basalcells were predominantly observed in the distal corpusand proximal cauda epididymidis.

In addition, from PNW9 to PNW12, we observed agradual decrease in the height of clear cells in the distalcorpus epididymidis from values of 29.1G1.5 mm atPNW9 and 27.5G0.7 mm at PNW10 to 24.5G0.6 mmat PNW12, representing a 12% decrease at PNW12 vsPNW9 (P!0.05). However, we observed no significantchanges in the number of basal cells per unit length ofcircumference and in the percentage of basal cells or

Reproduction (2013) 146 455–469

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

dCPS

dCD

A

B

Figure 7 Epididymis at PNW8. (A) At PNW8, in the distal corpus,almost all the basal cells, labeled for KRT5 (red), have a triangularshape and are located beneath adjacent epithelial cells and very fewcells have a body projection that extends toward the lumen (arrow).(B) In the distal cauda, a considerable number of basal cells have aluminal-reaching projection (arrows). Nuclei are labeled with DAPI inblue. Scale bars: 50 mm.

462 W W C Shum and others

clear cells per total number of epithelial cells per unitlength of circumference (data not shown). These resultsshowed that while the epididymal tubule has reached itsfinal cell composition at PNW9, basal cells and clearcells were still undergoing postpubertal remodeling.

Regulation of the structural plasticity of basal cells byandrogens in adulthood

Our results indicate that basal cells exhibit dynamicmorphological plasticity during postnatal development.Epididymal development and maturation are androgendependent, and we next determined whether theformation of apical-reaching body extensions in basalcells is modulated by androgens. We were particularlyintrigued by the plasticity of basal cells during the last

Reproduction (2013) 146 455–469

stages of maturation, between PNW9 and PNW12. We,therefore, treated 10-week-old rats with flutamide for aperiod of 2 weeks and compared their basal cells withthose of 12-week-old control rats. The effectiveness offlutamide treatment was confirmed by a significantdecrease in the intensity of AQP9 in the principal cellsof the cauda epididymidis (data not shown), as reportedpreviously (Pastor-Soler et al. 2002). We also detected asignificant reduction in the weight of epididymis andaccessory sex glands in the flutamide-treated ratscompared with control rats (Table 1). A shown inFig. 10, a marked reduction in the number of luminal-reaching basal cells was observed in the distal corpusepididymidis of the flutamide-treated rats (Fig. 10B, B 0,and B 00) compared with control rats (Fig. 10A, A 0, and A 00).This observation was confirmed by quantitative analysis(Fig. 10C) (P!0.01, two-way ANOVA). We found nochange in the total number of basal cells per unit lengthof circumference, in the percentage of basal cells orclear cells vs total number of epithelial cells, or in theheight of clear cells after flutamide treatment comparedwith control rats (data not shown).

Vas deferens

Examination of the VD during the postnatal andperipubertal periods showed that by day 10 most basalcells had lost their contact with the luminal border andthat they maintained a dome-shaped appearance all theway through adulthood (Fig. 11A B, C, D, E, F, G, H,and I). During this time period, they remained abundantand formed a continuous layer at the base of theepithelium, consistent with previous reports (Atanassovaet al. 2005). By contrast, a progressive decrease in thenumber of V-ATPase-labeled clear cells was observedfrom PNW5 to PNW12. No cell debris was observed inthe VD lumen from day 10 to PNW4. At PNW5, severalV-ATPase-positive vesicles were detected in the VDlumen (Fig. 11D, arrows), whereas no vesicles weredetected in the lumen of the epididymis at this time point(data not shown). At PWN6 and PNW7, cell debris andentire cells – several being weakly labeled for V-ATPase –were detected in the lumen (Fig. 11E and F). By PNW8,the VD lumen was devoid of cell debris, and sperma-tozoa were still absent in this segment (Fig. 11G). AtPNW9, spermatozoa with normal morphology weredetected in the lumen, but dome-shaped basal cellsremained. As the epithelium matured from day 10 toPNW7, its overall height became progressively smaller(Fig. 11A, B, C, D, E, and F). Treatment of rats fromPNW10 to PNW12 with flutamide did not alter themorphology of basal cells or the overall height of theepithelium (measured as the height of clear cells; 17.4G1.3 mm for the control group vs 15.3G0.9 mm for theflutamide-treated group, from at least 15 cells per rat forfour rats; PZ0.22).

www.reproduction-online.org

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

Week 9, dCPS Week 12, dCPS

A B C

D60

50

40

30

20

10

0pCPS dCPS

**

*

**

*

mCPS pCPS dCPSmCPS

***

******

*

Per

cent

age

of b

asal

cel

lsre

achi

ng th

e ap

ical

pol

e

40

30

20

10

0P

erce

ntag

e of

bas

al c

ells

reac

hing

the

apic

al b

orde

r

Week 10, dCPS

Week 9 Week 10 Week 12

Figure 8 Epididymis at PNW9, 10, and 12. In the distal corpus, the number of basal cells (labeled for KRT5; red) with luminal-reaching extensionsincreases progressively from PNW9 (A) to PNW10 (B) and PNW12 (C). Some basal cells show a body projection that reaches the apical pole of theepithelium (arrowheads) (defined as the region located above the nuclei of adjacent V-ATPase-negative principal cells (negative for the clear cellmarker V-ATPase (green)). In some basal cells, the extension reaches all the way to the apical border of the epithelium (arrows). Nuclei are labeled withDAPI in blue. Scale bars: 25 mm. (D) Left: quantitative analysis showing that the number of basal cells with body projections reaching the apical pole ofthe epithelium (defined as the region located above the nuclei of adjacent epithelial cells) progressively increases from PNW9 to PNW12 in theproximal, middle, and distal corpus. At PNW12, more luminal-reaching cells can be detected in the middle and distal corpus than in the proximalcorpus. Right: the number of basal cells that project their body extensions all the way to the apical border of the epithelium also increased from PNW9to PNW12 in all the corpus regions examined. These numbers were normalized for the total number of basal cells present in each tubule sectionexamined. *P!0.05, **P!0.01, and ***P!0.001 by two-way ANOVA. pCPS, proximal corpus; mCPS, middle corpus; dCPS, distal corpus.

Postnatal regulation of basal cell plasticity 463

Discussion

In this study, we found that basal cells in the rat VD andepididymis appear within the epithelia in a spatio-temporal manner during the first 5 PNWs. These resultsare in agreement with those of previous studies showingthat these epithelia are immature at birth and thatepithelial cells acquire their differentiated charac-teristics over a period of several PNWs (Sun & Flickinger1979, Hermo et al. 1994, Breton et al. 1999,Pastor-Soler et al. 2001, 2010, Rodriguez et al. 2002,Da Silva et al. 2006). In particular, using the basal cellmarker KRT5, we found that basal cells first appear inthe VD at PNW1, then in the cauda at PNW2, andsubsequently in more proximal segments of the ratepididymis during PNW3–5. In the mouse epididymis,a similar pattern of retrograde maturation has also beenreported (Seiler et al. 1998), although p63-positive basalcells have been detected as early as postnatal day 5(Murashima et al. 2011). By the end of PNW5, basalcells formed a uniform basal layer in all the epididymalregions, except in the initial segments, where thisoccurred at later time points. As described previouslyin adulthood (Veri et al. 1993, Robaire & Viger 1995),basal cells then sent lateral extensions underneath

www.reproduction-online.org

adjacent epithelial cells and they formed a loosenetwork at the base of the epithelium.

Interestingly, the number of basal cells remained highin the VD from PNW1 to PNW12, while the number ofclear cells progressively decreased during the PNW5–8period. The reduction in the number of clear cells wasobserved together with the appearance of severalV-ATPase-labeled cells in the lumen, indicating selectiveclear cell shedding. This phenomenon is similar to theremoval of proton-secreting V-ATPase-expressing inter-calated cells, which also occurs in part via luminalshedding in the developing rat kidney (Kim et al. 1996).

Origin of basal cells

It has been proposed that during postnatal developmentepididymal basal cells originate either from an extra-tubular source (Holschbach & Cooper 2002) or from thedifferentiation of immature columnar epithelial cells(Sun & Flickinger 1982). In the present study, we foundthat at the onset of basal cell differentiation (e.g., atPNW1 in the VD, during PNW2 in the cauda andcorpus epididymidis, and at PNW3 in the caput), thenewly detectable KRT5-positive basal cells exhibit

Reproduction (2013) 146 455–469

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

A

B

Figure 9 Basal cells send projections toward the epididymal lumen.3D reconstructions of corpus epididymal sections of a 12-week-oldrat labeled for KRT5 (red) and ZO1 (green). (A) A basal cell extendingits body projection to reach the tight junction at the intersectionbetween three epithelial cells (arrow; see Supplementary Movie S4).This cell is not in contact with the lumen. (B) A basal cell (labeled inred for KRT5) having its narrow body projection in contact with theluminal side of the epithelium. The tricellular corner formed by thetight junction (labeled in green for ZO1) located at the point ofcontact of the basal cell extension is open (arrow), while othertripartite junctions (arrowheads) are tightly connected(see Supplementary Movie S5). Colored arrows in panels A and Bindicate the 3D X (green), Y (red), and Z (blue) axes, and 3Dspatial representation is illustrated by the cross-hatched sphere.DNA in sperm and nuclei is visualized with DAPI in blue.Scale bars: 10 mm.

464 W W C Shum and others

Reproduction (2013) 146 455–469

morphological features similar to adjacent columnarepithelial cells. At this early stage, their nucleus islocated at the same level as that of adjacent cells andtheir cell body occupies the entire height of theepithelium, and they are, therefore, in contact withboth the luminal and basolateral sides. These resultsindicate that basal cells may originate from the non-differentiated columnar epithelial cells, rather than froman extratubular source. In the study carried out byHolschbach & Cooper (2002), one of the criteria used toidentify basal cells was the location of their nucleus atthe base of the epithelium. The proposal that basal cellsoriginate from a non-epithelial source was based onseveral observations including the absence of observedmitotic figures at the base of the epithelium and the factthat after BrdU incorporation ‘division of epithelial cellsinto daughter cells was circumferential (expected ifprincipal cells were to give rise to daughter principalcells) and not radial (expected if principal cells were togive rise to basal cells). However, the availability of newbasal cell-specific markers allowed us to determine thatepididymal basal cells cannot be solely identified by thelocation of their nucleus, especially during earlypostnatal development when they have their nucleuslocated at the same height as adjacent epithelial cells. Inaddition, we have previously reported that dendriticcells, the nuclei of which are also located at the base ofthe epithelium, populate the epididymal tubule (Da Silvaet al. 2011). These cells express several markers,including ITGAX (CD11c) and CX3CR1, and a sub-population also expresses the macrophage marker EMR1(F4/80). This protein has been used in previous studies toidentify basal cells in the murine epididymis (Seiler et al.1998, 2000), illustrating the difficulty in identifying theexact cell types that populate the base of the epididymaltubule. In addition, we observed no migratory KRT5-positive cells crossing the basement membrane into theVD and epididymal epithelia, a typical phenomenon forcells coming from the circulation such as dendritic cellsand macrophages. We, therefore, propose that basalcells arise from the differentiation of the primitivecolumnar epithelial cells, as has also been suggested in

Table 1 Absolute and relative weights of the reproductive organs inadult control and flutamide-treated rats.

Group Control Flutamide treated

Body weight (g) 394G9 414G7Testes (g/100 g BW) 0.403G0.011 0.395G0.021Epididymis (g/100 g BW) 0.1288G0.0067 0.1120G0.0074*Ampullary gland (g/100 g BW) 0.0128G0.0007 0.0101G0.006†

Coagulatory gland (g/100 g BW) 0.0700G0.0066 0.0499G0.0055*Prostate (g/100 g BW) 0.2554G0.0215 0.1932G0.0197†

Seminal vesicle (g/100 g BW) 0.2943G0.0097 0.1971G0.0190‡

Values are means derived from five rats (GS.E.M.) and determined asreported previously (Pastor-Soler et al. 2002). *P!0.05, †P!0.01,and ‡P!0.001 by unpaired t-test, significantly different from thecontrol value.

www.reproduction-online.org

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

Postnatal regulation of basal cell plasticity 465

other studies (Martan 1969, Sun & Flickinger 1979,1982, Francavilla et al. 1987).

Morphological plasticity of basal cells duringpostnatal development

Throughout postnatal development, basal cells exhibitvariable morphological features. When they first appearwithin the epithelium (e.g., in the distal cauda at PNW1and in the caput at PNW2), basal cells resemble adjacentundifferentiated columnar cells and they are in contactwith both the apical and basolateral sides of theepithelium. At a later stage when more basal cells (andclear cells) are visible (e.g., in the corpus and caput atPNW3), the apical pole of basal cells becomes narrower,but their nucleus is still located at the same height as

Control

Flutamide

A′′

A′A

B′B

70C

60

50

40

30

20

Per

cent

age

of b

asal

cel

ls

10

0Reachingapical pole

Control+Flutamide

382(1214)

84(1214)

546(1172)

233(1172)

**

***

Reachingapical border

B′′

www.reproduction-online.org

adjacent cells. As basal cells progress toward theirmature phenotype (from PNW4 and onward in theepididymis), they position their nucleus at the base ofthe epithelium. During PNW5, 6, and 7 the vast majorityof epididymal basal cells have a hemispherical shape andonly rare ‘luminal-reaching’ body extensions are detect-able. No obvious morphological difference was detectedin basal cells at the time of the arrival of spermatozoa inthe lumen of the epididymis and VD, i.e., at pre- andperipubertal time periods. Thus, the luminal-reachingability of basal cells does not seem to be regulated by thepresence of luminal spermatozoa or other luminalfactors that might be delivered to the lumen at the timeof puberty.

At birth, the plasma concentration of androgens in ratsdeclines sharply from the high fetal levels to a value as lowas 1 nM, and it remains low during the first 3 PNWs. Thetesticular androgen concentration, while still relativelyhigh at birth, also declines by 100-fold during the first5 PNWs (Lee et al. 1975, Paz et al. 1980, Corpechot et al.1981). Thus, our observation that basal cells undergomajor morphological changes prior to puberty indicateseither that low androgen levels are sufficient to initiate andmaintain epithelial cell differentiation or that thesedifferentiation processes are regulated by factors otherthan androgens. We have previously shown that theexpression of two markers of epididymal cell differen-tiation, V-ATPase in clear cells and AQP9 in principal cells,is significantly reduced in the epididymis of 25-day-oldrats that had been treated during the perinatal period with

Figure 10 The apical-reaching property of basal cells is androgendependent. (A) A mosaic image showing the distal corpus epididymidisof a control male rat at PNW12. At this age, numerous basal cells(labeled for KRT5) extend their body projections toward the lumen.(A 0 and A 00) Higher magnification of the distal corpus epididymidisshowing many basal cells (labeled for KRT5, red) extending their longbody projections toward the lumen. Clear cells are labeled in green forV-ATPase and nuclei are visualized with DAPI. (B) A mosaic imageshowing the distal corpus epididymidis of a 12-week-old rat that hadbeen treated for 2 weeks with the androgen antagonist flutamide at adose of 50 mg/kg. A significant decrease in the number of basal cellsextending their body projections toward the lumen was observed.(B 0 and B00) Higher magnification image showing that the distal corpusepididymidis has much fewer apical-reaching basal cells afterflutamide treatment, while the total number of basal cells per tubule isnot affected. Clear cells are labeled in green for V-ATPase and nucleiare visualized with DAPI. Scale bars: 250 mm in A and B; 50 mm inA 0 and B 0; and 10 mm in A 00 and B 00. (C) Quantitative analysis showingthat in the distal corpus epididymidis 48G3% of basal cells reachthe apical pole of the epithelium and 22G5% of them can actuallyreach the luminal border. After flutamide treatment, the number ofbasal cells showing body projections reaching the apical pole andreaching the luminal border was significantly decreased to 30G3 and7G2% respectively. Data were obtained from 10-mm stacks of Z-seriesimages acquired at 2-mm intervals using a 90i NIKON microscopewith a 20! objective. Images were processed using Volocity 5.1.Data represent those obtained from the analysis of four integratedimages from five rats per group. **P!0.01 and ***P!0.001 bytwo-way ANOVA.

Reproduction (2013) 146 455–469

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

Day 10 Week 3 Week 4

Week 5

A B C

D E F

G H J

I

Week 6 Week 7

Week 8 Week 9 Week 12, Control

Week 12, Flutamide

Figure 11 Proximal VD from postnatal day 10 toadulthood. Basal cells, labeled for KRT5 in red, in theproximal VD have a triangular shape and form acontinuous layer at the base of the epithelium frompostnatal day 10 and onward. V-ATPase-labeled clearcells (green) are abundant from postnatal day 10 toPNW4 (A, B, and C). From PNW5 to PNW12, aprogressive reduction in the number of clear cells isobserved. Several V-ATPase-labeled vesicles can bedetected in the VD lumen at PNW5, and luminal cellsand cell debris are observed at PNW7 (F). Several ofthese luminal cells are weakly labeled for V-ATPase.At PNW8 (G), sperm are still absent in the lumen, butthey are abundant at PNW9 (H). Flutamide treatmentfrom PNW10 to PNW12 did not alter the structuralfeatures of basal cells or the height of the epithelium(J) of 12-week-old rats compared with control rats (I).DNA in sperm and nuclei is visualized with DAPI inblue. Scale bars: 25 mm in A, B, I, and J and 50 mm inC, D, E, F, G, and H.

466 W W C Shum and others

a gonadotrophin-releasing hormone (GNRH) antagonist(GNRHa), the estrogenic compound diethylstilboestrol(DES), or flutamide (Fisher et al. 2002, Pastor-Soler et al.2010). The expression of these markers was maintained atcontrol levels when testosterone was administeredtogether with DES. We concluded that the effects elicitedby DES were due to the inhibition of androgen stimulationsecondary to the elevation of estrogen levels. By contrast, aprevious study has shown that neonatal administration ofGNRHa or flutamide alone did not affect the numberof basal cells in the cauda epididymis and VD of rats,whereas the appearance of basal cells is impaired byneonatal exposure of DES (Atanassova et al. 2005).Interestingly, testosterone administration also maintainedthe number of basal cells at control values in rats treatedwith DES, and it was concluded that disruption of thenormal androgen–estrogen balance during the neonatalperiod mediated the delayed basal cell differentiationinduced by DES. During embryological development,androgens play an essential role in driving the differen-tiation of the Wolffian duct to epididymis (Ezer & Robaire2002, Rodriguez et al. 2002, Robaire et al. 2007, Welshet al. 2009). Deletion of the androgen receptor in theWolffian duct significantly impaired basal cell differen-tiation and reduced the number of basal cells in theepididymis of 5-week-old rats. This implied that androgensignaling during the embryonic period, a time ofsubstantial androgen production, is required for thesubsequent postnatal differentiation of epididymal basalcells (Murashima et al. 2011).

Regional regulation of basal cell plasticity

Interestingly, between PNW8 and PNW12, several basalcells were observed with a body projection that extended

Reproduction (2013) 146 455–469

toward the luminal border of the epithelium. Theseluminal-reaching basal cells reappeared first in the distalcauda epididymidis at PNW8 and then progressivelybecame more abundant from PNW9 to PNW12 in themiddle and distal corpus and proximal cauda regions, butonly rare luminal-reaching basal cells were detected inother epididymal segments and in the VD.

A significant increase in serum and testicular testoster-one levels occurs during this time period (Lee et al. 1975,Paz et al. 1980, Scheer & Robaire 1980, Corpechot et al.1981, Turner et al. 1985), indicating that the epididymiscontinues to undergo significant remodeling at these latertime points and that testosterone might play an importantrole in the regulation of basal cell plasticity. Indeed,we observed a significant reduction in the number ofbasal cells with luminal-reaching projections in12-week-old rats that had been treated for 2 weekswith flutamide compared with control rats. Thesechanges were detected mainly in the corpus region,where luminal-reaching basal cells are normally moreabundant. No obvious morphological changes in basalcells in other regions were observed, and they remainedhemispherical in nature. The weight of the epididymisand accessory glands and the expression of AQP9 inepididymal principal cells were significantly reduced,demonstrating the effectiveness of flutamide treatment,as reported previously (Pastor-Soler et al. 2002). Thus,our results show that the apical-reaching property ofbasal cells is regulated by androgen.

Why do basal cells have luminal-reaching projectionsonly in the corpus and proximal cauda of epididymis andnot in other epididymal regions in adult rats? Androgensare key regulators of epididymal function in adults, andseveral androgen receptor effectors have been identified,including insulin-like growth factor (IGF), fibroblast

www.reproduction-online.org

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

Postnatal regulation of basal cell plasticity 467

growth factor (FGF), transforming growth factor, vascularendothelial growth factor, and epidermal growth factor(EGF) (Robaire & Viger 1995, Ezer & Robaire 2002,Henderson et al. 2006, Robaire et al. 2007, Robaire &Hamzeh 2011). Basal cells in the epididymis and prostateexpress EGF and FGF receptors (Cotton et al. 2008, Dubeet al. 2012), but the expression of the other growthhormone receptors in these cells has not been fullyestablished, although some have been detected in theepididymal epithelium (Leheup & Grignon 1993). Inter-estingly, IGF1 and FGF10 share region-specific expressionprofiles that would be compatible with their potential rolein the modulation of basal cell luminal projections in theepididymis – i.e., they are expressed at low levels in thecaput but at high levels in the corpus (Henderson et al.2006). In rats treated with inhibitors of 5a-reductase toprevent DHT formation, the most potent androgen agonistin the epididymis, the expression of both IGF1 and FGF10has been shown to be significantly downregulated(Henderson et al. 2006). This effect occurred only inthe corpus and cauda regions, indicating that thesereceptors might be involved in the region-specificandrogen-dependent plasticity of basal cells in adult rats.

Are basal cells involved in the renewal ofthe epithelium?

Basal cells have been considered as the source of cellrenewal in adult pseudostratified epithelia, such as theairway tract, the mammary gland, and the prostate(Rizzo et al. 2005, Chen et al. 2009, Rock et al. 2009,2010, Cole et al. 2010, Van Keymeulen et al. 2012).In this study, we found that while clear cells appearedsimultaneously with basal cells in the different epididy-mal regions during postnatal development, they weredetected at a discrete distance from each other at theironset of differentiation. No cells transiently expressingboth KRT5 and V-ATPase or KRT5 and AQP9 (data notshown) were observed during the first 2 PNWs. Theseresults are consistent with those of a previous studyshowing that basal cells are not required to direct orsupport the differentiation of principal and narrow/clearcells in the murine epididymis (Murashima et al. 2011).Thus, the present study confirms that basal cells are notthe progenitors of clear cells or principal cells duringearly development. However, a previous study hassuggested that normal timing of basal cell developmentduring the perinatal period is a key determinant for theappropriate development of the cauda epididymis andVD (Atanassova et al. 2005). Our data do not exclude thepossibility that basal cells might renew epididymalepithelial cells under particular conditions.

Summary

This study showed that epididymal basal cells originatefrom undifferentiated columnar epithelial cells. Basal cell

www.reproduction-online.org

differentiation occurs in an ascending pattern, starting inthe VD during the first PNW and appearing progressivelyin more proximal regions from PNW2 to PNW5. ByPNW6, basal cells form a network that covers a large areaof the base of the epithelium of all the epididymalsegments. Basal cells exhibit significant structuralplasticity during differentiation. First, they are similar toadjacent columnar cells. They then become triangular inshape with a narrow apical pole that is still in contact withthe luminal side of the epithelium. As they progress towardmaturation, their nucleus moves to the basal pole and theirapical body becomes narrower. At around PNW5, mostbasal cells would have retracted their narrow body andthey nestle underneath epithelial cells. Between PNW8and PNW12, many basal cells in the corpus and caudaextend a narrow body projection that reaches toward thelumen. These extensions were significantly decreased in12-week-old rats that had been treated with flutamide for2 weeks. Overall, our results indicate that the apical-reaching property of basal cells is not triggered by thepresence of spermatozoa in the lumen of the epididymis,but that it is regulated by androgen in adult rats. Thus,androgens are necessary for basal cells to perform theluminal-sensing function that we previously uncovered inthe epididymis of adult rats (Shum et al. 2008).

Supplementary data

This is linked to the online version of the paper at http://dx.doi.org/10.1530/REP-12-0510.

Declaration of interest

The authors declare that there is no conflict of interest thatcould be perceived as prejudicing the impartiality of theresearch reported.

Funding

This work was supported by NIH grants DK085715 andHD040793 to S Breton and by an MGH ECOR Fund forMedical Discovery Award to W W C Shum. The MicroscopyCore facility of the MGH Program in Membrane Biologyreceives support from the Boston Area Diabetes and Endo-crinology Research Center (DK57521) and the Center for theStudy of Inflammatory Bowel Disease (DK43341). S Breton is arecipient of the Charles and Ann Sanders Research ScholarAward at MGH.

References

Atanassova N, McKinnell C, Fisher J & Sharpe RM 2005 Neonataltreatment of rats with diethylstilboestrol (DES) induces stromal–epithelialabnormalities of the vas deferens and cauda epididymis in adulthoodfollowing delayed basal cell development. Reproduction 129 589–601.(doi:10.1530/rep.1.00546)

Reproduction (2013) 146 455–469

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

468 W W C Shum and others

Breton S & Brown D 2007 New insights into the regulation of V-ATPase-dependent proton secretion. American Journal of Physiology. RenalPhysiology 292 F1–F10. (doi:10.1152/ajprenal.00340.2006)

Breton S, Smith PJ, Lui B & Brown D 1996 Acidification of the malereproductive tract by a proton pumping (HC)-ATPase. Nature Medicine2 470–472. (doi:10.1038/nm0496-470)

Breton S, Tyszkowski R, Sabolic I & Brown D 1999 Postnatal developmentof HCATPase (proton-pump)-rich cells in rat epididymis. Histochemistryand Cell Biology 111 97–105. (doi:10.1007/s004180050339)

Brown D & Breton S 2000 H(C)V-ATPase-dependent luminal acidificationin the kidney collecting duct and the epididymis/vas deferens: vesiclerecycling and transcytotic pathways. Journal of Experimental Biology203 137–145.

Brown D, Lydon J, McLaughlin M, Stuart-Tilley A, Tyszkowski R & Alper S1996 Antigen retrieval in cryostat tissue sections and cultured cells bytreatment with sodium dodecyl sulfate (SDS). Histochemistry and CellBiology 105 261–267. (doi:10.1007/BF01463929)

Chen H, Matsumoto K & Stripp BR 2009 Bronchiolar progenitor cells.Proceedings of the American Thoracic Society 6 602–606. (doi:10.1513/pats.200907-078RM)

Cheung KH, Leung GP, Leung MC, Shum WW, Zhou WL & Wong PY 2005Cell–cell interaction underlies formation of fluid in the male reproductivetract of the rat. Journal of General Physiology 125 443–454.(doi:10.1085/jgp.200409205)

Cole BB, Smith RW, Jenkins KM, Graham BB, Reynolds PR & Reynolds SD2010 Tracheal basal cells. A facultative progenitor cell pool. AmericanJournal of Pathology 177 362–376. (doi:10.2353/ajpath.2010.090870)

Cornwall GA 2009 New insights into epididymal biology and function.Human Reproduction Update 15 213–227. (doi:10.1093/humupd/dmn055)

Corpechot C, Baulieu EE & Robel P 1981 Testosterone, dihydrotestosteroneand androstanediols in plasma, testes and prostates of rats duringdevelopment. Acta Endocrinologica 96 127–135.

Cotton LM, O’Bryan MK & Hinton BT 2008 Cellular signaling by fibroblastgrowth factors (FGFs) and their receptors (FGFRs) in male reproduction.Endocrine Reviews 29 193–216. (doi:10.1210/er.2007-0028)

Da Silva N, Silberstein C, Beaulieu V, Pietrement C, Van Hoek AN,Brown D & Breton S 2006 Postnatal expression of aquaporins inepithelial cells of the rat epididymis. Biology of Reproduction 74427–438. (doi:10.1095/biolreprod.105.044735)

Da Silva N, Cortez-Retamozo V, Reinecker HC, Wildgruber M, Hill E,Brown D, Swirski FK, Pittet MJ & Breton S 2011 A dense network ofdendritic cells populates the murine epididymis. Reproduction 141653–663. (doi:10.1530/REP-10-0493)

De Miguel MP, Marino JM, Martinez-Garcia F, Nistal M, Paniagua R &Regadera J 1998 Pre- and post-natal growth of the human ductusepididymidis. A morphometric study. Reproduction, Fertility, andDevelopment 10 271–277. (doi:10.1071/R98059)

Dube E, Dufresne J, Chan PT & Cyr DG 2012 Epidermal growth factorregulates connexin 43 in the human epididymis: role of gap junctions inazoospermia. Human Reproduction 27 2285–2296. (doi:10.1093/humrep/des164)

Ezer N& Robaire B 2002 Androgen regulation of the structure and functionsof the epididymis. In The Epididymis: from Molecular to ClinicalPractice. A Comprehensive Survey of the Efferent Ducts, the Epididymisand the Vas Deferens, pp 297–316.

Fisher JS, Pastor-Soler N, Sharpe RM & Breton S 2002 Modulation of theonset of postnatal development of H(C)-ATPase-rich cells by steroidhormones in rat epididymis. Biology of Reproduction 67 1106–1114.(doi:10.1095/biolreprod67.4.1106)

Francavilla S, Moscardelli S, Properzi G, De Matteis MA,Scorza Barcellona P, Natali PG & De Martino C 1987 Postnataldevelopment of epididymis and ductus deferens in the rat. A correlationbetween the ultrastructure of the epithelium and tubule wall, and thefluorescence-microscopic distribution of actin, myosin, fibronectin,and basement membrane. Cell and Tissue Research 249 257–265.(doi:10.1007/BF00215508)

Gregory M, Dufresne J, Hermo L & Cyr D 2001 Claudin-1 is not restrictedto tight junctions in the rat epididymis. Endocrinology 142 854–863.(doi:10.1210/en.142.2.854)

Reproduction (2013) 146 455–469

Gusterson BA, Ross DT, Heath VJ & Stein T 2005 Basal cytokeratins andtheir relationship to the cellular origin and functional classification ofbreast cancer. Breast Cancer Research 7 143–148. (doi:10.1186/bcr1041)

Hayashi T, Yoshinaga A, Ohno R, Ishii N, Kamata S & Yamada T 2004Expression of the p63 and Notch signaling systems in rat testes duringpostnatal development: comparison with their expression levels in theepididymis and vas deferens. Journal of Andrology 25 692–698.

Henderson NA, Cooke GM& Robaire B 2006 Region-specific expression ofandrogen and growth factor pathway genes in the rat epididymis and theeffects of dual 5a-reductase inhibition. Journal of Endocrinology 190779–791. (doi:10.1677/joe.1.06862)

Hermo L & Robaire B 2002 Epididymal cell types and their functions.In The Epididymis: from Molecular to Clinical Practice. A Comprehen-sive Survey of the Efferent Ducts, the Epididymis and the Vas Deferens,pp 81–102. Eds B Robaire and BT Hinton. New York: KluwerAcademic/Plenum Publishers.

Hermo L, Papp S & Robaire B 1994 Developmental expression of theYf subunit of glutathione S-transferase P in epithelial cells of the testis,efferent ducts, and epididymis of the rat. Anatomical Record 239421–440. (doi:10.1002/ar.1092390409)

Hermo L, Krzeczunowicz D & Ruz R 2004 Cell specificity of aquaporins 0,3, and 10 expressed in the testis, efferent ducts, and epididymis of adultrats. Journal of Andrology 25 494–505.

Holschbach C & Cooper TG 2002 A possible extratubular origin of epidi-dymal basal cells in mice. Reproduction 123 517–525. (doi:10.1530/rep.0.1230517)

Imperato-McGinley J, Sanchez RS, Spencer JR, Yee B & Vaughan ED 1992Comparison of the effects of the 5a-reductase inhibitor finasteride andthe antiandrogen flutamide on prostate and genital differentiation: dose–response studies. Endocrinology 131 1149–1156. (doi:10.1210/en.131.3.1149)

Kim J, Cha JH, Tisher CC & Madsen KM 1996 Role of apoptotic andnonapoptotic cell death in removal of intercalated cells from developingrat kidney. American Journal of Physiology 270 F575–F592.

Lee VW, de Kretser DM, Hudson B & Wang C 1975 Variations in serumFSH, LH and testosterone levels in male rats from birth to sexual maturity.Journal of Reproduction and Fertility 42 121–126. (doi:10.1530/jrf.0.0420121)

Legare C, Thabet M & Sullivan R 2004 Expression of heat shock protein 70in normal and cryptorchid human excurrent duct. Molecular HumanReproduction 10 197–202. (doi:10.1093/molehr/gah027)

Leheup BP & Grignon G 1993 Immunohistochemical localizationof insulin-like growth factor I (IGF-I) in the rat epididymis. Journal ofAndrology 14 159–163.

Leung GP, Cheung KH, Leung CT, Tsang MW & Wong PY 2004 Regulationof epididymal principal cell functions by basal cells: role of transientreceptor potential (Trp) proteins and cyclooxygenase-1 (COX-1).Molecular and Cellular Endocrinology 216 5–13. (doi:10.1016/j.mce.2003.10.077)

Li Z, Sun ZJ, Liao CG, Ma L, Ma BF & Zhang YQ 2010 Regulated uponactivation normal T-cell expressed and secreted originating from theepididymis differentially associates with viable and defective spermatozoa.Fertility and Sterility 93 2661–2667. (doi:10.1016/j.fertnstert.2010.01.053)

Martan J 1969 Epididymal histochemistry and physiology. Biology ofReproduction 1 (Suppl 1) 34–54.

Marty MS, Chapin RE, Parks LG & Thorsrud BA 2003 Developmentand maturation of the male reproductive system. Birth Defects Research.Part B, Developmental and Reproductive Toxicology 68 125–136.(doi:10.1002/bdrb.10015)

Murashima A, Miyagawa S, Ogino Y, Nishida-Fukuda H, Araki K,Matsumoto T, Kaneko T, Yoshinaga K, Yamamura K, Kurita T et al.2011 Essential roles of androgen signaling in Wolffian duct stabilizationand epididymal cell differentiation. Endocrinology 152 1640–1651.(doi:10.1210/en.2010-1121)

Nilnophakoon N 1978 Histological studies on the regional postnataldifferentiation of the epididymis in the ram. Anatomia, Histologia,Embryologia 7 253–272. (doi:10.1111/j.1439-0264.1978.tb00800.x)

O’Connor JC, Frame SR & Ladics GS 2002 Evaluation of a 15-day screeningassay using intact male rats for identifying antiandrogens. ToxicologicalSciences 69 92–108. (doi:10.1093/toxsci/69.1.92)

www.reproduction-online.org

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access

Postnatal regulation of basal cell plasticity 469

Orgebin-Crist MC, Danzo BJ & Davies J 1975 Endocrine control of thedevelopment and maintenance of sperm fertilizing ability in the epididymis.In Hand-book of Physiology, Sec 7, Endocrinology, pp 319–339.

Pastor-Soler N, Bagnis C, Sabolic I, Tyszkowski R, McKee M, Van Hoek A,Breton S & Brown D 2001 Aquaporin 9 expression along the malereproductive tract. Biology of Reproduction 65 384–393. (doi:10.1095/biolreprod65.2.384)

Pastor-Soler N, Isnard-Bagnis C, Herak-Kramberger C, Sabolic I,Van Hoek A, Brown D & Breton S 2002 Expression of aquaporin 9 inthe adult rat epididymal epithelium is modulated by androgens. Biologyof Reproduction 66 1716–1722. (doi:10.1095/biolreprod66.6.1716)

Pastor-Soler N, Pietrement C & Breton S 2005 Role of acid/basetransporters in the male reproductive tract and potential consequencesof their malfunction. Physiology 20 417–428. (doi:10.1152/physiol.00036.2005)

Pastor-Soler NM, Fisher JS, Sharpe R, Hill E, Van Hoek A, Brown D &Breton S 2010 Aquaporin 9 expression in the developing rat epididymisis modulated by steroid hormones. Reproduction 139 613–621. (doi:10.1530/REP-09-0284)

Paz GF, Winter JS, Reyes FI & Faiman C 1980 Developmental pattern oftestosterone production by the rat testis. Steroids 36 675–688. (doi:10.1016/0039-128X(80)90050-1)

Poulton TA, Everard D, Baxby K & Parslow JM 1996 Characterisation of asperm coating auto-antigen reacting with antisperm antibodies ofinfertile males using monoclonal antibodies. British Journal of Obstetricsand Gynaecology 103 463–467. (doi:10.1111/j.1471-0528.1996.tb09774.x)

Rizzo S, Attard G & Hudson DL 2005 Prostate epithelial stem cells. CellProliferation 38 363–374. (doi:10.1111/j.1365-2184.2005.00356.x)

Robaire B & Hamzeh M 2011 Androgen action in the epididymis. Journal ofAndrology 32 592–599. (doi:10.2164/jandrol.111.014266)

Robaire B & Hermo L 1988 Efferent ducts, epididymis, and vas deferens:structure, functions, and their regulation. In The Physiology ofReproduction. Eds DW Hamilton and RO Greep. Washington: AmericanSociety of Physiologists.

Robaire B & Viger RS 1995 Regulation of epididymal epithelial cellfunctions. Biology of Reproduction 52 226–236. (doi:10.1095/biol-reprod52.2.226)

Robaire B, Seenundun S, Hamzeh M & Lamour SA 2007 Androgenicregulation of novel genes in the epididymis. Asian Journal of Andrology 9545–553. (doi:10.1111/j.1745-7262.2007.00316.x)

Rock JR, Onaitis MW, Rawlins EL, Lu Y, Clark CP, Xue Y, Randell SH &Hogan BL 2009 Basal cells as stem cells of the mouse trachea and humanairway epithelium. PNAS 106 12771–12775. (doi:10.1073/pnas.0906850106)

Rock JR, Randell SH & Hogan BL 2010 Airway basal stem cells: aperspective on their roles in epithelial homeostasis and remodeling.Disease Models & Mechanisms 3 545–556. (doi:10.1242/dmm.006031)

Rodriguez CM, J L K & B T H 2002 The development of the epididymis.In The Epididymis: from Molecular to Clinical Practice. A Comprehen-sive Survey of the Efferent Ducts, the Epididymis and the Vas Deferens,pp 251–267. Eds B Robaire and BT Hinton. New York: KluwerAcademic/Plenum Publishers.

Scheer H & Robaire B 1980 Steroid delta 4–5a-reductase and 3a-hydroxysteroid dehydrogenase in the rat epididymis during development.Endocrinology 107 948–953. (doi:10.1210/endo-107-4-948)

Seiler P, Wenzel I, Wagenfeld A, Yeung CH, Nieschlag E & Cooper TG 1998The appearance of basal cells in the developing murine epididymis andtheir temporal expression of macrophage antigens. InternationalJournal of Andrology 21 217–226. (doi:10.1046/j.1365-2605.1998.00116.x)

Seiler P, Cooper TG &Nieschlag E 2000 Sperm number and condition affectthe number of basal cells and their expression of macrophage antigen inthe murine epididymis. International Journal of Andrology 23 65–76.(doi:10.1046/j.1365-2605.2000.00210.x)

www.reproduction-online.org

Shum WW, Da Silva N, McKee M, Smith PJ, Brown D & Breton S 2008Transepithelial projections from basal cells are luminal sensors inpseudostratified epithelia. Cell 135 1108–1117. (doi:10.1016/j.cell.2008.10.020)

Shum WW, Da Silva N, Brown D & Breton S 2009 Regulation of luminalacidification in the male reproductive tract via cell–cell crosstalk.Journal of Experimental Biology 212 1753–1761. (doi:10.1242/jeb.027284)

ShumWW, Ruan YC, Da Silva N & Breton S 2011 Establishment of cell–cellcrosstalk in the epididymis: control of luminal acidification. Journal ofAndrology 32 576–586. (doi:10.2164/jandrol.111.012971)

Sun EL & Flickinger CJ 1979 Development of cell types and of regionaldifferences in the postnatal rat epididymis. American Journal of Anatomy154 27–55. (doi:10.1002/aja.1001540104)

Sun EL & Flickinger CJ 1982 Proliferative activity in the rat epididymisduring postnatal development. Anatomical Record 203 273–284.(doi:10.1002/ar.1092030209)

Turner TT 1991 Spermatozoa are exposed to a complex micro-environment as they traverse the epididymis. Annals of the New YorkAcademy of Sciences 637 364–383. (doi:10.1111/j.1749-6632.1991.tb27323.x)

Turner TT 1995 On the epididymis and its role in the development of thefertile ejaculate. Journal of Andrology 16 292–298.

Turner TT 2002 Necessity’s potion: inorganic ions and small organicmolecules in the epididymal lumen. In The Epididymis: from Molecularto Clinical Practice. A Comprehensive Survey of the Efferent Ducts,the Epididymis and the Vas Deferens, pp 131–150. Eds B Robaire andBT Hinton. New York: Kluwer Academic/Plenum Publishers.

Turner TT, Ewing LL, Jones CE, Howards SS & Zegeye B 1985 Androgens inmale rat reproductive tract fluids: hypophysectomy and steroidreplacement. American Journal of Physiology 248 E274–E280.

Van Keymeulen A, Rocha AS, Ousset M, Beck B, Bouvencourt G, Rock J,Sharma N, Dekoninck S & Blanpain C 2012 Distinct stem cellscontribute to mammary gland development and maintenance. Nature479 189–193. (doi:10.1038/nature10573)

Veri JP, Hermo L & Robaire B 1993 Immunocytochemical localization ofthe Yf subunit of glutathione S-transferase P shows regional variation inthe staining of epithelial cells of the testis, efferent ducts, and epididymisof the male rat. Journal of Andrology 14 23–44.

Welsh M, Sharpe RM, Walker M, Smith LB & Saunders PT 2009 Newinsights into the role of androgens in Wolffian duct stabilization in maleand female rodents. Endocrinology 150 2472–2480. (doi:10.1210/en.2008-0529)

Wong PY, Gong XD, Leung GPH & Cheuk BLY 2002 Formation of theepididymal fluid microenvironment. In The Epididymis: from Molecularto Clinical Practice. A Comprehensive Survey of the Efferent Ducts, theEpididymis and the Vas Deferens, pp 119–130. Eds B Robaire andBT Hinton. New York: Kluwer Academic/Plenum Publishers.

Yeung CH, Nashan D, Sorg C, Oberpenning F, Schulze H, Nieschlag E &Cooper TG 1994 Basal cells of the human epididymis – antigenic andultrastructural similarities to tissue-fixed macrophages. Biology ofReproduction 50 917–926. (doi:10.1095/biolreprod50.4.917)

Zondek LH & Zondek T 1980 Normal and abnormal development of theepididymis of the fetus and infant. European Journal of Pediatrics 13439–44. (doi:10.1007/BF00442401)

Received 20 December 2012

First decision 7 February 2013

Revised manuscript received 2 August 2013

Accepted 15 August 2013

Reproduction (2013) 146 455–469

Downloaded from Bioscientifica.com at 05/20/2020 11:57:24AMvia free access