RE S EARCH COMMUN ICAT ION

Immunolocalization of aquaporin 7 in human sperm and its relationshipwith semen parameters

Elena Moretti1,2, Gaia Terzuoli1, Lucia Mazzi1, Francesca Iacoponi1, and Giulia Collodel1,2∗

1Department of Biomedical Sciences, Applied Biology Section, 2Interdepartmental Centre for Research and Therapy of MaleInfertility, University of Siena, Italy

Aquaporins (AQPs) are a family of 13 small hydrophobic trans-membrane proteins expressed in numerous tissues and cells.Some AQPs work as strict water channels, others are permeableto a range of substances, including glycerol. In the male repro-ductive system their localization in testis, efferent ducts, epidi-dymis, and spermatozoa has been described. We studied thedistribution of AQP7 in ejaculated human sperm and therelationship between AQP7 labeling and sperm characteristics.Semen samples from 33 men were examined by light and trans-mission electron microscopy (TEM). TEM data were quantifiedusing a mathematical formula that calculates a fertility index(FI) and the percentages of sperm apoptosis, immaturity, andnecrosis. Immunocytochemistry with a polyclonal antibodyanti-AQP7 was performed on the sperm samples. Normalsperm were labeled in the pericentriolar area, midpiece, equa-torial segment, and weakly in the tail (grade 1). Abnormalsperm showed a diffuse low intensity of fluorescence evidentin the cytoplasmic residues, coiled tails, in the entire head,and acrosome (grade 2). A high number of motile spermobtained by swim up were labeled in a dotted manner in themitochondria. A significant positive correlation was foundbetween the spermatozoa with AQP7 grade 1 labeling andthe percentage of normal form (P < 0.008), progressive motilityand FI (P < 0.005); a negative correlation was noted with thepercentages of cytoplasmic residues (P < 0.010) and immaturity(P < 0.006) and coiled tails (P < 0.012). The link between AQP7distribution and sperm morphology and the particular dottedlabeling in swim up selected motile sperm are novel anddeserve additional studies.

Keywords aquaporin 7, human spermatozoa,immunocytochemistry, swim up, transmission electronmicroscopy

Introduction

Aquaporins (AQPs) consist of a family of 13 small hydro-phobic trans-membrane proteins expressed in numerous

tissues and cells, such as kidney, lung, pancreas, brain,gastrointestinal tract, immune system, skin, adipose tissue,uterus, and testis, in which water movement is essentialand physiologically crucial [Huang et al. 2006]. SomeAQPs work as strict water channels, yet others are permeableto a wide range of substances, including glycerol [Ishibashiet al. 2009]; for this reason they are called aquaglyceroporins,such as AQP7, first identified in rat testis by Ishibashi et al.[1997]. The function and the different distribution of AQPsin male and female reproductive systems were extensivelyrevised by Huang et al. [2006] and they encompass uterineimbibitions mechanisms, ovum transport, oviductal fluidbalance, follicle maturation, blastocyst formation, embryoimplantation, and spermatogenesis. Regarding the malereproductive system, Hermo and Smith [2011] describedthe localization of AQPs 5, 7, 9, 10, and 11 in differentareas, such as testis, efferent ducts, and epididymis, revealingunique associations of these AQPs with specific membranedomains in a cell type and region specific manner. Themassive and specific localizations of such membrane pro-teins indicate their important role in water movementacross testicular cell membranes.

Germ cells are known to express AQPs, mainly isoforms 7and 8. AQP8 was observed in all germ cell lines of rat testicularepithelium [Calamita et al. 2001], whereas Yeung et al. ident-ified this protein channel exclusively in early and late sperma-tids inmouse testis [2009] and in all germ cells of human testis[2010]. AQP7 has been localized in germ cells and ismostly ex-pressed in post meiotic male germ cells. In particular, it hasbeen identified in the plasma membrane of elongated sperma-tids and maturing spermatozoa from rat testis [Ishibashi et al.1997; Suzuki-Toyota et al. 1999;Calamita et al. 2001;Kageyamaet al. 2001] and even in the residual bodies remaining in theseminiferous epithelium [Suzuki-Toyota et al. 1999; Calamitaet al. 2001]. In addition, Hermo and Smith [2011] observedan AQP7 immunoreaction over the whole cytoplasm of lateelongating spermatids located near the lumen of rat seminifer-ous epithelium. AQP7 is expressed in human spermatids and

∗Address correspondence to Giulia Collodel, Department of Biomedical Sciences, Applied Biology Section, University of Siena, Policlinico LeScotte, Viale Bracci, 14 53100 Siena. E-mail: collodel@unisi.it

Received 16 June 2011; accepted 24 September 2011.

Systems Biology in Reproductive Medicine, 2012, 58: 129–135Copyright © 2012 Informa Healthcare USA, Inc.ISSN 1939-6368 print/1939-6376 onlineDOI: 10.3109/19396368.2011.644385

129

Syst

Bio

l Rep

rod

Med

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

McG

ill U

nive

rsity

on

11/0

7/12

For

pers

onal

use

onl

y.

sperm at the tail level [Saito et al. 2004; Yeung et al. 2010]. Inhuman ejaculated sperm, Saito et al. [2004] reportedimmunolabeling in the midpiece region and the anteriorportion of the tail, whereas Yeung et al. [2010] showed the

presence of AQP7 along the entire flagellum, with an intensitythat varied from cell to cell. Although this protein seems to playa role in late spermatogenesis and tobe involved inmechanismsofmale fertility, the function of AQP7 in themale reproductive

Figure 1. A-M) Ultraviolet micrographs of ejaculated human sperm treated with anti- AQP7 rabbit polyclonal antibody (A, B, C, E, G, I, J, and L)and phase contrast light microscopy pictures (D,F,H,K, andM showing the same part of the C,E,G,J, and L fluorescent micrographs, respectively).A,B) Normal sperm show a clear fluorescent labeling in the pericentriolar area, at the midpiece level and in the equatorial segment (A); sometimes amore diffuse staining along the tail (B) is present. C,E,G) Abnormal sperm show a different staining modality: a diffuse general low intensity offluorescence is located in the cytoplasmic residues and in the coiled tails (C, E), in the entire head and in the acrosome (G). I) Sperm frommotile fraction obtained after swim up. A high number of sperm highlight a dotted label, particularly evident in the mitochondrial region(arrows). J) Sperm from the lowermost fraction of swim up rich in sperm labeled in a diffuse, non-dotted manner. L) Sperm from a patient withdysplasia of the fibrous sheath and abnormalities of head-neck attachment, a diffuse AQP7 labeling is located in the short, thick tail. A-I) Bar 5µm; J-M) Bar 4 µm.

130 E. Moretti et al.

Systems Biology in Reproductive Medicine

Syst

Bio

l Rep

rod

Med

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

McG

ill U

nive

rsity

on

11/0

7/12

For

pers

onal

use

onl

y.

system is still controversial. Current findings seem to indicate amajor role of AQP7 in glycerol metabolism and a possible roleof AQP8 in water movement for sperm volume regulation[Yeung 2010]. Recently, AQP11 was identified in the residualcytoplasm of rat elongating spermatids and in the distalregion of the sperm flagellum, suggesting a function in recy-cling the excess of cytoplasmic components [Yeung andCooper 2010] required to make the sperm slim enough to beable to swim fast and reach the oocyte.

The present paper was aimed at exploring the distributionof AQP7 in ejaculated human sperm and in the motilesperm population obtained by swim up, in order to investi-gate the behavior of the protein after selection. We alsoexplored the relationship between AQP7 labeling andsperm characteristics such as sperm concentration, motility,and morphology and the scores obtained by means ofmathematical elaboration of data obtained by transmissionelectron microscopy sperm analysis.

Results

AQP7 localization in ejaculated human spermatozoaIn all of the 33 investigated semen samples, we observed AQP7binding to the sperm surface with different intensity andmodality. In particular, semen with normal parameters(sperm concentration, sperm motility percentage, and percen-tage of spermwith normalmorphology in the range of normal-ity reported byWHO guidelines [1992 and 1999]) were rich insperm showingAQP7 labeling as a brilliant spot in the pericen-triolar region of the neck associated with staining at the mid-piece level, sometimes concomitant with an evidentfluorescent area in correspondence with the equatorial regionof the acrosome and a more diffuse staining along the tail(classified as grade 1; Fig. 1A and B). In contrast, we notedthat semenwith abnormal parameters (oneormore semenpar-ameters among sperm concentration, sperm motility percen-tage, and percentage of sperm with normal morphologybelow the normal range reported by WHO guidelines[1992;1999]) showed an increased percentage of sperm witha low intensity of fluorescence and with a different stainingmodality: diffuse AQP7 immunostaining has been noted incytoplasmic residual bodies (Fig. 1C-F), in the entire head(Fig. 1G and H), and faintly in the tail (classified as grade 2).

All controls for each experiment were negative for staining.No labeling was observed when the human spermatozoawere treated with the non immune rabbit serum (negativecontrol, data not shown).

AQP7 localization in swim up selected fractions of humanspermAQP7 immunolabeling was evaluated in two sperm frac-tions, separated by swim up, of semen from 5 men randomlychosen among the ejaculates with sperm motility ≥ 30%(Fig. 2) in order to ensure retrieval of a sufficient amountof sperm to perform the experiments. The sperm character-istics and AQP7 grade 1 labeling before swim up arereported in Table 1. After swim up, the fraction with themotile and well shaped sperm showed a high number ofcells stained in a dotted manner, particularly evident in themitochondrial region (Fig. 1I). This number was signifi-cantly higher than that found in the remaining fraction(75 vs. 30 P < 0.008; Fig. 2) rich in sperm labeled in adiffuse, non-dotted manner (Fig. 1J and K).

AQP7 labeling and semen characteristicsThe distribution of the 33 subjects involved in the study isshown in Figure 3. One patient out of 33 individuals wasexcluded from the statistical analysis because his spermdisplayed a mosaic of flagellar alterations, which consist in

Figure 2. Box plot of the percentages of sperm with a dotted AQP7staining in the mitochondrial region calculated in the uppermostswim up fraction (motile sperm) and lowermost fraction (immotilesperm).

Table 1. Mean values (SD) of the semen variables and of the percentage of sperm with grade 1 AQP7 localization in the total of the 32 subjectsconsidered in this study, in the 5 out of 32 ejaculates treated with swim up method, and in the 15 out of 32 samples processed for TEM (TEM indices:Fertility index, % Immaturity, % Necrosis, and % Apoptosis).

Variable N = 32 N = 5 N = 15

Volume/ml 3.53 (1.31) 2.8 (0.91) 3.73 (1.42)Number sperm x106/ml 81.03 (49.29) 132.63 (15.63) 62.75 (44.72)% Motility (a + b) 39.47 (16.94) 53.75 (17.86) 32.60 (19.15)% Normal sperm 27.16 (8.17) 37.25 (6.19) 25.27 (9.04)% Sperm grade 1 AQP7 stained 45.38 (18.29) 58.25 (15.88) 39.00 (18.94)% Cytoplasmic residue 11.81 (6.78) 8.75 (3.95) 14.27 (6.79)% Coiled tail 6.91 (7.25) 4 (4.32) 7.20 (8.72)Fertility Index - 3445406 (7812189)% Immaturity - 64.60 (14.53)% Necrosis - 35.33 (16.32)% Apoptosis - 6.00 (3.82)

AQP7 in human sperm 131

Copyright © 2012 Informa Healthcare USA, Inc.

Syst

Bio

l Rep

rod

Med

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

McG

ill U

nive

rsity

on

11/0

7/12

For

pers

onal

use

onl

y.

dysplasia of the fibrous sheath and abnormalities of head-neck attachment, two defects of possible genetic origincharacterised by TEM [Moretti et al. 2011]. Diffuse AQP7labeling was located in the short, thick tail where themarked hypertrophy and hyperplasia of the fibrous sheathwere the common characteristics (Fig. 1L and M).

The mean values and standard deviations (SD) of thesperm parameters, including the percentages of cytoplasmicresidues and coiled tails, related to 32 out of 33 subjects arereported in Table 1. Eight of 32 cases showed normal semenparameters, 3 men were moderately teratozoospermic (per-centage of normal sperm 25-26%), 3 men were asthenozoos-permic, 14 men asthenoteratozoospermic, and 4 menoligoasthenoteratozoospermic.

As shown in Table 1 the sperm variables of the 5 of 32ejaculates before swim up treatment and of the 15 of 32 eja-culates used for TEM analysis are described. For this lastgroup the TEM indices (fertility index, immaturity, necrosis,and apoptosis), derived from mathematical elaboration ofTEM data, are reported. Each variable of the 32 subjectsand TEM indices from 15 of 32 individuals were correlatedwith the percentage of sperm with grade 1 AQP7 staining.

The percentages of sperm with grade 1 labeling showed asignificant positive correlation with the percentages ofnormal morphology (r = 0.78, P < 0.008), progressive moti-lity (r = 0.87, P < 0.005), and FI (r = 0.80, P < 0.005) and anegative correlation with percentages of cytoplasmic residues(r = -0.66, P < 0.010), coiled tails (r = -0.44, P < 0.012), andimmaturity (r = -0.85, P < 0.006).

Discussion

The above has documented the presence of the AQP7protein in human sperm. It should be noted that a differentpattern of expression among men has been observed using

FACS analysis and Western blot [Saito et al. 2004; Yeunget al. 2010]. Our study, focused on the localization ofAQP7 in ejaculated human sperm. Various specimensshowed a different distribution of fluorescent label. It wasevident that semen with normal parameters show anincreased percentage of stained sperm, defined as grade 1,with clear fluorescent labeling at the pericentriolar area,midpiece, equatorial segment, and more diffuse stainingalong the tail. In contrast, samples with reduced semenquality displayed more diffuse AQP7 staining and in differ-ent localizations, such as the sperm head, the acrosome, andcytoplasmic residual bodies. AQP7 staining along the entiretail of human ejaculated sperm [Yeung et al. 2010] and local-ization in the midpiece region and anterior portion of the tail[Saito et al. 2004] have been described by other groups.The discrepancy in AQP7 localization could be due tothe use of different antibodies, cell preparation, or themethods employed for immunostaining.

Validation of the observations by the analysis of AQP7staining in the two sperm fractions obtained by swim upthen followed. In addition to confirming the different stain-ing intensities in the two separate sperm populations, thismethod enabled us to detect AQP7 almost exclusively dis-tributed in a dotted manner in the midpiece of the spermthat had been recovered after swim-up. The results may indi-cate a probable rearrangement of AQP7 in distinct domainsduring the process of selection of a motile sperm populationand the possible involvement of this protein in controllingsperm motility. Since AQP7 enables the transport of waterand glycerol, a possible role of this protein in providing gly-cerol as an energy substrate can be postulated, as demon-strated in cardiomyocytes by Hibuse et al. [2009]. Duringsperm selection, even by the swim up procedure, spermare prepared to undergo capacitation and, a remarkableremodelling of the lipid and protein architecture of sperm

Figure 3. Flow chart indicating the distribution of the subjects (N) and the sperm analysis performed in the study of AQP7 localization in humansperm.

132 E. Moretti et al.

Systems Biology in Reproductive Medicine

Syst

Bio

l Rep

rod

Med

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

McG

ill U

nive

rsity

on

11/0

7/12

For

pers

onal

use

onl

y.

plasma membrane takes place, which consists in a release ofcholesterol, requisite for acrosome reaction and subsequentsperm-egg interaction. The cholesterol release determinesthe partial disassembly of plasma membrane lipidic rafts,modifying the activity of several enzymes [Travis andKopf, 2002; Cross 2004]. In addition, the sperm membranerafts, heterogeneous, highly dynamic microdomainsenriched of sterol and sphingolipids, are dynamically redis-tributed during capacitation; membrane rafts may representplatforms for the organization of proteins involved in fertili-zation [Nixon et al. 2010]. Moreover Buffone et al. [2008]reported that low-motility spermatozoa are unable torespond to capacitation with the necessary changes in mem-brane fluidity.

It is probably premature to affirm that AQP7 may beconsidered a marker of capacitation even though this topicdeserves attention. Other studies, able to validate thepresent results, should be performed by labeling AQP7 insperm selected by a density gradient centrifugation,another procedure enabling the sperm capacitation.

In order to formalize our observations, we analyzedsperm characteristics in human semen samples followingWHO criteria [1992; 1999] and AQP7 localization. Tofurther validate our results, a number of the semensamples were processed for TEM analysis and the obtaineddata were elaborated by mathematical methods based onprobability calculation for numerical indices able toexpress the examined semen quality in numbers: FI(number of sperm probably free of ultrastructural defects),probability percentage of sperm pathologies as immaturity,apoptosis, and necrosis.

The percentage of sperm with specific and shining grade1 AQP7 labeling showed a significant correlation with spermmotility, as reported by other groups [Saito et al. 2004;Yeung et al. 2010] and with sperm morphology, as detectedby TEM. A positive correlation was found with the percen-tage of normal sperm morphology and FI and a negative cor-relation was noted with percentages of cytoplasmic residues,coiled tails, and immaturity. Among the sperm alterationsobserved during evaluation of PAP stained slides, we notedthe percentage of cytoplasmic residues, since AQP7 hasbeen localized in this cell structure [Hermo and Smith2011] and coiled tails, as they are often concomitant withthe presence of cytoplasmic residue. Perhaps AQP7 isinvolved in the reduction of spermatid volume during sper-miogenesis by mediating the efflux of water [Suzuki-Toyotaet al. 1999; Calamita et al. 2001; Kageyama et al. 2001] and itmay persist in this diffuse localization in the membrane ofcytoplasmic residue of abnormal sperm.

The population studied is rather small, however it shouldbe considered that the relationship between the distributionof AQP7 and sperm morphology is supported by statisticalanalysis of many variables obtained by light and electronmicroscopy. Another clue that seems to confirm thisrelationship is the diffuse, weak AQP7 labeling observed inthe sperm with serious morphological alterations, such asdysplasia of the fibrous sheath concomitant with abnormal-ities of head-neck attachment. Although a role of AQP7 in

sperm motility cannot be ruled out [Saito et al. 2004;Yeung et al. 2010] and its relationship with sperm modelingappears to be possible, it should be considered that AQP7knockout mice showed normal sperm and fertility [Soharaet al. 2007] and that a subject homozygous for a nonfunc-tional mutation of AQP7 has been reported as fertile[Kondo et al. 2002], probably excluding an indispensablefunction of AQP7 in the regulation of fertility.

Since the Aqp7 null mutation in mice induces an up-regu-lation of spermatid Aqp8 and an enhancement of watertransport [Yeung et al. 2009], a synergic effect of the two iso-forms could be hypothesized; thus a conditional knockoutanimal model for all AQPs expressed in germ cell epitheliummight help to clarify this issue.

Although the role of AQP7 in male infertility is not clearyet, a relationship between its distribution and sperm moti-lity and morphology has been demonstrated. In particular,the link between the distribution of AQP7 and sperm mor-phology and the observation that swim up selected motilesperm show a dotted labeling at mitochondrial level arenovel and deserve attention and additional studies.

Materials and Methods

Semen samplesSemen samples were obtained from 33 consecutive menrecruited at the Interdepartmental Centre for Research andTherapy of Male Infertility, University of Siena. Semensamples were obtained from men with fertility problemsand from those who only wanted to be tested for fertility.All participants in this study signed a written informedconsent for the use of ejaculates in the research.

Semen samples were collected by masturbation after 3-5days of sexual abstinence and examined after liquefactionfor 30 min at 37°C. Volume, pH, concentration, and motilitywere evaluated according to WHO [1999] guidelines. Spermmorphology was assessed by the Papanicolaou (PAP) testmodified for spermatozoa. Ejaculated human sperm werewashed twice in phosphate buffered saline (PBS), smearedon pre-cleaned glass slides, air dried, and fixed in an equalvolume of ethanol 95% and ether for 5-15 min. The fixedslides were stained using the protocol in the WHO guide-lines [1992]. Sperm with malformations of head, midpiece,or tail were classified as ‘abnormal’.

Separation of sperm fractions: swim upFive of 33 ejaculates showing sperm motility ≥ 30% weretreated with the swim up technique in order to separatetwo different fractions: one fraction containing sperm withimproved motion characteristics and the other one withimmotile sperm. One ml of each semen sample was placedin a sterile conical centrifuge tube and gently layered with1.2 ml of Quinn’s ® Sperm Washing Medium (Sage, Invitro fertilization, Inc., Trumbull, CT, USA). The tubes,inclined at an angle of 45°, were incubated for 45 min at37°C with 5% CO2. One ml of the uppermost medium wasthen recovered and it was found to contain highly motile

AQP7 in human sperm 133

Copyright © 2012 Informa Healthcare USA, Inc.

Syst

Bio

l Rep

rod

Med

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

McG

ill U

nive

rsity

on

11/0

7/12

For

pers

onal

use

onl

y.

sperm cells; the lowermost fraction was represented by im-motile, often immature sperm.

Immunofluorescence of AQP7 in human spermThe ejaculated sperm from 33 individuals and the two spermfractions recovered by swim up of 5 semen samples werewashed in phosphate-buffered saline (PBS), smeared oncleaned glass slides, and air dried. The slides were thenrinsed in PBS, saturated for 20 min with PBS-bovine serumalbumin (BSA) 1% containing 5% normal goat serum(NGS) and incubated overnight at 4°C with an anti-AQP7rabbit polyclonal antibody (Santa Cruz Biotechnology Inc.,CA, USA) diluted 1:50 in PBS/0.1% BSA/1% NGS. The reac-tion was revealed by an anti-rabbit FITC conjugate antibodyraised in goat (Sigma-Aldrich, Milan, Italy), diluted 1:300.

Finally, the samples were washed three times in PBS andmounted with Vectashield (Vector Labs, Burlingame, CA,USA). Incubation with primary antibody was omitted incontrol samples. A negative control of the procedure wasperformed using a non-immune rabbit serum (Sigma-Aldrich) diluted 1:50 revealed by an anti-rabbit FITC conju-gate antibody raised in goat (Sigma-Aldrich), diluted 1:300.

Fluorescence was observed with a Leitz Aristoplan lightmicroscope (Leica, Wetzlar, Germany) equipped with a flu-orescence apparatus. Images were acquired using Leica QFluoro Standard, Leica Chantal software. Based on immuno-fluorescence results, AQP positive sperm were classified intotwo grades following the intensity and the localization oflabeling. Grade 1 showed clear staining in the pericentriolararea, at the midpiece level, often at the equatorial segmentand, weakly, along the tail. Grade 2 showed a diffuse labelingof lower intensity located in the cytoplasmic bodies, in theentire tail, and/or head. A minimum of 300 sperm werescored for each sample, with a total of 10,020 spermanalyzed.

Mathematically quantified transmission electron microscopyFor transmission electron microscopy (TEM), 16 out of 33sperm samples were fixed in cold Karnovsky fixative andmaintained at 4°C for 2h. Fixed semen was washed in 0.1mol/l cacodylate buffer (pH 7.2) for 12h, postfixed in 1%buffered osmium tetroxide at 4°C for 1h, dehydrated, andembedded in Epon Araldite. Ultra-thin sections were cutwith a Supernova ultramicrotome (Reickert Jung, Vienna,Austria), mounted on copper grids, stained with uranylacetate and lead citrate, and observed and photographedwith a Philips EM208 TEM (Philips Scientifics, Eindhoven,The Netherlands). Three hundred ultra-thin sperm sectionswere analyzed for each patient. Major submicroscopiccharacteristics were recorded by trained examiners whowere blind to the experiment. TEM data were quantifiedby a method that has been used in our laboratory formany years, i.e., a mathematical formula [Baccetti et al.1995] that considers 16 selected submicroscopic character-istics of sperm organelles, each of which can be present ina normal state or in one or more different abnormal states,or defects. These characteristics, in a normal or abnormalstate, were recorded during TEM examination and

subsequently elaborated by means of a Bayesan technique(probability calculation) which is able to quantify the dataobtained by TEM analysis by calculating the number of sper-matozoa free of structural defects in a semen sample (the fer-tility index, FI). The lowest fertility index assuring normalfertility was observed to be slightly lower than two millionof sperm free of ultrastructural defects [Collodel andMoretti 2008]. In addition, this method enables us toobtain the percentages of three main sperm pathologies:immaturity, necrosis, and apoptosis [Collodel and Moretti2008]. Sperm pathologies are defined by typical ultrastruc-tural characteristics: altered acrosomes, round or ellipticalnuclei with uncondensed chromatin, cytoplasmic dropletsfor immaturity, marginated chromatin, cytoplasmic residueswith translucent vacuoles and swollen and badly assembledmitochondria for apoptosis, broken plasma membrane,reacted or absent acrosome, misshaped nuclei with disruptedchromatin, and poor axonemal and periaxonemal cyto-skeletal structures for necrosis. FI, immaturity, necrosis,and apoptosis values are able to express the sperm qualityof each examined ejaculate in numbers. These scores areclosely interconnected because an increase in these pathol-ogies is concomitant with a decrease in the FI.

Statistical analysisThe data are reported as mean and standard deviation (SD)or as median and interquartile range (IQR: 25°-75° percen-tile) if the distributions were respectively normal or nonnormal. Normality of distribution was evaluated using theKolmogorov-Smirnov test. The Pearson correlation coeffi-cient was used to estimate the linear relationship betweenthe percentage of sperm with AQP7 staining of grade 1and the considered variables. The difference of the percen-tages of sperm with a dotted AQP7 staining in the motileand the immotile fractions obtained by swim up was calcu-lated with the Mann-Whitney U test. A P value < 0.05 (two-tailed) was considered statistically significant. All analyseswere performed using SPSS statistical software version 13(SPSS Inc., Chicago, IL, USA).

Declaration of interest: The authors have no declarations ofinterest to report.

ReferencesBaccetti, B., Bernieri, G., Burrini, A.G., Collodel, G., Crisa, N., Mirolli,

M., et al. (1995) Notulae seminologicae. 5. Mathematical evaluationof interdependent submicroscopic sperm alterations. J Androl16:356–371.

Buffone, M.G., Doncel, G.F., Calamera, J.C. and Verstraeten S.V.(2008) Capacitation-associated changes in membrane fluidity inasthenozoospermic human spermatozoa. Int J Androl 32:360–375.

Calamita, G., Mazzone, A., Bizzoca, A. and Svelto, M. (2001) Possibleinvolvement of aquaporin-7 and -8 in rat testis development andspermatogenesis. Biochem Biophys Res Commun 288:619–625.

Collodel, G. and Moretti, E. (2008) Morphology and meiotic segre-gation in spermatozoa from men of proven fertility. J Androl29:106–114.

Cross, N.L. (2004) Reorganization of lipid rafts during capacitation ofhuman sperm. Biol Reprod 71:1367–1373.

134 E. Moretti et al.

Systems Biology in Reproductive Medicine

Syst

Bio

l Rep

rod

Med

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

McG

ill U

nive

rsity

on

11/0

7/12

For

pers

onal

use

onl

y.

Hermo, L. and Smith, C.E. (2011) Thirsty Business: Cell, Region andMembrane Specificity of Aquaporins (AQPs) in the Testis, EfferentDucts and Epididymis and Factors Regulating Their Expression.J Androl 32:565–575.

Hibuse, T., Maeda, N., Nakatsuji, H., Tochino, Y., Fujita, K., Kihara,S., et al. (2009) The heart requires glycerol as an energy substratethrough aquaporin 7, a glycerol facilitator. Cardiovasc Res83:34–41.

Huang, H.F., He, R.H., Sun, C.C., Zhang, Y., Meng, Q.X. and Ma, Y.Y.(2006) Function of aquaporins in female and male reproductivesystems. Hum Reprod Update 12:785–795.

Ishibashi, K., Kuwahara, M., Gu, Y., Kageyama, Y., Tohsaka, A.,Suzuki, F., et al. (1997) Cloning and functional expression of anew water channel abundantly expressed in the testis permeableto water, glycerol, and urea. J Biol Chem 272:20782–20786.

Ishibashi, K., Hara, S. and Kondo, S. (2009) Aquaporin water channelsin mammals. Clin Exp Nephrol 13:107–117.

Kageyama, Y., Ishibashi, K., Hayashi, T., Xia, G., Sasaki, S. and Kihara,K. (2001) Expression of aquaporins 7 and 8 in the developing rattestis. Andrologia 33:165–169.

Kondo, H., Shimomura, I., Kishida, K., Kuriyama, H., Makino, Y.,Nishizawa, H., et al. (2002) Human aquaporin adipose (AQPap)gene. Genomic structure, promoter analysis and functionalmutation. Eur J Biochem 269:1814–1826.

Moretti, E., Geminiani, M., Terzuoli, G., Renieri, T., Pascarelli, N. andCollodel, G. (2011) Two cases of sperm immotility: amosaic of flagel-lar alterations related to dysplasia of the fibrous sheath and abnorm-alities of head-neck attachment. Fertil Steril 95:1787.e19–23.

Nixon, B., Mitchell, L.A., Anderson, A., McLaughlin, E.A., O’BryanM.K. and Aitken, R.J. (2010) Proteomic and functional analysisof human sperm detergent resistant membranes. J Cell Physiol.226:2651–2665.

Saito, K., Kageyama, Y., Okada, Y., Kawakami, S., Kihara, K., Ishiba-shi, K., et al. (2004) Localization of aquaporin-7 in human testisand ejaculated sperm: possible involvement in maintenance ofsperm quality. J Urol 172:2073–2076.

Sohara, E., Ueda, O., Tachibe, T., Hani, T., Jishage, K., Rai, T., et al.(2007) Morphologic and functional analysis of sperm and testesin Aquaporin 7 knockout mice. Fertil Steril 87:671–676.

Suzuki-Toyota, F., Ishibashi, K. and Yuasa, S. (1999) Immunohisto-chemical localization of a water channel, aquaporin 7 (AQP7), inthe rat testis. Cell Tissue Res 295:279–285.

Travis, A.J. and Kopf, G.S. (2002) The role of cholesterol efflux in reg-ulating the fertilization potential of mammalian spermatozoa. JClin Invest 110:731–736.

World Health Organization. (1992) WHO Laboratory Manual for theExamination of Human Semen and Sperm-Cervical Mucus Inter-action. Cambridge University Press, Cambridge, United Kingdom.

World Health Organization. (1999) WHO Laboratory Manual for theExamination of Human Semen and Semen-Cervical Mucus Inter-actions. 4th ed. Cambridge University Press, Cambridge, UnitedKingdom.

Yeung, C.H. (2010) Aquaporins in spermatozoa and testicular germcells: identification and potential role. Asian J Androl 12:490–499.

Yeung, C.H., Callies, C., Rojek, A., Nielsen, S. and Cooper, TG. (2009)Aquaporin isoforms involved in physiological volume regulation ofmurine spermatozoa. Biol Reprod 80:350–357.

Yeung, C.H. and Cooper, T.G. (2010) Aquaporin AQP11 in the testis:molecular identity and association with the processing of residualcytoplasm of elongated spermatids. Reproduction 139:209–216.

Yeung, C.H., Callies, C., Tüttelmann, F., Kliesch, S. and Cooper, T.G.(2010) Aquaporins in the human testis and spermatozoa - identifi-cation, involvement in sperm volume regulation and clinical rel-evance. Int J Androl 33:629–641.

AQP7 in human sperm 135

Copyright © 2012 Informa Healthcare USA, Inc.

Syst

Bio

l Rep

rod

Med

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

McG

ill U

nive

rsity

on

11/0

7/12

For

pers

onal

use

onl

y.