Use of a highly sensitive quantitative telomerase assayin intracytoplasmic sperm injection programmesfor the treatment of 47,XXY non-mosaic Klinefelter men*

Y. Yamamoto1, N. Sofikitis1,2, A. Kaponis1, J. Georgiou3, D. Giannakis3, CH. Mamoulakis3,D. Loutradis1, X. Yiannakopoulos2, Y. Mio3, I. Miyagawa1 and A. Chatzikyziakidou2

1Tottori University School of Medicine, Tottori, Japan; 2Ioannina University School of Medicine, Ioannina, Greece;3MFC Clinic, Yonago, Japan

Key words. ICSI–Klinefelter’s syndrome—spermatozoon—telomerase—testis

Summary. We evaluated the role of the sensi-tive quantitative telomerase assay (SQTA) in themanagement of men with non-mosaic Klinefelter’ssyndrome (KS). Diagnostic testicular biopsy (DTB)was performed in 24 men with KS. A part of theDTB was stained and the remaining fragment wasprocessed for the SQTA. After 3–18 months, atherapeutic testicular biopsy (TTB) was performedin the same testicle and the recovered specimenswere processed to identify spermatozoa. Men with aSQTA outcome equal to 0.00 Units lg)1 protein(n ¼ 7) demonstrated therapeutic testicular biopsymaterial that was negative for spermatogenic cells.In five men with a SQTA outcome of 8.11–38.03 Units lg)1, the most advanced germ cell wasthe spermatogonium ⁄ primary spermatocyte. In theremaining 12 men, the most advanced spermato-genic cell in the TTB was the spermatozoon. Inthese men, the SQTA outcome was equal to 25.76–92.68 Units lg)1 protein. Using 39.00 Units lg)1

protein as a cut-off value, the accuracy of theSQTA in identifying men positive for spermatozoawas 91.6%. It appears that the SQTA has a role foridentifying non-mosaic KS men who have testicularspermatozoa.

Introduction

The introduction of ooplasmic injections of testicu-lar spermatozoa or spermatids has revolutionizedthe treatment of nonobstructed azoospermic (NOA)men (Silbert et al., 1995; Silbert, 1996; Sofikitis et al.,1998a). One of the most perplexing problems inassisted reproduction programs dealing with NOA-men is the lack of parameters that can be used topredict the presence or absence of foci of testicularspermatozoa. It is widely accepted that the histo-logical results of the diagnostic testicular biopsy(DTB; tissue stain), the peripheral serum levels offollicle stimulating hormone (FSH) and the testicularsize are not highly accurate markers in predictingthe presence or absence of spermatozoa in thetesticles of NOA men (for review see Sofikitis et al.,1998a). Occasionally, men with small testicles, highperipheral serum FSH profiles, and histologicalimages of Sertoli cell–only syndrome are positive fortesticular foci of spermatozoa (Sofikitis et al., 1998a).It is of great clinical importance to find newparameters that can predict the presence ⁄ absenceof foci of advanced spermatogenesis in a therapeutictesticular biopsy [tissue cutting ⁄mincing and pro-cessing for assisted reproduction trials; therapeutictesticular biopsy (TTB)] sample. Evaluation of suchparameters may have a role in the therapeuticmanagement of men with Klinefelter’s syndrome(KS) known to have small testicles (Foss & Lewis,1971; Laron et al., 1982; Terzoli et al., 1992; Bourneet al., 1997; Tournaye et al., 1997a; Nodar et al.,1998; Palermo et al., 1998; Reubinoff et al., 1998;Ron-El et al., 1999). In these men, prediction of theabsence of spermatozoa in a TTB is of paramountimportance in avoiding unnecessary TTB that

Correspondence: Prof. Nikolas Sofikitis, Department of Urol-ogy, Tottori University School of Medicine, Nishimachi 36–1,Yonago 683, Japan. Tel.: +81 859 348119; Fax: +81 859348074; e-mail: akrosnin@hotmail.com*A part of this study was presented at the 55th Annual Meetingof The American Society of Reproductive Medicine in SanFrancisco, CA, October 4–9, 1998.

andrologia 34, 218–226 (2002) Accepted: January 10, 2002

U.S. Copyright Clearance Center Code Statement: 0303-4569/2002/3404-0218 $ 15.00/0 www.blackwell.de/synergy

would further reduce testicular volume. Consideringthat (i) KS is a common genetic disorder found in1 : 1000 live born males and in 3.1% of infertilemen (Sharara, 1998), and (ii) few pregnancies havebeen achieved with assisted reproduction in coupleswith KS (for review see Sharara, 1998); it is notsurprising that men with KS often appear in assistedreproduction centres. These men currently repre-sent a significant subpopulation of those infertilemen for whom an assisted reproduction trial is theonly hope of fathering their own children. Aprognostic parameter that would accurately indicatethose men with KS who were negative for testicularspermatozoa and could not be candidates forassisted reproduction programs would avoid theunnecessary expense, and associated risks, of ovar-ian stimulation for the female partners.

Telomeres are specialized structures at the endsof eukaryotic chromosomes that appear to functionin chromosome stabilization, positioning, and rep-lication (Blackburn, 1991). Telomeres stabilizenatural chromosome ends and inhibit aberrantfusions and rearrangements that occur on brokenchromosomes (Blackburn, 1991). The length of thetelomere, which contains TTAGGG repeats, pro-gressively decreases with cell division (Morin, 1989;Prowse et al., 1993). Telomere repeats are synthes-ized de novo onto chromosome ends by the enzymetelomerase. Telomerase, an RNA-dependent DNApolymerase, is the only enzyme to compensate forthe telomeric losses of DNA that occur at each celldivision (Hisatomi et al., 1997, 1999). It is wellknown that telomerase activity is expressed inimmortal cell lines and most human tumours, butis inactive in most normal somatic cells, tissuesadjacent to tumours, or benign growths (Kim et al.,1994; Kim, 1995). Telomerase activity has alsobeen demonstrated in human germ lines andtesticles (Wright et al., 1996; Yamamoto et al.,1999a, 2000a). Some investigators have reportedpositive telomerase activity in mouse, rat, andhuman spermatogonia ⁄ primary spermatocytes, sec-ondary spermatocytes, and round spermatids,whereas testicular and epididymal spermatozoaare negative for telomerase activity (Prowse et al.,1993; Eisenhauer et al., 1997; Yamamoto et al.,1999a,b). Telomerase is inactivated during sper-miogenesis (Yamamoto et al., 1999b). The telom-erase hypothesis suggests that telomerase activity isgreater in embryonic cells and decreases in somatictissues during development and differentiation(Eisenhauer et al., 1997). In a previous study, thesensitive quantitative telomerase assay (SQTA) wasused in DTB material collected from men withSertoli cell–only syndrome (Yamamoto et al.,1999a) and showed a high sensitivity, specificity,positive predictive value, and negative predictive

value, for the identification of Sertoli cell–onlysyndrome men with foci of haploid cells in theTTB. That study muted an important role for theSQTA in DTB for the prediction of testicular fociof haploid cells in the TTB of NOA men. Ourhypothesis that severe defects in spermatogenesisare accompanied by a reduction in testicular tissuetelomerase activity is consistent with the findingthat men with testicles with active spermatogenesis(i.e. men with obstructive azoospermia with normalspermatogenesis) have significantly larger SQTAprofiles than men with Sertoli cell–only syndrome(with or without foci of haploid cells) (Yamamotoet al., 1999a). SQTA was performed to quantifytelomerase activity (Hisatomi et al., 1997, 1999;Yamamoto et al., 1999a) and has been proven to bea highly sensitive and quantitative assay (Hisatomiet al., 1997; Yamamoto et al., 1999a,b, 2000a,b).

The objective of the present study was toevaluate the role of the SQTA in predicting thepresence or absence of spermatozoa in TTBmaterial recovered from men with non-mosaicKS. The SQTA is the gold standard for quantifyingtelomerase activity and this is the first publishedstudy to use SQTA for assessing the testicles of menwith KS. Additionally, we assessed the fertilizingpotential and the reproductive capacity of sperma-tozoa obtained from the testicular tissue of non-mosaic KS men. However, the results of the assistedreproduction are reported in another communica-tion (Yamamoto et al., 2001).

Participants and methods

Patients

Twenty-four men with non-mosaic KS werereferred to our facilities (age 23–45; FSH:14–56 IU L)1). Normal levels of FSH for males ofreproductive age in our facilities ranged from3-11 IU l)1. All these men were negative forspermatozoa in centrifuged semen samples. Theirwives were 20–41 years old. A DTB was per-formed. A part of the DTB material (5–8 mg) wasprocessed for haematoxylin-eosin staining. Theremaining piece (5 mg) was frozen and processedfor the SQTA. After 3–18 months, all men under-went a TTB (in the same testicle that had undergoneDTB) during assisted reproduction trials.

The TTB material (148–194 mg) was processedfor mincing, filtering and dispersion ⁄ extraction ofspermatogenic cells. During mincing ⁄ filtering of theTTB material, fractions of recovered cells wereobserved via a confocal scanning laser microscope-computer assisted system (CSLM-CAS). Othertesticular cellular fractions (from the TTB material)

Klinefelter’s syndrome and telomerase 219

ANDROLOGIA 34, 218–226 (2002)

were processed for fluorescent in situ hybridization(FISH). Recovered spermatozoa were processed forassisted reproduction and cryopreservation.

Highly sensitive quantitative telomerase assay (SQTA)

Details on the application of the SQTA have beendescribed previously (Hisatomi et al., 1997, 1999;Yamamoto et al., 1999a,b, 2000a). Previous experi-ments have shown that the SQTA is a quantitativeassay (Yamamoto et al., 1999a). Twenty aliquots ofextracts (each containing 1 lg protein) from each5-mg piece of testicular tissue from each participantwere prepared and processed for 20 TRAP assays(Yamamoto et al., 2000a). Thus, 20 TRAP assayswere performed for each participant. The averagefor each participant was then calculated.

The SQTA was performed by The JapaneseSpecial Reference Laboratory (Matsue, Japan) aspreviously described (Yamamoto et al., 1999a). Analiquot of extract containing 1 lg protein was usedfor each TRAP assay. Aliquots of the extracts wereincubated with 0.1 ng Cy-5 labeled TS primer(TRAP-eze). Following a 30-min incubation at30�C, a polymerase chain reaction (PCR) wasperformed at 94�C (30 s), 60�C (30 s) and 72�C(45 s) for 30 cycles (Yamamoto et al., 2000a). Theexternal control, a TSR8 (TRAP-eze), was used as apositive control (Hisatomi et al., 1997). The productswere diluted with an equal volume of formamide dyesolution, heated at 94�C for 5 min and applied(5 ll lane)1) to a 10% denaturing gel containing 6 m

urea fitted to an automatedDNAsequencer (ALF redTM DNA Sequencer: Pharmacia Biotech, Uppsala,Sweden). The temperature of the gel was kept at45�C during electrophoresis at 45 W. The data fromtheALF red TMDNA Sequencer were collected andanalyzed automatically by Fragment Manager V1.1(Pharmacia Biotech, Osaka, Japan). Each peak wasquantified in terms of size, peak height and peak area.The quantification of telomerase activity was deter-mined by the mathematic formula described byHisatomi et al. (1997) as we previously reported(Yamamoto et al., 1999a, 1999b). The SQTA out-come values were expressed as TPG (Total ProductGenerated) Units lg)1 protein. Quantification oftelomerase activity was determined by the followingformula (· 100):

fmeasured total area of telomerase activity

(50 bp, 56 bp, 62 bp, 68 bp ...) g=fmeasured

area of internal control (36 bp)g of the sample

fmeasured total area of telomerase activity (50 bp,

56 bp, 62 bp, 68 bp...)g=fmeasured area of internal

control (36 bp)g of the external control TSR8

Processing of therapeutic testicular biopsy material

The testicular tissue was washed three times withnormal saline. The seminiferous tubules were thenwashed in Dulbecco’s phosphate buffered saline(DPBS; Sigma Co., St. Louis, MI) containing5.6 mm glucose and 5.8 mm sodium lactate (modi-fied DPBS; Sofikitis et al. 1998a) and subsequentlyminced into small pieces. Samples were observedvia a dissecting microscope (Olympus SZ-STS;Olympus, Tokyo, Japan) during the tissue mincingprocess. The overall mincing process lasted 1 h.The samples were then centrifuged at 500 g for30 min, the sedimented pieces of tissue and cellswere resuspended in modified DPBS, and thesamples were passed through a filter paper of30–40lm pore size (Whatman Co., New York, NY)(Sofikitis et al., 1998b). The total volume of testicu-lar cell suspension, which passed through the filter,was 8–18 ml. The filtration procedure was per-formed at room temperature and lasted 15–25 min.A minor part of the filtrate (Fraction-1) wascollected, centrifuged at 750 g for 30 min, and thesedimented cells were resuspended in modifiedDPBS. Some of the sedimented cells were thenobserved via CSLM-CAS. The remaining sedi-mented cells were processed for FISH. The majorpart of the filtrate was filtered again (secondfiltration; via an 11–12 lm pore size filter). Filtra-tion via the 30–40 lm-pore size filter and the12 lm-pore size filter removed the small pieces ofseminiferous tubuli and most of the Sertoli cellfragments, debris, tissue fragments and spermato-gonia ⁄ primary spermatocytes (if present) from thefinal (second) filtrate. Thus in men with KS positivefor testicular foci of round spermatids and sperma-tozoa, the final filtrate contained mainly spermatidsand spermatozoa.

A small part (Fraction-2) of the final filtrate wasprocessed for observation via CSLM-CAS andFISH techniques. The major part of the finalfiltrate was processed for sperm identification.

Confocal-scanning laser microscope-computer assistedsystem

CSLM-CAS is powerful in the field of microscopy(Sofikitis et al., 1994, 1996a, 1996b; Yamanaka et al.,1997; Sofikitis et al., 1998a, b, c; Yamamoto et al.,2001). The capacity of the CSLM-CAS to accu-rately identify subtypes of germ cells has beenpreviously confirmed using FISH or transmissionelectron microscopy as a control (Yamanaka et al.,1997; Sofikitis et al., 1998a, c; Yamamoto et al.,1999a, 2001). The stage of the most advancedspermatogenic cell in Fraction-1 or Fraction-2 noted

220 Y. Yamamoto et al.

ANDROLOGIA 34, 218–226 (2002)

during observation (and quantitative morphometricanalysis) via CSLM-CAS was recorded.

Fluorescent in-situ hybridization techniques

FISH techniques were applied as an additionalmethodology to define the most advanced germ cellin Fractions-1 and -2 of the minced testicularsamples. Three colour FISH was performed usingpreviously described methodology (Harper et al.,1994; Sofikitis et al., 1998c; Yamamoto et al.,1999a). The probe kit was a combination of Yspectrum Red ⁄Pink, X spectrum Yellow, and 18spectrum Green ⁄Blue fluorescently labelled DNAprobes specific for chromosomes Y, X and 18,respectively.

Spermatogonia ⁄ primary spermatocytes were46-XY1818-germ cells or 47-XXY1818 germcells. Secondary spermatocytes were consideredX18-(2C)-DNA cells or Y18-(2C)-DNA cells (2Cindicates a diploid amount of DNA; Alberts et al.,1994; Sofikitis et al., 1998c) and round spermatidswere characterized X18-(C)-DNA cells or Y18-(C)-DNA cells (C indicates a haploid amount ofDNA; Alberts et al., 1994). Only clear fluores-cence signals of integral nuclei were analyzed.Two signals of the same colour were consideredto represent two individual chromosomes whenthe distance between these two signals was largerthan one signal’s diameter. Details on theperformance of FISH techniques for germ cellidentification and their outcome (normal andabnormal subtypes of germ cells) in men withKS are presented in another communication(Yamamoto et al., 2001).

As none of the participant’s secondary spermato-cytes were the most advanced germ cells in theminced TTB sample (all men positive for secondaryspermatocytes also had testicular spermatids andspermatozoa), the term chromosomally �haploid� cellrefers to spermatids and spermatozoa in thismanuscript. No attempt was made to distinguishbetween spermatogonia and primary spermatocytes.

Statistical analysis

Statistical analysis was performed using the non-parametric Wilcoxon’s test for comparisonsbetween groups. Values were expressed asmean ± SD. Considering that: (i) telomerase datahad been expressed as mean ± SD in previousreports evaluating testicular telomerase activity(Yamamoto et al., 1999a, b), and (ii) data is oftenexpressed as mean ± SD in the internationalliterature, even if the data distribution is notnormal, and so parametric indications of the data(i.e., mean ± SD) were used in the current study. A

probability of P < 0.05 was considered to bestatistically significant.

Results

The DTB results confirmed the Sertoli cell–onlysyndrome pathophysiology, hyalinization of semin-iferous tubules and hyperplasia of Leydig cells(Fig. 1) that is characteristic of KS in all participants.The TTB were negative for germ cells in seven men.However, 17 men were positive for foci of germ cellsin the TTB material. Results of CSLM-CAS andFISH techniques in Fraction-1 and Fraction-2indicated that the most advanced spermatogeniccell in the TTB was the primary spermatocyte orspermatozoon in five and 12 men, respectively(Tables 1 and 2; Fig. 2). All the men withouttesticular foci of spermatozoa had a SQTA outcomeof < 39.00 TPG Units lg)1 protein. (range: 0.00–38.03 TPG Units lg)1 protein.). Men in whom themost advanced germ cell in the TTB was theprimary spermatocyte had an SQTA outcome of8.11–38.03 TPG Units lg)1 protein. Among the12 men positive for spermatozoa, two men had aSQTA outcome <39.00 TPG Units lg)1 protein(equal to 25.76 and 30.91 TPG Units lg)1 protein),whereas the remaining 10 men had a SQTA out-come of > 39.00 TPG Units lg)1 protein (range:49.35–92.68 TPG Units lg)1 protein). The 12 menwho were positive for spermatozoa had a signifi-cantly greater mean SQTA outcome (P < 0.05)than the 12 men who were negative for spermatozoa(Table 2). Using 39.00 or 42.00 TPG Units lg)1

protein (the latter value was used as a cut-off value ina previous communication; Yamamoto et al., 1999a)as a cut-off, the sensitivity, specificity, positivepredictive value, and negative predictive value ofthe SQTA assay in DTB for prediction of thepresence ⁄ absence of spermatozoa in TTB materialwere 83.3, 100, 100, 85.7%, respectively. This cut-off value results in the highest specificity and positivepredictive value. Employment of any SQTA profilegreater than 38.03 and less than 49.35 TPGUnits lg)1 protein as a cut-off value results insensitivity, specificity, positive predictive value, andnegative predictive value of 83.3, 100, 100, and85.7%, respectively (Fig. 2). The overall accuracy ofthe SQTA in DTB for identification of men positivefor spermatozoa in a TTB was 91.6%. All sevenmen without germ cells in the TTB had an SQTAoutcome in the DTB equal to 0.00 TPGUnits lg)1 protein. Using a value of 25.00 TPGUnits lg)1 protein (Fig. 2) as a cut-off, the sensitiv-ity, specificity, positive and negative predictive valueof the SQTA assay in DTB for the prediction of thepresence ⁄ absence of spermatozoa in TTB material

Klinefelter’s syndrome and telomerase 221

ANDROLOGIA 34, 218–226 (2002)

were 100, 91.7, 92.3, and 100%, respectively. Thiscut-off value results in the highest sensitivity andnegative predictive value. Employment of anySQTA profile greater than 17.68 (Fig. 2) and lessthan 25.76 TPG Units lg)1 protein (Fig. 2) as acut-off value results in sensitivity, specificity, positivepredictive value, and negative predictive value equalto 100, 91.7, 92.3, and 100%, respectively.

Spermatogonia ⁄ primary spermatocytes, secon-dary spermatocytes, spermatids and spermatozoa

were found in Fraction-1 of the 12 men who werepositive for testicular spermatozoa. In these men,Fraction-2 contained mainly spermatids and sper-matozoa. Fraction-2 of the seven men who werenegative for germ cells in the TTB and the five menin whom the most advanced germ cell in the TTBwas the primary spermatocyte did not contain anygerm cells. Fraction-1 of the latter five men or theformer seven men contained spermatogonia ⁄ pri-mary spermatocytes mainly or no germ cells,respectively.

Chromosomal abnormalities observed via FISHin men with KS in the current study are describedand discussed extensively in another communica-tion (Yamamoto et al., 2001). In brief, within the

Table 2. Sensitive quantitative telomerase assay (SQTA)

outcome in subpopulations of men with non-mosaic Klinefelter

syndrome

Most advanced

spermatogenic

cell in TTB

n SQTA (TPG Units

lg)1 protein)

(mean ± SD)

No germ cells or primary

spermatocytes

12 7.23 ± 11.54a

Spermatozoon 12 62.53 ± 20.46b

a vs. b: P < 0.001, statistically significant.

TTB: Therapeutic testicular biopsy.

Table 1. Focal active spermatogenesis up to the stage of

primary spermatocyte (PS) or spermatozoon (most advanced

germ cell) in therapeutic testicular biopsy (TTB) material of

non-mosaic Klinefelter men

Tissue ⁄ cellprocessing

Number of men

without germ

cells

Number of men

with PS-arrest

(TTB)

Number of

men with

sperms (TTB)

DTB 24

TTB 7 5 12

DTB: Diagnostic testicular biopsy.

Figure 2. Relationship between quantitative telomerase assay indiagnostic testicular biopsy and the most advanced spermatogenic cellin therapeutic testicular biopsy material (as seen using confocalscanning laser microscope-computer assisted system or FISH). PS:Primary spermatocyte; SZ: Spermatozoon.

Figure 1. Haematoxylin-eosin stain oftesticular tissue of a man with Klinefelter’ssyndrome (original magnification · 400;image from a TV monitor).

222 Y. Yamamoto et al.

ANDROLOGIA 34, 218–226 (2002)

group of 12 men who were positive for testicularspermatozoa, the percentage of round spermatidsthat were haploid and normal was greater than93% (in Fraction-2).

Application of assisted reproduction trials for thetherapeutic management of KS–men resulted indelivery of five newborns.

Discussion

The role of SQTA in prediction of the presence ⁄ absenceof testicular spermatozoa in men with non-mosaic KS

When assisted reproduction programs are consid-ered for the therapeutic management of men withKS it is of paramount importance to recognize thesubpopulation of men negative for testicularspermatozoa, as the testicles of men with KS aresmall and therefore repeated testicular biopsiesduring assisted reproduction trials may have detri-mental consequences on testicular function. Inaddition, performance of assisted reproductiontrials (i.e. testicular biopsy-ovarian stimulation) inKS couples negative for testicular spermatozoaexposes the spouse to needless ovarian stimulationand all the associated risks. The DTB is of real valuein distinguishing between obstructive and nonob-structive azoospermia. In contrast, the importanceof DTB histology in the therapeutic management ofnonobstructed azoospermic men has been ques-tioned as a significant percentage of men withhistology in DTB indicating Sertoli cell–only syn-drome or spermatogenic arrest at the primaryspermatocyte stage occasionally demonstrate testi-cular foci of active spermatogenesis up to the roundspermatid or spermatozoon stage in TTB (forreview seeYamanaka et al., 1997; Sofikitis et al.,1998a). Devroey (1998) showed that the sensitivityand specificity of DTB histology to predict theoutcome of a TTB in nonobstructive azoospermiaare not high. Thus, histology in a DTB cannotaccurately predict the men with KS who are positivefor testicular spermatozoa who can subsequentlyparticipate in assisted reproduction programs. Inaddition, it is wellknown that the peripheral serumconcentrations of FSH and testicular size are notreliable predictors for successful testicular spermrecovery in TTB specimens (Tournaye et al., 1997b;Sofikitis et al., 1998a). In a previous study, wedemonstrated that the performance of the SQTA inDTB has a role in predicting the presence ⁄ absenceof testicular foci of spermatozoa in TTB of Sertolicell–only syndrome men (Yamamoto et al., 1999a).The study demonstrated that those men withhistological diagnosis of Sertoli cell–only syndromewith testicular foci of active spermatogenesis up to

the spermatozoon stage have significantly largertesticular SQTA outcome in a DTB than menwithout testicular foci of spermatozoa (Yamamotoet al., 1999a). The significant difference in theSQTA outcome between men with KS positive ornegative for testicular spermatozoa suggests thattesticles with foci of active spermatogenesis up to thespermatozoon stage have a larger SQTA outcomethan testicles without foci of spermatozoa. Thisdifference may be attributable to the fact that themen with KS who are positive for spermatozoa havemore active spermatogenesis focally and subse-quently a larger number of testicular stem cellmitoses, greater stem cell renewal, and enhancedgeneration of primary spermatocytes comparedwith men with KS who are negative for germ cellsor those with foci of spermatogenesis up to theprimary spermatocyte stage. The overall result isthat men with KS with testicular foci of spermato-zoa have a larger number of primary spermatocytesper tissue weight unit (Yamamoto et al., 2000b) andsubsequently, a larger SQTA outcome than menwith KS negative for testicular foci of spermatozoa.Primary spermatocytes are considered to be themain source of testicular telomerase activity (Yama-moto et al., 1999b). In addition, men with KS withtesticular foci of spermatozoa have a larger SQTAoutcome due to the presence of round spermatids(and telomerase positive cells) in their testicles(Eisenhauer et al., 1997; Yamamoto et al., 1999b).Furthermore, as men with KS with foci of sperma-togenesis arrested at the primary spermatocyte stagehave a defect in the primary spermatocyte nuclearcapacity to undergo the first meiotic division, theprobability that an additional qualitative or quan-titative defect is present in the primary spermatocytenucleus telomerase cannot be ruled out. Thus, thereduced SQTA outcome in men with KS positivefor primary spermatocytes but negative for sperm-atids ⁄ spermatozoa compared to those positive forspermatids ⁄ spermatozoa may also be due to thelower telomerase activity per primary spermatocyte.The latter speculation is supported by a previousstudy showing that highly purified fractions ofprimary spermatocytes recovered from healthy miceconsistently have larger SQTA outcomes thanpurified fractions of primary spermatocytes recov-ered from mice with primary testicular damage(Yamamoto et al., 1999b). In addition, preliminaryexperiments in animals have shown that highlypurified fractions of primary spermatocytes recov-ered from varicocelized rats with damage atspermatogenesis have significantly smaller SQTAprofiles than fractions of primary spermatocytesrecovered from control rats.

The minimal values of the SQTA outcome(0.00 TPG Units) in men with KS who are negative

Klinefelter’s syndrome and telomerase 223

ANDROLOGIA 34, 218–226 (2002)

for foci of germ cells in a TTB and the larger valuesof testicular SQTA outcome in men with KS whoare positive for spermatozoa than in men with KSwho are negative for spermatozoa raised theprobability of using the SQTA in DTB as a novelassay to recognize the men with KS with foci ofspermatozoa in TTB. Taking into considerationthat (i) all men with KS without testicular foci ofspermatozoa had an SQTA outcome <39.00 TPGUnits lg)1 protein, and (ii) 10 of the 12 men withKS with testicular foci of spermatozoa had a SQTAoutcome > 39.00 TPG Units lg)1 protein, weevaluated the role of 39.00 TPG Units lg)1 pro-tein as a cut-off value for the prediction of foci ofspermatozoa in TTB material of men with KS.Thus, using the latter value as a cut-off it was foundthat the sensitivity, specificity, positive predictivevalue, and negative predictive value of the SQTA inDTB for recognizing the subpopulation of menwith KS with foci of active spermatogenesis up tothe spermatozoon in TTB were 83.3, 100, 100 and85.7%, respectively. This cut-off value results in thehighest specificity and positive predictive value.Employment of any SQTA profile greater than38.03 and less than 49.35 as a cut-off value alsoresults in sensitivity, specificity, positive predictivevalue, and negative predictive value equal to 83.3,100, 100 and 85.7%, respectively (Fig. 2). Theoverall accuracy of the SQTA in DTB to predictthe presence of spermatozoa in TTB was high(91.6%). Using 25.00 TPG Units lg)1 protein as acut-off value, the sensitivity, specificity, positivepredictive value, and negative predictive value ofthe SQTA assay in DTB were 100, 91.7, 92.3 and100%, respectively. This cut-off value results in thehighest sensitivity and negative predictive value.Employment of any SQTA profile greater than17.68 and less than 25.76 as a cut-off value alsoresults in sensitivity, specificity, positive predictivevalue, and negative predictive value equal to 100,91.7, 92.3 and 100%, respectively.

Tournaye et al. (1997b) have reported that histo-pathology is not an accurate parameter to predictthe presence ⁄ absence of spermatozoa in TTB innonobstructed azoospermic men with peripheralserum FSH concentrations > 12 IU l)1 and testicu-lar volume < 15 ml (such values are almost alwaysfound in men with non-mosaic KS). It appears thatthe accuracy of the SQTA in DTB is higher orsuperior to the accuracy of histopathology and thusthe SQTA has a major role in the detection of menwith KS with foci of spermatozoa in TTB. Men withKS with an SQTA outcome of > 38.03 or 39.00TPG Units lg)1 protein in DTB may be encour-aged to undergo TTB and participate in assistedreproduction programs (see above; all men whowere negative for testicular spermatozoa had an

SQTA outcome of less than 39.00 TPG Units lg)1

protein). In contrast, men with KS with a SQTAoutcome of 0.00 TPG Units lg)1 protein will notbe successful in assisted reproduction and if theychoose to participate, their wives would undergounnecessary ovarian stimulation because the assis-ted reproduction cycle would be cancelled due tolack of testicular spermatozoa. As the SQTA is ahighly sensitive quantitative assay (Yamamoto et al.,1999a, 2000a), a very small amount of DTBmaterial (equal or less than 5 mg) is necessary tocalculate the testicular telomerase activity. It is ofclinical importance that a minor amount of DTBcan ascertain, with high accuracy, the men with KSwho could be candidates for assisted reproductiontechniques. We speculate that a SQTA can beperformed in cellular samples recovered by percu-taneous testicular aspiration. Such a study iscurrently under way in our facilities.

One of the smallest cut-off values that resulted in100% specificity and positive predictive value in thecurrent study was 39.00 TPG Units lg)1 protein(Fig. 2). There is a small difference between thiscut-off value and that of the SQTA (42.00 TPGUnits lg)1 protein) defined in another study(Yamamoto et al., 1999a) when the SQTA inDTB was applied to predict the presence ⁄ absenceof testicular spermatozoa in a general group of menwith Sertoli cell–only syndrome. However, it shouldbe emphasized that when the SQTA value of 42.00TPG Units lg)1 protein was employed as a cut-ofvalue in the current study, the sensitivity, specificity,positive predictive value and negative predictivevalue were same as those found when a cut-off valueof 39.00 TPG Units lg)1 protein was employed.All these observations suggest that SQTA, in asmall amount of DTB, is a promising, novel assay,important for the therapeutic management of menwith KS. Additional studies are necessary toconfirm the importance of the SQTA in DTB inthe therapeutic management of a general popula-tion of nonobstructed azoospermic men.

Assays to evaluate telomerase activity in testicular tissue

Telomerase activity is present in testicular roundgerm cells up to the stage of the round spermatid(Eisenhauer et al., 1997; Yamamoto et al., 1999b).Telomerase is expected to remain active inprimary and secondary spermatocytes to ensurethe transmission of full-length chromosomes to theround spermatid and subsequently to the sper-matozoon. We have previously shown (Yamamotoet al., 1999b) that telomerase activity is inhibitedduring transformation of a round spermatid to aspermatozoon. To assess testicular tissue telom-erase activity, a qualitative assay (Eisenhauer et al.,

224 Y. Yamamoto et al.

ANDROLOGIA 34, 218–226 (2002)

1997; Yamamoto et al., 1999b), a relative telo-merase assay (Yamamoto et al., 1999b) and aPCR ⁄ELISA assay evaluating colour changes (Fu-jisawa et al., 1998) have previously been applied.The above three assays are not very sensitive andrelatively subjective (Yamamoto et al., 1999b). Thecurrent study is the second report in the interna-tional literature applying the SQTA for quantify-ing telomerase activity in human testicular tissue.The SQTA is highly sensitive and is considered tobe the gold standard for telomerase quantification(Hisatomi et al., 1997, 1999; Yamamoto et al.,1999a). In the current study, the accuracy of theSQTA in predicting the testicular meiotic activitywas further increased by performing the assay 20times for each participant’s piece of DTB; theaverage was then calculated. Schrader et al. (2000)performed an ELISA-based testicular tissue telom-erase evaluation and concluded that testiculartelomerase profile is a highly specific marker ofgametogenesis.

CSLM-CAS microscopy and FISH techniques resultin consistent findings

Observation of different fractions of TTB samplesvia CSLM-CAS and FISH for identification of themost advanced spermatogenic cells resulted inconsistent findings in each participant. It appearsthat both FISH and CSLM-CAS are accurate andof great importance for evaluating the stages ofspermatogenic cells present in men with non-obstructive azoospermia.

Homogenously and nonhomogenously distributedspermatogenesis in the testicle

Two men positive for spermatozoa in the TTBsample had a relatively low SQTA outcome in theirDTB samples. As SQTA is a highly sensitivequantitative assay that can accurately predict thepresence of spermatozoa in the testicle, we believethat foci of active spermatogenesis up to thespermatozoon stage were present in the largeTTB sample but not in the small DTB piece(Fig. 3). This may be the cause of the relativelysmall SQTA profiles in the DTB of the above twomen. In these men, foci of active spermatogenesisresulting in the generation of spermatozoa mayhave been nonhomogenously distributed in thetesticular tissue and therefore may have been absentfrom the very small piece of DTB material. As theTTB samples were relatively large, the probabilityof including foci of nonhomogenously distributedspermatogenesis was higher than in the limitedDTB samples. The remaining 10 men that werepositive for testicular spermatozoa in TTB with

large SQTA values in DTB may have hadhomogenously distributed foci of active spermato-genesis ⁄ spermatozoa in the testicle that may havebeen apparent in the large TTB samples and in thesmall DTB samples. In the above 10 men, thepresence of foci of active spermatogenesis up to thespermatozoon stage within the limited DTB ma-terial resulted in high values for the SQTA.

The current study includes the largest number ofmen with non-mosaic KS evaluated in one singlestudy in the international literature. In addition, ourstudy reports the delivery of five newborns fatheredby men with non-mosaic KS. Application of thehighly sensitive SQTA in DTB in the current studyshowed that (i) men with KS with testicular foci ofspermatozoa demonstrate larger SQTA outcomesthan men with KS negative for spermatozoa, (ii)men with KS without foci of germ cells in TTB hada low SQTA outcome in DTB, and (iii) when the39.00 TPG Units-value was used as a cut-off value,the SQTA in DTB has a high sensitivity, specificity,positive predictive value, and negative predictivevalue for the identification of men with KS with fociof spermatozoa in a TTB. Considering that theamount of DTB necessary for performing theSQTA is very small it appears that application ofthis assay in DTB has a role in the therapeuticmanagement of men with KS.

Acknowledgement

Part of the cost of this study was covered by aclinical grant from Tottori University Hospital.

Figure 3. (a) Testicle with homogenous distribution of active ⁄advanced spermatogenesis up to the spermatozoon stage. Note thatfoci of spermatozoa (red spots) are not only present in the large TTBsample but also in the limited DTB specimen. (b) Testicle withnonhomogenous distribution of foci of active spermatogenesis up to thespermatozoon stage. A result of this nonhomogenous distribution is thatthe small piece of DTB is negative for foci of spermatozoa. The largeTTB material is positive for foci of spermatogenesis up to thespermatozoon stage.

Klinefelter’s syndrome and telomerase 225

ANDROLOGIA 34, 218–226 (2002)

References

Alberts B, Dennis B, Lewis J et al. (1994) Molecular Biology of theCell. Garland Publishing Inc, New York, pp 1–1273.

Blackburn EH (1991) Structure and function of telomerase.Nature 350:569.

Bourne H, Stern K, Clarke G (1997) Delivery of normal twinsfollowing the intracytoplasmic injection of spermatozoa froma patient with 47,XXY Klinefelter’s syndrome. HumReprod 12:2447–2450.

Devroey P (1998) Clinical application of new micromanipula-tive technologies to treat the male. Hum Reprod 13 (Suppl.3):112–122.

Eisenhauer KM, Gerstein RM, Chiu CP (1997) Telomeraseactivity in female and male rat germ cell undergoing meiosisand in early embryos. Biol Reprod 56:1120–1126.

Foss GL, Lewis FJ (1971) A study of four cases with Klinefelter’ssyndrome, showing motile spermatozoa in their ejaculates.J Reprod Fertil 25:401–408.

Fujisawa M, Tanaka H, Tatsumi N (1998) Telomerase activityin the testes of infertile patients with selected causes. HumReprod 13:1476–1479.

Harper JC, Coonen E, Ramaekrs FC, Delhanty JD, HandysideA, Winsten RM, Hopman AH (1994) Identification of thesex of human preimplantation embryos in two hours usingan improved spreading method and fluorescent in-situhybridization. Hum Reprod 9:721–724.

Hisatomi H, Nagao K, Kanamaru T (1999) Levels of telo-merase catalytic subunit mRNA as a predictor of potentialmalignancy. Int J Oncol 14:727–732.

Hisatomi H, Nagao K, Komatsu H (1997) Quantification oftelomerase activity in human liver tissue by fluorescence-based TRAP analysis. Hepatol Res 7:35–42.

Kim NW (1995) Telomerase dynamics and telomerase acti-vation in tumor progression: prospects for prognosis andtherapy. Oncol Res 7:121–125.

Kim NW, Piatyszek MA, Prowse KR (1994) Specific associ-ation of human telomerase activity with immortal cells andcancer. Science 266:2011–2013.

Laron Z, Dickerman Z, Zamir R (1982) Paternity inKlinefelter’s syndrome – a case report. Arch Androl 8:149–151.

Morin G (1989) The human telomerase terminal transferase isa ribonuclear protein that synthesizes TTAGGG repeats.Cell 59:521–523.

Nodar F, De Vincentiis S, Olmedo SB (1998) Birth of twinmales with normal karyotype after intracytoplasmic sperminjection with use of testicular sperm from a non-mosaicpatient with Klinefelter’s syndrome. Fertil Steril 71:1149–1152.

Palermo GD, Schlegel PN, Sills ES (1998) Births after intra-cytoplasmic injection of sperm obtained by testicularextraction from men with nonmosaic Klinefelter’s syndrome.N Engl J Med 338:588–590.

Prowse KR, Avilion AA, Greider CW (1993) Identification of anonprocessive telomerase activity from mouse cells. ProcNatl Acad Sci USA 92:4818–4820.

Reubinoff BE, Abeliovich D, Werner M (1998) A birth in non-mosaic Klinefelter’s syndrome after testicular fine needleaspiration, intracytoplasmic sperm injection and preim-plantation genetic diagnosis. Hum Reprod 13:1887–1892.

Ron-El R, Friedler S, Strassburger D (1999) Birth of a healthyneonate following the intracytoplasmic injection of testicularspermatozoa from a patient with Klinefelter’s syndrome.Hum Reprod 14:368–370.

Schrader M, Muller M, Heicappell R (2000) Telomeraseactivity and expression of telomerase subunits in the

testicular tissue of infertile patients. Fertil Steril 73:706–711.

Sharara FI (1998) Klinefelter’s syndrome: critical review andnew developments. Mid East Fertil J 3:109–113.

Silbert SJ (1996) Sertoli cell–only syndrome. Hum Reprod11:229–233.

Silbert SJ, Van Steirteghem A, Devroey P (1995) Sertoli cellonly syndrome revisited. Hum Reprod 10:1031–1032.

Sofikitis N, Miyagawa I, Zavos PM (1994) Confocal scanninglaser microscopy of morphometric human sperm para-meters: correlation with acrosin profiles and fertilizingcapacity. Fertil Steril 62:376–386.

Sofikitis N, Toda T, Miyagawa I (1996a) Beneficial effects ofelectrical stimulation before round spermatid nuclei injec-tions into rabbit oocytes on fertilization and embryonicdevelopment. Fertil Steril 65:176–185.

Sofikitis N, Miyagawa I, Incze P, Yamamoto Y (1996b)Detrimental effects of left varicocele on the reproductivecapacity of the early haploid male gamete. J Urol 156:267–270.

Sofikitis N, Yamamoto Y, Loutradis D, Miyagawa I (1998a)Micro- and macro- consequences of ooplasmic injections ofearly haploid male gametes. Hum Reprod Update 4:197–212.

Sofikitis N, Yamamoto Y, Kanakas N, Miyagawa I (1998b)Ooplasmic injections of elongating spermatids for thetreatment of non-obstructive azoospermia. Hum Reprod13:709–714.

Sofikitis N, Mantzavinos T, Loutradis D, Miyagawa I (1998c)Ooplasmic injections of secondary spermatocytes for thetreatment of non-obstructive azoospermia. Lancet 351:1177–1178.

Terzoli G, Lalatta F, Lobbiani A (1992) Fertility in a 47,XXYpatient: assessment of biological paternity by deoxyribonu-cleic acid fingerprinting. Fertil Steril 58:821–822.

Tournaye H, Camus M, Vandervorst M (1997a) Surgicalsperm retrieval for intracytoplasmic sperm injection. Int JAndrolsupplement 3 (20):69–73.

Tournaye H, Verheyen G, Nagy P (1997b) Are there anypredictive factors for successful testicular sperm recovery inazoospermic patients? Hum Reprod 12:80–86.

Wright W, Piatyszek M, Rainey E (1996) Telomerase activityin human germ line and embryonic tissues and cells. DevGenet 18:173–177.

Yamamoto Y, Sofikitis N, Mio Y (1999a) Highly sensitivequantitative telomerase assay of diagnostic testicular biopsymaterial predicts the presence of haploid spermatogenic cellsin therapeutic testicular biopsy in men with Sertoli cell–onlysyndrome. Hum Reprod 14:3041–3047.

Yamamoto Y, Sofikitis N, Mio Y, Miyagawa I (2001)Morphometric and cytogenetic characteristics of testiculargerm cells and Sertoli cell secretory function in men withnon-mosaic Klinefelter’s syndrome. Hum Reprod in press.

Yamamoto Y, Sofikitis N, Miyagawa I (2000a) Telomeraseactivity in testicular tissue material. Hum Reprod 15:2058–2059.

Yamamoto Y, Sofikitis N, Miyagawa I (2000b) Cytogeniccharacteristics of testicular germ cells from non-mosaicKlinefelter men. J Urol 163 (Suppl.):296.

Yamamoto Y, Sofikitis N, Ono K, Miyagawa I (1999b) Post-meiotic modifications of spermatogenic cells are accom-panied by inhibition of telomerase activity. Urol Res 27:336–345.

Yamanaka K, Sofikitis N, Miyagawa I, Mekras G (1997)Ooplasmic round spermatid nuclear injection procedures asan experimental treatment for nonobstructive azoospermia.J Assist Reprod Genet 14:55–62.

226 Y. Yamamoto et al.

ANDROLOGIA 34, 218–226 (2002)