Human Reproduction vol.15 no.4 pp.840–845, 2000
Relationship between semen quality and the seminal
plasma components carnitine, alpha-glucosidase,
fructose, citrate and granulocyte elastase in infertile
men compared with a normal population
A.Zöpfgen1, F.Priem2, F.Sudhoff1, K.Jung1,
S.Lenk1,3, S.A.Loening1 and P.Sinha2
1Department
of Urology and 2Institute of Laboratory Medicine and
Pathobiochemistry, University Hospital Charité, Humboldt
University, Berlin, Germany
3To
whom correspondence should be addressed at: Department of
Urology, University Hospital Charité, Humboldt University Berlin,
Schumannstrasse 20/21, D-10117 Berlin, Germany
The seminal plasma components neutral α-glucosidase,
carnitine, fructose, citrate, and polymorphonuclear granulocyte (PMN) elastase in 253 men were determined. The
seminal plasma of 221 infertile men, a control group with
proved fertility and 13 patients after vasectomy were
investigated. The concentrations of free carnitine (212
versus 521 µmol/l, n ⍧ 219, P < 0.001), total carnitine (437
versus 743 µmol/l, n ⍧ 219, P < 0.001), and the activity of
neutral α-glucosidase (15.1 versus 23.4 IU/l, n ⍧ 236, P <
0.05) were significantly reduced in the infertile patient
group as compared to the fertile control group, the concentration of PMN elastase (102 versus 48 µg/l, n ⍧ 234, P <
0.05) being significantly increased in the infertile patients.
In the patients after vasectomy the activity of neutral αglucosidase was the only epididymal marker that was
significantly reduced (4.3 versus 9.8 IU/l, n ⍧ 35, P ⍧
0.002) in comparison with the patients with testicular
azoospermia. At a limit of 6 IU/l the sensitivity of the
method was 92% and the specificity was 72%. Altogether,
the epididymal markers were reduced in subfertile patients
compared with the control group. For the differential
diagnosis of azoospermia only the determination of the
neutral α-glucosidase activity is useful.
Key words: biochemical markers/carnitine/α-glucosidase/male
infertility/seminal plasma
Introduction
In the seminal plasma, after separation of the cellular components, various chemical substances are to be found, which can
be assigned to specific organs or compartments of the male
genital system. Those substances can thus serve as diagnostic
indicators [World Health Organization (WHO), 1993].
Neutral α-glucosidase, carnitine, and glycerolphosphocholine are markers of epididymal function (Cooper et al., 1990;
WHO, 1993). Fructose and prostaglandins are mainly generated
in the seminal vesicles, and the determination of citrate, zinc,
and of the prostate-specific acid phosphatase enables the
prostatic function to be evaluated (WHO, 1993). Polymorpho840
nuclear granulocyte elastase (PMN elastase) in the seminal
plasma, which is released by degranulation of polymorphonuclear granulocytes, can be used for diagnosing and observing
the course of a clinically quiescent infection of the spermatic
vessels (Jochum et al., 1986; Reinhardt et al., 1997).
Although clinical evaluations are available with respect to
the individual parameters (Wetterauer, 1986; Cooper et al.,
1988; Wolff et al., 1991; Reinhardt et al., 1997), these
components were rarely compared or if so, the comparison
was based on a small number of biochemical measurements.
Therefore, it was the aim of the present study to compare the
concentrations of neutral α-glucosidase, total carnitine and
carnitine fractions, free carnitine and acetylated carnitine,
fructose, citrate, and PMN elastase in seminal plasma of
infertile patients with the respective values measured in a
control group and thus to determine the diagnostic validity of
these markers as additional tools of the classic spermiogram.
Materials and methods
Study subjects
Samples of seminal plasma of a total of 253 men with an average
age of 33.2 years (18–59) were obtained. In all, 221 men presented
at our department because they had failed to achieve conception in
⬎2 years. In addition seminal plasma of 13 patients after vasectomy
and of 19 patients with proven fertility was investigated. All 13 men
after vasectomy had azoospermia (obstructive azoospermia). The
spermiograms of the 19 fertile men in the control group were normal.
The 221 infertile men were subdivided according to their
spermiograms (as recommended by WHO, 1993): normozoospermia,
oligozoospermia, asthenozoospermia, teratozoospermia, OAT (oligoasthenoteratozoospermia) syndrome, azoospermia. Testicular azoospermia was proved by testicular biopsy in 22 of the patients with
azoospermia.
Retrieval and preparation of ejaculates
Retrieval, analysis and classification of the ejaculates were performed
according to the WHO recommendations (WHO, 1993).
Samples were obtained by masturbation into a sterile plastic vessel
with a preparatory period of at least 3 days. After physical examination
of the ejaculate (pH, volume, consistency, aspect), sperm motility
after 60 min was determined microscopically at 400-fold magnification
from the native sample and assigned to the categories (i) vividly
motile, (ii) moderately motile, and (iii) immotile. Sperm count and
concentration of round cells were determined in the counting chamber
Thoma (depth 0.100 mm, volume 0.0025 mm3) after dilution with a
bicarbonate, formalin, and gentian violet solution. The number of
peroxidase positive cells was determined using a conventional staining
technique with benzidine and cyanosine (Ludwig et al., 1996). The
vitality of the spermatozoa was determined by staining with eosin.
The differentiation between normal and pathological spermatozoa
© European Society of Human Reproduction and Embryology
Seminal plasma components related to semen quality
followed the WHO recommendations (WHO, 1993). Spermatozoa
were considered normal if there were neither defects of the head
(length 4.0–5.5 µm, breadth 2.5–3.3 µm, shape oval, ratio of length
to breadth 1.5–1.75, acrosome easily distinguishable) nor defects of
the neck, tail or centre part. One hundred spermatozoa were counted.
The seminal plasma was centrifuged at 3000 g for 15 min within
2 h after sampling. The supernatant was carefully removed and stored
at –80°C before biochemical analysis.
Biochemical analysis
Neutral α-glucosidase, fructose, and citric acid were determined with
commercially available test kits from Boehringer Mannheim GmbH
(Mannheim, Germany). For determining neutral α-glucosidase, 4nitrophenyl-α-D-glucopyranoside was converted into 4-nitrophenol
and α-D-glucopyranoside, and 4-nitrophenol were photometrically
measured. The acid isoenzyme was inhibited by addition of 1%
sodium dodecyl sulphate and pH 6.8 phosphate buffer, so that only
the neutral isoform was measured (Cooper et al., 1990). Fructose
was determined according to the hexokinase method (Kunst et al.,
1984) and citric acid with the ultraviolet method using the citrate
lyase catalysed reaction (Möllering, 1985).
Granulocyte elastase was measured with the homogeneous immunoassay ECOLINE® PMN elastase (Merck, Darmstadt, Germany). For
this purpose, seminal plasma was mixed with latex particles coated
with antibody fragments [F(ab⬘)2] against human PMN elastase,
and after agglutination the opacity proportional to the elastase
concentration was photometrically measured.
Free, acetylated and total carnitine were determined according to
a radiometrical method (McGarry and Foster, 1976). With this
technique, carnitine, after addition of [14C]acetylcoenzyme A and
carnitine acetyltransferase, was completely transformed into acetylcarnitine. Subsequent treatment of the reaction mixture with anionexchange resin Dowex 1-X10 removed the highly negatively charged
[14C]acetyl-CoA. The positively charged [14C]acetylcarnitine
remained in solution and was determined using an LKB-Wallac 1410
liquid scintillation isotope counter. A reference curve was used to
obtain the carnitine content of the sample by calibration. Total
carnitine was determined after previous saponification with 200 mmol/
l NaOH.
Statistical evaluation
Data were statistically evaluated with the Statistics Package for Social
Sciences (SSPS) 7.5 program for Windows (SPSS Inc., Chicago,
USA). The Kolmogorov–Smirnov test was used to test the goodness
of fit to the normal or non-normal distribution of values. The
significance of the differences within the individual groups of the
test population was evaluated with the Kruskal–Wallis test for nonnormally distributed parameters and with the analysis of variance
(ANOVA) test for normally distributed values. To answer one-sided
questions, the one-sided significance was given. In the rest of the
cases the two-sided significance was determined. For detecting a
statistical correlation between two variables Spearman’s coefficient
of correlation of ranked data was calculated. Receiver operating
characteristics (ROC) curve analysis was used to identify criterion
values and to determine the discrimination power between two groups
(Kairisto and Poola, 1995). The measured values were given as
median values and interquartile range.
Results
Biochemical parameters in seminal plasma of the control
group and of infertile patients
Figure 1 shows the data of the individual groups in the form
of box plots. As not all patients in every subgroup could be
measured for each parameter, the real number of patients
investigated in each subgroup was given at the bottom of the
figure. Total carnitine was not included in the figure because
it behaved almost the same way as free carnitine. The results
can be summarized as follows:
1. When comparing infertile men having normozoospermia
(n ⫽ 40) with the control group (n ⫽ 9), we found
significantly reduced concentrations of free carnitine (295
versus 521 µmol/l, P ⬍ 0.001) as well as total carnitine
(513 versus 743 µmol/l, P ⬍ 0.001) and significantly
increased concentrations of PMN elastase (144 versus
48 µg/l, P ⬍ 0.05) were found in the infertile group (Figure
1). The rest of the parameters showed no differences. This
was also true especially for the standard markers fructose
and citrate.
2. In the group of infertile patients, striking differences were
detected only for neutral α-glucosidase, for total and for
free carnitine, depending on the semen classification. In
patients with oligozoospermia (n ⫽ 103) and azoospermia
(n ⫽ 22) these parameters were significantly reduced
(P ⬍ 0.01) compared with the values measured in patients
with normozoospermia. It was striking that in the patients
with asthenozoospermia (n ⫽ 17) significantly increased
(P ⬍ 0.001) concentrations of α-glucosidase, total and free
carnitine were measured in comparison with the rest of
patients. In eight of 40 infertile patients with normozoospermia, PMN elastase concentrations ⬎1000 µg/l were
measured.
The activity of neutral α-glucosidase was significantly
reduced in the group with obstructive azoospermia (n ⫽ 13,
patients after vasectomy) compared with the patients with
testicular azoospermia (4.3 versus 9.8 IU/l, P ⬍ 0.01), whereas
the rest of the parameters indicating the secretory capacity
of the epididymis (free carnitine, total carnitine, acetylated
carnitine) were not significantly reduced in concentration in
the patients with obstructive azoospermia (Table I).
Altogether, the epididymal markers were reduced in subfertile patients compared with the control group, the reduction of
free carnitine and total carnitine being the most distinct. For the
differential diagnosis of azoospermia, only the determination of
the neutral α-glucosidase concentration was useful.
All patients in the group with obstructive azoospermia
showed glucosidase activities ⬍9 IU/l or 26 mIU/ejaculate.
ROC curve analysis showed a maximum diagnostic efficiency
at the threshold value of 6 IU/l to differentiate between
testicular azoospermia and obstructive azoospermia. At this
limit, the sensitivity was 92% and the specificity 72% for
diagnosing an obstruction. The positive predictive value was
66% and the negative predictive value 94%.
Correlation between biochemical parameters and spermiogram findings
In Tables II and III, the correlations between the biochemical
parameters and the quality criteria of the spermiogram are
summarized.
Between the biochemical parameters of the epididymal
function, free carnitine, total carnitine, and neutral α-glucosid841
A.Zöpfgen et al.
Figure 1. Biochemical markers in the seminal plasma of controls and in the infertile men. The infertile men were classified into subgroups
according to semen parameters. Symbols: contr ⫽ controls; nor ⫽ normozoospermia; ol ⫽ oligozoospermia; azo ⫽ azoospermia; ast ⫽
asthenozoospermia; ter ⫽ teratozoospermia; OAT ⫽ oligoasthenoteratozospermia syndrome; PMN ⫽ polymorphonuclear granulocyte. The
central box in the box-and-whisker plot represents the interquartile range, the whiskers represent the ranges without the outliers and the
middle line in the box represents the median.
ase, a close statistical relationship with a correlation coefficient
⬎0.5 was shown, while between the other parameters only
single correlations of not more than 0.25 occurred (Table II).
For the concentration of free carnitine the strongest positive
correlations were demonstrated in the spermiogram to the
842
number of spermatozoa as well as to motility, vitality, and
morphology, whereas these correlations were less intensive for
total carnitine and for the activity of neutral α-glucosidase
(Table III).
The correlation between the concentration of PMN elastase
Seminal plasma components related to semen quality
Table I. Epididymal markers in cases of testicular and obstructive azoospermia. Values are given as medians
and interquartile range in parentheses
Neutral alpha-glucosidasea (IU/l)
Free carnitine (µmol/l)
Total carnitine (µmol/l)
aSignificant
Testicular azoospermia (n ⫽ 22)
median (interquartile range)
Obstructive azoospermia (n ⫽ 13)
median (interquartile range)
9.8 (4.4–16.5)a
139 (78–239)
375 (352–478)
4.3 (3–4.7)
153 (59–213)
289 (230–470)
difference between the groups (P ⫽ 0.002).
Table II. Rank correlations between the biochemical parameters in seminal plasma
Free
carnitine
Free carnitine (µmol/l)
Total carnitine (µmol/l) 0.747a
Acetylcarnitine (µmol/l) –0.139
Neutral
0.634a
α-glucosidase (IU/l)
Fructose (mmol/l)
0.011
Citrate (mmol/l)
0.153b
PMN elastase (µg/l)
0.038
aSignificant
bSignificant
Total
Acetylcarnitine Neutral
Fructose
carnitine
α-glucosidase
Citrate
0.747a
0.153b 0.038
0.061
0.013
–0.126b –0.028
0.151b 0.040
–0.139b
0.477a
0.477a
0.548a
–0.007
0.116
0.061
0.013
0.158b
–0.126b
–0.028
0.634a
0.548a
–0.007
–0.085
0.151b
0.040
0.011
0.116
0.158b
–0.085
–0.072
–0.072
–0.021
PMN
elastase
–0.021
–0.240a
–0.240a
at the level P ⬍ 0.001.
at the level P ⬍ 0.05.
Table III. Rank correlations between semen parameters and the concentrations of free carnitine, total
carnitine, neutral α-glucosidase, fructose, citrate, and PMN elastase in seminal plasma. The number of
correlations is given in parentheses
Free carnitine (µmol/l)
Total carnitine (µmol/l)
Acetylcarnitine (µmol/l)
Neutral α-glucosidase (IU/l)
Fructose (mmol/l)
Citrate (mmol/l)
PMN elastase (µg/l)
aSignificant
bSignificant
Sperm count
(% ⫻ 106)
Motility
(progressive %)
Morphology
(normal %)
Vitality
(vital %)
0.51a (245)
0.42a (245)
–0.02 (245)
0.50a (263)
–0.15b (261)
–0.06 (264)
0.07 (257)
0.33a (164)
0.16b (164)
–0.09 (164)
0.16b (176)
–0.1 (176)
–0.09 (177)
0.04 (174)
0.31a (156)
0.12 (156)
–0.07 (156)
0.1 (163)
–0.08 (169)
–0.06 (170)
0.02 (167)
0.25a (157)
0.20b (157)
0.08 (157)
0.14 (162)
–0.02 (169)
–0.15b (170)
0.03 (168)
at the level P ⬍ 0.001.
at the level P ⬍ 0.05.
in seminal plasma and the number of peroxidase-positive white
blood cells in the ejaculate was clearly positive (r ⫽ 0.717,
P ⬍ 0.001). There was no statistical relationship between the
microbiological findings in the ejaculate and the PMN elastase
concentration in seminal plasma.
Discussion
The biochemical parameters in the seminal plasma of infertile
and of fertile men were seldom compared and, if so, had led
to variant results. Kohengkul et al. (Kohengkul et al., 1977)
were able to prove a reduced acetylcarnitine concentration and
Soufir et al. (Soufir et al., 1984) demonstrated a reduced
concentration of free carnitine in the seminal plasma of infertile
men. Jeyendran et al. (Jeyendran et al., 1989) compared 18
different biochemical components in the seminal plasma of
men after successful in-vitro fertilization and the seminal
plasma of infertile men after therapy failure. The seminal
plasma components did not reveal any significant differences
between the groups, except for reduced glycerylphosphorylcholine. In another publication, higher concentrations of free
carnitine, total α-glucosidase and citrate in infertile men with
normozoospermia were described than for the values measured
in fertile men (Cooper et al., 1991).
In this study, clearly reduced (P ⬍ 0.001) concentrations of
free carnitine (212 versus 521 µmol/l) and total carnitine
843
A.Zöpfgen et al.
(437 versus 743 µmol/l) were demonstrated in infertile men,
compared with controls. When only infertile men with normozoospermia were compared with the control group, those
markers were still significantly reduced (P ⬍ 0.001). The
activity of neutral α-glucosidase was lower in the infertile
men than in the controls (15.1 versus 23.4 IU/l, P ⬍ 0.05).
The reduction of neutral α-glucosidase in infertile men is to
be attributed to the large proportion of patients with reduced
sperm count and/or azoospermia in this group. Consequently,
the comparison exclusively of patients with normozoospermia
with controls showed no difference in the activity of the
enzyme.
An association between a large number of leucocytes and
limited male fertility has been described previously (Caldamone et al., 1980; Talbert et al., 1987). The present study
revealed a slightly significant (P ⫽ 0.034) increase of the
PMN elastase concentration in the presence of male sterility
factor. Significantly increased values of PMN elastase (P ⫽
0.015) were measured in patients with fertility problems and
normozoospermia in comparison with the controls (48 versus
144 µg/l).
In eight of 40 infertile patients PMN elastase concentrations
were ⬎1000 µg/l, while in the control group no value ⬎400
µg/l was measured. The results suggest that particularly in
infertile men with normozoospermia, the reduced fertility may
be due to clinically quiescent infections of the male genital
tract. In these cases, antibiotic therapy would be indicated,
with measurement of the PMN elastase concentration in the
follow-up (Reinhardt et al., 1997).
The spectrum of clinical pictures involving reduced quantities of fructose and fertility problems ranges from androgendependent insufficiency of the seminal vesicles over general
diseases, drug effects, infections of the male reproductive tract
to pathological anatomical damage of the spermatic ducts
(Schill, 1976). No differences in concentrations of fructose
and citrate in infertile patients and controls were observed,
nor were there any positive effects of these parameters on the
semen quality. The results concur with the values measured
by Lewis-Jones et al. (1996), who found no correlation between
fructose and semen quality. The value of the markers of the
accessory sexual glands has thus to be questioned and can be
considered obsolete. Not to be confused with this is the use
of fructose measurement in seminal plasma in individual cases,
for instance for diagnosing dysfunction or hypoplasia of the
seminal vesicles (Aumüller and Riva, 1992).
The importance of the determination of neutral α-glucosidase
as a sensitive and non-invasive method for diagnosing an
obstruction in patients with azoospermia was demonstrated by
a number of examiners (Cooper et al., 1990; Garcia Diez
et al., 1992; Mahmoud et al., 1998). The results of the current
investigations were similar to theirs. Of the biochemical
markers, only the activity of neutral α-glucosidase was significantly reduced in patients with azoospermia compared with
patients with testicular azoospermia (P ⫽ 0.002), the sensitivity
and specificity of measuring in the current study being comparable to that of Mahmoud et al. (Mahmoud et al., 1998).
The measurement of glucosidase proved to be useful for the
differential diagnosis of azoospermia, whereas the determina844
tion of carnitine (free carnitine, total carnitine, acetylcarnitine)
is unsuitable for differentiating between a testicular and an
obstructive azoospermia.
Before starting medical reproductive measures, the reasons
for male infertility must be precisely diagnosed. The measuring
of PMN elastase concentrations (infection) and of the activity
of neutral α-glucosidase (occlusion, testicular) in seminal
plasma gives additional information.
From the results of this study, it is concluded that only αglucosidase hints at an obstruction of the seminal ducts and
that concealed inflammation of the male reproductive tract is
often associated with male infertility. Therefore, in excluding
an infection of the seminal ducts in subfertile men, measuring
the PMN elastase concentration is useful only in addition to
the spermiogram according to WHO criteria; in cases of
azoospermia of unknown origin the concentration of α-glucosidase should be measured additionally.
Acknowledgements
This work was supported in part by grants from the Fund of the
German Chemical Industry (to K.J.; No. 400770). The study contains
part of the doctoral thesis of A.Z.
References
Aumüller, G. and Riva, A. (1992) Morphology and function of the human
seminal vesicle. Andrologia, 24, 183–196.
Caldamone, A.A., Emilson, L.B.V., Al-Juburi, A. et al. (1980) Prostatitis:
prostatic secretory dysfunction affecting fertility. Fertil. Steril., 34, 602–603.
Cooper, T.G., Yeung, C.-H., Nashan, D. and Nieschlag, E. (1988) Epididymal
markers in human infertility J. Androl., 9, 91–101
Cooper, T.G., Yeung, C-H., Nashan, D. et al. (1990) Improvement in the
assessment of human epididymal function by the use of inhibitors in the
assay of alpha-glucosidase in seminal plasma. Int. J. Androl., 13, 297–305.
Cooper, T.G., Jockenhövel, F. and Nieschlag, E. (1991) Variations in semen
parameters from fathers. Hum. Reprod., 6, 859–866.
Garcia Diez, L.C., Esteban Ruiz, P.E., Villar, E. et al. (1992) Enzyme and
hormonal markers in the differential diagnosis of human azoospermia. Arch.
Androl., 28, 181–194.
Jeyendran, R.S., van der Ven, H.H., Rosecrans, R. et al. (1989) Chemical
constituents of human seminal plasma: relationship to fertility. Andrologia,
21, 423–428.
Jochum, M., Pabst, W. and Schill, W.-B. (1986) Granulocyte elastase as a
sensitive diagnostic parameter of silent male genital tract inflammation.
Andrologia, 18, 413–419.
Kairisto, V. and Poola, A. (1995) Software for illustrative presentation of
basic clinical characteristics of laboratory tests – GraphROC for Windows.
Scand. J. Clin. Lab. Invest., 55, 43–60.
Kohengkul, S., Tanphaichitr, V., Muangnum, V. et al. (1977) Levels of Lcarnitine and L-O-acetylcarnitine in normal and infertile human semen: A
lower level of L-O-acetylcarnitine in infertile semen. Fertil. Steril., 28,
1333–1337.
Kunst, A., Draeger, B. and Ziegenhorn, J. (1984) D-Glucose/D-fructose. In
Bergmeyer H.U. (ed.), Methods of Enzymatic Analysis. Verlag Chemie,
Weinheim, 3rd edn, vol. VII, pp. 321–327.
Lewis-Jones, D.I., Aird, I.A., Biljan, M.M. et al. (1996) Effects of sperm
activity on zinc and fructose concentration in seminal plasma. Hum. Reprod.,
11, 2465–2467.
Ludwig, L., Frick, J. and Rovan, E. (1996) Praxis der Spermatologie. Springer,
Berlin, pp. 42–43.
Mahmoud, M., Geslevich, J., Kint, J. et al. (1998) Seminal plasma alphaglucosidase activity and male infertility. Hum. Reprod., 13, 591–595.
McGarry, D.J. and Foster, D.W. (1976) An improved and simplified
radioisotopic assay for the determination of free and esterified carnitine.
J. Lipid Res., 17, 277–281.
Möllering, H. (1985) Citrate. In Bergmeyer, H.U. (ed.), Methods of Enzymatic
Analysis. Verlag Chemie, Weinheim, 3rd edn, vol. VII, pp. 2–12.
Seminal plasma components related to semen quality
Reinhardt, A., Haidl, G. and Schill, W.-B. (1997) Granulocyte elastase
indicates male genital tract inflammation and appropriate anti-inflammatory
treatment. Andrologia, 24, 187–192.
Schill, W.B. (1976) Fructosebestimmung im Spermaplasma. Med. Klin., 24,
1031–1041.
Soufir, J.C., Doucot, B., Marson, J. et al. (1984) Levels of seminal free
carnitine in fertile and infertile man. Int. J. Androl., 7, 188–197.
Talbert, L.M., Hammond, M.G., Halme, J. et al. (1987) Semen parameters
and fertilization of human oocytes in vitro: a multivariable analysis. Fertil.
Steril., 48, 270–273.
Wetterauer, U. (1986) Recommended biochemical parameters for routine
semen analysis. Urol. Res., 14, 241–246.
WHO (1993) In Rowe, P.-J., Comhaire, F.H., Hargreave, T.B. and Mellows,
H.J. (eds), Manual for the Standardized Investigation and Diagnosis of the
Infertile Couple. Cambridge University Press, Cambridge.
Wolff, H., Bezold, G., Zebhauser, M. et al. (1991) Impact of clinically silent
inflammation on male genital tract organs as reflected by biochemical
markers in semen. J. Androl., 12, 331–334.
Received on August 4, 1999; accepted on December 6, 1999
845