ANDROLOGIA 23,42 1-425 ( 1991 ) r- ACCEPTED: SEPTEMBER 25, 1991

Adenosine triphosphate (ATP) in human spermatozoa

C. Gottlieb, K. Svanborg and M. Bygdeman

Key words. Spermatozoa - semen - motility - ATP - human.

Summary. The seminal adenosine triphosphate (ATP) content was determined by biolumin- escence after treatment with trichloroacetic acid (TCA) in 81 semen samples 1.5 h after ejaculation obtained from men attending our fertility clinic, and selected to contain either 207" or less sperma- tozoa with good progressive motility ( n = 2 2 ) , or 60% or more spermatozoa with good progressive motility (n=59) (Study I), and in 18 semen samples from fertile men 30 min and 3.5 h after ejaculation (Study 11). The latter samples were divided into 2 equally large groups according to sperm motility. In Study I the mean sperm ATP concentration was significantly higher in the semen samples with bad motility (0.63 nmol per living spermatozoa x lo-') than in semen samples with good motility (0.39 nmol per living spermatozoa x lo-'; P<O.Ol). In Study I1 the ATP concentration per living spermatozoa was also lower in the group with the best motility in comparison with the spermatozoa with lower motility (P<O.Ol), both 30 min and 3.5 h after ejaculation. During the 3-5 h incubation the sperm ATP concentration decreased by 21 7" (P<O.Ol) in the former group of samples but remained unchanged in the latter group. The results indicate that, in semen samples with highly motile spermatozoa, the consumption of ATP is higher than in semen samples with impaired sperm motility. I t is therefore essential that the time between ejaculation and ATP measurement is as short as possible to obtain comparable results. Repeated ATP measurements in combi- nation with an analysis of the number of living spermatozoa, may provide further information on the fertilizing capacity of spermatozoa.

Introduction

Adenosine triphosphate (ATP), mainly generated by glycolysis (Peterson & Freund 1970), is the immediate source of energy for the movements of the spermatozoa (Satir, 1974; Suter et al., 1979). Sperm motility is the characteristic within the standard semen analysis which correlates best with the fertilizing potency of the ejacdate (Tal- bert et al., 1987) and measurement of ATP has been considered a valuable tool in assessing this capacity (Comhaire et al., 1983; Irvine & Aitken, 1985; Chan & Wang, 1987). I t is therefore disap- pointing that in several studies concerning the relation between seminal ATP and different sperm variables the results have been contradic- tory (Levin et al., 1981; Calamera et al., 1982; Comhaire et al., 1983; Singer et al., 1983; Schlegel et al., 1985; Pousette et al., 1986; Mieusset et al., 1989). A direct correlation between sperm motility and seminal ATP content was, for instance, reported by Chan and Wang (1987), but could not be confirmed by Pousette et al. ( 1986).

To minimize methodological errors, as a poss- ible reason for the variations in the results obtained, a modified technique for the measure- ment of seminal ATP has been developed and thoroughly evaluated (Gottlieb et al., 1987). The aim of this study was to use this technique to re- evaluate the possible relation between the sperm ATP content and sperm movement and the effect of storage on seminal A'TP content in relation to the number of living spermatozoa.

Material and methods

Department of Obstetrics and Gynecology, Karolinska Hos- pital, Stockholm, Sweden

Correspondence: Claes Gottlieb, Department of Obstetrics and Gynecology, Karolinska Hospital, S-104 01 Stockholm, Sweden.

Semen samples were obtained by masturbation in our laboratory. Sperm count, -viability, -mor- phology, and -motility were assessed microscopi- cally by one trained technician using the criteria proposed by WHO (Belsey et al., 1980). Sperm

422 C. GOTTLIEB ET AL.

velocity was determined by using an objective video micrography technique, as described else- where (Gottlieb et al., 1988). I n short, 10 p of the semen sample were introduced into a Makler chamber (Makler, 1980) and sperm movements were recorded on a video tape. The mean pro- gressive sperm velocity was calculated from the time needed for 50 spermatozoa, randomly chosen from a still video picture, to traverse a 100 mm' square as judged from a slow motion playback. Simultaneously, the proportion of progressively motile spermatozoa was determined.

The ATP content in the spermatozoa was measured by a bioluminescence assay after extrac- tion with TCA (Gottlieb et al., 1987). All the measurements were performed in triplicate and a pilot experiment was made prior to each assay to ensure that the total ATP content (with or without dilution) in the sample (including addition of the internal ATP standard) ranged within the straight part of the standard curve. The coefficient of variation for the ATP assays was less than 3%.

A number of semen samples from the male partners in couples attending our fertility clinic were analysed for sperm motility, -viability, -count, -morphology, and -ATP content I h 30 min after ejaculation. Among these, samples were selected for inclusion in the present study (Study I ) , which had 20% or less spermatozoa (Group A; n = 22), or 60% or more spermatozoa (Group B; n = 59) with good progressive motility (Table 1). The correlation and regression between the ATP content and the number of living spermatozoa in eac,h group and the differ- ence in ATP content per living spermatozoa x (ATP:LS) between the two groups were calculated.

Semen samples from 18 healthy volunteers were analysed 30 min and 3.5 h after ejaculation (Study 11). The semen samples were divided into two equally large groups according to the initial progressive sperm velocity. The semen samples were analysed as above.

The ATP content was analysed in 10 semen samples before and after centrifugation. Immedi- ately after liquefaction the semen samples were

Table 1. Mean + SD.

Semen variables in 81 semen samples (study I);

Group A Group B

Sperm concentration 53.2+ 15.3 59+ 14.2 Motile sperm, yo 50k12.1 59.2k11.1 Morphologically normal, yo 42 f 14 48+ 13 Living sperm, yo 6 0 k 13 6 2 k 14

divided into two aliquots. One aliquot from each sample was centrifuged at 500 g for 10 min and the upper half of the supernatant was immedi- ately removed. The other aliquot was not pro- cessed. The ATP content in the non-centrifuged samples and in the supernatants after centrifug- ation was analysed.

The statistical calculations were performed using the Wilcoxon rank signed test and the Mann Whitney U-test (Mann & Whitney, 1947).

Results

study I In the 81 semen samples selected, the correlation between ATPml- ' semen and the total number of spermatozoa ml-' semen was low (r=0.25). The coefficient of correlation was increased when ATPml- ' was correlated with the number of living spermatozoa ml-' ( r = 0.51). When the two groups of semen samples were compared (good and bad motility; see Material and methods) the coefficients of correlation between ATP ml- ' and the number of living spermatozoa were still higher (Group A; r=0.75; P<O.OOl; n=22, Group B; r=0.85; P < O . O O l ; n=59). In addition the coefficient of regression differed in the two groups (bz0.52 in Group A and 6=0.37 in Group B). The mean ATP concentration per lo6 living spermatozoa was lower in semen samples with good motility (Group B), 0.39 nmol LS-' (range 0.26-0.78), than in those with poor motility (Group A), 0.63 nmol LS-' (range 0.27-1.42). The difference was significant (P < 0.01). The regression line differed significantly from zero but the coefficient of correlation was only r=0.52. Although the two groups of semen samples differed, no direct correlation between the per- centage of good progressively motile spermatozoa in the individual samples and the ATP content per living spermatozon was found. Sperm mor- phology and the proportion of defective or abnor- mal sperm tails were not correlated to the ATP content in these semen samples.

study 11 The regression coefficient between the ATP con- tent perm1 semen and the number of living spermatozoa per ml semen 30 min after ejacu- lation was b = 0.40 ( r = 0.93; n = 9) in the group of samples with the best velocity (see Material and methods) and b=O.60 (r=0.81; n=9) in the group with lower velocity. The corresponding figures 3.5 h after ejaculation were b=0.55 ( r = 0.80) and b=0.31 (r=0.83). The mean ATP:LS

ANDROLOGIA 23,421-425 (1991)

ATP IN HUMAN SPERMATOZOA 423

ratio was higher in the semen samples from the second group than in those with the best motility on both occasions (P<O.OI; Table 2). On the other hand, there was no correlation between the sperm motility in each sample and ATP:LS 30 min or 3.5 h after ejaculation ( r = -0.40 and r = - 0.12, respectively). During the 3 h storage of semen at room temperature the proportion of living, motile, and progressively motile spermato- zoa decreased but the average mean sperm veloc- ity of the latter spermatozoa increased (Table 3). The ATP content in nmol decreased both when calculated per ml semen and per living spermato- zoa x lop6 (ATP:LS). In the semen samples containing spermatozoa with a good motility the mean ATP:LS decreased by 21% (P<O.OI) dur- ing the 3 h storage and in those with an initially less good motility the decrease in ATP:LS was not so marked (1 1 yo) and was not significant.

ATP could not be detected in the supernatants from 10 centrifuged samples. In the correspond-

ing non-centrifuged aliquots of whole semen the ATP content was at a mean 32.3 nmol ml-' (4.2- 82.0 nmol ml-I).

Discussion

In the semen samples which were handled accord- ing to the routines a t our laboratory (Study I) the ATP content per living spermatozoa was higher when sperm motility was impaired than in samples with normal sperm motility. The angle of the regression line between ATP perm1 and the number of living spermatozoa per ml was accordingly greater in the former group of samples than in the latter one. This finding was not expected since previously reported investi- gations have claimed that semen samples with normal sperm motility contain more ATP than do samples with impaired sperm motility (Levin et al., 1981; Orlando et ad., 1982; Comhaire et al.,

Table 2. ATP content (nmol/living spermatozoa x mean and (range)".

Sperm velocity

Time after ejaculation Number of Change (70) samples 3 0 3h30

2 25 pm sec- 9

< 25 pm sec- 9

0.47 -P<O.Ol- 0.36 21 (0.36-0.68) (0.25-0.54)

P<O.Ol P<O.Ol

(0.46-1.65) (0.42-1.33) 0.81 -NS- 0.70 11

NS not significant "Study I

I Table 3. Semen variables 3 0 and 3h30 after ejaculation in healthy volunteers; mean and (range); n= 18.

Semen variable Time after ejaculation

30' 3h30 Significance of differenceb

Sperm concentration, 106 ml-I 91.2 91.2 (21-198) (21-192)

Living spermatozoa, yo 64 52 <0.01 (4 1 -79) (36-76)

Motile spermatozoa, yo 57 49 < 0.05

Progressively motile 52 40 < 0.05 (40-70) (30-65)

spermatozoa, yo" (30-90) (20-80)

velocity, mm s - ' (16.7-30.6) (18.3-31.9) Mean progressive sperm 21.7 24.4 <0.01

ATP, nmol ml- I 36.0 26.2 CO.01 (4.6-96.0) (3.3-59.0)

ATP, nmol/living 0.65 0.53 <0.05 spermatozoa x (0.36-1.65) (0.25-1.29)

I "Progressively motile spermatozoa x 100~'/motile spermatozoa. bWilcoxon paired rank signed test.

ANDROLOGIA 23,421-425 (1991)

424 C. GOTTLIEB ET AL.

1983; Chan & Wang, 1987), or a lack of corre- lation between these two parameters (Levin et al., 1981; Caldini et al., 1982; Irvine & Aitken, 1985; Pousette e l al., 1986; Bilgeri et al., 1987; Megory et al., 1987). Only Schlegel e l al. (1985) reported an inverse correlation between semen and ATP and sperm motility. However, all these authors calculated the ATP content per ml semen or per spermatozoon without taking into consideration the viability of the spermatozoa. Another reason for the divergence in results may be that the importance of the duration between ejaculation and ATP analysis was not considered.

A prospective study was initiated (Study 11) in which semen samples were analysed both 30 min and 3.5 h following ejaculation. The results confirmed those from Study I. The mean ATP content per living spermatozoa was lower in the 9 samples with the best sperm motility than in the remaining samples with lower sperm motility, and the slope of the regression line between ATP ml-' and the number of living spermatozoa ml-' had a greater angle in the samples in the second group than in samples in the first group on both occasions.

During the 3 h storage of semen there was an increase in the average progressive sperm velocity in both groups. Similar results have previously been reported by Makler et al. (1979). At the same time, the proportion of living, motile, and progressively motile spermatozoa decreased (Table 3) . The increase in the average progressive velocity may therefore be due, at least in part, to a more pronounced decline in the number of progressively motile spermatozoa with a lower velocity in comparison to the number of sperma- tozoa having a higher velocity.

The total ATP content in semen decreased during the 3 h period of storage, in accordance with previously reported results (Caldini et al., 1982; Singer et al., 1983; Pousette et al., 1986; Gottlieb et al., 1987; Dolci et al., 1988). Since ATP is present only in living spermatozoa as indicated by the present results and as reported previously by Gottlieb et al. (1987), a decrease in the ATP content concomitantly with the known reduction in number of living spermatozoa with time is to be expected. However, the ATP content expressed as ATP:LS also decreased during stor- age. In addition, semen samples with the best motility contained less ATP:LS than did those with a lower motility both 30 min and 3.5 h after ejaculation. The decrease during storage was more pronounced in the samples containing sper- matozoa with an initially good sperm motility than in those with an initially lower sperm motility. The ATP content in semen immediately

after the ejaculation could not be measured since the liquefaction process must have occurred before analysis. Because of the difference in the rate of decrease in ATP:LS between the two groups of samples, the difference in ATP concen- tration per living spermatozoa may have been less marked at ejaculation than that found 30 min after ejaculation.

The reported findings would indicate that the consumption of ATP was higher than the new synthesis in samples with highly motile spermato- zoa, whereas the difference between ATP syn- thesis and consumption was less marked in samples with less motile spermatozoa. The results also indicate that the interval between ejaculation and ATP measurement must be taken into account when examining sperm ATP content. It is preferable that ATP is measured as soon as possible after ejaculation.

The regulation of sperm motility is multifactor- ial (Mann & Lutwak-Mann, 1982). The avail- ability of ATP and its turnover is only one such factor (Suter et al., 1979). A linear regression between sperm movement and ATP:LS can not therefore be expected. This was confirmed in the present study in which the range in the calculated ATP:LS was wide. Therefore the ATP:LS ratio in individual samples could not be used to predict the degree or quality of sperm motility. Only when the semen samples were grouped according to sperm motility was a difference found in the mean ATP content per living spermatozoa between the groups.

Thus it may be concluded that a single measurement of the ATP content per living sper- matozoa is not directly related to sperm motility. However, a significant decrease in ATP concen- tration in relation to the number of spermatozoa that have survived after a couple of hours indi- cates good sperm movement and may add infor- mation to the routinely used subjective motility assessment. Assessment of the sperm capacity to fertilize the human ovum is the ultimate goal of semen analysis. Sperm motility (Talbert et al., 1987) and the outcome of the hamster oozyte penetration test are indicators of sperm fertilizing ability in man. However, the relation between sperm ATP content and the outcome of the hamster penetration test is disputed (Comhaire et al., 1983; Irvine & Aitken, 1985; Chan & Wang, 1987) and a correlation to the outcome in the IVF program could not be shown (Mieusset et al., 1989). The question of whether ATP deter- minations may provide additional information on sperm function besides that of sperm motility still remains to be answered.

ANDROLOGIA 23,421-425 (1991)

ATP IN HUMAN SPERMATOZOA 425

Acknowledgements

This study was supported by the Swedish Medical Research Council (Project No. 17X-05696). We are grateful to Astrid Haggblad, Solveig Johans- son and Eva Lilliehook for their technical assist- ance and to Zoe Walsh, MD, for revision of the English language.

References

Belsey MA, Eliasson R , Gallegos AS, Moghissi KS, Paulsen CA, Prasad MNR (1980) Laboratory manual for the examination of human semen and semen-cervical mucus interaction. Singapore Concern Press, Singapore

Bilgeri YR, Winckelmann A, Berzin M, Lyons G, Reinach SG (1987) Adenosine triphosphate levels in human sperm- atozoa. Arch Androl 18:183-188

Calamera JC, Brugo S, Vilar 0 (1982) Relationship between motility and adenosine triphosphate (ATP) in human spermatozoa. andrologia 14:239-241

Caldini AI, Orlando C, Fiorelli G, Cuomo S, Serio M (1982) ATP and ADP content of human ejaculated sperm- atozoa. 11: Time-dependent changes after ejaculation. Int J Androl 5:579-585

Chan S, Wang C (1987) Correlation between semen adenos- ine triphosphate and sperm fertilizing capacity. Fertil Steril 47:717-719

Comhaire F, Vermeulen L, Ghedira K, Mas J, Irvine S, Callipolitis G (1983) Adenosine triphosphate in human semen: Quantitative estimate of fertilizing potential. Fertil Steril 40:500-504

Dolci S, Manna EA, Jannini E, Piccione E, Pasetto N (1988) ATP-content and kinetics of acrosome reaction in human spermatozoa: Influence of various culture media and incubation time. andrologia 20: 169-1 72

Gottlieb C, Svanborg K, Eneroth P, Bygdeman M (1987) Adenosine triphosphate in human semen-A study on conditions for a bioluminescence assay. Fertil Steril

Gottlieb C, Svanborg K, Eneroth P, Bygdeman M (1988) Effects of prostaglandins on human sperm function in vitro and seminal ATP content. Fertil Steril 49:322-327

Irvine DS, Aitken RJ (1985) The value of adenosine tri- phosphate (ATP) measurements in assessing the fertilizing ability of human spermatozoa. Fertil Steril 44:806-8 13

Levin RM, Shofer J, Wein AJ, Greenberg SH (1981) ATP

47~992-999

concentration of human spermatozoa: Lack of correlation with sperm motility. andrologia 13:468-472

Makler A ( 1980) The improved ten-micrometer chamber for rapid sperm count and motility evaluation. Fertil Steril 33:337-338

Makler A, Zaidise I, Paldi E, Brandes J M (1979) Factors affecting sperm motility. I. In vitro change in motility with time after ejaculation. Fertil Steril 3 1 : 147- 154

Mann T, Lutwak-Mann C (1982) Biochemistry ofspermato- zoa: chemical and functional correlations in ejaculated semen: andrological aspects. In: Male Reproductive Func- tion and Semen, Mann T, Lutwak-Mann C (eds). Springer, Berlin pp 196

Mann HB, Whitney DR (1947) O n a test of whether one of two random variables is stochastically larger than the other. Ann Math Stat 28:50

Megory Em Shoham (Schwartz) Z, Madgar I, Lunenfeld B, Modan M, Weissenberg R (1987) ATP content in human semen and sperm quality. Arch Androl 19:243-247

Mieusset R, Parinaud J, Chap H, Pontonnier G (1989) Absence of correlation between the levels of ATP and other seminal compounds in semen and the results of human i n - v i h fertilization. Int J Androl 12:346-353

Orlando C, Caldini AL, Fiorelli G, Cuomo S, Seriou M (1982) ATP and ADP content of human ejaculated sperm- atozoa. I. Relationship with semen physical parameters in normal donors and oligozoospermatic patients. Int J Androl 5:497-502

Petersson RN, Freund M ( 1970) ATP synthesis and oxidative metabolism in human spermatozoa. Biol Reprod 3:47-54

Pousette A, Akerlof E, Lundin A, Rosenborg L, Fredricsson B (1986) O n the use of adenosine triphosphate for esti- mation of motility and energy status in human spermato- zoa. Int J Andrology 9:331-340

Satir P (1974) How cilia move. Sci Am 231:44-52 Schlegel W, Kusseler R, Nieschlag E (1985) Significance of

adenosine triphosphate in human sperm as clinical param- eter. Arch Androl 14: 17 1 - 176

Singer R, Barnet M, Sagiv M, Allalouf D, Landau B, Servadio C (1983) ATP contents of human semen, seminal plasma and isolated sperm at time intervals after ejacu- lation. Int J Biochem 15:105-109

Suter I, Chow PYW, Martin ICA (1979) Maintenance of motility in human spermatozoa by energy derived through oxidative phosphorylation and addition of albumin. Biol Reprod 20:505-510

Talbert LM, Hammond MG, Halme J , O’Rand M, Fryer JG, Ekstrom R D (1987) Semen parameters and fertilizing of human oocytes in vitro: A multivariable analysis. Fertil Steril 48:270-273

ANDROLOGIA 23,421-425 (1991)