Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
7 Biologic Markers of Human Male Reproductive Health and Physiologic Damage NEEDS FOR BIOLOGIC MARKERS OF HUMAN MALE REPRODUCTIVE HEALTH The uncertainties associated with risk extrapolation from animal data argue strongly for the development of validated and sensitive methods for measuring ger- minal and reproductive damage directly in people. In males, reproductive damage can be divided into two broad types: patho- physiologic and genetic. This chapter reviews the markers of reproductive health suitable for detecting physiologic dam- age. Physiologic damage may reduce the chance of successfully fertilizing an egg. Human markers of germinal genetic toxicity and heritable mutations are reviewed in Chapter 9. Currently available methods used in the evaluation of the reproductive health of human males are in three broad classes: · Personal history. · Physical examination. · Laboratory analyses. The roles of personal history and physical examination in fertility assessment was described earlier (Chapter 3~. Laboratory analyses include testicular biopsy, hormonal analyses, and semen analyses. Testicular biopsy and hormonal 83 analyses were described earlier with em- phasis on animal studies (Chapters 4- 6~. Table 7-1 lists the categories of biolog- ic markers of physiologic damage to male reproduction discussed in this chapter. These markers are grouped by the source of tissue and data required: testicular tissue, semen, blood, surveys and medical records, and maternal urine. They differ in the numbers and kinds of assays avail- able to measure them, in the degree of quan- titation attainable so far, in the extent to which underlying mechanisms are under- stood, and in their feasibility for human studies. The development and validation of markers of human male reproductive health require a multidisciplinary ap- proach that includes basic research in animal and human reproductive biology, engineering and statistical development of automated and quantitative procedures, clinical studies of human factors that affect variation and of the predictive value of individual markers, and epidemio- logic studies of populations exposed to xenobiotic agents. This chapter begins with a short review of the kinds of epidemiologic studies that have been performed to evaluate fertility effects of human males exposed to reproduc- tive toxicants. It then discusses in de- tail markers currently available as well
84 "4LE REPRODUCTIVE TOXICOLOGY TABLE 7-1 Biologic Markers of Physiologic Damage to Human Male Reproductions Tissue or Data Required Markers of Testis (or biopsy)b Seminal sperm Other seminal parameters Bloodb Survey and medical records Maternal urine Histopathology Sperm number StructureC MotilityC Double F bodies Viability Agglutination Penetration and egg interaction: cervical mucus hamster eggs nonl~g human eggs Internal and surface domains Chromatin structure Physical characteristics Immature germ cells Non-germ cells Chemical composition: normal and xenobiotic constituents Sertoli cell, I~ydig cell, and accessory gland function Hormone levels Fertility status: standardized fertility ratio time to conception Indicators of early pregnancy - ~Available markers that have been used to detect effects of ioniz~ng radiation or chemicals in exposed men and markers on which human baseline data are available or under investigation. See Table 7-2 for further details. bDiscussed In Chapter 4. CAutomated methods are under development. as the status of new areas of research that may lead to new, useful biologic markers. Several features of human semen are evalu- ated including some physical characteris- tics of the ejaculate, presence of nonsperm cells, sperm number, sperm motility, sperm viability, sperm structure, and various aspects of sperm function and penetration. In addition, the chemical composition of semen has been evaluated to a limited extent as an indicator of reproductive function and toxicity. EPIDEMIOLOGIC STUDIES OF HUMAN SPERM PRODUCTION AND FERTILITY Interview data, medical records, and demographic birth records, as well as semen and blood hormone analyses have been used to evaluate the effects of radiation or chemicals on human fertility and spermato- genesis. Although surveys are not gener- ally considered markers of reproductive health, they are a means of evaluating the fertility status of well-characterized groups and thus provide important bench- marks in the development and validation of markers related to fertility. Epidemiologic Surveys of Fertility Status Interviews and the use of medical rec- ords are indirect approaches to assess- ing fertility; epidemiologically, they measure the lack of an event (the event being the birth of a healthy child). Reduc- tion in fertility is measured by comparing birth rates or intervals between births or pregnancies to assess couples' ability to procreate. The male component is evalu- ated by comparing the average fertility of a reference group with the fertility of a group of couples in which all males
MARKERS OF REPRODUCTIVE HEALTH have a common exposure to an environmental or occupational agent. Wong, Levine, and coworkers (Won" et al., 1979; Levine, 1983) developed the standardized fertility ratio (SFR) meth- od, which compares the numbers of births observed with the numbers expected on the basis of the number of person-years of observation, considering such factors as age, race, marital status, and parity. Statistically and epidemiologically, the SFR has advantages and limitations in interpretation (Tsai and Wen, 1986~. The SFR method was used to evaluate several male occupational exposures, including exposures to ethylene dibromide and DBCP (Levine et al., 1980; 1983~. The interval between births or pregnan- cies is a second indirect measure of fer- tility (Baird et al., 1986~. Several fac- tors affect this measure, including use of contraceptives and birth order. Birth order needs to be controlled, because the interval between births increases with increasing parity. Births within a family are not independent events, because of the presence of social, genetic, and health factors. Statistical methods that account for these factors must be used in analyses. The intervals between births and between pregnancies were increased in smoking mothers (Baird and Wilcox, 1985~. Further work and additional approaches are needed to improve the quantitation of survey- based indicators of human fertility status. Markers of Early Pregnancy Early pregnancy monitoring may provide a very sensitive method for measuring male- mediated effects on fertility and preg- nancy. It has been difficult to detect the early days of human pregnancy, and most women don't suspect that they are pregnant until a menstrual period is missed. How- ever, early pregnancy is a time of elevated frequencies of embryo loss, although neither the precise nor the portion that can be attributed the human male (i.e., via the sperm) are known. Sensitive meth- ods are being developed to detect early pregnancy; for example, the immunological detection of §-chorionic gonadotropin 85 in the maternal urine can detect pregnancy by about 10 days after conception (Canfield et al., 1987~. Methods such as these might be useful for evaluating male fertility and for identifying male factors that may affect the frequency of early embryo loss. Further research is encouraged to deter- mine human baseline variations for early pregnancy loss and of male factors that might alter these rates and to develop methods that detect even earlier pregnan- c~es. Cohort Studies with Semen and Blood Samples Sperm number, motility, and structure have been used to evaluate the effects of exposures to physical and chemical agents. Detrimental effects of increased scrotal temperatures and ionizing radiation and exposure to over 50 therapeutic, occupa- tional, and environmental chemicals have been identified (Wyrobek et al., 1983a). Cohort studies or case reports of exposure-related changes in sperm char- acteristics provide no direct information on fertility effects, because specific "sperm-characteristics versus fertility" relationships and within- laboratory standards for normal fertility are usually not presented in these studies. Primarily, two types of study designs have been used with semen analyses: cross- sectional and longitudinal. Cross-sec- tional studies require the analyses of single semen samples from exposed men and from men of at least one reference group (Wyrobek et al., 1982~. The refer- ence group can consist of unexposed men sampled concurrently or historical with- in-laboratory values from unexposed men. In rare cases, workers can be strati- fied by exposure to assess dose-effect relationships, as exemplified by Lancran- jan et al. (1975) for lead workers. They reported that proportions of men with olig- ospermia, asthenospermia, and teratosper- mia increased with increased peripheral blood lead concentrations. More typical- ly, exposure is difficult to quantify, and studies are limited to group compari- sons of exposed and unexposed men. The statistical evaluation of confounding
86 factors is crucial in cross-sectional studies; abstinence time, illness, smok- ing habits, and other chemical or drug exposures are common confounding factors. Among-person variations in sperm num- ber, motility, and structure scores can be used to estimate statistical power of studies and to calculate sample sizes required to detect changes with the cross- sectional design. For example, one labora- tory (Wyrobek et al., 1982) calculated that detection of a 20% change in mean be- tween the exposed and unexposed groups required analyses of semen from approxi- mately 800 men for sperm concentration (400 exposed and 400 controls) and 80 men for sperm structure. Considering realis- tic participation rates, more men would have to be identified in each group to achieve the required numbers of semen samples. The insensitivity of the cross-section- al design plagues human semen studies, because it is difficult and expensive to find, gain cooperation of, and evaluate large numbers of men. The large sample re- quirements also limit the kinds of factor- ies and agents that can be investigated. The design is nevertheless commonly used, probably because it permits rapid collec- tinn and analysis of semen samples (in comparison with the longitudinal design described below). Studies of occupational exposures to DBCP (Whorton et al., 1977; Whorton et al., 1979; Babich et al., 1981), carbaryl (Wyrobek et al., 1981), and wastewater treatment (Rosenberg et al., 1985) and others (Wyrobek et al., 1983a) have used this design. The longitudinal design, requires analysis of at least two sequential semen samples from each man (Sherins et al., 1977~. The samples should be separated by at least several months (and more than two repeat samples are preferred). Sample collection times are selected in relation to exposure. For example, the first sample might be collected before exposure begins, and the second about 6 months after it be- gins (this would allow for two to three spermatogenic durations between sam- pling). The advantage of this design is based on the observation that within-per- son variation of sperm concentration, motility, and structure is generally MALE REPRODUCTIVE TOXICOLOGY smaller than among-person variation. Thus, smaller numbers of men are required for detection of a specific exposure-in- duced change in group mean. For example, a longitudinal study to detect a 20% change in group mean would require approximately 80 men for sperm concentration, but only about 10 men for sperm structure (A. Wyro- bek, Lawrence Livermore National Labora- tory, unpublished data, 1989~. Further- more, this design does not, in principle, require an unexposed cohort for reference, although an unexposed cohort might be help- ful in controlling factors closely corre- lated with exposure. The longitudinal design with semen analyses has been used primarily to evaluate the effects of ex- posure to drugs, e.g., AMSA (da Cunha et al.. 19821. colchicine, various chemo- therapeutic agents, and antifertility drug candidates (see review by Wyrobek et al., 1983a). Further studies are needed to evaluate the relative utilities of lon- gitudinal and cross-sectional designs in exposed populations. Blood hormone levels have been evaluated as surrogate measures of spermatogenesis. The levels of follicle stimulating hor- mone (FSH) are elevated in men with very low or zero sperm numbers, as seen with DBCP and certain cancer chemotherapeutic com- pounds (see review bv Sever and Hessol. 1985~. However, this method is ~insen- sitive," and hormone measurements have not been used commonly in studies of environmental exposures. ~ , HUMAN SPERMATOGENESIS AND DEVELOPMENT OF SEMEN-BASED MARKERS OF MALE REPRODUCTIVE HEALTH Human sperm production is unique among animals in several ways: · In most animals (laboratory, as well as domestic), sperm are produced in great excess over what is needed for adequate fertility. Even in mice, among the smallest mammals, males produce some 10 times the sperm required for fertility. In comparison, the mean sperm concentra- tion man is a smaller proportion of levels associated with subfertility.
MARKERS OF REPRODUCTIVE HEALTH · Individual variation in semen quality is much greater in men than in laboratory and domestic animals. For example, approx- imately 10% of men within a normal fertile group might be at or below 20 million sperm per milliliter (MacLeod and Wang, 1979; Whorton and Meyer, 1984~. Such variation would be virtually nonexistent among fer- tile laboratory and domestic animals. · There is more variation among the types of cells and cell structure in the typical human ejaculate than in animals' ejaculate. · Humans are genetically heterogeneous, so responses (i.e., pharmacokinetics and metabolism) to particular chemicals can differ unpredictably. · The minimal number of sperm required for adequate and minimal fertility might depend on species. For humans, the number for adequate fertility is generally con- sidered to be 20 million sperm per milli- liter of ejaculate or more. Some reports suggest that the minimum can be any number greater than zero (Barfield et al., 1979; Clark and Sherins, 1986~. Little is known of how animals and humans compare on this point. · Human seminiferous epithelium is unique among mammals in several ways (Hel- lerand and Clermont, 1964; Clermont, 1972), including kinetics of stem cell renewal, capacity to regenerate after toxic insult, effects of chromosomal ab- normalities on spermatogenic differentia- tion, cellular association patterns, and the duration of differentiation. In addition to the above features of human spermatogenesis, semen is an unusual body fluid-a person's ability to produce it has no direct bearing on his health or longevity. Semen is also a complex fluid. Although sperm are considered its primary constitu- ent, it also contains numerous other ger- minal and somatic cell types, chemical constituents from germinal and supporting somatic tissues, hormones, and probably xenobiotic agents. Unlike most cells found in other body fluids, seminal sperm are not terminally differentiated in one major respect. That is, from the perspective of the human germ 87 line and its generational cycles, seminal sperm are in a more or less central location of each cycle, which begins with the ger- minal stem cells in an adult and ends with the germinal stem cells in offspring. Thus, in principle, sperm could provide two types of markers: those which are re- trospective, in the sense that they monitor differentiation events that occurred earlier in the testis and epididymis, and those which are predictive of the fertility status of the adult and of the genetic health of his offspring. The fertility status of an adult is judged by his ability to produce sperm that successfully fertil- ize a female egg; markers of fertility status are a major topic of this chapter. (Markers for assessing genetic damage in sperm and genetic health of offspring are reviewed in Chapter 9.) Both retrospective and predictive sperm markers are needed for evaluating the reproductive health of males. Semen analysis is a common component of fertility examinations (see Chapter 3~. It has had a very long history; sperm are thought to have been one of the first kinds of cells analyzed under the micro- scope by its inventors, Leeuwenhoek and his student Hamm. The biology of the human spermatozoa has been the topic of several comprehensive reviews (Zaneveld, 1979; Olson, 1982) Three applications of semen markers have been proposed: as markers of sperm production and function, as indicators of fertility status, and as indicators of exposure to a reproductive toxin. In this chapter, we distinguish among these three applications because we still lack the data to understand the quantita- tive relationship among them. With sperm count, for example, there remains uncer- tainty in how to interpret a 10% reduction in group mean value even though all would agree that a 100% reduction (i.e., no sperm in the ejaculate) assures infertility. The use of semen markers to discriminate the effects of environmental, therapeu- tic, and occupational exposures does not necessarily require that a marker be as- sociated with fertility status. As dis- cussed earlier in this chapter, semen from exposed and unexposed cohorts can be com-
88 pared with each other and with historical controls to identify and evaluate the ef- fects of hazardous agents. Markers of sperm production and function reflecting both testicular and accessory organ function are needed for understand- ing the molecular mechanisms of human sper- matogenesis and will provide candidate markers for fertility and toxicologic evaluations. No semen marker has emerged as a defini- tive indicator of male reproductive health. It is generally agreed that a bat- tery of sperm characteristics should be measured jointly to assess male reproduc- tive health, and repeat measurements are very useful. The underlying hypothesis is that male fertility is multifactorial, requiring the normal function of numerous molecular and physical aspects of sperm and seminal fluid (see, for example, Amann and Berndtson, 1986~. Table 7-2 assigns the biologic markers of male reproductive toxicology to the following groups: (1) markers that have been successfully used to detect the human male reproductive effects of expo- sure to radiation or chemical agents and (2) markers for which no toxic effects data are available but for which human baseline data have been obtained or are under inves- tigation. Also listed are new research concepts based on modern cellular, molecu- lar, and recombinant DNA techniques that promise the next generation of biologic markers of human male reproductive health. For all markers, baseline data in normal individuals usually are considered pre- requisite for detecting abnormal events and for determining the sample sizes re- quired to detect changes. Markers that already have been used to detect the ef- fects of toxic exposure are not necessarily further in their development than markers for which baseline data are under inves- tigation. Both categories of markers need continued research to improve their sen- sitivity, investigate underlying mechan- isms, determine their quantitative rela- tionship to changes in human fertility, and other aspects as detailed in the next sections. ABLE REPRODUCTIVE TOXICOLOGY PHYSICAL CHARACTERISTICS OF THE HUMAN EJACULATE Human semen coagulates shortly after ejaculation, and prostatic enzymes li- quefy it within about 30 minutes, although in some men this may take hours. The coagu- lation and increased semen viscosity are due to prostatic and seminal vesicle com- ponents added during ejaculation. The specific relationship between semen YiS- cosity and fertility remains unclear. The color of the ejaculate varies among shades of white, yellow, and gray. The white cloudy component is thought to be concentrated sperm, but the relevance of semen color to fertility remains obscure. The pH of normal semen is between 7 and 8 (low pH might be due to obstruction of the ejaculatory ducts, a rare disorder in men), but the relationship of subtle changes in semen pH to fertility also is unclear. No reports are available on the effects of exposure to xenobiotic agents on coagulation or color (Wyrobek et al., 1 983a). Recently, Welch et al. ( 1988) reported a very slight elevation in semen pH of shipyard painters exposed to ethyl- ene glycol ethers. The volume of human ejaculate is usually about 2 to 5 ml (Hargreave and Nilsson, 1983), although samples of less than 1 ml and greater than 6 ml are not uncommon in large surveys. Semen volume can be affect- ed by abstinence time and has been reported as increasing by 0.4 ml/day and reaching a plateau by about 5 days (Clark and Sher- ins, 1986~. Human semen volume is usually not affected by exposure to xenobiotics (Wyrobek et al., 1983~. However, metha- done treatment has been reported to lead to an apparent increase in sperm concen- tration by decreasing semen volume (Cicero etal., 1975~. PRESENCE OF NONSPERM CELLS IN SEMEN Four types of cells other than sperm can be found in human semen: microorgan- isms, white blood cells, duct cells, and immature germ cells (Amelar and Dubin, 1977; Belsey et al., 1980; Amann, 1981; Eliasson, 1981; Alexander, 1982~. White
AL4RKERS OF REPRODUCTIVE HEALTH blood cells and pathogens, such as bacter- ia, are indicative of reproductive tract infections, which are usually treated accordingly. Duct-lining cells and imma- ture germ cells are normal components of the human ejaculate. Immature cell types can be distinguished morphologically on smears with Papanicolaou~s stain or Har- ris-Shorr technique (Auroux et al., 1985~. However, immature germ cells are not com- monly scored as markers of reproductive health, because their identification is highly subjective and their relevance to fertility status is not established. Im- mature germ cells are an unexplored and promising area for future marker develop- ment. SPERM NUMBER Sperm number refers to the number of sperm in the ejaculate, expressed as total sperm or numbers per milliliter of semen. Objective measurements are easily made with a hemocytometer. Electronic scoring with a Coulter counter is an alternative to the use of a hemocytometer, but has reduced accuracy at low sperm concentra- tions (under 10 million per milliliter) (Gordon et al., 1965, 1967~. Also, debris can clog the measurement orifice, and other cells or fragments can produce electronic measurement artifacts. As Marker of Sperm Production The presence of sperm in semen is com- pelling evidence of active sperm produc- tion. Its absence can reflect the inac- tivity of the seminiferous epithelium or a post-testicular tubular obstruction that can be resolved by testicular biopsy and Lasography. Clinically, men without sperm in their ejaculates are termed azoo- spermic; men with some sperm, but fewer than 20 million per milliliter are general- ly termed oligospermic; and men with higher concentrations are termed normospermic. Men who cannot produce semen are termed aspermic. To allow a continuum of sperm production that can reach approximately 100 million per day, there are probably 4 to 6 million divisions of stem cells to form new stem 89 cells and committed spermatogonia each and every day. Sperm concentrations vary markedly among men; group-average values in fertile men are typically about 60- 100 million per milliliter; about 1% of men are azoospermic and 10% are oligo- spermic (MacLeod and Wang, 1979; Whorton and Meyer, 1984~. Sperm concentration also varies among ejaculates of a given person. Katz et al. (1981) evaluated the sources of variation in sperm concentra- tion and reported 73% of the total varia- tion to be due to "among-donor effects and only 27% to be due to Within-donor effects. A large source of the variability is due to the interval of sexual absti- nence. In mice, the number of mature epididymal sperm is correlated with the number of spermatogenic stem cells (Meistrich, 1982), but a similar relationship has not been established for human sperm. As Marker of Fertility Status There is little agreement about the spe- cific relationship between sperm number and fertility status, nor is there agree- ment on whether total number per ejaculate or sperm concentration (number per ml) are more relevant. Using group data from several infertility clinics, Meistrich and Brown (1983) evaluated the mathema- tical relationships between sperm con- centration and likelihood of infertility. On the basis of group averages, men with sperm counts below approximately 20 mil- lion per milliliter show an increased like- lihood of infertility. However, at higher sperm concentrations, there was no apparent correlation between concen- tration and fertility status. Very high sperm concentrations (over 200 million per milliliter) also might increase the likelihood of infertility (Niendorf, 1964~. The biologic basis for the in- creased likelihood of infertility at both extremes of sperm concentration is not well understood. (The reason that millions of sperm are required to fertilize a single human egg is unknown.) The relevance to individuals of rela- tionships based on group data is uncertain, for several reasons. First, total semen
9o volume and sperm number differ among people, irrespective of their fertility (MacLeod and Gold, 1951; Smith et al., 1977; Zukerman et al., 1977; Homonnai et al., 1980a). Second, sperm number fluc- tuates within each person; for example, a man with an initial sperm concentration of 100 million per milliliter could have a true mean sperm concentration, based on six subsequent samples, of 50-230 mil- lion per milliliter (Schwartz et al., 1979~. Third, the sperm number threshold for subfertility remains uncertain and might depend on the couple. Meistrich and Brown's (1983) analyses suggest that a person's fertility status is of a probabil- istic nature below the value of approxi- mately 20 million per milliliter; the lower the sperm count, the less likely a man would be to impregnate his partner. Similarly, the duration required to achieve fertili- zation might increase with lower sperm counts (Bostofte et al.. 1982a: Collins et al., 1983~. - However, the numerical relationship for individuals is unknown. Because of the large variation, some laboratories suggest that a person's true mean sperm concentration can be as- sessed only from repeated semen samples (three to six, or more) collected over periods of many months. In spite of these efforts, sperm concentration might have only minor utility as a marker of a person's fertility status, although it seems to be useful for comparing group effects. As Indicator of Exposure to Toxic Agents Germ cell killing is a common conse- quence of exposure to agents that reduce fertility or produce germinal mutations. The relationships among testicular ex- posure, time-course of sperm-concentra- tion changes, and time-course of fertility changes are understood best for ionizing radiation (Searle and Beechey, 1974; Oak- berg, 1975~. In irradiated mice, approxi- mately 10-fold reductions in sperm con- centration are required for a noticeable reduction in fertility, and the duration of the induced sterile period is dose- dependent (e.g., 8-10 Gy led to a 10- to 11-week-long sterile period) (Oakberg, 1975~. Human data relating dose to sperm MALE REPRODUCTIVE TOXICOLOGY concentration changes show that men might be more sensitive to germ cell killing than mice and that the effects are longer-lived; acute exposures as low as 15 reds induced 4-fold reductions, and 4-6 Gy led to azoo- spermia for 5 or more years (Rowley et al., 1974; Clifton and Bremner, 1983~. As sta- ted earlier, induced reductions in sperm concentration might be more critical in men than in most animals. In men, a 4- fold reduction in mean sperm concentration would lower it into the range of oligosper- mia; larger reductions are required to do that in most animals. Also, as stated earlier, sperm concentration varies greatly from person to person. Therefore, even small reductions would bring some men into the oligospermic range and could increase their likelihood of becoming infertile. In addition, as we see with DBCP, chemical exposure of a group of men can shift the sperm concentration distri- bution to lower values and increase the proportion of azoospermics. Of the indicators listed in Tables 7- 1 and 7-2, sperm number has been used most often in studies of human male reproductive toxicity; 87 of the 89 chemical exposures surveyed in the last major review on this topic (Wyrobek et al., 1983b) and all semen studies of male reproductive effects of occupational exposure surveyed included it. Exposure to any of 57 agents led to detrimental effects on human sperm produc- tion. However, very few of the studies used rigorous study designs, including power calculations and sample size estimates (see the discussion of epidemiologic studies above). Other Factors That Affect Sperm Number Several factors not related to chemical exposure decrease sperm concentration, including short abstinence period, some illnesses, some viral infections, in- creased scrotal temperature, and some genetic and chromosomal disorders. Ab- stinence period is one of the best-under- stood factors (Schwartz et al., 1979; Baker et al., 1981; Mortimer et al., 1982~. As abstinence period increased from 1 day to 1 week, sperm concentration increased
MARKERS OF REPRODUCTIVE HEALTH by 10-15 million per milliliter per day, and total sperm count increased by 50- 90 million per day (semen volume increases by 0.4 ml per day). The progressive in- crease in sperm concentration is thought to be due to accumulation of sperm within the epididymis and vas deferent, which are referred to as the extragonadal re- serves. The capacity of these reserves is considerable (Amann and Howards, 1980; Johnson et al., 1980a; Tyler et al., 1982~. Frequent (daily) ejaculation reduces sperm concentration to about 25% of normal; that suggests that the 4-fold higher sperm numbers seen after a week of abstin- ence are recruited from extragonadal re- serves. After 2 weeks of abstinence, there might be up to a 10-fold increase in sperm concentration, compared with the values after frequent ejaculation (Johnson, 1982). In addition, the means and location of semen collection may affect its quality. It is generally agreed that col- lection of the sample in the physician's office or clinic is superior to bringing the sample from home. On the other hand, the volume and total number of sperm in the sample might be influenced by the degree of sexual arousal that could be greater for samples collected at home, though there are few studies directly addressing this point. SPERM STRUCTURE Sperm structure is the study of sperm shape and size. Numerous systems have been developed for classifying human sperm into categories based on head, midpiece, and tail features. Scoring methods generally rely on observer judgment and visual cri- teria (e.g., Hotchkiss et al., 1938; Mac- Leod, 1964; Wyrobek et al., 1982; Eliasson, 1983~. Sperm with differing features are assigned to shape-abnormality classes, such as double, tapered, narrow, and ir- regular (the specific numbers and types of categories depend on the classification system used). Sperm structure has been used as a marker of sperm quality and fertility sta- tus and to monitor exposure to reproductive toxins in animals and humans. However, 91 its value in assessing sperm quality and fertility status remains controversial. In part, that is due to the lack of consis- tency among scoring methods and the associ- ated difficulties in making interlabora- tory comparisons (Freund, 1966~. The prob- lem is dramatized by a comparison of the distributions of the proportions of abnor- mal sperm among groups of men attending major fertility clinics in Europe (Fred- ricsson, 1979~; the mean proportions of abnormally shaped sperm ranged from ap- proximately 20% to 70% among clinics. Some laboratories have high within-laboratory reliability by relying on one highly trained scorer or only a few, by using de- cision-tree logic in assigning sperm to structural categories, or by using refer- ence slides for setting and maintaining quantitative standards. However, the problems associated with visually deter- mined sperm structure remain a major im- pediment to large-scale evaluation of the utility of sperm structure as a marker of male reproductive health, and further work is needed to automate these measures. As Marker of the Quality of Sperm Production Sperm structure is a measure of the qual- ity of sperm produced by the testis and of alterations occurring in the efferent ducts and accessory glands. The effects of genetic and environmental factors on sperm nuclear structure have been inves- tigated in detail in laboratory animals (see review by Wyrobek et al., 1983b). In inbred and hybrid lines of mice, adult males produce proportions and types of sperm-head shape abnormality that are characteristic of their genotype. The proportions of sperm with shape abnormali- ties in unexposed mice are determined by multiple genetic loci, including both Y-chromosomal and autosomal factors. Specific recessive mutations and chromo- somal abnormalities have also been associ- ated with abnormal sperm-head shapes, and the length of midpiece also depends on the strain of mouse studied. Abnormally shaped sperm are less likely to reach the oviduct and site of fertilization (Nestor and Handel, 1984~.
92 o ._ Cal o Q Ct Cal r o Ct Ct C> ·~0 o - o ·_ o Cal A> Ct a> ·_ - o ·_ m o U. Ct Cal m Cat o .O ~ O to ~ .s ~ no V ~ lo: .s Ce C;, D ~ °a ~ :: U) U) U. CO U) O O ._ ~ .= O ~ .5 o ~ =m U) Ct C) Ct C) C) Ct a o C' c: ~ , ' ~ ~ 9 ~ O I ~ ~ ~ a S ~ ~ ~, 4, ~ ~ ~ ~ ~ o ~ ~ ~ ~ ._ o ~ ~ ct ~ O ~ ~ O Ct c' C_ .D ~ _ C. .= ~ ~ ~ C ~ ~ O ~ 0 ~ C o g 4, .2? 3 ~ · - ,= 0 ~ ~ e ·_ p~ h_ E~ . O ~ _ . O ;> . ~ ~.= C~ ~ ~ O ~ ~a' ~ ~ ~ C~ 00 0 ·^ O ~ ~ DO O ~ ~ =~ =: Ce =.D ~ C C U, ~ ~: ~ ~ ~ =.5 ~ ~> ~ ~ _ o e .- ·~ ~ 2 ~ ~ C~ O ~
hE4RKERS OF REPRODUCTIVE HEALTH Human data on the genetic component con- trolling sperm-shape . ~ . . . . ~ . abnormalities are Incomplete, cut evade evidence is consistent with animal findings. Men with some genetic diseases and chromosomal disorders show increased proportions of sperm-shape abnormalities (see review by Wyrobek et al., 1983a). Some cases of human sterility have been associated with specific types of sperm-shape abnormali- ties, such as round-headed sperm. Human sperm structural characteristics also might have a familial component. Men show little change in their sperm structure distributions over periods of many years (Wyrobek et al., 1983a). At the average site in the seminiferous epitheli- um, it is estimated that approximately 23 spermatogenic stem cell renewals occur per year. Thus, the constancy seen in the proportion and types of sperm ab- normalities over time might reflect genet- ic determination like that reported for mice. Further research is required to elucidate the genes and protein functions responsible for human sperm shaping. As Marker of Fertility Status Evidence from both animals and men links increased proportions of abnormal sperm forms with reduced fertility. Human sperm with abnormal head shapes are less motile in vitro than normally shaped sperm, and sperm structure has been correlated with poor hamster-egg penetration (Shalgi et al., 1985~. In general, as the propor- tion of morphologically abnormal sperm increases, fertility decreases. The rela- tionship appears to be nonlinear but has not been well described. There are case reports of specific sperm-shape abnor- malities associated with sterility (Weissenberg et al., 1983~. Although it is not common to find a single type of sperm defect indicative of infertility, this has been seen in bulls (e.g., Dag ef- fect, where the head and tail are sepa- rated). In addition, Bostofte et al. ( 1 982b) found a correlation among time to pregnancy, number of children, and pro- portion of morphologically abnormal sperm. Partners of men with higher propor- tions of abnormal sperm took longer to get pregnant, and the men generally had fewer children. 93 As Indicator of Exposure to Toxic Agents Sperm structure has been used to assess the effects of ionizing radiation and chem- icals on animal and human sperm production. For mice and several domestic animals ex- posed to nonsterilizing doses of ionizing radiation, two-component time responses were generally observed-a large transient increase in the proportion of morpholo- gically abnormal forms shortly after ex- posure in mice treated with x-rays, begin- ning within 1 week after treatment, peaking at week 5 to 6, and returning to near back- ground by 11 weeks after end of treatment. This is followed, depending on genotype, by a smaller but persistent increase above background (e.g., Bruce et al., 1974~. In mice, both the transient and persistent increases were dose-dependent and have been observed also after chemical treat- ments (Wyrobek et al., 1983b, Meistrich etal., 1985~. The induction of sperm-shape abnormali- ties is highly indicative of exposure to a male reproductive toxin. A sperm mor- phology assay was developed and used to evaluate the germ cell effects of ionizing radiation and chemicals (Wyrobek et al., 1983a,b). Classifying sperm by their structure has statistical characteristics that make it well suited for use in both cross-sectional and longitudinal study designs. Earlier studies in mice reported cor- relations between agents' abilities to induce sperm-shape abnormalities and their germ cell genotoxicity, as measured by tests for dominant lethality, heritable translocations, and seven specific- locus mutations (discussed in the next chapter of the report and reviewed by Wyro- bek et al., 1983b). Because these correla- tions were based on small numbers of highly selected agents and the molecular mechan- isms underlying them have not been iden- tified, more data are required to resolve this point. Furthermore, for several agents known to induce sperm-shape ab- normalities in mice, no mutational damage has been detected. Also, in a retro- spective human study of 534 pregnancies, Homonnai et al. (1980b) found no relation- ship between sperm quality and adverse pregnancy outcome.
94 The available evidence suggests that the transient induction of sperm-shape abnormalities after exposure represents physiologic damage, rather than mutation- al damage to germ cells. The correlations observed between sperm structure and muta- genicity in mice might reflect the fact that chemical mutagens generally also damage cells physiologically. Two aspects of the sperm-structure response deserve closer scrutiny in regard to muta- tional damage: evaluations in the propor- tion of sperm-shape abnormalities that persist long after exposure, and the multi- generational inheritance of sperm-shape defects after mutagen treatment of male mice (e.g., Timourian et al., 1983~. The effect of ionizing radiation on human sperm structure has not been as fully investigated as the effect on sperm concentration. The results indicate that ionizing radiation induces sperm-shape abnormalities in exposed men. In addition, 44 of the 89 (49%) chemical exposures in the above-mentioned survey (Wyrobek et al., 1983a) used sperm structure as one of the markers of spermatogenic damage. Where data are available, the time course of induced sperm structural changes ap- proximate the time course of reduction in sperm number. The chemical exposures studied included occupational, therapeut- ic, and environmental chemicals. Other Factors That Affect Sperm Structure As in other species, human sperm struc- ture is relatively unaffected by abstin- ence time and several of the technical factors that influence sperm concentra- tion and motility (Amann, 1981~. However, sperm structure is sensitive to testicular temperature. Febrile diseases and severe allergic reactions might lead to increased sperm-shape abnormalities (MacLeod, 1964~. Increased temperatures-such as those found in saunas or hot baths and those experienced by professional truck driv- ers-could also induce sperm-shape abnor- malities (see review by Wyrobek et al., 1983b). Although major increases in tes- ticular temperature are clearly associ- ated with increased sperm abnormalities, IL4LE REPRODUCTIVE TOXICOLOGY the effects of small and irregular in- creases (as in the occasional use of hot tubs or in common illnesses, such as sea- sonal colds and influenza) have not been well studied, and their impact on human sperm production and fertility remains uncertain. Any factor that raises the core temperature of the testis or reduces heat dissipation is a candidate for heat-in- duced injury to spermatogenesis. Also, sperm structure is usually assessed on stained sperm smear and thus might be af- fected by slide preparation and staining conditions. However, the lack of consis- tent scoring criteria is by far the largest single factor affecting the assessment of sperm structure; efforts are under way to rectify this based on image analyses. SPERM MOTILITY Sperm motility is a measure of the. "movement" characteristics of sperm. Sperm are flagellate cells propelled by tails equipped with contractile proteins whose movements relative to each other control tails' characteristic wavelike motion. The contractile proteins are ar- ranged in longitudinal organelles within the sperm tail and include the coarse outer fibers, subfilaments, and microtubules. Although sperm motility is crucial in fer- tilization, sperm motility probably is not the primary method for transporting sperm within the female reproductive tract. That seems to be accomplished pri- marily by muscular contractility and cili- ary activity of the female reproductive tract. Sperm motility is important for traversing several key junctures within the female tractthe cervix and the utero- tubular junctionand might be essential for sperm penetration of cumulus cells and zonae pellucidae. Sperm motility is highly sensitive to extracellular conditions both within and outside the body. Sperm are immotile while in the lumen of the seminiferous epithelium in the efferent ducts and in the proximal and middle portions of the epididymis; they do not attain their motility potential until they reach the distal epididymis. It is hypothesized that the sperm cell mem- brane contains specific chemical recep-
MARKERS OF REPRODUCTIVE HEALTH tore. These sites could regulate the repe- titive depolarization and depolarization cycles and might function to coordinate the molecular events of the beating of the tail. Within ejaculates there is great diver- sity in the speed, direction, and type of sperm motion, which may be in part due to postejaculation technical factors, such as control of temperature (Phillips, 1972; Makler et al., 1 979a). Little is known about the motility selection processes that exist during transport within the female; only a few hundred highly motile sperm reach an ovum, and probably only one activates it. Once these processes are better understood on the molecular level, diversity in motility of ejaculated sperm could become more useful for identifying fertile and infertile sperm and semen. The visual procedures for measuring human sperm motility used in most clinical laboratories are imprecise and subjective (Sherins and Howards, 1986~. Typically, laboratory normal values depend heavily on method and scorer. The subjectivity of motility scoring has plagued interla- boratory and interscorer comparisons. Efforts are under way to use computer- assisted image analyses to provide objec- tive measures of sperm motility. As Marker of Sperm Function The presence of motile sperm in the ejac- ulate is strong evidence of normal sperm production and function. However, immo- tile sperm are not uncommon and arise because of a variety of technical and bio- logic factors. Clinically, persons who produce only immotile sperm are termed asthenospermic. Defects in sperm struc- ture and motility might be correlated, inasmuch as abnormally shaped sperm have poorer sperm motility than do normally shaped sperm (Katz et al., 1982~. As Marker of Fertility Status The importance of sperm motility in fer- tility has been well established; numerous studies have demonstrated a correlation between motility and fertility (Freund, 1968; Eliasson, 1975; Hargreave and Etton, 95 1983). Motility is probably not directly coupled with fertilizing capacity. For example, freezing causes sperm fertiliz- ing capacity to be lost before motility is lost. Although sperm motility might be a good correlate of fertility, a man with highly motile sperm could be infertile. However, men who repeatedly produce im- motile sperm are generally not fertile. The subjectivity of most clinical motility measurements has made it difficult to es- tablish standardized criteria for motili- ty and to investigate possible quantita- tive relationships between specific as- pects of motility and fertility. As Indicator of Exposure to Toxic Agents Exposure of humans to some toxic agents can affect their sperm motility. In a sur- vey of the effects of chemical exposures on human semen (Wyrobek et al., 1983a), 59 of 89 agents were evaluated for sperm motility; 22 showed significant decreases in exposed men. None of the studies meas- ured sperm motility quantitatively, and most did not use control or comparison groups. Furthermore, none of the agents surveyed reduced sperm motility without also decreasing sperm number (Wyrobek et al., 1983a; Ratcliffe et al., 1987~. Sperm motility is difficult to evaluate in human field studies. In fertility clin- ics, men are usually motivated to provide semen samples on site: but participants in environmental or occupational exposure studies often do not show such willingness. More typically, such participants prefer to collect samples at home and to deliver them to the laboratory at their conveni- ence. That requires careful protocol prep- aration, patient instruction, adequate equipment for transporting the semen to the laboratory, and attention to tempera- ture fluctuations and to the time between collection and analyses. Video-equipped microscopes have been used to collect mov- ing images in the field for subsequent analyses in the laboratory, thus reducing time between semen collection and mo- tility measurement.
96 Other Factors That Affect Sperm Motility Sperm motility is relatively insensi- tive to duration of abstinence. However, several endogenous male factors have been reported to affect sperm motility, includ- ing donor age, extent of sperm maturation, and surface-active agents (antibodies and agglutination factors). Motility is also sensitive to exogenous factors, in- cluding viscosity, osmolality, pH, tem- perature, ionic composition, nature of the suspending fluids, and presence of chemical modifiers, such as Inorganic ions, hormones, cyclic nucleotides, kin- ins, prostaglandins, and immunologic agents. Compared with sperm count and structure, sperm-motility measurement is especially sensitive to postcollection factors, especially time and temperature. Collection procedures can vary among clin- ical laboratories and those factors can affect motility in viva (within the male and female tracts) and during semen handling and analysis. Thus all interla- boratory comparisons of visually deter- mined motility data are problematic. As discussed in the next section, automation itself does not circumvent the variability caused by postejaculation factors, and controlling them will continue to be im- portant as clinics turn to more automated sperm-motility measurements. · SPERM VIABILITY Viability refers to the membrane in- tegrity of sperm. Dyes that are excluded by live cells but - incorporated by dead cells permit the determination of the pro- portion of live cells (Hargreave and Nils- son, 1983~. That determination is parti- cularly useful for distinguishing between live immotile cells and dead cells. A method for evaluating the membrane integ- rity of sperm is based on measurement of a cell's resistance to hypo-osmotic shock (Jeyendran et al, 1984~. Such methods are sensitive to postejaculation factors, and baselines in normal men and the effects of intrinsic and extrinsic factors have not been fully evaluated. . AL4LE REPRODUCTIVE TOXICOLOGY SPERM FUNCTION Sperm number, structure, and, to some extent, motility measure the physical aspects of human sperm, but indicate little about sperm function (i.e., ability to travel the female tract and to fertilize the ovum). Numerous attempts have been made to assess the functional aspects of sperm. The following are two in vitro meas- ures of sperm function. Cervical-Mucus Penetration The ability to traverse cervical mucus is one of the first major requirements of fertile sperm within the female tract (Moghissi, 1976~. Several laboratory methods have been proposed to evaluate sperm in cervical mucus (Kremer, 1965; Ulstein, 1972; Katz et al., 1980; Bergman et al., 1981~. Usually, the ability of sperm to enter the mucus, and their motil- ity within the mucus, are measured as well as their viability after penetration. The postcoital test is included in this group. The data provided by these tests are highly variable, owing in part to na- tural variations in the quality of human cervical mucus. Normal changes occur in cervical mucus during the menstrual cycle and test results are affected by the time at which the mucus is collected. In addi- tion, the mucus of different women differs in other ways that make it difficult to establish objective criteria for mucus penetration and to assess the quantitative relationship between mucus penetration and fertility (Hargreave and Nilsson, 1983~. Sperm-Oocyte Interaction For fertilization to occur, one of the several hundred sperm that reach the peri- phery of an ovum must traverse the cumulus cells and zone pellucida to penetrate it successfully. Ideally, living human ova would be used to assess the ovum-penetra- tion ability of human sperm, but ethical considerations bar these types of analyses for diagnostic purposes. Two alternate techniques have been developed: the zona- free hamster egg penetration test, and
at4RKERS OF REPRODUCTIVE HEALTn the nonliving, human egg penetration test. The zone-free hamster egg penetration test scores the proportion of enzymatical- ly denuded hamster eggs that are success- fully penetrated by human sperm (Yanagima- chi et al., 1976; Barros et al., 1978; Rog- ers et al., 1979; Hall, 1981; Chang and Albertson, 1984~. Penetration is scored as the proportion of eggs that contain sperm undergoing nuclear Recondensation. However, there is large variability in results, and only samples with very low penetration rates (e.g., less than 10%) are considered potentially abnormal. Even repeat samples from the same men show large variability. The utility of this test as an indicator of fertility remains highly controversial. Furthermore, the underly- ing molecular mechanisms are unknown, and it remains unknown exactly what the hamster assay is measuring. Using nonliving human eggs and human zone pellucida circumvents some of the disadvantages of using hamster eggs (Overstreet et al., ~ 980~. However, this technique is not widely available, because of the difficulties in obtaining a supply of human eggs. OTHER SPERM MEASUREMENTS Sperm Agglutination Techniques for detecting antisperm antibodies in blood, semen, and cervical mucus are available (Rumke and Hellinga, 1959; Halpern et al., 1967; Rumke, 1968; Haas et al., 1980; Mathur et al., 1981~. They are used clinically where incompati- bility between a man and a women is sus- pected (for example, poor cervical mucus penetration). The presence of sperm an- tibodies is a concept closely related to sperm domains, which is discussed later in this chapter. Nuclear Chromatin The nucleus of the differentiating germ cell undergoes dramatic changes during spermatogenesis in chromatin structure and in the constitution of basic proteins. The mammalian sperm nucleus is highly com- pact, rigid, and of high specific gravity; 97 its DNA is genetically inactive and rela- tively inert in response to xenobiotic chemicals. During meiosis, several his- tones peculiar to meiosis appear in the nucleus with residual variants common to somatic cells. After meiosis, the meiotic histories are replaced by a series of tran- sition proteins that are finally replaced with the basic sperm protein called prota- mine. Protamine is thought to facilitate the dense compaction of the sperm nucleus, making it genetically inactive and rela- tively inert to chemical exposure and im- parting nuclear rigidity required for fertility. A single protamine is found in sperm of most mammals. In some mam- mals-including men, mice, and hamsters sperm also contain a second protamine that is more variable in length and sequence than that of the first protamine. In human sperm, about 15% of the nuclear protein is histones. It is not known whether the protamine and histone content of human sperm is related to nuclear structure or function. Sperm nuclei naturally decon- dense on entry into the ovum in fer- tilization. The following are two approaches for evaluating the ability of sperm to decon- dense in vitro: · Chemical Recondensation of sperm nuclei in vitro. Sperm nuclei can be Recondensed in vitro with invertebrate egg extract, in high-salt solutions, with reducing agents, with detergents, or with special salts (Huret, 1986~. The susceptibility of ejaculated human sperm nuclei to these agents in vitro might be indicative of the extent of chromatin condensation or nuclear protein cross-linkage. However, that has not been confirmed experimental- ly. Wildt and coworkers (1983) suggested that lead-exposed workers have an in- creased susceptibility to Recondensation with SDS. · Colorimetric measurements of sperm chro- matin. Evenson and colleagues (1980) de- veloped a flow-cytometric method for quan- tifying the ratio of single- and double- stranded sperm DNA. The method measures the fluorescent dye acridine orange, which distinguishes strandedness of DNA. In mutagen-exposed mice, the fluorescence
98 ratio was highly correlated with the degree of induced sperm-shape abnormalities. Infertile bulls seem to have higher fluore- scence ratios than fertile bulls. Data on humans are insufficient for evalua- tion. The method requires further study. Double F Bodies When stained with the fluorescence dye qu~nacrine and viewed in a fluorescent microscope, human sperm fluoresce over their entire nucleus. Some sperm have a brighter spot within the nucleus, thought to be the Y chromosome (Barlow and Vosa, 1970~. The Y chromosome is known to fluor- esce brightly in stained somatic nuclei and metaphase spreads of male cells. However, contrary to expectation, the proportion of sperm with one spot seldom reached the 50% value expected if the spots truly represented all sperm with Y chromo- somes. Also, Kapp and associates (1979) suggested that the rare sperm with two spots could be due to Y-chromosomal aneu- ploidy and proposed that the OFFS test be an indicator of induced aneuploidy. However, no data support that contention; in fact, mass measurements and comparisons with chromosomal analyses of human-ham- ster hybrid chromosomes suggest that sperm with two bright spots do not necessarily represent aneuploid sperm. The proportion of sperm with double fluorescence (double F bodies) increases with exposure to DBCP, adriamycin, and several other clinical agents (Kapp et al., 1979) and might be an indicator of exposure. However, the mole- cular and cellular mechanisms underlying this observation are unknown. . . CHEMICAL COMPOSITION OF SEMINAL FLUID Human seminal fluid is a biochemically diverse solution containing peptide, lipid, carbohydrate, glycoprotein, and salt components derived from the testis, efferent ducts, epididymis, and accessory glands. Normal seminal fluid constituents were described in detail in Chapters 4- 6. Considered a research tool, chemical analysis of semen is not commonly used in fertility evaluation. AL4LE REPRODUCTIVE TOXICOLOGY Semen can contain xenobiotic agents, including metals and chlorinated organic substances (Dougherty et al., 1981~. How- ever, it is not easy to determine whether xenobiotics in semen represent testicular or accessory sex organ exposures. It is also unclear how xenobiotics from acces- sory glands can affect sperm and their fate in the female tract; the events are likely to be agent-dependent. Because fertile sperm are removed from the seminal fluid early during their transit to the fertili- zation site, it is likely that only agents that are tightly bound to or internalized in sperm will be present at the fertiliza- tion site (e.g., pesticides that may be solubilized in the plasma membrane). A1- ternatively, xenobiotics in semen might affect fertilization and development by indirect exposure via the mother's cir- culatory system. However, exposure to the egg and embryo via maternal circulation is likely to be very small. The role of seminal fluid components on fertilization and development warrants further basic research. Reliable markers of the normal and xenobiotic constituents of semen are needed. PROMISING RESEARCH CONCEPTS The following are selected research concepts (see Table 7-2) that have promise of yielding semen-based markers of male reproductive health. Most are still too early in their development to evaluate their utility (1) for assessing male fer- tility, (2) as indicators of sperm produc- tion and function, or (3) as indicators of exposure to agents that interfere with male reproduction. Some approaches, how- ever, are already so advanced that human baselines are being established (e.g., computer-assisted image analyses of motility and structure). Automated Sperm Measurements Automation is a general concept that may eventually be considered in the devel- opment of all sperm markers. At present, it is best exemplified by machine-based measurement of sperm motility and struc- ture. However, a sizable gap remains be-
M`4R=RS OF REPRODUCTIVE HEALTH tween the capabilities of research instru- ments and their commercial availability. Machine- Based Sperm -Motility Measurements Considerable progress has been made toward developing automated methods for measuring sperm motility (Walker et al., 1982~. With early quantitative ap- proaches, sperm motion was observed di- rectly under the microscope or was deter- mined from photomicrographs of strobo- scopically illuminated sperm. The most widely used quantitative approaches in- clude multiple-exposure photomicrography (Makler et al., 1 979b) and videomicro- graphy. Some of these methods have been used on thousands of men (e.g., Katz and Overstreet, 1981; Overstreet et al., 1981) as part of routine clinical evaluations to determine percent sperm motility, mean swimming speed, and percent progressive motility. With videomicrography, micro- scopic images of multiple sperm fields can be permanently recorded for later analysis with manual techniques, such as the placement of graduated overlays on the video image (Overstreet et al., 1981), or with automated methods. Modern automated methods generally rely on image analysis to locate the sperm and to quantify their motion. Edge-con- trasting methods convert the visual images of sperm into binary masks from which the central coordinates (centroid) of each head are determined. By following the trace of the centroid in time, one can compute both rectilinear and curvilinear movements to evaluate sperm progressive- ness. However, sperm motility is complex, and only a few aspects of sperm movement have been evaluated so far. Clearly, addi- tional descriptors of sperm head and fla- gellar movement are needed. Furthermore, most automated methods provide distribu- tional data and, statistical analyses should include descriptors of central tendency, dispersion, as well as distribu- tional outliers. Automation of sperm-motility measure- ment is a young field. To date, there have been few detailed evaluations of any auto- mated method or motility characteristic. 99 Also, there is still no systematic study of normal values for any sperm-motility characteristic In a well defined popula- tion. It is important to emphasize that, for each new motility measure proposed, its range for normal values. sources of variation, relationship with fertility, and utility as an indicator of exposure must be established anew as part of the validation process for diagnostic pur- poses or exposure monitoring. At least three automated motility-meas- uring systems are available commercially. Extensive validation is needed before their utility can be thoroughly evaluated. Quantitative Sperm Structure As with sperm motility, subjectivity plagued all attempts at interlaboratory comparisons of visual sperm structure assessments (Freund, 1966~. Disagreement persists regarding the precise shape and size of fertile sperm, in part because different laboratories use different classification criteria (Fredricsson, 1979), sperm sizes are known to be donor- dependent, and no one has actually measured the structure of fertilizing sperm. In addition, fertile donors can differ mark- edly in the proportion of sperm in various shape classes. Several decades ago, MacLeod and Gold (1951) developed a widely recognized vis- ual scoring system; his success was due in large part to the high level of internal standardization he obtained by scoring smears himself. Later efforts to improve visual sperm classification have included complex assignment schemes that consider head, midpiece, and tail structure sepa- rately (David et al., 1975) and the use of decision-tree logic and reference slides (Wyrobek et al., 1982~. Recently, objective methods have been proposed for measuring sperm size and shape. Katz and Overstreet (1981) clas- sified human sperm with graduated overlays on previously recorded video images. Schmassman et al. (1982) used an image- analysis system to measure sperm nuclear size and reported that the sperm of infer- tile men had larger nuclear area than the sperm of fertile men. In another evalua-
100 lion of image analyses, 27 aspects of nu- clear size, shape, orientation, and stain content were measured for each sperm nucle- us and were evaluated statistically to identify parameter groupings that most accurately assigned sperm to 1 of 10 visu- ally based shape classes (Moruzzi et al., 1988~. An 86% overall classification ac- curacy was obtained with measurements, including basic morphometric characteris- t~cs, indicators of stain content, and measures of nuclear inhomogeneity. Those preliminary examples of the utility of automated sperm-structure analysis are encouraging, but much work is needed before these methods can be validated for clinical use. Additional research is needed to improve cell-staining procedures for image analysis, to strains computers to recognize specific shape classes, and to develop methods for classifying sperm in real time. Cooperation will be required among researchers and clinicians to define individual sperm-shape classes with mor- phometric characteristics rather than visual criteria. Alternatively, morphometric charac- teristics could be used to describe sperm samples independently of visual classifi- cation systems. For example, new classifi- cation systems could be based on distribu- tional data for such measures as nuclear area, length, and width. Evaluations could be multivariate and whenever possible should consider descriptors that repre- sent central tendencies (e.g., average nuclear area), dispersion (e.g., variance of nuclear area), and distributional out- liers (e.g., proportions of cells above or below threshold areas). General Considerations Regarding Automated Methods Automation is very attractive because it provides a degree of precision and ob- jectivity in measuring sperm motility and structure not attainable with previous visual methods. However, there are several factors that should be considered in the application of automated methods for any sperm parameter, not limited to sperm motility or structure: MALE REPRODUCTIVE TOXICOLOGY 1. Automated methods do not in themselves circumvent the fundamental problems of semen collection, handling, and pre- paration that especially affect sperm motility. 2. Machines do not make diagnoses; they simply provide quantitative data. The nature of data depends on the sophistica- tion of the machine, the program language used, and the ingenuity of the programer and operator. These instruments typical- ly provide voluminous data, and strate- gies for data handling must be developed concurrently. 3. Quantitative methods require refer- ence or normal values for interpretation and statistical methf`~ for making comparisons. 4. Each new machine method end especially each new proposed sperm measure will re- quire new investigations of normal values among fertile men and of the effects of confounding factors. There are few data to establish normal values for any of the quantitative sperm measurements in well- defined populations. 5. Also, the relationship of each charac- teristic to fertility must be evaluated before its clinical relevance can be ascer- tained. At present we can only guess which characteristics will be useful for pre- dicting fertility status and which will be sensitive to environmental exposures. Until methods and sperm characteristics are validated, they must be considered research tools, rather than clinical instruments. _ _ _ O These features of automated sperm analyses do not diminish the benefits of these technologic developments. Rath- er, they should serve as caution against overinterpretation and as guidelines for future needs. Markers of Sperm Function Sperm progression through the female tract is a multistage process, and our understanding of the underlying molecular components that control sperm capacita- tion and sperm penetration through the cervical junction, uterotubular junction, cumulus cells, zone pellucida, and plasma
MARKERS OFREPRODUCT~E HEALTH membrane of the egg is growing slowly. Proposed markers of sperm penetration are either mucus-based or chemical-based; both show promise. Change in androgen binding also deserves further study as a possible marker of sperm-penetration potential. There are few promising approaches for measuring sperm capacitation directly in sperm, because underlying mechanisms are not clear. Some sperm surface compo- nents are lost during capacitation, in- cluding antigens (Vernon et al., 1985) and an acrosomal stabilizing factor (Eng and Oliphant, 1978~. After capacitation, an antigen initially detected only on the posterior sperm tail becomes localized predominantly on the midpiece (Myles and Primakoff, 1984~. Further research is encouraged to define the key molecular aspects of capacitation. Fertile sperm require normal function- ing of many enzyme systems. In vitro assays of enzyme functions related to energy me- tabolism, sperm motility, sperm penetra- tion, and fertilization would provide valuable insight into the biology of repro- duction and might have clinical applica- tion in establishing the biochemical basis of infertility. An example of a promising approach is the recent observation that carooxymetny~ase activity is decreased in the semen of men with immotile sperm (Gagnonetal., 1982; 1986~. Available tests of sperm-egg interac- tion typically require a ready supply of fresh eggs and are plagued by both poor reproducibility and poor quantitation. Improvements and alternatives are needed. Optimistically, surrogate methods based on understanding specific molecular as- pects of normal sperm-egg interaction might yield markers that are sperm-based but do not require mammalian eggs. Immunologic Reagents for Studies of Semen and Spermatogenesis Antibodies are highly specific for de- tecting single antigenic determinants, can be available in virtually unlimited quantities, and are usable in a wide varie- ty of physiologic, morphologic, biochemi- cal, and molecular studies. They recognize 101 specific antigenic determinants on pro- teins and glycoproteins present in single domains on living cells, and they can be measured with sensitive and highly specif- ic enzyme-linked immunosorbent assays (ELISA) and radio-immunoassay (RIA) pro- cedures. Immunologic reagents can be used to investigate the molecular arrangement of both surface and internal components of human sperm, and they provide an ap- proach for elucidating male reproductive abnormalities on a biochemical and molecu- lar level. This section is organized by sperm components and well as surface do- mains (acrosome, tail and midpiece, and nucleus). In addition, the need of immuno- logic reagents for identifying and measur- ing natural and foreign components of sem- inal plasma, for determining hormonal status (not discussed in this section), and for use with recombinant-DNA gene- expression vectors is emphasized. Antibodies can be used to study either cells or molecules by ELISA and RIA tech- niques. For applications requiring the analysis of single cells, cytometry can provide detailed measurement of antibody distribution on individual cells and cell- to-cell variations with a precision not possible through microscopic inspection. Flow cytometry and image analysis provide powerful complementary analytic capabili- ties. In addition, flow cytometry can be used to sort individual cells for micro- scopic and biochemical analysis (Van Dilla and Mendelsohn, 1979~. Typically, cyto- metric analysis of antibody-labeled cells requires well-controlled fluorescence staining techniques. Sperm Surface Domains The sperm plasma membrane is divided into distinct regions or domains, each of which contains characteristic marker antigens, overlies a major structural component of the sperm, and has well-de- fined boundaries (Myles et al., 1981~. The plasma membrane of the sperm head in- cludes the anterior acrosome, equatorial segment, and postacrosomal domains. A variety of probes have been used to charac- terize the domains, but monoclonal anti- bodies have been used most effectively
102 to map the location of sperm surface do- mains, identify antigens in particular domains, define the origin of such anti- gens, and determine the role of specific sperm surface antigens in reproductive processes (Eddy, 1987~. For example, during the acrosome reac- tion, the plasma membrane of the anterior acrosome domain fuses with the underlying acrosomal membrane, releasing enzymes from the acrosome and allowing the sperm to penetrate investments of the egg. A monoclonal antibody to a protein in the anterior acrosome domain of mouse sperm blocks the acrosome reaction and prevents fertilization (Sating, 1986~. In addi- tion, when the sperm reaches the egg sur- face, the membrane of the posterior part of the sperm head fuses with the membrane of the egg, allowing the sperm to enter the egg cytoplasm. In the mouse, a monoclonal antibody to a protein in the equatorial segment domain blocks this process and prevents fertilization (Saling et al., 1985~. A result of the acrosome reaction is that most of the plasma membrane is lost from the anterior acrosome and equatorial segment domains. At the same time, part of the acrosomal membrane becomes con- tinuous with the plasma membrane at the anterior edge of the equatorial segment, forming a new domain of the plasma membrane over the anterior sperm head. On fertil- ization, the sperm plasma membrane is in- serted into the egg plasma membrane. Some sperm surface components appear to remain in a small patch in the egg plasma membrane (Gabel et al., 1979~; others diffuse over the entire egg surface (Gaunt, 1983~. Monoclonal antibodies have been used to identify some of the antigens added to the sperm surface in the epididymis (Orge- bin-Crist and Fournier-Delpech, 1982; Sating, 1982; Eddy et al., 1985~. One such antigen is secreted by the epithelium in a discrete region of the mouse epididymis and attaches to the midpiece and distal tail portion, apparently by binding to specific acceptor sites in these domains (Vernon et al., 1982~. Surface alterations also occur during ejaculation, when pro- teins secreted by the accessory glands of the reproductive system bind to sperm (Irwin et al., 1983; Isaacs and Coffey, 1984~. MALE REPRODUCTIVE TOXICOLOGY Additional changes take place in the plasma membrane as sperm undergo ca- pacitation in the female reproductive tract. That process must occur if sperm are to complete the acrosome reaction and fertilize the ovum. Some sperm surface components are lost during capacitation, including antigens (Vernon et al., 1985) and an acrosomal stabilizing factor (Eng and Oliphant, 1978) that are acquired in the epididymis. The boundaries and con- tents of domains are also modified. An antigen initially detected only on the posterior tail domain of guinea pig sperm becomes localized predominantly on the midpiece domain after capacitation (Myles and Primakoff, 1984~. Another anti- gen migrates from the postacrosomal seg- ment domain into the acrosomal domain after the acrosome reaction (Myles and Primakoff, 1984). Such studies indicate that sperm plasma membrane domains are dynamic, undergo structural and functional changes throughout the life of the cell, and con- tain antigens that serve vital roles in reproduction. The establishment of domains requires synthesis of specific components in ap- propriate quantities, delivery of the components to the cell surface, and segre- gation of the components to specific re- gions of the sperm surface. Toxic agents could perturb domain formation, composi- tion, or maintenance by acting directly on those processes or indirectly by alter- ing spermatogenesis, Sertoli cell func- tion, or endocrine processes. Later modi- fications of the sperm surface in the male reproductive tract depend on the nor- mal metabolic and secretory functions of the epididymis and accessory glands. Toxic agents might affect those modifica- tions directly (by interfering with proc- esses at the sperm surface required for the addition, removal, or alteration of domain components) or indirectly (by per- turbing biochemical or physiologic ac- tivities of the ducts or glands of the male reproductive tract). Such effects of toxic agents might be expected to alter the com- position and distribution of domains and to be detrimental to male fertility. It is recommended that studies to deter- mine the usefulness of monoclonal antibod-
MARKERS OF REPRODUCTIVE HEALTH ies to sperm surface components for detect- ing toxic effects on the male reproductive system be accorded high priority. They have considerable potential to identify the site and mechanism of action of toxic agents. These studies can be initiated with antibodies that are now available, but larger numbers of antibodies to rodent and human sperm should be prepared specifi- cally for this purpose. If toxic agents are found to have detectable effects on sperm surface domains in experimental animals, monoclonal antibodies can proba- bly be used for detecting toxic effects on the human male reproductive system. Components in the Sperm Several biochemical components of the human acrosome have been identified (Abyholm et al., 1981; Eliasson, 1982), including acrosin, hyaluronidase, corona- penetrating enzyme (CPE), ATPase, acid phosphatase, aspartyl amidase, and beta- glucuronidase. These components sea in sperm penetration of the zone pellucida (acrosin), in penetration of cervical mucus (acrosin) (Beyler and Zaneveld, 1979), in sperm penetration of the cumulus oophorus (hyaluronidase), in sperm pene- tration of the corona radiate (CPE), and in the sperm capacitation process (acro- sin, CPE, and ATPase). The roles of acid phosphatase, aspartyl amidase, and beta- glucuronidase remain unknown. A monoclonal antibody for human acrosin recently was described (Elce et al., 1986~. The production of acrosome-specific mono- clonal antibodies would be instrumental in characterizing and measuring the acro- somal components of "fertile~ sperm and in identifying abnormalities. At the center and along the length of the tail is an arrangement of microtubular doublets, similar to that in cilia. Fur- ther research is warranted to catalog the structural constituents (including dynein and the tubulins) and to decipher the mole- cular mechanisms of energy conversion and the roles of motion-related proteins and carbohydrates of the sperm tail. Antibod- ies to such sperm molecules will help in this research and could lead to the devel- opment of molecular markers of sperm tail 103 and motility dysfunction. Monoclonal antibodies also could be valuable for investigating the nuclear constitution of sperm. Monoclonal anti- bodies specific for human protamines (Starker et al., 1 987a) and some sperm histones have been developed and charac- terized with ELISA and Western-blot meth- ods. These antibodies require further study in well-characterized populations of fertile and infertile men for evaluating the relationship between sperm nuclear constitution and fertility. In addition, sperm antibodies against transition pro- teins, meiotic histones, and other com- ponents of the sperm nucleus are needed for basic studies of sperm biochemical structure, identification of candidate reagents for clinical study, and evalua- tion of the sperm-nuclear effects of environmental exposures. Use of Antibodies for Studying Gene Expression With specially engineered phage and bacterial hosts (e.g., Riva et al., 1986), antibodies can be used for screening clones that contain genes producing tes- tis-specific proteins recognized by spe- cific antibodies. For example, human tes- tis cDNA libraries have been constructed into expression vectors, such as lambda gtl 1 and resulting bacterial plaques can be induced to produce human testis pro- teins. Isolated clones can be used as probes for the gene and to monitor sperm differentiation. Such probes also permit molecular characterization of chromosomal location, haploid copy number, and genetic controlling elements. In principle, any human protein or peptide (provided that it is antigenic) can be used to prepare monoclonal antibodies to screen for and clone the corresponding human gene. Thus, the development of antibodies against sperm components provides the useful first step for identifying and characterizing the genes that control that aspect of human sperm production. Antibodies for Characterizing Semen In addition to sperm, the human ejaculate
104 contains several other cell types de- scribed in Chapter 4. Monoclonal antibod- ies specific for each cell type would fa- cilitate the objective characterization and measurement of these seminal compo- nents. In principle, antibody methods could measure the chemical composition of semen. Antibodies are needed for nearly all the normal seminal constituents, in- cluding hormones, enzymes, proteins, carbohydrates, and lipids. Once devel- oped, the antibodies might be used to de- termine the concentrations of those normal constituents in groups of fertile and in- fertile men and in men exposed to reproduc- tive toxins, to evaluate their utility. Semen is a complex mixture derived from several glands; coupled with animal re- search, these antibodies could be used to determine the glandular source of each constituent. Research on monoclonal an- tibodies for detecting and measuring nor- mal chemical components of semen should receive high priority. In addition, antibodies could be used to detect chemicals not normally found in semen, such as environmental and occupa- tional chemicals and drugs. Antibody meth- ods have advantages over spectrophotomet- ric and chromatographic methods: they cost less to perform, are highly specific, and can have excellent sensitivity. The sen- sitivity of the antibody approach for de- tecting xenobiotics is illustrated by the recent demonstration that antibodies can detect small chlorinated carbon molecules at concentrations of less than 1 ppm in soil samples (Starker et al., 1987b; Vanderlaan et al., 1987~. Such methods have not yet been applied to semen. SEMEN MARKERS OF SERTOLI CELL AND LEYDIG CELL FUNCTION Semen provides a natural window for eval- uating retrospectively the function of the major somatic cells that support sperm production: Sertoli and Leydig cells, for evaluating retrospectively some aspects of epididymal function, and for evaluating prostatic and seminal ves- icle function. As discussed in Chapter 4, those cells and organs contribute spe- cific constituents to semen. For example, ABLE REPRODUCTIVE TOXICOLOGY semen is made up of a minute amount of sperm- dense epididymal fluid mixed at ejacula- tion with the secretions of the accessory glands (Lilja et al., 1987~. The major structural protein in the coagulated com- ponent of the ejaculate is a high-molecu- lar-weight protein from the seminal vesi- cles termed HMW-SV protein, or semenogel- in. Its transformation into three subunits during liquefaction results in a series of basic low-molecular-weight proteins. The seminal gel liquefies through proteo- lysis by prostatic kallikreinlike serine protease (also known as prostate-specific antigen). Fibronectin, an "adhesives glycoprotein, is also part of the seminal gel where it is linked to semenogelin. The seminal vesicles are thought to excrete lactoferrin, a metal-chelating protein that adheres to sperm. Easy and reliable quantitative methods are needed to monitor semen for those and other specific secretory products. The molecular func- tion of most of those products, as well as of the factors that modulate seminal amounts and activities, remains to be eval- uated. In addition, the normal baselines as well as the relationship to fertility of semen markers of the Sertoli cell, Ley- dig cell, and accessory organ function remain to be determined. · . RECOMBINANT-DNA METHODS FOR STUDY OF HUMAN SPERMATOGENESIS AND SEMEN As reviewed by Hecht (1987a), spermato- genesis is an ideal differentiating system for investigation of the control of gene expression as related to normal protein and cell function. Its morphology and kinetics have been well described, there is only a single cell product, and our un- derstanding of biochemical mechanisms is rapidly increasing. The availability of DNA probes for spermatogenic proteins derived from both messenger RNA and genomic DNA is increasing rapidly. As a result, the genetic factors that control . . sperm difI~erentiat~on are beginning to unfold. These recombinant-DNA probes are discussed in detail in Chapters 4 and 9. Recombinant-DNA probes promise at least two types of markers that might
MARKERS OF REPRODUCTIVE HEALTH 105 become useful for assessing male the proportion of sperm with specific ge- reproductive health. netic lesions and investigating the le- First, some recombinant-DNA probes signs in detail. However, this applica- might provide a means for evaluating lion has many additional technologic hurd- male infertility at the gene level. For les, because methods for probing DNA of example, in conjunction with other tech- single cells are still in their infancy. nologies, such as cellular staging under Although chromosome-specific repetitive the microscope and antibodies against DNA probes have been used to assess the spermatogenic proteins and other cell ploidy of individual cells, methods are markers, recombinant probes could be use- not yet sensitive enough to detect unique ful for determining whether transcription sequences in single cells. is altered in infertile men and, if so, what In spite of present limitations, re- search and development of recombinant- DNA probes for spermatogenic genes are progressing rapidly, and they are expected to have broad applications for assessing male reproductive health. Recombinant techniques are also applicable to the study of Sertoli cell, Leydig cell, and accessory gland function, and continu- ing research in this field is strongly encouraged. spermatogenic cell stages are involved. In addition, analysis of genomic DNA from carefully selected men with similar sper- matogenic arrest patterns could lead to the identification of characteristic genetic lesions that cause infertility. Second, as described by Hecht (1987b), recombinant probes could also provide a means to analyze the DNA of individual sperm. That could be useful in measuring
