The relevant questions here are: (i) is current conventional semen evaluation sufficient; and (ii) have we reached a level where current spermatological laboratory methods allow us to disclose the fertility of a male and/or his ejaculate, processed or raw? Conventional analysis of semen is most relevant for the initial investigation of male infertility but, as concluded by this author 10 years ago, only of limited value if prediction of fertility is the aim (Rodriguez-Martinez, 2003). Have we now made sufficient progress towards this goal? What will the situation be in 10 or even 20 more years? The present chapter does not intend to forecast the future, but it does aim to critically review advances in semen assessment methodologies and the capacity that different assays have to prognose fertility (or, perhaps, infertility). Although far from exhaustive, the review attempts to provide a comparative view of the different species for which clinical andrology and modern seminology interact with breeding strategies and the application of assisted reproduction technologies (ARTs). Particular attention is paid to new methods for determining DNA and transcript intactness, as well as to those biomimetic in vitro assays that, by resembling events during sperm transport, storage and interaction with the female genital tract and the oocyte, best provide clues for sperm selection and the role of sperm sub-populations in the ejaculate. Finally, it attempts to identify those techniques that might be available for clinicians and breeders 20 years from now.

Clinical Andrology and Diagnostic Seminology: Can they Prognose Fertility – or Just the Risk Level of Never Reaching Fertile Status?

The clinical andrological examination of a sire, including the screening of his semen, aims not only to determine the normality of the function of his testes and genital tract, but also his capacity to mate. Such an examination provides an andrological diagnosis with a final estimation of his potential capacity to procreate, which, among animal breeders, is often known as a breeding soundness evaluation (BSE). The BSE (as described in accompanying chapters), combines a general clinical examination, a detailed genital evaluation, a testing of the sire’s capacity to mate and laboratory evaluation of the ejaculate(s) (Kastelic and Thundathil, 2008). The BSE, mostly applied to production animal species, has helped us to identify clear-cut cases of infertility, or even of potential sub-fertility, thus facilitating the removal of sub-fertile males from both natural and assisted breeding (AI in particular).

In species that usually have a comparatively ‘less good’ semen picture, for example the feline, canine or equine, semen evaluation is directed towards the identification of whether the problem is related to sperm numbers or sperm dysfunction, as ‘treatment’ opportunities might be present even without hormones. Low sperm numbers in the ejaculate denote a low spermatogenic output, which might be optimized by the establishment of an individual ‘rhythm’ of semen collections, taking care of maintaining sperm maturation thresholds and the best possible sperm output per ‘abstinence’ (or noncollection) period (time elapsed from last collection). Often, this procedure might improve chances of fertility by the implementation of ‘controlled mating intervals’. Assessment of the type of sperm dysfunction that may be present is most relevant here so that the most appropriate type of assisted reproduction technology can be determined, e.g. lowsperm dose AI, in vitro fertilization (IVF) or even intracytoplasmic sperm injection (ICSI), and, if possible, applied. This is relevant for some pet breeds and particularly for horses, where sub-fertility is widespread among stallions and results with conventional IVF are, as yet, suboptimal.

The situation is far from similar in animals (mostly production animals) that have been selected for their fertility. Semen evaluation in breeding sires of production species (such as the bovine, ovine or porcine) often goes beyond the primary diagnostic purpose (early diagnosis of subfertility, so that the sire is promptly removed to avoid unnecessary costs), and extends to prognose levels of their ‘expected’ fertility before AI is used and their selection as top sires is decided upon. Thus, early identification and removal of the ‘notso-good’ sires is a priority.

Despite this, during clinical andrological examinations, fertilizing capacity is not monitored by a simple spermiogram, and is often only assessed after breeding by natural mating or AI of a certain number of females. The recordings of fertility are, therefore, of utmost importance if a male is to be downgraded, from ‘normal’ fertility output to subfertility or, in serious cases, to permanent infertility (sterility). The registering of male fertility varies in stringency among species, mostly depending on the number of females involved. In some species (e.g. the equine and canine), sires breed relatively few females and, owing to the innate variation present between ejaculates and males, they yield fertility results of less analytical reliability than livestock species in which sires breed with large numbers of females. In the equine, as stallions might breed with the same females over several oestrus cycles per season, the resulting cumulative fertility outcome ought to be properly expressed (as per oestrus or mating/AI, or per season) for reliable comparisons.

Estimations of in vitro fertility depend on the type of test used to study a heterogeneous suspension of pluri-compartmental, terminally differentiated cells: the spermatozoa. Current assessment of ejaculated spermatozoa focuses almost solely on the number of spermatozoa (per unit of volume, i.e. concentration or as total per ejaculate) and their motility (sometimes including their kinematic patterns, if a CASA instrument is used), although other aspects, such as pH and ejaculate volume are also included. Few studies are published on the relation between neat semen and fertility because after collection (when it is at 35°C), semen is always extended and then either chilled or frozen (from room temperature at +20°C, or fridge temperature at +5°C, or anything in between), and used thereafter for AI; these processes introduce other variables into the assessment of semen (its extension, dilution of the SP, additive effects, cooling, time, etc) – considerations that will be discussed later. The same considerations apply to sperm numbers, which implies that each sperm AI dose contains a specific number of spermatozoa. However, as handling (including cryopreservation and other manipulations) causes cell death, shortens lifespan and modifies the fertilizing capacity of various proportions of spermatozoa, sperm numbers in an AI dose are often excessive so that breeders are safeguarded in providing fertility to their customers. Only when viable sperm numbers are decreased to possible ‘threshold numbers’ do we see a proven relation between sperm numbers and fertility (Tardif et al., 1999; Christensen et al., 2011).

Morphological abnormalities are always present in any ejaculate, although various forms of abnormality differ in their impact on fertility. Some are specific defects that hamper fertilization, while others, such as the pyriform or pear-shaped sperm head deviation, impair proper embryo development (Rodriguez-Martinez and Barth, 2007), thus calling for frequent (intervals of 2 months) detailed assessments of sperm morphology in AI stud sires using wet or stained smears (Plate 63a,b). The reliability of such analyses requires large numbers of spermatozoa to be assessed per sample, i.e. 200 per wet-smear and 500 for stained sperm heads; the latter allows for determinations of defects that have clear relationships with fertility due to their uncompensable nature, such as the pyriform sperm head shape (as an expression of a defective chromatin condensation during spermiogenesis) (Al-Makhzoomi et al., 2008). Software for the ASMA system has been developed since the 1980s, and has now reached an acceptable reliability for the analyses of sperm head dimensions, though other types of sperm abnormalities cannot yet be determined (Auger, 2010), and nor are there clear relationships of sperm head morphometry with fertility (Saravia et al., 2007b; Gravance et al., 2009).

Does the Exploration of Other Sperm Attributes, including Testing Sperm Function In Vitro, Prognose Fertility?

The most advantageous use of fluorophores is when they are combined. For instance, subtle changes in the plasma membrane such as modifications in permeability can be disclosed by combining several fluorescent probes, such as SNARF-1, YO-PRO-1 and ethidium homodimer, the so-called triple stain (Peña et al., 2005a, 2007), as exemplified in Plate 65. Viable, membrane-stable spermatozoa are able to pump out small probes such as YO-PRO-1, which always enters living cells. Sub-viable cells, i.e. those with incipient membrane destabilization, become permeable and unable to pump out YO-PRO-1, which then stains the DNA bright green and can easily identify the sub-viable cells before they show other signs of degeneration (Peña, 2007). Membrane permeabilization is followed by modifications of the position of phospholipids such as phosphatidylserine (PS), which can be mapped by the fluorophore combination of Annexin-V/PI (Januskauskas et al., 2003; Peña et al. 2003, 2005a, 2007; Saravia et al., 2007a). Earlier changes in the membrane, related to the destabilization of the lipid layer, can also be followed using a triple combination of fluorophores: Merocyanine-540 (Mero-540), YO-PRO-1 and Hoechst 33342. Plate 66a–b shows typical readings obtained using two different probe combination for the estimation of sperm asymmetry. This approach has proven effective in disclosing defective spermatozoa in several species, and gives good correlations with fertility (Hossain et al., 2011a).

Besides energy, sperm mitochondria produce by-products of the metabolism of oxygen, including superoxide. This superoxide converts into the damaging hydrogen peroxide, a reactive oxygen species (ROS), which is mostly, but not completely converted to oxygen and water by the enzymes catalase or superoxide dismutase (also known as antioxidants or scavengers). A certain level of ROS is essential for sperm function, including fertilizing capacity, but only when it is kept at optimal levels by the antioxidant capacity of SP (Awda et al., 2009; Mancini et al., 2009; Am-in et al., 2011). When excessive numbers of leucocytes are present in the ejaculate or the semen is subjected to oxidative stress (as during cooling in the absence of SP or other natural scavengers), increased ROS generation, either extrinsic (leucocytes) or intrinsic (sperm neck cytoplasm in immature or morphologically abnormal mitochondria), act to deteriorate sperm motility (Guthrie et al., 2008), sperm membranes – through lipid peroxidation (LPO), and DNA integrity – via breakage and cross linking of the chromatin (Aitken and West, 1990; Koppers et al., 2008). ROS levels are very variable, making their proper determination difficult, though this is possible using the probe hydroethidine. Plate 69 shows such an FC mapping of sperm ROS obtained by loading the sperm with hydroethidine and Hoechst 33258. A useful alternative, which measures oxidative stress and ROS effects indirectly, is to determine the levels of LPO in the membrane lipid bilayer by using the 5-iodoacetamidofluorescein probe family (BODIPY-C₁₁^®) (Aitken et al., 2007; Guthrie and Welch, 2007; Ortega Ferrusola et al., 2009a). A representative scanning in stallion spermatozoa is shown in Plate 70a-d.

Mammalian spermatozoa have tightly compacted eukaryotic DNA, established by transformations during spermiogenesis in which the sperm chromatin replaces histones, first by transient proteins and then by protamines (Oliva and Castillo, 2011). The DNA strands are highly condensed by these protamines and form the basic packaging unit of sperm chromatin: the toroid. Toroids are even further compacted by intramolecular and intermolecular disulfide cross links between protamine cysteine residues (Schulte et al., 2010). There are differences in compaction between species; those with less compacted chromatin (equines or canines, for instance) are theoretically more prone to DNA damage. Protamine deficiency in loosely packed chromatin can be monitored by the fluorochrome chromocanmycin-A3 (CMA3) (Tavalaee et al., 2010).

Sperm chromatin can show different abnormalities related to compaction: from damage to the actual DNA physical integrity – as single- or double-stranded DNA strand breaks, nuclear protein defects interfering with histone or protamine conversion and DNA compaction – to chromatin structural abnormalities such as defective tertiary chromatin configuration. The latter scenario implies that defects will occur in the decondensation of the nucleus, thus impairing fertilization, whereas the other defects can jeopardize embryonic development as the oocyte (albeit being able to repair a limited amount of sperm DNA damage) would not be able to correct such damage (Johnson et al., 2011). Sperm DNA disorders also include mutations, epigenetic modifications, base oxidation and DNA fragmentation (which is also related to sperm handling). In accord with knowledge of its increasing relevance, the evaluation of DNA integrity has increased over recent years (Barratt et al., 2010).

Several methods are used to determine DNA damage: (i) the single-cell gel electrophoresis assay (COMET); (ii) the terminal deoxynucleotidyl transferase-mediated fluoresceindUTP nick-end labelling (TUNEL) assay; (iii) the acridine orange test (AOT); (iv) the tritium-labelled 3H-actinomycin D (3H-AMD) incorporation assay; (v) the in situ nick translation (ISTN) assay; (vi) the DNA breakage detection-fluorescence in situ hybridisations (DBD-FISH) assay; (vii) the Sperm Chromatin Dispersion test (SCD Halo); or (viii) the evaluation of the degree of induced denaturation of the DNA (the so-called ‘sperm chromatin structure assay’, SCSA) (Fraser, 2004; Evenson and Wixon 2006; Tamburrino et al., 2012). Examples of spermatozoa examined with some of these techniques are illustrated in Plate 71a–d. Most of the above methods can use fluorescence microscopy, but SCSA and TUNEL are usually explored via FC.

Although SCSA has been extensively used, the outcome has provided conflicting relationships with fertility (Rodriguez-Martinez and Barth, 2007; Christensen et al., 2011). The conflict has arisen from a single study relating the percentage of spermatozoa with denatured DNA (through an index of DNA fragmentation, DFI), to high, moderate or very low fertility values in humans, when DFI thresholds of ~0–15%, 16–29% and >30%, respectively, were recorded as in direct relation to the pregnancy outcome using ART (Evenson and Jost, 2000). Such relationships have been questioned (Collins et al., 2008). A threshold for sub-fertility in bulls and boars has also been calculated (as percentage DFI), albeit using semen from young, unproven bulls with low fertility (40–60% of 90 day NRRs) and boar semen without any fertility recordings (Rybar et al., 2004). These calculated threshold values (20% for bulls and 18% for boars) were much lower than those obtained with human semen. The lack of reliable fertility data when calculating these limits also leaves these values highly questionable, especially in the light of other, wellcontrolled studies of AI-sires with proven and varied fertility, in which neither proven stud bulls nor stud boars in AI-breeding programmes reached these high values. For instance, sperm DFI values (even as a percentage of cells with high COMP a_t values (i.e. cells outside the main population with a high proportion of single-stranded DNA over single + double-stranded DNA) in breeding sires are low (<5%) (Boe-Hansen et al., 2005; De Ambrogi et al., 2006; Christensen et al., 2011), even in frozen–thawed semen, where the processing ought to facilitate DNA damage (Januskauskas et al., 2001, 2003; MartinezPastor et al., 2004; Hallap et al., 2005a; Hernández et al., 2006; Chistensen et al., 2011). Conversely, equine sires unselected for sperm quality had higher DFIs and these lacked a clear relationship with equine fertility (Morrell et al., 2008).

SCSA does not specifically identify the amount of DNA damage, unlike TUNEL, but rather its susceptibility to harsh treatment. The TUNEL method is more cumbersome though, and this has led to modifications of the method. For example, a TUNEL/PI procedure is now available that combines the accuracy of TUNEL and the differentiation of two sperm populations depending on PI intensity, of which one is probably participating in fertilization as the damage has no relationship with either sperm motility or morphology (Muratori et al., 2008). Alternatively, the use of dithiothreitol (DTT) to decondense sperm nuclei and inclusion of a stain for dead cells provides a higher accessibility to the TdT (terminal deoxynucleotidyl transferase) enzyme of TUNEL, in combination with the detection of DNA fragmentation in live spermatozoa (Mitchell et al., 2011). Taking all the above considerations into account, TUNEL appears to be a more sensitive method to predict infertility than SCSA, as determined in a recent meta-analysis (Zini et al., 2008).

Functional Assays, Higher Discriminatory Value?

The structural integrity attributes of the sperm that have already been described are a prerequisite for sperm function, provided it is proven that they work in a concerted manner. For instance, it is logical to consider that the determination of membrane integrity in ejaculated spermatozoa is linked with sperm motility, as well as with the capacity of viable spermatozoa to undergo sperm capacitation and change their activated motility to a hyperactivated pattern, as occurs in vivo (Rodriguez-Martinez, 2007a). Sperm capacitation is, moreover, a stepwise process that includes the removal of epididymal and SP-adsorbed proteins, loss of membrane cholesterol, an increase in membrane fluidity due to lipid modifications, an influx of Ca²⁺ to the sperm perinuclear and neck regions and flagellum, the generation of controlled amounts of ROS, as well as the phosphorylation of protein residues (Gadella and van Gestel, 2004; Harrison and Gadella, 2005; O’Flaherty et al., 2006; Tulsiani et al., 2007; Fabrega et al., 2011). These are steps that can be measured in vitro and, ultimately, associated with the fertility of the males.

The earliest stages of sperm capacitation, those related to bicarbonate and Ca²⁺ triggered phospholipid scrambling, can be mapped using Merocyanine-540 or Annexin-V, as these measure general or specific destabilization of the plasmalemma and both are related to the fertility of the semen assayed, either fresh (Harrison, 1997) or frozen–thawed (Peña et al., 2004, 2007; Januskauskas et al., 2005; Hallap et al., 2006b), when a bicarbonate challenge has been used (Bergqvist et al., 2006; Saravia et al., 2007a).

Calcium influx elicits tyrosine phosphorylation, and apparently initiates sperm capacitation and hyperactivation (Suarez, 2008a,b). Therefore, mapping of low intracellular Ca²⁺ levels in spermatozoa should be a good parameter for discriminating between semen samples (or even individual boars) (Hossain et al., 2011b; Waberski et al., 2011). Ca²⁺ influx and displacement within the spermatozoon have been followed throughout capacitation using Indo-1 acetoxymethylester or Fluo-loaded spermatozoa (Hossain et al., 2011a,b). Most changes in relation to capacitation are registered during the later stages of this process, and can be indirectly visualized by the incubation of viable spermatozoa with the fluorescent antibiotic chlortetracycline (CTC), which monitors the displacement of Ca²⁺ in the sperm head, in preparation for the occurrence of the acrosome reaction (AR) (Rodriguez-Martinez, 2007a). In AI bulls with known fertility, the proportion of noncapacitated spermatozoa (CTC stable, unreacted) in AI semen samples was significantly related to fertility (Thundathil et al., 1999; Gil et al., 2000). The capacitation of spermatozoa, following Ca²⁺ and bicarbonate signalling, activates adenylate cyclase and the resulting cAMP in turn activates intracellular protein kinases that enrich the equatorial subsegment in tyrosine-phosphorylated proteins (Piehler et al., 2006; Jones et al., 2008). There is now an anti-phosphotyrosine antibody, and FC can estimate the amount of global protein residue phosphorylation in the sperm plasma membrane, thereby providing evidence of the real-time proportion of spermatozoa that might have engaged in the process of capacitation (usually 5–10% of all cells at a given time) (Sidhu et al., 2004).

Spermatozoa have also been explored for their ability to bind to the epithelium of the oviduct in vitro, an ability present in vivo, when potentially fertile spermatozoa reside in the oviductal sperm reservoir (Rodriguez-Martinez et al., 2005). The binding of spermatozoa to homologous oviductal epithelial cells in vitro has been found to prolong their life (Lefebvre and Suarez, 1996), thus indicating that sperm co-culture with oviductal explants might mimic the capacity of a semen sample to colonize the tubal reservoir and would, therefore, indicate its potential fertilizing capacity, because uncapacitated spermatozoa are preferentially bound (Fazeli et al., 1999). It should be noted here that the outcomes of such tests only provide reliable correlations when high-quality sperm samples are tested (De Pauw et al., 2002), and have shown only marginal relationships with fertility (Waberski et al., 2005, 2006).

Can we mimic sperm selection in the oviduct? Probably so, but only to some extent, because we still do not know if or how the female modulates all of the steps involved behind the restrictions in vivo. We do know, however, that ejaculated spermatozoa can be separated in vitro into sub-populations by differences in motility kinetics (Abaigar et al., 1999; Cremades et al., 2005; Peña et al., 2005b, 2006; Holt et al., 2007; Martinez-Pastor et al., 2011), membrane integrity (Perez-Llano et al., 2003), head morphometry (Núñez-Martinez et al., 2005; Peña et al. 2005b, 2006), sustainment of handling and cryopreservation (Gil et al., 2005; Peña et al., 2006; Roca et al., 2006), and the ability to respond to specific stimuli such as bicarbonate exposure (Holt and Harrison, 2002; Gadella and van Gestel, 2004; Holt and Van Look, 2004; Satake et al., 2006), or to specific proteins isolated from the oviduct (Coy et al., 2009). Several of these studies were based on individual sperm attributes, and some considered the fluids surrounding the spermatozoa, their composition and even their rheological features, with viscosity being among these. The ultimate goal would be to find a method that, by mimicking what occurs within the female genital tract, could preselect spermatozoa with primary competence for fertilization (Rodriguez-Martinez, 2006).

Methods for the in vitro