Human Reproduction Update Advance Access originally published online on March 31, 2006
Human Reproduction Update 2006 12(4):417-435; doi:10.1093/humupd/dml009
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Protamines and male infertility
Human Genetics Laboratory, Genetics Unit, Department of Ciències Fisiològiques I, Faculty of Medicine, University of Barcelona and Hospital Clínic, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain
To whom correspondence should be addressed at: Human Genetics Laboratory, Department of Ciencies Fisiològiques I, University of Barcelona, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Casanova 143, 08036 Barcelona, Spain. E-mail: roliva{at}ub.edu
Submitted on November 23, 2005; resubmitted on February 8, 2006; accepted on February 15, 2006
 |
Abstract |
|---|
Abstract
IntroductionProtamines are the major nuclear sperm proteins. The human sperm nucleus contains two types of protamine: protamine 1 (P1) encoded by a single-copy gene and the family of protamine 2 (P2) proteins (P2, P3 and P4), all also encoded by a single gene that is transcribed and translated into a precursor protein. The protamines were discovered more than a century ago, but their function is not yet fully understood. In fact, different hypotheses have been proposed: condensation of the sperm nucleus into a compact hydrodynamic shape, protection of the genetic message delivered by the spermatozoa, involvement in the processes maintaining the integrity and repair of DNA during or after the nucleohistone–nucleoprotamine transition and involvement in the epigenetic imprinting of the spermatozoa. Protamines are also one of the most variable proteins found in nature, with data supporting a positive Darwinian selection. Changes in the expression of P1 and P2 protamines have been found to be associated with infertility in man. Mutations in the protamine genes have also been found in some infertile patients. Transgenic mice defective in the expression of protamines also present several structural defects in the sperm nucleus and have variable degrees of infertility. There is also evidence that altered levels of protamines may result in an increased susceptibility to injury in the spermatozoan DNA causing infertility or poor outcomes in assisted reproduction. The present work reviews the articles published to date on the relationship between protamines and infertility.
Key words: chromatin / genome / mutation / protamine / spermatozoa
|
Introduction |
|---|
Protamines and DNA were isolated and discovered from the sperm more than a century ago by Friedrich Miescher (Miescher, 1874
; Kossel, 1928; Felix, 1960; Dixon and Smith, 1968; Dahm, 2005). They are the most abundant sperm nuclear proteins in many species and act by packaging the paternal genome (Bloch, 1969; Ando et al., 1973; Calvin, 1976; Mezquita and Teng, 1977, 1978; Subirana, 1983; Oliva and Dixon, 1991a; Aoki and Carrell, 2003; Lewis et al., 2003a). They are proteins with a high content of positively charged amino acids, particularly arginine (48% in human protamines; Figure 1).
|
In mammals, two types of protamines are known: the P1 protamine and the family of P2 proteins. The P1 protamine is present in all species of vertebrates studied (McKay et al., 1985, 1986; Gusse et al., 1986; Balhorn et al., 1987; Bellvé et al., 1988; Oliva and Dixon, 1991a; Chauvière et al., 1992; Yoshii et al., 2005). Protamine P2 is formed by the P2, P3 and P4 components, and it is only present in some mammalian species including human and mouse (Balhorn et al., 1977, 1987; McKay et al., 1985, 1986; Gusse et al., 1986; Bélaïche et al., 1987; Bower et al., 1987; Bellvé et al., 1988; Oliva and Dixon, 1991a; Yoshii et al., 2005).
Several functions have been proposed for the protamines (reviewed by Oliva and Dixon, 1991a). The most obvious would be:
- Generation of a condensed paternal genome with a more compact and hydrodynamic nucleus. The spermatozoa with the most hydrodynamic nucleus would move faster, being able to fertilize the oocyte first. Therefore, the most condensed and hydrodynamic sperm would transmit the advantageous trait to future generations through a marked Darwinian selection.
- Protecting the paternal genetic message delivered by the spermatozoa through making it inaccessible to nucleases or mutagens potentially present in the internal or in the external media. This hypothesis could be supported by recent observations in assisted reproduction linking defects in protamination with injured spermatozoal DNA, compatible with fertilization of the oocyte but precluding subsequent embryo development. However, of relevance to the understanding of the mechanisms leading to infertility, the presence of protamines may also recruit and/or potentiate the effect of certain toxins or heavy metals in the testis or spermatozoa.
- Competition and removal of transcription factors and other proteins from the spermatid resulting in a blank paternal genetic message, devoid of epigenetic information, therefore allowing its reprogramming by the oocyte.
- Involvement in the imprinting of the paternal genome during spermatogenesis. Also protamines themselves could confer an epigenetic mark on some regions of the sperm genome, affecting its reactivation upon fertilization.
In addition to the above potential functions, it has also been proposed that (v) protamines could be part of a checkpoint during spermiogenesis and (vi) they could have a role in the fertilized ova.
The present review focuses on the available evidence between protamines and male infertility. Thus, it complements and updates more extensive previous reviews on the nucleohistone–nucleoprotamine transition (Mezquita, 1985; Poccia, 1986; Ward and Coffey, 1991; Oliva and Dixon, 1991a; Dadoune, 1995, 2003; Wouters-Tyrou et al., 1998; Raukas and Mikelsaar, 1999; Braun, 2001; Aoki and Carrell, 2003; Meistrich et al., 2003; Kierszenbaum and Tres, 2004; Hogarth et al., 2005). The role of histones (His), histone modifications, remodelling factors and epigenetic changes during spermatogenesis have also been elegantly reviewed by different groups (Sassone-Corsi, 2002; Lewis et al., 2003b; Govin et al., 2004; Caron et al., 2005; Horsthemke and Ludwig, 2005; Kimmins and Sassone-Corsi, 2005; Morgan et al., 2005; Rousseaux et al., 2005). To cover the subject of the DNA-repair mechanisms, oxidative stress and sperm DNA integrity and male infertility, the reader is referred to recent articles and reviews (McPherson and Longo, 1993; Aitken and Krausz, 2001; Baarends et al., 2001; Kierszenbaum, 2001; Oehninger et al., 2003; Sakkas et al., 2003; O’Brien and Zini, 2005; Seli and Sakkas, 2005; Silva and Gadella, 2005; Erenpreiss et al., 2006; Muratori et al., 2006).
The following section of this review has been included to provide a brief synthetic summary of the protamine genes, their evolution, expression and involvement in the nucleohistone–nucleoprotamine transition. This initial section is not comprehensive but has been included to focus the subject and to facilitate reading of the rest of the review. In contrast, the rest of the review is intended to be comprehensive, for all articles published to date concerning protamines and infertility in man. The articles considered for inclusion were selected from the results of Medline and Journal Citation Report (ISI Web of Knowledge) searches with the keyword ‘protamine’ alone or combined with other keywords (‘infertility’, ‘human’ + ‘sperm’ and ‘human’ + ‘testis’).
|
Summary of protamine structure and function |
|---|
Evolution of the protamines
Protamines are proteins that have increased the number of positively charged residues in evolution allowing the formation of a highly condensed complex with the paternal genomic DNA, which has a strong negative charge (Oliva and Dixon, 1990, 1991a; Retief et al., 1993; Oliva, 1995; Queralt et al., 1995; Lewis et al., 2003a). In addition, protamines of different species incorporate cysteines (Cys) in their sequence allowing the formation of disulphide bonds between adjacent protamine molecules, therefore strongly stabilizing the nucleoprotamine complex (Saowaros and Panyim, 1979; Balhorn et al., 1992; Lewis et al., 2003a; Vilfan et al., 2004). Evidence already exists that protamines may have evolved from histone H1 ancestors (Ausió, 1999; Lewis et al., 2004; Eirin-Lopez et al., 2006). Another characteristic of the protamines is that they are among those proteins with one of the highest rates of evolutionary variation (Oliva and Dixon, 1991a; Oliva, 1995; Lewis et al., 2003a). It has been proposed that one cause of this rapid evolution rate could be a positive Darwinian selection (Rooney and Zhang, 1999; Clark and Civetta, 2000; Wyckoff et al., 2000). This proposal is supported by the observation, when comparing the sequence of protamines from different species, that the ratio of non-synonymous substitutions (the nucleotide changes resulting in a change of amino acid) per residue to synonymous substitutions is greater than 1 and also that the protamine exons evolve faster than the protamine intron (Rooney and Zhang, 1999; Wyckoff et al., 2000). However, a closer examination revealed an unusual form of purifying selection, where the overall number of arginine residues is maintained at about 50% in mammals, but the total number of amino acids and the positions of the arginine residues have changed considerably (Rooney et al., 2000). It has been proposed that the driving forces for this arginine-rich selection could be (i) the DNA-binding function of the protamine P1 resulting in a more compact sperm nucleus and (ii) the interaction and strong activation of oocyte creatine kinase II by protamine (Ohtsuki et al., 1996; Rooney and Zhang, 1999). While the evolution of protamines is providing important clues towards the understanding of their function, this aspect is not covered further here, so the reader is referred to other reviews and articles for a more in-depth analysis of this topic (Oliva and Dixon, 1991a; Ausió, 1999; Clark and Civetta, 2000; Wyckoff et al., 2000; Torgerson et al., 2002; Lewis et al., 2003a; Eirin-Lopez et al., 2005).
Genomic organization and transcription of the protamine genes
Humans have one copy of the protamine 1 gene (PRM1) and one copy of the protamine 2 gene (PRM2) per haploid genome, located on chromosome 16 (Figure 1; Krawetz et al., 1989; Reeves et al., 1989; Domenjoud et al., 1990; Oliva and Dixon, 1990, 1991a; Engel et al., 1992; Nelson and Krawetz, 1993, 1994; Queralt et al., 1993; Schlüter et al., 1996). Both genes contain a single intron (Figure 1). The genomic sequences of the PRM1 and PRM2 genes are organized in the form of a loop domain together with the transition protein 2 gene (TNP2) and a sequence called gene4 (Figure 1; Engel et al., 1992; Choudhary et al., 1995; Schlüter and Engel, 1995; Schlüter et al., 1996; Kramer and Krawetz, 1998; Wykes and Krawetz, 2003; Martins et al., 2004). This spatial organization may allow a co-ordinated expression of these genes during spermiogenesis. However, while the protamine (PRM1 and PRM2) and transition protein (TNP2) genes are expressed at high levels and their function has been extensively studied, the potential role of gene4 is more controversial and is expressed at very low levels, if at all, in humans (Schlüter and Engel, 1995; Schlüter et al., 1996; Kramer and Krawetz, 1998). Further studies should clarify whether or not gene4 is a pseudogene in humans. The gene4 sequence has also been called protamine 3 (Prm3; or gene4/Prm3), based on some evidence that it may have originated by duplication of the PRM1 gene (Schlüter et al., 1996; Kramer and Krawetz, 1998). However, the name Prm3 is misleading since its predicted amino-acid sequence is not at all related to protamines, as it lacks arginine clusters and, instead, is rich in glutamic acid. Therefore, gene4/Prm3 is not likely to bind DNA and should not be called protamine.
The positioning of nucleosomes in the protamine 1 gene has been assessed in vivo and in vitro using the rat as a model (Adroer and Oliva, 1998). The identification of regulatory elements and the expression of the protamine genes have been studied using a variety of approaches including homology comparisons, transgenic or knockout mice and different in vivo and in vitro approaches (Tamura et al., 1992; Queralt and Oliva, 1993, 1995; Zambrowicz et al., 1993; Nelson and Krawetz, 1994; Choi et al., 1997; Stewart et al., 1999; Giorgini et al., 2001; Hummelke and Cooney, 2004; Aleem et al., 2005). For further information on this subject, the reader is referred to excellent reviews and articles on the transcriptional, molecular and cellular mechanisms in spermatogenesis (Iatrou and Dixon, 1978; Mezquita, 1985; Hecht, 1988, 1993; Perreault, 1992; Braun et al., 1995; Dadoune, 1995, 2003; Kramer and Krawetz, 1997; Siffroi et al., 1999; Steger, 1999, 2001; Steger et al., 2000, 2002; Grootegoed et al., 2000; Aoki and Carrell, 2003; Hebbar and Archer, 2003; Kleene, 2003; Dadoune et al., 2004; Kierszenbaum and Tres, 2004; Kimmins et al., 2004; Rockett et al., 2004; Krawetz, 2005; Miller et al., 2005; Tanaka and Baba, 2005). Despite substantial knowledge available on the fundamental aspects of the transcriptional mechanisms, so far there have been relatively few studies assessing the potential involvement of changes in protamine gene transcription factors in human male infertility (Sassone-Corsi, 2002; Blocher et al., 2003; Kimmins et al., 2004; Krausz and Sassone-Corsi, 2005). Because of the extensive evidence for deregulation of protamine expression in male infertility, this issue would deserve further attention in the future.
The protamine P1 is synthesized as a mature protein, whereas the components of the P2 family are generated by proteolysis from a precursor encoded by a single gene (Figure 1A and B; Mckay et al., 1986; Yelick et al. 1987; Sautière et al., 1988; Chauvière et al., 1992; Green et al., 1994; Queralt et al., 1995; Wouters-Tyrou et al., 1998). Members of the P2 family differ only by the N-terminal extension of 1–4 residues, although the P2 component is the most abundant (Figure 2; Gusse et al., 1986; McKay et al., 1986; Sautière et al., 1988; Martinage et al., 1990; Arkhis et al., 1991; Oliva and Dixon, 1991a; Bianchi et al., 1992; Alimi et al., 1993; Yoshii et al., 2005). The content of protamine P1 in the human sperm nucleus is similar to the content of protamine P2 (P1/P2 ratio of approximately 1; Balhorn et al., 1988; de Yebra et al., 1993; Bench et al., 1996; Corzett et al., 2002; Mengual et al., 2003a; Aoki et al., 2005a). However, despite this, their functions may differ. Arguments in favour of the hypothesis of a different function for P1 and P2 protamines could be that (i) unlike P1 protamine, P2 protamines are zinc-finger proteins with one Cys2–His2 motif (Bianchi et al., 1992), (ii) P2 proteins are expressed only in some mammals whereas P1 is invariably present in all mammals, indicating a more basic and conserved function for P1 and an accessory function for P2 protamines in some species and (iii) alterations of P1 or P2 protamines in infertile patients impact differently on the integrity of the DNA and in the assisted reproduction outcome (Aoki et al., 2005b).
|
Both protamines will undergo post-transcriptional modifications before binding to the DNA and generating the highly compact nucleoprotamine complex.
The nucleohistone–nucleoprotamine transition
In the final stage of spermatogenesis, the nucleosomal structure is progressively disassembled, then replaced by TNPs and finally by protamines (Figure 2; reviewed by Mezquita, 1985; Poccia, 1986; Oliva and Dixon, 1991a; Hecht, 1993; Green et al., 1994; Dadoune, 1995; Grootegoed et al., 2000; Meistrich et al., 2003; Kierszenbaum and Tres, 2004; Rousseaux et al., 2005). This transition is preceded by extremely marked changes in many chromatin activities (Puwaravutipanich and Panyim, 1975; Oliva et al., 1982; Mezquita, 1985; Oliva and Dixon, 1991a; Dadoune, 1995, 2003; Wouters-Tyrou et al., 1998; Fuentes-Mascorro et al., 2000; Braun, 2001; Govin et al., 2004; Kierszenbaum and Tres, 2004). One of the initial chromatin changes is the incorporation of histone variants (Prigent et al., 1996, 1998; reviewed by Churikov et al., 2004; Govin et al., 2004; Loppin et al., 2005; Tanaka et al., 2005). Another important early event is histone hyperacetylation that occurs during spermiogenesis before the nucleosome disassembly in vivo (Candido and Dixon, 1972; Oliva and Mezquita, 1982; Grimes and Henderson, 1984; Meistrich et al., 1992; Hazzouri et al., 2000; Marcon and Boissonneault, 2004). It was postulated that histone hyperacetylation and rapid turnover of acetyl groups could rapidly and reversibly expose binding sites in chromatin for subsequent binding of chromosomal proteins (Oliva and Mezquita, 1982). More recently, it was also shown in vitro that histone hyperacetylation facilitated nucleosome disassembly and histone displacement by protamines (Oliva and Mezquita, 1986; Oliva et al., 1987). Also, hyperacetylated nucleosomes were shown to appear in a more relaxed structure upon binding to electron microscopy grids (Oliva et al., 1990). It has been shown that the testis-specific bromodomain-containing protein (BRDT) binds to hyperacetylated histone 4 (H4) triggering a reorganization of the chromatin (Pivot-Pajot et al., 2003). Impaired H4 hyperacetylation has been detected in infertile patients (Sonnack et al., 2002; Faure et al., 2003).
Concomitant with nucleosome disassembly, the sperm DNA is extensively complexed with TNPs (Figure 2; Kierszenbaum, 2001; Meistrich et al., 2003). Transition proteins are then finally replaced by protamines to form a highly compact nucleoprotamine complex (Figure 2). It is known that protamines are phosphorylated before binding to DNA and that a substantial dephosphorylation takes place concomitant with nucleoprotamine maturation (Ingles and Dixon, 1967; Marushige and Marushige, 1978; Oliva and Dixon, 1991a; Papoutsopoulou et al., 1999). The dynamics of protamine binding to DNA have also been studied (Prieto et al., 1997; Brewer et al., 1999, 2003). After binding to the DNA, the formation of disulphide bonds between protamines further stabilizes the nucleoprotamine complex (Balhorn et al., 1992). Different models for the structure of the nucleoprotamine have been proposed (Balhorn, 1982; Allen et al., 1993, 1997; Hud et al., 1993; Raukas and Mikelsaar, 1999; Vilfan et al., 2004; Biegeleisen, 2006). However, despite the substantial amount of information available, our understanding of the molecular mechanisms governing the nucleohistone–nucleoprotamine transition is still in its infancy. For example, little information is available on what other proteins or structures interact with protamines and what their function is (Kierszenbaum and Tres, 2004; Mylonis et al., 2004).
Organization of DNA in the sperm nucleus
It is important to note that not all of the DNA in the sperm nucleus is organized into a nucleoprotamine structure, but some regions retain a nucleosomal structure (Figure 2). It has been shown that approximately 85% of the DNA in the sperm nucleus is associated with protamines and that 15% remains associated with histones or other proteins (Figure 2; Tanphaichitr et al., 1978; Ammer et al., 1986; Gusse et al., 1986; Gatewood et al., 1987, 1990; de Yebra et al., 1993; Zalensky et al., 2002). In addition, human sperm DNA has a heterogeneous structure with some regions and genes remaining associated with histones or with other proteins (Zalensky et al., 1995, 2002; Gardiner-Garden et al., 1998; Kramer et al., 2000; Zalenskaya et al., 2000; Zalenskaya and Zalensky, 2002; Wykes and Krawetz, 2003). It will be interesting to determine how these heterogeneous structures in the sperm nucleus relate to the establishment of epigenetic information in the male gamete and how they may affect subsequent embryo development (Rousseaux et al., 2005). The spatial architecture of chromosomal DNA has also been studied with data, supporting that the centromeres are organized in a chromocentre, positioned well inside the nucleus, whereas the telomeres forming dimers are positioned in the nuclear periphery (Zalensky et al., 1995; Solov’eva et al., 2004). For further information on the nucleohistone–nucleoprotamine transition, the reader is referred to different reviews (Ward and Coffey, 1991; Oliva and Dixon, 1991a; Dadoune, 1995, 2003; Wouters-Tyrou et al., 1998; Raukas and Mikelsaar, 1999; Braun, 2001; Aoki and Carrell, 2003; Meistrich et al., 2003; Kierszenbaum and Tres, 2004; Rousseaux et al., 2005). The extent to which the structural organization of the sperm DNA is altered in infertile patients remains relatively unexplored.
After fertilization, the highly packaged nucleoprotamine sperm genome must be decondensed (Figure 2). One of the first steps must be reduction of the protamine disulphide bonds to allow protamine removal and subsequent organization of the DNA in a nucleosomal structure (Figure 2). The chromatin changes and unpacking after fertilization potentially relevant to the function of protamines are reviewed elsewhere (Griveau et al., 1992; Perreault, 1992; Poccia and Collas, 1996; Colleu et al., 1997; Shimada et al., 2000, 2002; Braun, 2001; Esterhuizen et al., 2002; Nakazawa et al., 2002; Schultz, 2002; Lefievre et al., 2003; McLay and Clarke, 2003; Mudrak et al., 2005; Romanato et al., 2005). It is possible that differential marking of different sperm genomic DNA regions with P1 or P2 protamines or with histones, histone variants or with other proteins could contribute, after fertilization, to establish the order of paternal gene reactivation or even could be involved in setting up the appropriate imprinting of different paternal genes.
|
Transgenes and knockout models |
|---|
A great deal of information relevant to the function and involvement of protamines in male infertility has been obtained from transgenes and knockout models for protamines and TNPs. The first transgenic model for a protamine corresponded to the homologous mouse protamine 1 gene (Peschon et al., 1987
). The resulting mice correctly expressed the protamine construction in round spermatids, indicating the good recognition of the regulatory elements in the transgene (Peschon et al., 1987; Zambrowicz et al., 1993). In addition, the mice were fertile indicating that small variations in the levels of expression of the protamine 1 were compatible with an apparently normal function of the spermatozoa (Peschon et al., 1987). Lines of transgenes, generated using the promoter of the mouse protamine 2 gene coupled to a reporter gene, also supported the idea that the regulatory elements were correctly recognized by the endogenous factors resulting in the correct expression at round spermatid stage (Stewart et al., 1988). Subsequent models designed to express the protamine 1 gene prematurely or in excess resulted in premature condensation of the nuclear chromatin, anomalies in the morphology of the sperm head and incomplete processing of protamine P2 (Peschon et al., 1989; Lee et al., 1995). The first heterologous expression of a protamine corresponded to over-expression of the chicken protamine gene in transgenic mice, which resulted in a disruption of the chromatin in spermatozoa (Oliva and Dixon, 1989; Rhim et al., 1995). Unexpectedly, however, these mice turned out to be fertile, suggesting that a very precise packaging of the DNA in the germinal cell line was not essential for decondensation and pronuclear formation in the fertilized oocytes (Rhim et al., 1995). Subsequent studies characterized in detail the presence of different anomalies in the spermatozoa of these transgenic mice, also confirming their relative fertility (Maleszewski et al., 1998). In a different type of experiment, over-expression of the human protamine cluster in transgenic mice demonstrated a conservation of the temporal expression pattern, indicating that the human regulatory elements were recognized by the mouse transcription factors (Stewart et al., 1999). Also, transgenic mice containing the complete human protamine domain flanked by different configurations of nuclear matrix attachment regions (MARs) demonstrated the importance of the overall chromatin structure for correct expression and function of the domain (Martins et al., 2004).
Of importance, it was found that knockout mice for only one of the P1 or P2 alleles were sufficient to result in infertility (Braun et al., 1989; Oliva and Dixon, 1991a; Cho et al., 2001). Since protamines are expressed in the haploid phases of spermatogenesis (Hecht, 1988; Oliva et al., 1988; Choudhary et al., 1995; Steger 1999, 2001; Dadoune et al., 2004), it could be thought that the disruption of only one allele should not affect the expression of the protamine gene in the other half of cells having the normal gene. But it is also known that cytokinesis is incomplete in the spermatogenic cells, which are connected by cytoplasmic bridges that can allow spermatids to share mRNA (Braun et al., 1989; Oliva and Dixon, 1991b). A few years later, the presence of damaged DNA in sperm cells of these knockout infertile mice was detected (Cho et al., 2003). Of relevance, these authors also observed that, if ICSI was used, it was possible to activate the oocytes but that few could progress to the blastocyst stage (Cho et al., 2003). It is also important to note that a similar phenomenon has been described in many infertile patients, with injured DNA under ICSI treatment (Tesarik et al., 2004; Greco et al., 2005).
Another extensively studied model is the knockout mouse for TNP1 or TNP2 (Yu et al., 2000; Adham et al., 2001; Zhao et al., 2001, 2004a,b; Meistrich et al., 2003; Shirley et al., 2004; Suganuma et al., 2005). In the double-knockout mice (for both TNPs), the remodelling of nuclear morphology, the repression of transcription, the disappearance of histones and the deposition of protamines were relatively normal. However, it was observed that condensation of the chromatin was irregular, that protamine P2 was not processed and that many of the elongated spermatids had DNA breaks (Zhao et al., 2004a). Interestingly, it has been found that there is an increase in structural anomalies in these mice, as revealed by acridine orange (AO) staining, during epididymal passage and that fertility declines, as revealed by ICSI (Suganuma et al., 2005).
|
Alterations in protamine content of spermatozoa in infertile patients |
|---|
Direct determination of protamines by electrophoresis
The first evidence of anomalies in the protamine content of spermatozoa was described in a study, which did not detect protamines, but did detect histones, in the spermatozoon of diverse infertile patients (Table I; Silvestroni et al., 1976). Subsequently, an independent group described an anomalous protein pattern in different patients, which was characterized by the presence of additional proteins (Chevaillier et al., 1987). However, in this work no reference was made to the protamines. One of the first complete studies that analysed the protamines in a series of fertile controls (n = 17) and compared the data with that of patients (n = 7) detected an increased P1/P2 ratio in six of the seven patients studied (Balhorn et al., 1988). A more heterogeneous protamine fraction was also observed in patients with altered seminal parameters as compared with samples with normal parameters (Lescoat et al., 1988). Subsequently, it was found that the percentage of protamines in fertile men was the same as that in infertile patients with normal seminal parameters, but that it varied in the patients with abnormal seminal parameters (Bach et al., 1990). Another independent group found that in patients with morphologic anomalies in the spermatozoa, characterized by the presence of a round head, the spermatozoa contained less protamines and more histones than normal spermatozoa (Blanchard et al., 1990).
|
The decrease in protamine P2 level and the increased P1/P2 ratio were confirmed a few years later (Belokopytova et al., 1993). But it was not until a report of the first extended series of patients (n = 116) that it was recognized that an important proportion of the patients (3.4%; n = 4) had a marked reduction in protamine P2 (de Yebra et al., 1993; de Yebra and Oliva, 1993), whereas the rest of the patients had a normal P1/P2 ratio (22.4%) or a slightly altered ratio (74.1%). In addition, it was noticed that a large proportion of the samples with an altered P1/P2 ratio also had increased levels of proteins with a mobility similar to histones and to intermediate proteins (de Yebra et al., 1993). More recently, an increase in histone H2B in infertile patients has been confirmed using immunocytochemistry (Zhang et al., 2006).
All these observations raised the question of the origin of the reduction of protamine P2 levels relative to those of protamine P1 in some of the patients. The detection of increased protamine P2 precursors in patients with an increased P1/P2 ratio narrowed the possible origin to an abnormal processing of the protamine P2 precursor (de Yebra et al., 1998). It should be noted that detectable levels of P2 precursors are also present in the mature sperm nucleus in the mouse and rat (Stanker et al., 1992; Debarle et al., 1995). This reduction in protamine content in patients was consistent with results of the analysis of the phosphorus and sulphur contents in individual spermatozoa by particle-induced X-ray emission (PIXE; Bench et al., 1998). In addition, the protamine P1/P2 ratio varied in samples taken from the same patients at different times (Bench et al., 1998). Another explanation for the altered P1/P2 ratio detected in different infertile patients is that it could be the consequence of a general failure in the replacement of histones by protamines during spermiogenesis. The detection of increased amounts of histones and intermediate proteins in patients with decreased protamines or altered P1/P2 ratio would support this hypothesis (Blanchard et al., 1990; de Yebra et al., 1993; Zhang et al., 2006).
All these initial works were carried out by analysing the semen samples without fractionation. It is well known that, even in a normal human ejaculate, populations of abnormal spermatozoa coexist with morphologically normal spermatozoa. Therefore, it was considered if the anomalies detected in the P1/P2 ratio affected all the cells in the sample or, instead, reflected a mixture of a normal population plus a population with an altered P1/P2 ratio. Percoll gradient centrifugation allowed separation of spermatozoa according to morphology and mobility, and fractions with a higher density were shown to be enriched in less-intermediate proteins and contain more mature protamine 2 (Colleu et al., 1996). However, the separation of cells in individual ejaculates from infertile patients and controls using a Percoll gradient, and the subsequent determination of the P1/P2 ratio in each of the fractions, detected only small differences in P1/P2 ratio between fractions despite the presence of marked differences in the morphology and mobility (Mengual et al., 2003a). Nevertheless, marked differences in the P1/P2 ratio were detected when comparing oligozoospermic and asthenozoospermic patients to controls (Mengual et al., 2003a). It will be interesting to test other separation methods, such as swim-up (Colleu et al., 1996; Sakkas et al., 2000), electrophoresis (Ainsworth et al., 2005) or cell sorting (Ziyyat et al., 1999), and the use of immunocytochemical methods (Zhang et al., 2006) to test whether levels of the protamines and other proteins do indeed vary among the different cells of an ejaculate and may correlate with DNA integrity or assisted reproduction outcomes.
Radical differences in protamine content in two siblings associated with different ICSI outcomes were also reported (Carrell et al., 1999). A recent article reporting the analysis of 272 infertile patients and 87 donors described a new type of anomaly in some patients, characterized by the presence of a decreased P1/P2 ratio (Aoki et al., 2005a). A summary of all articles measuring protamines directly after extraction and electrophoresis is given in Table I.
In addition to the above studies in infertile patients, the expression of protamines has also been determined in response to thermal stress in normal testicles (Love and Kenney, 1999; Evenson et al., 2000). Thermal stress in stallion testicle is associated with decreased formation of disulphide bridges in protamines (Love and Kenney, 1999). This aspect has also been studied in humans by Evenson et al. (2000), who measured protamine levels in a patient just after an episode of hyperthermia, induced by the influenza, and reported the appearance of protamine P2 precursors, detected by electrophoresis, between 33 and 39 days post-hyperthermia. These authors also showed that the P1/P2 ratio remained within the normal range, whereas the ratio between histones and protamines increased slightly between 33 and 39 days post-hyperthermia. Expression of the gene-encoding protamine P2 was also altered concomitant to induced thermal stress in the mouse testicle (Iuchi et al., 2003).
Indirect assessment of sperm chromatin structure by histochemical procedures
In all of the above studies, the protamine content was measured directly through protamine extraction and polyacrylamide gel electrophoresis (PAGE). Indirect methods of assessing the amount of protamines or measuring chromatin structure based on different staining procedures or fluorochromes have also been used (Bianchi et al., 1996; Lolis et al., 1996; Bizzaro et al., 1998; Sakkas et al., 1998; Franken et al., 1999; Esterhuizen et al., 2002; Zubkova et al., 2005). For example, in situ competition between protamine and chromomycin A3 (CMA3) indicated that CMA3 staining inversely correlated with the protamination state of spermatozoa (Bizzaro et al., 1998). Interestingly, CMA3 staining has been shown to be increased in the sperm cells of infertile patients (Lolis et al., 1996; Franken et al., 1999; Razavi et al., 2003; Nasr-Esfahani et al., 2004a,b, 2005). Correlations between CMA3 staining in sperm and assisted reproduction outcome have also been found (Nasr-Esfahani et al., 2004a, 2005). However, CMA3 staining cannot distinguish whether the potential protamine deficiency is due to a lack of P1, P2 or a combination of both. Another very popular test has been the sperm chromatin structure assay (SCSA) based on the AO red–green shift to differentiate double- versus single-stranded DNA (Evenson et al., 1980; Virro et al., 2004; Evenson and Wixon, 2005). A large amount of information correlating results from this indirect test, mainly intended to infer the presence of DNA breaks, with infertility or assisted reproduction outcome has accumulated over the years (Virro et al., 2004; Evenson and Wixon, 2005).
Another indirect approach has been the use of aniline blue staining to detect the presence of histones and therefore indirectly infer the presence of lower amounts of protamines in the sperm nucleus (Chevaillier et al., 1987; Colleu et al., 1988). An increase in the percentage of aniline blue cells was found in asthenozoospermic as compared with normozoospermic samples (Colleu et al., 1988). Acidic aniline blue was also correlated with differences in sperm nuclear morphology in sperm donors and in infertile patients (Auger et al., 1990). A decreased resistance to chromatin decondensation by treatment with sodium dodecyl sulphate (SDS) and dithiothreitol (DTT) in abnormal sperm compared with normal sperm has also been taken as evidence for lower protamine S–S stability and chromatin packaging (Bustos-Obregón and Leiva, 1983; Le Lannou et al., 1986; Jager, 1990). The accessibility of the fluorescent dye ethidium bromide to DNA has also been correlated to IVF outcomes (Filatov et al., 1999).
Other new sperm chromatin structure tests based on sperm chromatin dispersion are also being proposed (Silvestroni et al., 2004; Evenson and Wixon, 2005; Fernández et al., 2005; Schlegel and Paduch, 2005). The interpretation of the results of all these indirect tests is difficult since they depend on the sperm chromatin composition, structure, accessibility and integrity of the DNA (Schlegel and Paduch, 2005; Erenpreiss et al., 2006). Thus, changes in the overall amount of protamines, degree of protamine cross-linking, P1/P2 ratio, presence of P2 precursors, proportion of histones and other proteins, protein modifications, topological state of the DNA and double- or single-DNA breaks may all result in measurable changes. So, at present, direct protamine extraction and electrophoresis are still the gold standard to directly quantify protamines (Balhorn et al., 1988; de Yebra et al., 1993; de Yebra and Oliva, 1993; Mengual et al., 2003a; Aoki et al. 2005a). However this direct approach was more complex and time consuming that indirect staining procedures (Mckay et al., 1986; Yelick et al., 1987; Sautière et al., 1988). A systematic assessment of the factors involved in protamine recovery led to drastic reduction in the time involved and complexity of the methods used, so that routine clinical application is now easier (de Yebra et al., 1993; de Yebra and Oliva, 1993; Mengual et al., 2003a).
The use of antibodies to P1, P2 or to the protamine P2 precursor increases the sensitivity but should be further elaborated to allow fast routine clinical use (Stanker et al., 1992, 1993; Le Lannic et al., 1993; de Yebra et al., 1998). Also, because of the clinical use of protamines as drugs, there is pharmaceutical interest in developing more sensitive protamine detection methods (Lochmann et al., 2004; Shvarev and Bakker, 2005) and new proteomic approaches based on liquid fractionation mass spectrometry or new fluidic devices that have the potential to make protamine quantification even easier and faster in the near future.
|
Anomalies in protamine content and IVF potential |
|---|
 TOPAbstractIntroductionSummary of protamine structure...Transgenes and knockout modelsAlterations in protamine content... |
|---|