Introduction
According to the World Health Organization (WHO) (2004) estimation, infertility may now affect one couple of five in western countries. Male reproductive failure is thought to be the cause of at least one-half of the total number of failures. During the past half century and before the advent of large scale in vitro fertilization, routine semen analysis was oriented towards numeration, concentration, motility and morphology. Rapidly it has become obvious that a better appreciation of sperm chromatin is needed. A ‘transitory’ interest in nuclear shape and chromatin texture, i.e. the organization of chromatin condensation, measured by aniline blue (AB) staining (Auger et al., Reference Auger, Mesbah, Huber and Dadoune1990) or by chromomycin A3 (Sakkas et al., Reference Sakkas, Urner, Bianchi, Bizzaro, Wagner, Jaquenoud, Manicardi and Campana1996) has emerged. But overall, interest has been focused on sperm DNA fragmentation (Evenson et al., Reference Evenson, Darzynkiewicz and Melamed1980), especially in relation to oxidative stress. The large waste observed for embryos conceived by assisted reproductive technology (ART) has led to the common consensus that fertilization and embryo development constitute only a small proportion of sperm development capacity, as it is also true that the quality of the first embryo cleavage is under paternal effect (Eid et al., Reference Eid, Lorton and Parrish1994, Ward et al., Reference Ward, Rizos, Corridan, Quinn, Boland and Lonergan2001). True sperm capacity is defined as the ability to sustain embryo development to term. Progress has been made in the global determination of DNA strand fractures (i.e. fragmentation; Evenson et al., Reference Evenson, Darzynkiewicz and Melamed1980) and the identification of specific insults such as oxidation products and adducts (Gaspari et al., Reference Gaspari, Chang, Santella, Garte, Pedotti and Taioli2003, Badouard et al., Reference Badouard, Ménézo, Panteix, Ravanat, Douki, Cadet and Favier2008, Ji et al., Reference Ji, Yan, Wu, Liu, Wang, Zhang, Shi and Gu2013), essentially those originating from pollution. Interest towards the tertiary structure of the nucleus, i.e. its condensation, is re-emerging (Kim et al., Reference Kim, Kang, Kim, Oh, Kim, Ku, Kim, Moon and Choi2013, Sellami et al., Reference Sellami, Chakroun, Ben Zarrouk, Sellami, Kebaili, Rebai and Keskes2013). The importance of the structural interactions between DNA and the packaging proteins is no longer a matter for debate: failures in condensation induce delays in the first cell cycle with further detrimental consequences at some point for the developing embryo (Rousseaux et al., Reference Rousseaux, Reynoir, Escoffier, Thevenon, Caron and Khochbin2008; Ward, Reference Ward2010). This includes recurrent abortion (Kazerooni et al., Reference Kazerooni, Asadi, Jadid, Kazerooni, Ghanadi, Ghaffarpasand, Kazerooni and Zolghadr2009). This highly compacted structure is not inert. Some gene families that are highly important for early embryo development are associated with histones (in human spermatozoa: Hammoud et al., Reference Hammoud, Nix, Zhang, Purwar, Carrell and Cairns2009). If the oocyte has an important, yet finite, DNA repair capacity, it is largely efficient in relation to DNA strand breaks (Menezo et al., Reference Menezo, Russo, Tosti, El Mouatassim and Benkhalifa2007, Reference Menezo, Dale and Cohen2010). The oocyte capacity to improve tertiary structure is rather limited. Chromatin assembly factor 1 (CAF-1), and H2AFX (H2A histone family, member X) are correctly expressed in the oocyte; however it is not clear if these factors are active at the correct time, immediately at fertilization. This means that determination and correction (when possible; Menezo et al., Reference Menezo, Evenson, Cohen and Dale2014) of decondensation is of paramount importance in ART. There is no real consensus on the technique of evaluation of DNA fragmentation, but most of all for decondensation, which is usually totally neglected. In a first approach, we have demonstrated that chromomycin A and terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labelling (TUNEL) have overlapping characteristics: chromomycin A3 (CMA3) is a DNA dye, a guanine–cytosine-specific fluorochrome that competes with protamines for access to DNA. Aniline blue (AB) is a better marker of the sperm nucleus chromatin decondensation state (Belloc et al., Reference Belloc, Benkhalifa, Junca, Dumont, Bacrie and Ménézo2009). In this study we have compared, on the same samples, the DNA fragmentation index (DFI) measured by TUNEL in situ and the DNA decondensation state index (SDI) measured by AB on slides versus flow cytometry (FC) after staining with acridine orange.
Material and methods
The analyses were performed in a double-blinded manner in 88 patients, who were under consultation for infertility after at least one ART failure. The sperm was collected from patients by masturbation. After liquefaction (0.5 h), the sperm was aliquoted in 2 × 0.5 ml samples. Samples were frozen in liquid nitrogen until analysis for dosage using TUNEL on slides and FC analyses. For AB testing, the samples were directly analysed after liquefaction according to the technique of Hamadeh et al. (Reference Hamadeh, Al-Hasani, Stieber, Rosenbaum, Küpker, Diedrich and Schmidt1996) (see also Belloc et al. Reference Belloc, Cohen-Bacrie, Benkhalifa, Cohen-Bacrie, De Mouzon, Hazout and Ménézo2008). In total, 200 sperm cells were examined for each in situ determination.
DNA fragmentation with TUNEL
For DNA fragmentation with TUNEL, the classical terminal deoxy ribonucleotidyl transferase-mediated uridine 5′-triphosphate-biotin nick-end labelling test was performed. The fluorescein isothiocyanate–labelled deoxyuridine triphosphate (dUTP) kit (Roche Diagnostics Corporation, Mannheim, Germany) was used according to the instructions of the manufacturer for the in situ technique (see Cohen-Bacrie et al., Reference Cohen-Bacrie, Belloc, Ménézo, Clement, Hamidi and Benkhalifa2009).
Acridine orange staining and flow cytometry
For acridine orange, flow cytometry (FC), if the sperm concentration was lower than 0.5 × 106/ml, before acridine testing, the sample was washed/concentrated using a phosphate-buffered medium containing albumin at 10 g/l, in order to reach a final concentration of at least 1 × 106/ml (Menezo, 2003; unpublished data). After thawing, the sperm were diluted in Earles medium + 1% albumin. For either concentration or dilution, the presence of proteins was necessary in order to avoid agglutination. After dilution (1/10) the sperm sample was lysed in a 0.1% Triton solution in 1 M HCl. After mixing, the staining solution (acridine orange in a phosphate citrate buffer) is added. The suspension was shaken gently for 30 s before introduction into the cytometer. The number of cells analysed was roughly 8000. Sperm samples were thawed and prepared one by one at the last minute. This step is necessary to maintain reproducibility.
The samples were analysed on a flow cytometer FC500 Beckman Coulter equipped with a BioSense™ measurement cell (150 μM × 450 μM). The laser used was an Argon device 488 nM, with seven detection units, that was focused on a elliptic measurement area 10 μM × 80 μM. The flow reproducibility and reliability was controlled for each measurement using flow check microspheres. The optical fluorescence stability detection was controlled using Flowser microspheres. Quantification was realized by gathering the signals of sperm cells obtained at 525 nm for SDI and at 620 nm for the lesions of single DNA strands (DFI).
Statistical analysis
One of the comparison of the data obtained was performed using the Welch two-sample t-test; or an alternative two-sided test. The other was performed using the Wilcoxon test.
Results
DNA fragmentation (DFI)
DNA fragmentation (DFI): means observed TUNEL 18.07, acridine FC: 17.47.
• Welch two-sample t-test; alternative: two-sided T-test: array, confidence interval 95% [–1.9772; 3.1748]. P-value 0.65: no difference between the two techniques.
• Wilcoxon test: Z-score: –2.57. No difference between the two techniques. Asymptotic significance: 0.1.
Individual variations are shown in Fig 1, in situ image for the test is presented in Fig. 3.
Figure 1 In situ TUNEL versus acridine orange flow cytometry. Individual variations: blue, TUNEL; red, acridine orange.
Chromatin decondensation (SDI)
Chromatin decondensation (SDI): means observed AB: 9.94, acridine orange: 9.01.
• Welch two-sample t-test: confidence interval 95% [–0.63; 2.47]. P-value 0.24: no difference between the two tests.
• Wilcoxon test: Z-score: –2.64. Asymptotic significance 0.01. The tests cannot be considered as similar. Acridine orange gives a lower value for estimation of chromatin decondensation.
Individual variations obtained between the two techniques are shown in Fig. 2, in situ image for the test is presented in Fig. 4.
Figure 2 In situ aniline blue versus acridine orange flow cytometry. Individual variations: dark blue: aniline blue; light blue: acridine orange green fluorescence.
Figure 3 In situ determination of sperm DNA fragmentation with TUNEL. Pale green: fragmented DNA; blue: DNA presenting no fragmentation.
Figure 4 In situ determination of sperm decondensation index (SDI) using aniline blue. Top panel: light staining (SDI: 3%); bottom panel: heavy staining (SDI: 51%).
Discussion
In situ analysis has two main disadvantages/two obstacles: the number of cells analysed and the operator dependence. It has also to be said that sperm samples can be frozen before FC acridine orange testing for SDI; this is not the case for AB. The samples cannot be frozen before AB testing otherwise the sperm cells are fully (blue) coloured. Automated FC avoids these difficulties, comparing in situ TUNEL versus FC acridine orange (red fluorescence) on the one hand and in situ AB versus FC acridine orange (green fluorescence) on the other hand gives roughly the same results. This is a little bit less valid for AB: the Wilcoxon test give a significant lower quantification with FC acridine orange green fluorescence (SDI), whilst the Welch test leads to a similar quantification.
For sperm DNA fragmentation the distribution in our general population, is, on a yearly basis, the following: using TUNEL DFI <25%: 86%; DFI > 25%: 14% (1336 patients). But 53% of the patients have a DFI between 20–25%. An important percentage of the patients consulting for infertility is in this grey zone 20–25%. The pathological values limits are still controversial and may vary according to the techniques employed (Kazerooni et al., Reference Kazerooni, Asadi, Jadid, Kazerooni, Ghanadi, Ghaffarpasand, Kazerooni and Zolghadr2009, Sharma et al., Reference Sharma, Sabanegh, Mahfouz, Gupta, Thiyagarajan and Agarwal2010, Simon et al., Reference Simon, Lutton, McManus and Lewis2011, Lewis et al., Reference Lewis, Aitken, Conner, Iuliis, Evenson, Henkel, Giwercman and Gharagozloo2013). The threshold of 20–25% for sperm DNA fragmentation for decreased fertility is not far from the consensus: this could be considered as a grey zone where fertility problems start. The upper limit for having a healthy delivery can be considered at 35–40%. The success of ART also depends of the quality of the oocytes and especially the importance of the DNA repair possibilities, i.e. the mRNAs coding for this process (Menezo et al., Reference Menezo, Russo, Tosti, El Mouatassim and Benkhalifa2007). This is especially true for fragmentation: the repair mechanisms are complete and iterating in the healthy (young) oocyte. This is less valid for chromatin condensation, i.e. tertiary structure: the oocyte has a poor capacity to manage with an abnormal chromatin structure (Menezo et al., Reference Menezo, Russo, Tosti, El Mouatassim and Benkhalifa2007). In our overall population, 17% of patients have a SDI strictly over 25% (AB, 528 patients). The determination of the ‘risky’ threshold is not just a matter of semantics. SDI is a neglected parameter, even if it is now clear that chromatin structure is an important parameter for the capacity to produce live births. The negative impact of a high decondensation may occur immediately at the time of fertilization, with the ‘no fertilization’ syndrome in intra-cytoplasmic sperm injection (ICSI) (Junca et al., Reference Junca, Gonzalez Marti, Tosti, Cohen, De la Fontaine, Benkhalifa and Ménézo2012) and early developmental arrests (Dattilo et al., Reference Dattilo, Cornet, Amar, Cohen and Menezo2014), up to miscarriages (Kazerooni et al., Reference Kazerooni, Asadi, Jadid, Kazerooni, Ghanadi, Ghaffarpasand, Kazerooni and Zolghadr2009). Considering DFI, numerous observations have focused on the risk for the outcome and the next generation (Hemminki et al. Reference Hemminki, Kyyrönen and Vaittinen1999, Belloc et al., Reference Belloc, Cohen-Bacrie, Benkhalifa, Cohen-Bacrie, De Mouzon, Hazout and Ménézo2008, Fernández-Gonzalez et al., Reference Fernández-Gonzalez, Moreira, Pérez-Crespo, Sánchez-Martín, Ramirez, Pericuesta, Bilbao, Bermejo-Alvarez, de Dios Hourcade, de Fonseca and Gutiérrez-Adán2008). This includes an increased risk of cancer for the next generation. In a recent analysis of babies born after ICSI, Davies et al. (Reference Davies, Moore, Willson, Van Essen, Priest, Scott, Haan and Chan2012) demonstrated an increased risk of birth defects. The DFI increases with age and it is obvious that the patients seeking ART are older compared with the general population. Only one-third of our patients consulting for hypofertility seems totally ‘clean’ of an impairing DNA fragmentation. Moreover, in a recent paper we were able to demonstrate that both DFI and SDI are stable with time in the absence of any treatment (Menezo et al., Reference Menezo, Evenson, Cohen and Dale2014). In conclusion, FC acridine orange allows us to check simultaneously both sperm DNA fragmentation and decondensation. The determination of these parameters seems increasingly necessary: it has to be clearly said that ICSI does not cure all male fertility problems and may even be hazardous for the next generation, as oxidative stress and imprinting are also linked (Menezo et al., Reference Menezo, Lichtblau and Elder2013). Finally, it has to be pointed out that the repair capacity of the human oocyte cannot be improved and may be insufficient to overcome paternally borne damage. Therefore, a scientific approach for an in vivo improvement of spermatozoa, via the improvement of the one carbon cycle (Menezo et al., Reference Menezo, Evenson, Cohen and Dale2014) before ART remains of paramount importance.
Open access
