Sperm cephalic vacuoles: new arguments for their non acrosomal origin in two cases of total globozoospermia


  • N. Gatimel,

    1. Centre d'Assistance Médicale à la Procréation, Hôpital Paule de Viguier, Toulouse, France
    2. UPS; Groupe de Recherche en Fertilité Humaine (EA 3694, Human Fertility Research Group), Université de Toulouse, Toulouse, France
    Search for more papers by this author
  • R. D. Léandri,

    Corresponding author
    1. UPS; Groupe de Recherche en Fertilité Humaine (EA 3694, Human Fertility Research Group), Université de Toulouse, Toulouse, France
    • Centre d'Assistance Médicale à la Procréation, Hôpital Paule de Viguier, Toulouse, France
    Search for more papers by this author
  • B. Foliguet,

    1. Laboratoire de Microscopie Electronique, Faculté de Médecine de Nancy, Vandoeuvre-lès-Nancy, Nancy Cedex, France
    Search for more papers by this author
  • L. Bujan,

    1. UPS; Groupe de Recherche en Fertilité Humaine (EA 3694, Human Fertility Research Group), Université de Toulouse, Toulouse, France
    Search for more papers by this author
  • J. Parinaud

    1. Centre d'Assistance Médicale à la Procréation, Hôpital Paule de Viguier, Toulouse, France
    2. UPS; Groupe de Recherche en Fertilité Humaine (EA 3694, Human Fertility Research Group), Université de Toulouse, Toulouse, France
    Search for more papers by this author


Roger D. Léandri, Centre d'Assistance Médicale à la Procréation, Hôpital Paule de Viguier, 330 Avenue de Grande-Bretagne, 31059 Toulouse cedex 9, France. E-mail: leandri.r@chu-toulouse.fr


To add new arguments concerning the origin of the sperm-head vacuoles observed under high magnification with interference contrast microscopy, we carried out in two patients with total globozoospermia confirmed using transmission electron microscopy (TEM), a detailed sperm morphometric analysis with high magnification (×6000) under Nomarski contrast, an acrosomal status analysis (using fluorescent labelling with peanut agglutinin (PNA) lectins and anti-CD46 antibodies) and a nuclear status analysis (using terminal deoxynucleotidyl transferase-mediated dUTP nick-end labelling assay TUNEL, sperm chromatin structure assay SCSA and aniline blue staining). Our two patients with globozoospermia had relative sperm vacuole areas of 6.3% and 5%, similar to those observed in a reference population of 12 fertile men (5.9%). TUNEL and SCSA assays gave normal results in both patients, although the percentage of immature nuclei using aniline blue staining was increased (27 and 46% for patient 1 and 2 respectively). Cytofluorescence and TEM analysis evidence differences between the two patients: although no acrosomal neither Golgi residue could be detected in patient 1, patient 2 had positive PNA lectin labelling for 9% of spermatozoa and Golgi residues were seen using electron microscopy. Unlike patient 1, a live birth could be obtained after intracytoplasmic sperm injection (ICSI) for patient 2. This descriptive study of two patients with total globozoospermia confirmed using TEM argue in favour of a deep analysis of total globozoospermia before assisted reproductive technology and provides further information on the non-acrosomal origin of the sperm-head vacuoles observed under high magnification.


Since 2001, close examination of spermatozoa by high-magnification interference contrast microscopy or motile sperm organelle morphology examination has revealed sperm-head vacuoles that are not visible on standard magnification (Bartoov et al., 2001). This observation technique is used to optimize selection of the spermatozoon before microinjection in the oocyte by intracytoplasmic morphologically selected sperm injection (IMSI). Under certain conditions, IMSI may improve the pregnancy rate compared with conventional ICSI, in particular after several failures of ICSI (Antinori et al., 2008), but its value as first-line treatment has not been demonstrated.

The origin of these vacuoles is still somewhat debated. Nuclear vacuoles have been well described using electron microscopy as areas of uneven chromatin condensation (Zamboni, 1987), but what of the vacuoles observed by Nomarski interference contrast? Are they of nuclear or acrosomal origin, or are there several types of vacuoles? The majority of publications support a nuclear origin. Some authors link the presence of large vacuoles with an increase in fragmented DNA (Boughali, 2006; Franco et al., 2008; Garolla et al., 2008; Oliveira et al., 2010), although for others these large vacuoles are associated with chromatin condensation defects (Boitrelle et al., 2011; Perdrix et al., 2011; Franco et al., 2012) and are related to an increase in aneuploidy rates (Perdrix et al., 2011). However, other authors consider that the vacuoles are of acrosomal origin (Kacem et al., 2010).

Globozoospermia is a syndrome whose incidence is estimated at less than 0.1% in infertile men. Schirren et al.(Schirren et al., 1971) were the first to describe this syndrome, which in its complete form is defined by the presence of spermatozoa that all have round heads and lack an acrosome. In all the cases reported in the literature, the aetiology is not always elucidated but genetic causes have been found, notably in families with brothers with globozoospermia (SPATA16 gene (Dam et al., 2007b), PICK1 gene (Liu et al., 2010), and the DPY19L2 gene which appears to be the most frequently involved (Koscinski et al., 2011)). The pathophysiological mechanisms are poorly known, but several abnormalities of sperm remodelling during spermiogenesis have been described (Dam et al., 2007a) leading to formation defects or premature elimination of acrosomal structures. Fertilization rates after ICSI in patients with globozoospermia are low (Dam et al., 2007a) as 22 births have been reported in the literature after ICSI and one birth after IMSI (Sermondade et al., 2011).

We report the cases of two patients diagnosed with total globozoospermia using conventional sperm analysis. We studied sperm morphology using high-magnification interference contrast microscopy and electron microscopy, and assessed acrosomal status by fluorescent labelling with peanut agglutinin (PNA) lectins and anti-CD46 antibodies.

The total absence of acrosome in these patients provided a good study model to further investigate the origin of the vacuoles.


Patient 1, aged 30 years, and patient 2, aged 31 years, consulted for primary infertility in our centre. Their partners were aged 29 and 30 years. Their body mass index was normal and their medical histories were unremarkable except for controlled hypothyroidism in patient 2. On clinical examination, grade 1 bilateral varicocoele was found in patient 1. No consanguinity was reported. Patient 1 had a brother aged 36 years who had sought medical advice for infertility, but no further data were available. Spermiograms were carried out according to WHO guidelines 2010 (WHO, 2010)(Table 1). Spermiogram parameters were normal except for severe asthenozoospermia in patient 2, in agreement with the data of the literature which report sometimes asthenozoospermia alone in patients with globozoospermia (Dam et al., 2007a).

Table 1. Clinical data, spermiogram values, and ICSI outcomes in the two patients with globozoospermia
 Patient 1Patient 2
Age (years)3031
Female age (years)2930
Body mass index (kg/m2)27.124.8
Female body mass index (kg/m2)21.218.0
Female basal FSH (UI/l)7.27.5
Duration of abstinence (days)34
Sperm volume (mL)5.37.5
Sperm count (m/mL)17019
Round cells (m/mL)<1<1
Motility a+b/c/d %30/5/657/3/90
Vitality %6665
Number of metaphase two oocyte retrieved49
Number of 2PN zygotes01
Number of embryos01
Number of embryos transferred01
ICSI outcomeFailureBirth of healthy girl

Using conventional spermocytogram with modified Schorr staining, the typical morphological features were observed (100% round-headed spermatozoa lacking an acrosome) suggesting total globozoospermia in both patients.

Acrosomal or Golgi content was labelled with fluorescein-isothiocyanate (FITC)-conjugated peanut agglutinin (FITC; Sigma-Aldrich, St-Quentin-Fallavier, France), which is lost after acrosome reaction (Parinaud et al., 1993). Inner acrosomal membrane was labelled before membrane permeabilization with a mouse anti-CD46 monoclonal antibody and Texas Red-conjugated anti-IgG secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA), this labelling is acquired after acrosomal reaction. Therefore, acrosomal reaction is associated to both a loss of PNA labelling and an acquisition of CD46 labelling. Acrosomal status was assessed before and after acrosome induction with calcium ionophore A23187 (Sigma-Aldrich, St-Quentin-Fallavier) with counting of 200 cells per patient. In a fertile control, calcium ionophore induced acrosome reaction in 50% of spermatozoa as assessed by the decrease of PNA labelling and the increase of CD46 labelling (Table 2). In patient 1, PNA and anti-CD46 labelling were both negative before and after incubation with calcium ionophore. In patient 2, we observed homogeneous PNA labelling on the entire head of the spermatozoa (not characteristic of usual acrosome labelling) in 9% and 8% of spermatozoa before and after calcium ionophore, respectively. Anti-CD46 labelling was negative (Table 2, Fig. 1).

Figure 1.

Peanut agglutinin (PNA) lectin labelling, negative labelling (head arrow) in patient 1 (left) and positive labelling (arrow) in 9% of spermatozoa in patient 2 (right).

Table 2. Labelling of acrosomal content and of the internal acrosomal membrane in a fertile control and in the two patients with globozoospermia
 ControlPatient 1Patient 2
Calcium ionophoreBefore %After %Before %After %Before %After %
  1. a

    PNA labelling of the equatorial plate.











CD 46−


We carried out detailed morphometric analysis at ×6000 magnification by Nomarski interference contrast microscopy using a Leica DMI 6000 inverted microscope with a ×100 objective on whole washed sperm fixed in suspension in 3.4% PBS-formaldehyde in a different sample from that used on the day of the attempted microinjection. After image capture, morphometric analysis was carried out using Leica Application Suite software with measurement of area of each vacuole and area of the head in 100 spermatozoa per patient (Fig. 2). Patients 1 and 2 had a mean number of vacuoles per spermatozoon of 1 and 0.8, a mean vacuolar area of 6.3% and 5%, and a percentage of spermatozoa with large vacuoles (>13% of the head area) of 4% and 7.6%, respectively. All these values were within the 95% confidence interval of our reference population of 12 fertile men (Table 3).

Figure 2.

Spermatozoon from patient 1 at ×6000 magnification with 2 vacuoles measuring 0.47 and 0.37 μm2, and a relative vacuolar area of 12.4%.

Table 3. Vacuolar parameters at ×6000 magnification and study of nuclear material
 Patient 1Patient 2Reference populationa
  1. a

    Twelve fertile controls (Mean ± SD).

  2. Vacuoles were measured on 100 spermatozoa per sample, aniline blue staining on 200 spermatozoa per sample, TUNEL on 10 000 spermatozoa per sample, SCSA on 5000 spermatozoa per sample.

Mean number of vacuoles per spermatozoon1.00.81.3 ± 0.3
Vacuolar area (%) ± 2.3
% Spermatozoa with vacuolar area 0%82415 ± 10
% Spermatozoa with vacuolar area [0.1;6.5]%545252 ± 18
% Spermatozoa with vacuolar area [6.6;13] %341624 ± 12
% Spermatozoa with vacuolar area >13%489 ± 8
% Fragmented DNA (TUNEL technique, N <20%)211310.0 ± 6.3
DFI (SCSA technique, N < 30%)25227.7 ± 6.0
HDS (SCSA, < 15%)586.0 ± 1.7
% Immature nuclei (aniline blue staining, N < 20%)27466.5 ± 3.9

In patient 1, scanning electron microscopy (SEM) of more than 2000 gametes and transmission electron microscopy (TEM) of more than 500 gametes confirmed monomorphous teratospermia of total globozoospermia type, without any structure of acrosomal appearance. No structure suggested rudimentary (pre-acrosomal) Golgi vesicles (Fig. 3), which are sometimes present in globozoospermia (Dam et al., 2007a; Sermondade et al., 2011).

Figure 3.

Image from TEM (×18000) and SEM (×20000) for patient 1 and 2. (a) TEM for patient 1 showing a small and round head, without acrosomic or golgi residues and a densified nucleus with many redundancies of nuclear membrane. (b). TEM for patient 2 showing the presence of a saccule and preacrosomic golgi vesicles (arrow) with a low density content. (c). SEM for patient 2: round head, no acrosomic relief, and thin midpiece.

In patient 2, SEM was carried out on over 1000 gametes and TEM on 200 gametes. The cytoplasm sometimes contained rare Golgi sacs, with a clear content (which differed from the usual, denser acrosomal content), that resembled rudimentary acrosomal vesicles (Fig. 3b).

The percentage of fragmented DNA measured by the terminal deoxyribonucleotidyl transferase-mediated dUTP nick-end labelling assay (TUNEL assay) was 21% in patient 1 and 13% in patient 2 (the accepted threshold value with this technique is 20% (Sergerie et al., 2005). Assessment of chromatin quality using the SCSA test (Evenson et al., 2002) showed a DNA fragmentation index (DFI) and high DNA stainability (HDS) within normal ranges (Table 3). However, both patients had a higher DFI than our control population. In both patients the percentage of immature nuclei after aniline blue staining was increased at 27% and 46% (Table 3), reflecting a chromatin condensation defect.

The female partner of patient 1 had regular 28-day menstrual cycles, normal ovarian reserve and permeable fallopian tubes. Intra-couple ICSI was attempted in couple 1, four mature oocytes were injected but no fertilization was obtained. The failure of ICSI was followed by inseminations with donor spermatozoa, but these were unsuccessful.

The female partner of patient 2 had irregular menstrual cycles of 28–34 days and normal ovarian reserve. Intra-couple ICSI was attempted and nine mature ovocytes were injected, giving one top-quality embryo, which was transferred. After a pregnancy carried to term, a girl weighing 3620 g was born by vaginal delivery.


To provide new arguments to the controversy concerning the origin of sperm cephalic vacuoles (acrosomal, nuclear or others…) observed under high magnification, we analyse acrosomeless spermatozoa from two patients whose spermocytograms suggested total globozoospermia.

The absence of acrosome was confirmed in patient 1 by complete absence of PNA and CD46 fluorescent labelling and also by the absence of acrosomal and Golgi structures on electron microscopy. These findings could explain the total fertilization failure during the intra-couple ICSI attempt. In patient 2, 9% of spermatozoa showed an atypical fluorescent PNA labelling of the entire surface of the head. Under electron microscopy, as in patient 1, the very homogeneous appearance of the abnormalities observed and their high frequency confirmed generalized teratozoospermia of total globozoospermia type. However, in patient 2, Golgi elements (acrosome precursors) were observed by electron microscopy but were dysplasic, which could explain the absence of acrosome reaction to calcium ionophore and the low PNA positivity (9% of spermatozoa). We postulate that among the nine spermatozoa used for ICSI, the one that led to obtention of an embryo contained Golgi residues, whose presence seems necessary for oocyte activation, as is shown by the numerous failures or low fertilization rates reported in ICSI in total globozoospermias (Dam et al., 2007a).

Our full investigation of the two patients shows that so-called ‘total’ globozoospermias may have different profiles at a cellular level and suggests that the presence of Golgi elements could change the prognosis of assisted reproductive technology (ART) and the type of management proposed. In patient 2, the presence of Golgi residues was associated with positive results in ART within the couple. Some authors have suggested that IMSI should be used to select this type of spermatozoon, with a more oval shape (Sermondade et al., 2011). On examination at ×6000 magnification, we observed in patient 2 some sparse oval forms that were not found in patient 1. This finding leads us to believe that it seems important to better characterize globozoospermias by looking for the presence of Golgi residues in a larger population.

Abnormalities of chromatin condensation in our two globozoospermic patients are in agreement with previous studies (Carrell et al., 1999; Vicari et al., 2002) and reflect disorders during spermiogenesis, as the absence of acrosome formation does. On the contrary, we found no high increase of the DNA fragmentation rates in our two patients, in accordance with the two-step hypothesis for the origins of sperm DNA damages (Aitken & De Iuliis, 2010). This theory supposes in a first step, a disruption of spermiogenesis resulting in uncondensed and vulnerable chromatin, and in a second step, an oxidative stress condition leading to DNA fragmentation of this vulnerable chromatin. This could explain the difference in terms of DNA fragmentation between our two subjects, suggesting that patient 1 is exposed to higher oxidative stress conditions than patient 2. One of these conditions could come from its grade I bilateral varicocoele.

Another interesting feature of this study is the characterization of sperm-head vacuoles (their number and percentage area) by morphometric analysis. In our two patients in whom 100% of spermatozoa are acrosomeless, high-magnification examination revealed a number of vacuoles and a vacuolar area very similar to those of fertile controls. These findings further support the argument that these vacuoles, in particular the largest vacuoles, are not of acrosomal origin. However, we found higher level of chromatin condensation defects, but identical numbers and areas of cephalic vacuoles in comparison with fertile subjects. It seems therefore difficult to address these vacuoles as witnesses of chromatin decondensation.

Authors' roles

N.G. performed experiments, collected and analysed data, and drafted the manuscript. R.L. played a role in clinical care, designed the study and critically revised the manuscript. B.F. performed electron microscopy analysis. L.B. played a role in clinical care and critically revised the manuscript. J.P. played a role in clinical care, designed the study and critically revised the manuscript. All authors have read and approved the final manuscript.


The authors thank Linda Masson, Centre d'Assistance Médicale à la Procréation, Hôpital Paule de Viguier, Toulouse, France, for carrying out the microinjection and Camille Esquerré-Lamare, EA 3694, Human Fertility Research Group, Toulouse, France for performing SCSA and TUNEL analysis. The authors thank the patients for sharing their story and for providing the permission to publish it.