Defects in phospholipase C zeta cause polyspermy and low fertilization after conventional IVF: not just ICSI failure

Abstract

Phospholipase C zeta (PLCζ) is a key sperm-borne oocyte-activating factor that triggers Ca²⁺ oscillations and the subsequent block to polyspermy following gamete fusion. Mutations in PLCZ1, the gene encoding PLCζ, cause male infertility and intracytoplasmic sperm injection (ICSI) fertilization failure; and PLCζ expression and localization patterns are significantly correlated with ICSI fertilization rate (FR). However, in conventional in vitro fertilization (cIVF), whether and how sperm PLCζ affects fertilization remain unclear. Herein, we identified one previously reported and two novel PLCZ1 mutations associated with polyspermy in vitro that are characterized by excessive sperm–zona binding and a delay in pronuclei (PN) formation. Immunofluorescence staining and oocyte activation testing revealed that virtually all spermatozoa from patients lacked functional PLCζ and were thus unable to evoke Ca²⁺ oscillations. ICSI with an artificial oocyte activation treatment successfully rescued the polyspermic phenotype and resulted in a live birth. Furthermore, we analyzed PLCζ in an additional 58 males after cIVF treatment in the Reproductive and Genetic Hospital of CITIC-Xiangya (Changsha, China) between February 2019 and January 2022. We found that the proportion of spermatozoa that expressed PLCζ was positively correlated with both 2PN rate and total FR. The optimal cutoff value below which males were likely to experience low FR (total FR ≤30%) after cIVF was 56.7% for the proportion of spermatozoa expressing PLCζ. Our study expands the mutation and the phenotypic spectrum of PLCZ1 and further suggests that PLCζ constitutes a promising biomarker for identifying low FRs cases in cIVF due to sperm-related oocyte activation deficiency and that sperm PLCζ analysis may benefit the wider male population and not only men with ICSI failure.

INTRODUCTION

In mammals, fertilization refers to the fusion of an oocyte with a spermatozoon to initiate the development of a new individual. Successful fertilization is also a vital component of assisted reproductive technology (ART). Through the application of contemporary technology and sophisticated facilities, the modern in vitro fertilization (IVF) laboratory achieves fertilization rate (FR) close to 70%–80%. However, low, abnormal, or failed fertilization still occurs, greatly compromising zygote availability for fertility treatment and challenging the clinician and embryologist.

After fusion with an oocyte, the most significant function of the spermatozoon is activating the oocyte, which comprises a group of preprogrammed early biochemical events in the cytoplasm of the oocyte, such as cortical granule exocytosis, prevention of polyspermy, resumption of meiosis, and formation of pronuclei (PN).

In recent decades, researchers have been intent on searching for a “sperm factor” that was proposed to be released from the sperm head and responsible for inducing oocyte activation via Ca²⁺ signaling after gamete fusion.1 Since the identification of phospholipase C zeta (PLCζ) in spermatozoa,2,3 accumulating evidence has strongly supported PLCζ as the long sought-after sperm factor that evokes the Ca²⁺ oscillations in the mammalian oocyte leading to oocyte activation and embryonic development.2,3,4,5,6 In mouse and human spermatozoa, PLCζ is predominantly located in the acrosomal, equatorial, and postacrosomal regions of the perinuclear theca,7,8 a cytoskeletal element surrounding the sperm nucleus.

Since its first case report in humans,9 growing studies have linked PLCζ defects (reduced expression levels or the absence or abnormal location of PLCζ protein) with cases of male infertility and found that pathogenic mutations in PLCZ1, the gene encoding PLCζ, were the primary cause of total fertilization failure with intracytoplasmic sperm injection (ICSI).8,10,11,12,13,14

However, few studies have been published on the relationship between sperm PLCζ and conventional in vitro fertilization (cIVF), an insemination procedure involving the co-incubation of oocytes with hundreds of thousands of spermatozoa in a culture system. cIVF involves several key events that are bypassed in ICSI, such as sperm–zona binding and penetration, as well as polyspermy blocking. It is also possible that sperm PLCζ may affect fertilization in a different manner from that in ICSI.

Herein, using whole-exome sequencing (WES), we identified three PLCZ1 variants in two unrelated males characterized by infertility and polyspermy in cIVF. In vitro functional studies were employed to examine the effects of the variants on PLCζ protein. We also evaluated the correlation between PLCζ and fertilization outcomes in an additional 58 males who underwent cIVF.

PARTICIPANTS AND METHODS

Participants

Sixty couples who underwent cIVF in the Reproductive and Genetic Hospital of CITIC-Xiangya (Changsha, China) were recruited between February 2019 and January 2022. Our inclusion criteria were as follows: all starting semen parameters were at least or nearly at the World Health Organization (WHO) 2010 reference values,15 and a minimum of four oocytes were inseminated for each case. Exclusion criteria included severe oligozoospermia, abnormal sperm morphology (such as round- or pin-shaped head), or surgically retrieved spermatozoa. According to fertilization outcomes, 58 patients (except for two patients with PLCZ1 variants) were allocated to three groups for the PLCζ analysis: a multiple pronuclei (MPN) group (number of PN ≥3, MPN rate ≥50.0%, n = 9), low fertilization (LF) group (total FR ≤30.0%, n = 18), and normal fertilization (NF) group (2PN rate ≥50.0%, 1PN + MPN rate <30.0%, n = 31). The normal fertile controls were obtained from a cohort of fertile Chinese men with normal semen parameter values and who had fathered at least one healthy child. The study was approved by the Ethics Committee of the Reproductive and Genetic Hospital of CITIC-Xiangya (Approval No. LL-SC-2017-009), and all participants provided written informed consent for the use of their peripheral blood, semen, or oocytes in research.

Sperm preparation

All semen samples were collected as recommended by the WHO in 2010¹⁵ and were processed by density-gradient washing (DGW).11 After fixation with 4% paraformaldehyde (P0099, Beyotime, Shanghai, China) for 20 min at room temperature, aliquots intended for PLCζ analysis were then loaded onto 0.01% poly-L-lysine (P4832, Sigma-Aldrich, St. Louis, MO, USA)-precoated slides, air-dried, and stored at −80°C.

Immunofluorescence staining

For sperm PLCζ analysis, the slides were permeabilized for 30 min with 0.5% (v/v) Triton X-100 (T8787, Sigma-Aldrich) in phosphate-buffered saline (PBS) and then blocked for 1 h with 3% (w/v) bovine serum albumin (BSA; B2064, Sigma-Aldrich) in PBS at room temperature. The rabbit polyclonal antibody against human PLCζ (1:100; pab0367-P, Covalab, Villeurbanne, France) was applied overnight at 4°C, while 0.5% BSA in PBS was used instead of primary antibody as the negative control. Slides were subsequently incubated with Alexa Fluor™ 594 donkey anti-rabbit secondary antibody (1:1000; A21207, Invitrogen, Carlsbad, CA, USA) for 1 h at room temperature, with three washes of 0.05% Tween-20 (ST825, Beyotime) in PBS performed between all steps. The slides were ultimately counterstained with 4’,6-diamidino-2-phenylindole (DAPI; C1002, Beyotime) and observed at 100× through a high-resolution oil-immersion objective affixed to a fluorescence microscope (BX53, Olympus, Tokyo, Japan). At least 200 spermatozoa per sample were categorized into one of the four patterns according to PLCζ localization within the head: (1) acrosomal (Ac) region, sometimes accompanied by equatorial or postacrosomal expression; (2) equatorial (Eq); (3) postacrosomal (Pa), sometimes accompanied by equatorial expression; and (4) none. The proportions of spermatozoa that exhibited a PLCζ signal within the head, including the acrosomal, equatorial, and postacrosomal regions, were then recorded and expressed as percentage of PLCζ⁺ (PLCζ⁺%).

For penetrated sperm analysis, immunofluorescence staining was conducted as previously described.16 Briefly, to remove the zona pellucida (ZP), the MPN zygotes were incubated with acidic Tyrode’s solution (T1788, Sigma-Aldrich). Then, the mouse monoclonal antibody against acetylated tubulin (1:600; T6793, Sigma-Aldrich) was applied overnight at 4°C, followed by incubation with Alexa Fluor™ 488 donkey anti-mouse secondary antibody (1:1000; A21202, Invitrogen) for 1 h at room temperature. Fluorescent images were captured with an inverted fluorescence microscope (IX73, Olympus).

Fluorescence in situ hybridization (FISH)

An MPN zygote obtained from a patient with PLCZ1 variants (patient 1) was subjected to FISH, and performed in accordance with the manufacturer’s instructions. The FISH probes, including CEP18 (aqua), CEP X (green), and CEP Y (orange), were produced commercially by Abbott-Vysis (Downers Grove, IL, USA).

WES and Sanger sequencing analysis

Genomic DNA extracted from peripheral blood was subjected to WES, and data analysis was performed by the Beijing Grandomics Biosciences Company (Beijing, China) as described previously.16 We filtered and identified the candidate variants with the following criteria: (1) variants with a minor allelic frequency less than 1.0% in the 1000 Genomes, ExAC, and gnomAD databases; (2) coding small insertions or deletions, exonic nonsynonymous or splice-site variants; (3) variants with high or specific gene expression in the human testis; and (4) compound heterozygous variants or homozygous variants. Subsequently, using the primers listed in Supplementary Table 1, PLCZ1 exons containing the variant sites were amplified for bidirectional Sanger sequencing directly or after TA clone library construction.

Supplementary Table 1.

Genomic polymerase chain reaction primers used to amplify phospholipase C zeta 1 exons for Sanger sequencing

Exon	Primer sequence (5’–3’)	PCR size (bp)
2F	F: TTGTGGGTTTGGAGGTGGTC	885
2R	R: TGCCACCGAACACAGTTTCT
6F	F: TGGGTTGGGAGTAGGGATGT	771
6R	R: TTCCTGGCCATCATCCACAA

F: forward primers; R: reverse primers; PCR: polymerase chain reaction

Protein structure modeling

The three-dimensional (3D) structures of wild-type and mutant PLCζ (NP_033123.4, p.Cys196Ter, and p.Lys226Ile) were predicted using the I-TASSER modeling server (https://zhanggroup.org/I-TASSER/). Molecular graphics were generated and analyzed using PyMol software (https://pymol.org/2/).

Construction of expression vectors and immunoblotting

The full-length human PLCZ1 cDNA (NM_033123.4) was cloned and fused to the N-terminus of mCherry in the pmCherry-N1 vector (632523, Clontech, Mountain View, CA, USA). The Mut Express® II Fast Mutagenesis Kit (C214, Vazyme Biotech, Nanjing, China) was used to introduce the two variants (c.588C>A and c.677A>T).

Chinese hamster ovary (CHO-K1) cells transiently expressing the indicated plasmids were lysed with radioimmunoprecipitation assay (RIPA) lysis buffer (G2002, Servicebio, Wuhan, China) containing a cocktail of protease inhibitor (78429, Thermo Fisher Scientific, Waltham, MA, USA). The protein extracts were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Separated proteins were then transferred to polyvinylidene fluoride membranes (IPVH00010, EMD Millipore Corporation, Billerica, MA, USA). The target proteins on membranes were labeled with a primary antibody against mCherry (1:1000; ab21351, Abcam, Cambridge, UK) and an antibody against glyceraldehyde-3-phosphate dehydrogenase (GAPDH; 1:2000; WG329515, Thermo Fisher Scientific). The proteins were then labeled with secondary antibodies (1:2000; CoWin Biotech, Taizhou, China) and visualized by enhanced chemiluminescence detection (89880, Thermo Fisher Scientific).

Human oocyte activation test (HOAT)

Immature oocytes were collected from ICSI cycles of donors with written informed consent and cultured in vitro as described previously.17 Each metaphase II (MII) oocyte matured in vitro was preincubated with the Ca²⁺-sensitive fluorescent dye Fluo-3 AM (S1056, Beyotime) at 37°C for 30 min and then injected using an ICSI pipette with a spermatozoon from patient 1 or a fertile control. After ICSI, the oocytes were analyzed for intracellular Ca²⁺ in a Celldiscoverer 7 fluorescence microscope (Zeiss, Oberkochen, Germany). We executed fluorescence dynamic imaging every 5 s for 3 h, and PN were assessed 16–18 h after injection.

Artificial oocyte activation (AOA)

Immediately after ICSI, oocytes were treated with a solution of 10 μmol l⁻¹ A23187 (C9275, Sigma-Aldrich) for 5 min, followed by extensive washing. Then, fertilization and embryonic development were monitored in a time-lapse incubator at 37°C in 6% CO₂ and 5% O₂.

Statistical analyses

SPSS 26.0 software (IBM, Armonk, NY, USA) was used for statistical analyses, and P < 0.05 was considered statistically significant. Continuous data were presented as mean ± standard deviation (s.d.) or median (interquartile range). Intergroup differences were assessed using the Mann–Whitney U test and independent-sample t-test. Relationships between PLCζ and FRs were evaluated using the Pearson correlation coefficient. The receiver operating characteristic (ROC) curve and Youden’s index analysis were adopted to evaluate the performance of PLCζ⁺% and the optimal cutoff value in predicting a low FR after cIVF.

RESULTS

Description of two infertile males and indications of polyspermy in vitro

Two Chinese men with infertility of unknown cause were evaluated, and both exhibited normal semen parameter values (semen volume, sperm concentration, motility, and normal morphology) on analysis (Supplementary Table 2).

Supplementary Table 2.

Semen parameters of two males carrying phospholipase C zeta 1 variants

Items	Patient 1	Patient 2	Normal value range
PR (%)	72.3	41.7	≥32.0
PR + NP (%)	81.4	52.6	≥40.0
Volume (ml)	4.1	2.8	≥1.5
Concentration (×106 ml−1)	49.9	45.8	>15.0
Normal morphology (%)	4.2	4.2	≥4.0

PR: progressive; NP: nonprogressive

Patient 1 was a 34-year-old man with a 5-year history of primary infertility, while his wife manifested secondary infertility as part of her pregnancy history with an ex-boyfriend. The couple had one failed cIVF attempt (Table 1) in which 13 oocytes were retrieved, but none had formed a PN by 16 h postinsemination. When further cultured for 6 h, 10 of the oocytes unexpectedly showed MPN, indicating a delay in PN formation. Notably, all MPN zygotes showed numerous sperm heads tightly bound to or penetrating the ZP (Figure 1a).

Table 1.

Itemized in vitro fertilization data for two men carrying phospholipase C zeta 1 variants

Patient	Number of oocytes retrieved	Insemination method	Number of oocytes inseminated	Fertilization outcomes	Number of good-quality embryos	Number of blastocysts obtained	Number of blastocysts transferred	Number of pregnancies
Patient 1
Cycle 1	13	IVF	13	MPN: 10 MI/GV: 3	NA	NA	NA	NA
Cycle 2	10a	ICSI + AOA	5	2PN: 5	5	5	2	1
Patient 2
Cycle 1b	5	IVF	5	MPN: 3 0PN: 2	0	0	NA	NA
Cycle 2c	10	IVF	5	MPN: 4 0PN: 1	0	0	NA	NA
		ICSI	5	0PN: 5	0	0	NA	NA
Cycle 3d	4	ICSI + AOA	3	2PN: 1 1PN: 2	3	2e	0	NA
Cycle 4d	5	ICSI + AOA	3	2PN: 2 3PN: 1	2	2	1	1

^aTwo mature oocytes were cryopreserved. ^bSecond wife. ^cThird wife. ^dFourth wife. ^eOne blastocyst derived from a 2PN embryo was determined to be a complex aneuploid embryo, and the other from a 1PN embryo was determined to be a parthenogenetic embryo. IVF: in vitro fertilization; ICSI: intracytoplasmic sperm injection; AOA: artificial oocyte activation; PN: pronucleus; MPN: multiple pronuclei; MI: metaphase of meiosis I; GV: germinal vesicle; NA: not applicable

Figure 1.

Phenotypic features of patients with PLCZ1 variants. (a) Representative images of zygotes from a normal control and two patients. In the normal fertilized oocyte, we observed two pronuclei (black arrowheads) at 16–18 h postinsemination, while in oocytes fertilized by mutant spermatozoa, multiple pronuclei (MPN) were noted at 22–26 h postinsemination, along with numerous intracellular sperm tails (depicted as green staining with anti-acetylated α-tubulin) and a large number of spermatozoa binding to or penetrating the zona pellucida. Scale bar = 20 μm. (b) Localization patterns of PLCζ (red), including acrosomal (Ac), equatorial (Eq), postacrosomal (Pa), and no staining (None), by immunofluorescence staining. Scale bar = 2.5 μm. (c) The proportion of spermatozoa reflecting each PLCζ subcellular localization in normal controls and patients with PLCZ1 variants. (d) Profiles of Ca²⁺ responses induced by spermatozoa from a normal control and from patient 1. For oocytes (n = 2) injected with normal spermatozoa, 2–3 Ca²⁺ spikes were induced in 3 h (green line), and two pronuclei were formed 17 h post-ICSI (top right, black arrowheads); while in oocytes (n = 2) injected with mutant spermatozoa, we observed flat traces (gray line) and no pronucleus (bottom right). Scale bar = 20 μm. PLCZ1: phospholipase C zeta 1; ICSI: intracytoplasmic sperm injection; DAPI: 4’,6-diamidino-2-phenylindole.

MPN zygotes and excessive sperm–ZP binding imply polyspermy. To assess its chromosomal composition, an MPN zygote from patient 1 was analyzed with FISH with probes specific for chromosomes 18, X, and Y (Supplementary Table 3). A total of 21 pronuclei that exhibited different karyotypes were assessed, of which 14 were haploid (n, X or n, Y), three were diploid (2n, XY or 2n, XX), three were triploid (3n, XXY), and one was aneuploid. Moreover, using immunofluorescence staining with an antibody to acetylated α-tubulin, we observed several sperm tails within an MPN zygote (Figure 1a). These observations indicated that there were several spermatozoa entering the oocyte and that the formation of MPN was induced by polyspermy.

Supplementary Table 3.

Chromosomal constitution of an MNP zygote from patient 1

Number of PN	Ploidy and sex chromosomal constitution
9	n, X
5	n, Y
2	2n, XY
1	2n, XX
3	3n, XXY
1	Chaotic
Total=21

PN: pronucleus; MPN: multiple pronuclei

Patient 2 reported that he had previously fathered a child via natural conception with his first wife. Following that, the man showed unexplained infertility with three later wives over the next 20 years. He was referred to our hospital at the age of 40 years and then underwent two successive and failed cIVF attempts with his second and third wives, both of which exhibited a high proportion of MPN zygotes and excessive sperm–ZP binding (Table 1 and Figure 1a). Although half of the oocytes from the man’s third wife were subjected to ICSI, all resulted in total fertilization failure.

Identification of biallelic PLCZ1 variants in men who generated polyspermy on cIVF

The primary infertility of patient 1 and the repeatedly abnormal fertilization (i.e., MPN) after cIVF for patient 2 with different partners led us to consider a genetic origin for the phenomena. To determine the causative gene variants, we carried out WES and Sanger sequencing and identified biallelic PLCZ1 variants in both men (Supplementary Table 4 and 5). Patient 1 carried compound heterozygous variants at c.11+4T>C and c.588C>A (p.Cys196Ter), with the former present in his father. Patient 2 carried two heterozygous variants at c.588C>A (p.Cys196Ter) and c.677A>T (p.Lys226Ile). Since parental DNA was unavailable, we used TA-clone library sequencing and uncovered two heterozygous variants located on separate DNA strands, indicating that they had different parental origins (Figure 2a). The altered amino acid residues Cys196 and Lys226 were within the functional X catalytic domain, and these are highly conserved across species (Figure 2b and 2c).

Supplementary Table 4.

Filtering of whole-exome sequencing variants in the patients

Filtering criteria	Patient 1	Patient 2
Total variants	99 967	99 553
Rare variants (MAF <0.01) in public databasesa	4598	4581
Exonic nonsynonymous or splice-site variantsb, or coding INDELs	846	890
Genes highly or specifically expressed in human testis	32	37
Homozygous or compound heterozygous variants (excluding X and Y chromosomes)	2	6
Variants in a gene identified in both patients	2 (PLCZ1)	2 (PLCZ1)

^aPublic database includes 1000 genomes, the ExAC and gnomAD browsers; ^bA splice-site variant is defined if it is within 10 bp away from an exon/intron boundary. PLCZ1: phospholipase C zeta 1; MAF: minor allelic frequency; INDELs: insertions and deletions

Supplementary Table 5.

Overview of the phospholipase C zeta 1 variants observed in the patients

Patients	Genomic position on Chr12 (bp)	cDNA changea	Protein change	Mutation type	ExAC_easb	GnomAD_easb
1	18865902	c.588C > A	p. Cys196Ter	Stop gain	0.00023	0.00022
	18890291	c.11+4T > C	NA	Splicing	0	0.00005
2	18865902	c.588C > A	p. Cys196Ter	Stop gain	0.00023	0.00022
	18865813	c.677A > T	p. Lys226Ile	Missense	0	0

^aGeneBank NM_033123.4; ^bAllele frequency of corresponding variants in the East Asian population according to the gnomAD and ExAC browsers. NA: not applicable

Figure 2.

Genotypic features of the patients presenting with polyspermy in cIVF and effects of PLCZ1 variants on protein expression and structure. (a) Pedigrees of two families with PLCZ1 variants. The filled squares indicate infertile male individuals, and the arrows indicate the probands. The PLCZ1 genotype for each subject is shown, with WT indicating a normal allele and MT indicating the mutant allele. Sanger sequencing and TA-clone library sequencing confirmed the presence of biallelic PLCZ1 variants (MT 1–MT 3) in family 1 II-1 and family 2 II-1. (b) Locations of the variants are shown in the genomic structure (top) and domain organization (bottom) of PLCZ1. (c) Cys196 and Lys226 residues are highlighted in red boxes and conserved across the selected species. (d) Immunoblotting of transfected CHO-K1 cells probed with antibody against mCherry at the C-terminal of recombinant human PLCζ. (e) Three-dimensional structures of the PLCζ protein (upper left panel). PLCZ1 variants resulted in a C-terminal truncated protein (lower left panel) or disrupted the ion pair formed by wild-type PLCζ (right panels). PLCZ1: phospholipase C zeta 1; WT: wild-type; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; cIVF: conventional in vitro fertilization; MT: mutation; Cys: Cysteine; Ter: termination; Lys: Lysine; Ile: Isoleucine; EF: EF-hand domain; X: X domain; Y: Y domain; C2: C2 domain.

Effects of PLCZ1 variants on protein expression and structure

To test the effects of PLCZ1 variants, expression plasmids carrying wild-type or mutant PLCZ1 cDNA tagged with fluorescent mCherry at the C-terminus were transfected into CHO-K1 cells. In cells transfected with wild-type PLCZ1, immunoblotting for mCherry revealed a band of 102 kDa corresponding to the anticipated size for recombinant human PLCζ ([74 + 28] kDa mCherry) that was absent from untreated cells (Figure 2d). No signal was detected in cells transfected with p.Cys196Ter. In p.Lys226Ile-transfected cells, a decreased protein level was detected (Figure 2d and Supplementary Figure 1 ^{(35.9KB, tif)} ). From the predicted 3D structure of the PLCζ protein (Figure 2e, upper left panel), the substitution of Lys226 with Ile226 was predicted to disrupt the hydrogen bond to Glu210 (Figure 2e, right panels), which might have further effects on protein stability and activity. The variant c.11+4C>T was predicted to be disease causing by MutationTaster18 and deleterious by SpliceAI.19

Effect of variants on sperm PLCζ function and subsequent defects in evoked Ca²⁺ oscillations

PLCZ1 encodes an acknowledged sperm-borne oocyte-activating factor (SOAF) that is termed PLCζ, stimulates oocyte activation, and engages the mechanisms that block polyspermy by triggering Ca²⁺ oscillations. To investigate the effects of these variants in vivo, we evaluated PLCζ expression and localization in patients’ spermatozoa with immunofluorescence staining (Figure 1b and 1c). In normal fertile controls (n = 3), 84.1% ± 3.4% (mean ± s.d.) of spermatozoa showed PLCζ signals within the head, located principally in the acrosomal, equatorial, and postacrosomal regions. By contrast, 97.5% and 95.6% of spermatozoa from patients 1 and 2, respectively, showed a complete absence of PLCζ staining within the head. These results suggest that PLCZ1 variants lead to the absence of functional PLCζ in spermatozoa.

To examine oocyte activation ability further, each MII oocyte matured in vitro from an immature donor oocyte was injected using an ICSI pipette with a spermatozoon from a fertile control or from patient 1, and resultant Ca²⁺ oscillations were monitored. Spermatozoa with unmutated PLCZ1 genes triggered 2–3 Ca²⁺ spikes in 3 h postinjection, while flat traces were noted following injection of mutant spermatozoa (Figure 1d). Seventeen hours after ICSI, two PN formed in the two oocytes injected with nonmutant spermatozoa (Figure 1d, top right) but not in the oocytes injected with mutant spermatozoa, even after a 30-h observation period (Figure 1d, bottom right).

PLCζ-associated polyspermy and rescue by ICSI with AOA

After excluding any rare mutations in their wives’ PLCZ1, we conducted AOA in the subsequent ICSI attempts of both men. All five (100.0%) oocytes from patient 1 and three of the six oocytes from patient 2 were fertilized and developed to blastocysts that were cryopreserved (Table 1). After embryo selection, vitrified blastocysts were thawed and transferred in the following frozen embryo transfer cycles, resulting in a live birth for patient 1 and a singleton pregnancy for patient 2.

Correlation of the proportion of spermatozoa expressing PLCζ and fertilization rates after cIVF

We compared the expression and localization of sperm PLCζ among the three groups that showed differential fertilization outcomes after cIVF, i.e., NF, LF, and MPN groups.

The demographics and baseline values for the three groups are compared in Supplementary Table 6 and 7. The age and body mass index (BMI), as well as semen parameter values (sperm concentration, motility, normal morphology, and acrosin activity), did not differ among groups, except for the number of follicles >12 mm on the trigger day, which was less in the MPN group compared to the NF group (7.6 vs 11.3, respectively; P = 0.0199).

Supplementary Table 6.

Baseline data and semen parameters of males in groups with different fertilization outcomes after conventional in vitro fertilization

Characteristics	NF group	LF group	MPN group
Number of patients	31	18	9
Age (year)	33.4±4.1	33.0±2.8	34.4±6.2
BMI (kg m−2)	25.5±4.0	24.9±3.4	24.4±3.9
Semen parameters
Concentration (×106 ml−1)	63.0 (53.0–73.0)	68.0 (29.0–69.5)	61.0 (49.5–69.0)
PR (%)	33.9±10.0	30.4±10.3	29.6±16.0
Normal morphology (%)	4.0 (3.3–4.4)	3.9 (2.9–4.2)	3.7 (2.6–4.2)
Acrosin (U/L)	82.3±36.2	83.0±32.3	66.6±29.4

Values are expressed as percentages, means±s.d., or medians (IQR). IQR: interquartile range; PR: progressive; BMI: body mass index; NF: normal fertilization rate; LF: low fertilization rate; MPN: multiple pronuclei; s.d.: standard deviation

Supplementary Table 7.

Baseline data of females in groups with different fertilization outcomes after conventional in vitro fertilization

Characteristics	NF group	LF group	MPN group
Number of patients	31	18	9
Age (year)	32.3±3.8	30.8±3.1	32.8±5.5
BMI (kg m−2)	22.0 (20.4–23.8)	21.3 (19.0–22.6)	24.0 (18.4–25.6)
AMH (ng ml−1)	2.2 (1.7–4.1)	2.6 (1.9–5.2)	1.3 (0.6–2.8)
Trigger day
hLH (IU/L)	1.6 (1.3–2.1)	1.9 (1.6–2.6)	2.4 (1.8–3.2)
Estradiol/follicle (pg ml−1)	281.6±83.0	268.5±132.2	310.7±69.9
Progestin (ng ml−1)	0.6±0.4	0.7±0.4	0.7±0.3
Number of follicles ≥12 mm	11.3±4.2	11.9±4.0	7.6±4.0a
Number of oocytes retrieved	11.3±5.3	12.2±6.5	8.3±7.6

^aP=0.0199. Values are expressed as means±s.d. or medians (IQR). IQR: interquartile range; NF: normal fertilization rate; LF: low fertilization rate; MPN: multiple pronuclei; BMI: body mass index; AMH: anti-Mullerian hormone; hLH: human luteinizing hormone; s.d.: standard deviation

As shown in Figure 3a and Supplementary Table 8, the proportion of spermatozoa that expressed PLCζ within the head region (expressed as PLCζ⁺%) was significantly lower in the LF group (64.1%; 95% confidence interval [CI]: 46.8%–84.3%) than that in the NF group (82.2%; 95% CI: 73.1%–87.7%; P = 0.0176), but we noted no difference between the MPN group (89.6%; 95% CI: 78.0%–94.3%) and NF group (P = 0.1296). In addition, there were no statistical differences (P > 0.05) among the three PLCζ localization patterns (Figure 3b–3d). Moreover, we found a significant positive correlation between PLCζ⁺% and both the 2PN rate (P = 0.0005) and total FR (P = 0.0008), but no significant correlation between the MPN rate and any of the PLCζ parameters evaluated (P > 0.05), as shown in Supplementary Table 9. ROC curve analysis based on the PLCζ data from the NF and LF groups indicated that the optimal cutoff value associated with low FR after cIVF (total FR ≤30.0%) was 56.7% for the PLCζ⁺% (area under the ROC curve: 0.705; P = 0.0176; Figure 3e).

Figure 3.

Comparison of the expression and localization patterns of sperm PLCζ among the three groups with different fertilization outcomes after cIVF. (a) Comparison of the proportion of spermatozoa expressing PLCζ within the head. (b) Comparison of the proportion of spermatozoa exhibiting a PLCζ signal within the acrosomal region. (c) Comparison of the proportion of spermatozoa exhibiting a PLCζ signal within the equatorial region. (d) Comparison of the proportion of spermatozoa exhibiting a PLCζ signal within the postacrosomal region. (e) The accompanying ROC curve analysis. The red line and error bars represent the median value and 95% CI, respectively. NS: nonsignificant differences. The red dashed line in a indicates the cutoff value (56.7%) of the proportion of spermatozoa exhibiting PLCζ associated with the low fertilization rate. MPN group: number of PN ≥3, MPN rate ≥50.0%, n = 9; LF group: total FR ≤30.0%, n = 18; NF groups: 2PN rate ≥50.0%, 1PN + MPN rate <30.0%, n = 31. PLCζ: phospholipase C zeta; cIVF: conventional in vitro fertilization; ROC: receiver operating characteristic; CI: confidence interval; NF: normal fertilization rate; LF: low fertilization rate; MPN: multiple pronuclei; PN: pronucleus; FR: fertilization rate.

Supplementary Table 8.

Proportion (%) of spermatozoa that expressed phospholipase C zeta in groups with different fertilization outcomes after conventional in vitro fertilization

Localization of PLCζ	NF group	LF group	MPN group
PLCζ+ (Ac + Eq + Pa)	82.2% (73.1%–87.7%)	64.1% (46.8%–84.3%)a	89.6% (78.0%–94.3%)
Ac	70.8% (42.8%–82.2%)	52.4% (34.6%–75.0%)	84.6% (46.2%–90.5%)
Eq	2.6% (0.6%–6.6%)	2.4 (1.4%–9.5%)	1.4% (0.0%–21.7%)
Pa	5.1% (2.0%–20.0%)	3.5% (1.0%–5.6%)	1.8% (0.5%–3.7%)

^aP=0.0176 (LF group vs NF group). Values are expressed as medians (IQR). IQR: interquartile range; NF: normal fertilization rate; LF: low fertilization rate; MPN: multiple pronuclei; Ac: acrosomal; Eq: equatorial; Pa: post-Ac sperm segments; PLCζ: phospholipase C zeta

Supplementary Table 9.

Correlations between fertilization rates after conventional in vitro fertilization and the proportion of spermatozoa expressing phospholipase C zeta/localization patterns

	2PN ratea	Total FRa	MPN rateb
PLCζ+%	0.459c	0.438d	0.132
Localization patterns
Ac%	0.173	0.143	0.187
Eq%	−0.004	0.011	0.028
Pa%	0.222	0.247e	−0.213

^aAnalyses based on samples from NF and LF groups; ^bAnalyses based on samples from NF and MPN groups; ^cP=0.0005; ^dP=0.0008; ^eP=0.044, significant correlations. FR: fertilization rate; IVF: in vitro fertilization; PLCζ⁺%: the proportion of spermatozoa expressing phospholipase C zeta; LF: low FR; PN: pronucleus; MPN: multiple pronuclei; Ac: acrosomal; Eq: equatorial; Pa: post-Ac sperm segments

DISCUSSION

Our study identified PLCZ1 as a causative gene of polyspermy in humans, similar to the phenotype observed in Plcz1^−/− males.20,21,22 We have also established a significant and positive correlation between PLCζ⁺% and FRs after cIVF.

Structurally, PLCζ consists of two pairs of N-terminal EF-hand domains, a catalytic core that encompasses the X and Y domains that are essential for phosphatidylinositol 4,5-bisphosphate (PIP2) hydrolysis, and a C2 domain at the C-terminus.2 Notably, point mutations within the catalytic domain of PLCζ have been associated with a loss in enzymatic activity and consequent male infertility.10,13,14 We, herein, identified three PLCZ1 variants in two unrelated infertile males, two of which were within the X catalytic domain (p.Cys196Ter and p.Lys226Ile; Figure 2b). The nonsense variant p.Cys196Ter has also been reported in several infertile men with ICSI fertilization failure. Our studies in CHO-K1 cells indicated that the variant p.Cys196Ter contributed to either a truncated protein lacking the C-terminal domains or alternatively nonsense-mediated mRNA decay, and the variant p.Lys226Ile resulted in reduced protein level. More than 95% of the mutant spermatozoa from both patients lacked PLCζ expression within the head and were neither able to induce Ca²⁺ oscillations nor form PN after ICSI. Such inconsistent expression in culture cells and patients’ samples was not specific to these PLCZ1 variants. Hou et al.23 reported a frameshift mutation in the KASH domain containing 5 (KASH5) gene and detected a truncated KASH5 protein in cultured cells but not in the patient’s testis. In a study on male infertility and ICSI fertilization failure, four missense PLCZ1 mutations were identified in male probands from five independent families, as homozygous or compound heterozygous status with either a nonsense, splicing, or deletion mutation. Similarly, protein levels of PLCζ are nearly undetectable in semen samples from all patients by immunoblotting.14 These intriguing phenomena could be due to the presence of alterations in distal regulatory regions of PLCZ1 gene. Another explanation may be that, as in silico analysis suggested, the substitution of Lys226 with Ile226 at this key residue probably weakens the stability of the PLCζ protein to a greater extent in vivo than in high-expression cells. The underlying mechanisms are very attractive and warrant further study.

It is intriguing that the two infertile males in this study exhibited polyspermy after cIVF. A similar phenotype was reported in Plcz1 knockout (KO) mice, in which dramatically increased polyspermy and delayed PN formation were observed in oocytes fertilized by PLCζ-null spermatozoa.20,21,22 Loss of PLCζ abolishes the ability of spermatozoa to induce Ca²⁺ changes in ICSI, while, in contradistinction, PLCζ-null spermatozoa induced atypical patterns of Ca²⁺ spikes in cIVF, with significantly diminished amplitudes and numbers that were insufficient to trigger oocyte activation or to block polyspermy. However, as additional spermatozoa fuse with the oolemma, the numbers of Ca²⁺ spikes increase and multiple PN ultimately form when a threshold of intracellular Ca²⁺ spikes is attained.21 In this study, polyspermy and delayed PN formation exhibited by males carrying PLCZ1 variants were congruent with the reported phenotype of Plcz1 KO mice and further support the concept that PLCζ is also the physiological trigger of the Ca²⁺ oscillations that ensure monospermic fertilization in humans. AOA is currently the only available treatment option for oocyte activation deficiency (OAD), and as we expected, a combination of ICSI and AOA treatments successfully rescued the polyspermic phenotype and provided good-quality blastocysts for transfer in both couples – achieving a live birth and an ongoing pregnancy, respectively.

Unexpectedly, PLCζ analyses in an additional 58 males after cIVF revealed that a decline in PLCζ⁺% was significantly correlated with a low FR but not to a high rate of MPN. In PLCζ-null mice, activation failure (i.e., no PN but the presence of a fertilization cone[s]) was present at attenuated sperm concentrations, while the increase in PLCζ-null sperm concentration and polyspermy rose to approximately 80%.21 These observations imply that a high proportion of PLCζ-negative spermatozoa (such as that in PLCZ1-mutant patients) might be required to allow more spermatozoa to fuse with the oocyte and produce sufficient Ca²⁺ spikes to reach a threshold that triggers PN formation. Precise mechanisms that underlie the roles of PLCζ in cIVF outcomes remain to be fully elucidated by further study.

ROC analysis revealed that the appropriate cutoff value for PLCζ⁺% was 56.7%, below which males were likely to exhibit low FRs with cIVF protocols. Intriguingly, all 31 males in the NF group possessed a high PLCζ⁺% (over 56.7%), while eight males with a proportion below the cutoff value were all from the LF group. We attribute the latter’s low FR to sperm-related OAD, with AOA treatment recommended for subsequent ICSI cycles.24,25 Although it is worth noting that 10 of the 18 males in the LF group showed a PLCζ⁺% higher than 56.7%, reasons for cIVF failure are not restricted to PLCζ deficiency and may include oocyte factors such as inadequate nuclear and cytoplasmic maturation as well as other sperm factors. Such cases are likely to benefit from modified superovulation protocols that improve oocyte maturation and from ICSI treatment that bypasses the steps of sperm–oocyte interaction.24,26 Follow-up studies with larger numbers of patients are required to validate the sensitivity, specificity, and accuracy of PLCζ as a diagnostic biomarker in cIVF outcomes.

In conclusion, we present two unrelated males with infertility and cIVF polyspermy associated with PLCZ1 variants. Moreover, we also demonstrate that PLCζ⁺% is significantly and positively correlated with cIVF FR. These findings improve the understanding of the molecular basis of polyspermy in humans, highlighting the potential value of PLCζ as a fertility marker for cIVF outcomes.

AUTHOR CONTRIBUTIONS

CD and GL conceived and designed the study and critically commented on the manuscript. JFC and HXW conducted most of the experiments. CD, JFC, and HXW analyzed and interpreted the data and wrote the manuscript. SCZ reviewed the experimental data and helped draft the manuscript. YRW and JD participated in the immunofluorescence staining and HOAT experiments. DHC performed the FISH experiment. FG and GXL provided the clinical samples. All authors read and approved the final manuscript.

COMPETING INTERESTS

All authors declare no competing interests.

Supplementary Figure 1

Immunoblotting analysis of wild-type and mutant PLCζ-mCherry fusion proteins in CHO-K1 cells. All values are means ± s.d. from three independent experiments. PLCζ: phospholipase C zeta; s.d.: standard deviation