TRPM7 and Ca_V3.2 channels mediate Ca²⁺ influx required for egg activation at fertilization

Significance

At fertilization, repetitive calcium signals are induced in mammalian eggs that are essential for activating embryo development. These signals are supported by calcium influx from the surrounding medium. Two ion channels supporting calcium influx into eggs, Ca_V3.2 and TRPV3, were identified previously. Here we find that a third channel, TRPM7, is essential for this function. In the absence of both TRPM7 and Ca_V3.2, the calcium signals induced by the fertilizing sperm stop prematurely. Female mice carrying eggs lacking TRPM7 and Ca_V3.2 are subfertile. Because TRPM7 channel function is modulated by ions in culture medium, these findings highlight the importance of exact culture medium composition used during laboratory procedures for human assisted reproduction therapies.

Abstract

The success of mammalian development following fertilization depends on a series of transient increases in egg cytoplasmic Ca²⁺, referred to as Ca²⁺ oscillations. Maintenance of these oscillations requires Ca²⁺ influx across the plasma membrane, which is mediated in part by T-type, Ca_V3.2 channels. Here we show using genetic mouse models that TRPM7 channels are required to support this Ca²⁺ influx. Eggs lacking both TRPM7 and Ca_V3.2 stop oscillating prematurely, indicating that together they are responsible for the majority of Ca²⁺ influx immediately following fertilization. Fertilized eggs lacking both channels also frequently display delayed resumption of Ca²⁺ oscillations, which appears to require sperm–egg fusion. TRPM7 and Ca_V3.2 channels almost completely account for Ca²⁺ influx observed following store depletion, a process previously attributed to canonical store-operated Ca²⁺ entry mediated by STIM/ORAI interactions. TRPM7 serves as a membrane sensor of extracellular Mg²⁺ and Ca²⁺ concentrations and mediates the effects of these ions on Ca²⁺ oscillation frequency. When bred to wild-type males, female mice carrying eggs lacking TRPM7 and Ca_V3.2 are subfertile, and their offspring have increased variance in postnatal weight. These in vivo findings confirm previous observations linking in vitro experimental alterations in Ca²⁺ oscillatory patterns with developmental potential and offspring growth. The identification of TRPM7 and Ca_V3.2 as key mediators of Ca²⁺ influx following fertilization provides a mechanistic basis for the rational design of culture media that optimize developmental potential in research animals, domestic animals, and humans.

All animal species utilize a large rise in intracellular cytoplasmic Ca²⁺ {[Ca²⁺]_i} levels to trigger development of the fertilized egg (1). In mammals, prophase I, germinal vesicle (GV) stage oocytes prepare for fertilization during maturation by accumulating Ca²⁺ into endoplasmic reticulum (ER) stores (2, 3). This change accompanies nuclear maturation into a metaphase II (MII)-arrested oocyte, henceforth referred to as an “egg.” Introduction of PLCζ by the fertilizing sperm triggers production of inositol 1,4,5-trisphosphate (IP₃), which causes Ca²⁺ release from the egg ER stores by activating its cognate receptor (4–6). [Ca²⁺]_i typically remains elevated for several minutes before returning to baseline as a consequence of Ca²⁺ reuptake into ER stores and Ca²⁺ extrusion out of the egg. Repetitive Ca²⁺ release events follow that cause persistent transient increases in [Ca²⁺]_i, known as Ca²⁺ oscillations, which continue for several hours following fertilization and are essential for activating successful embryo development (7). However, persistence of these Ca²⁺ oscillations depends on the continued presence of Ca²⁺ in the extracellular medium to provide a source to replenish ER stores (8, 9). Hence, Ca²⁺ influx is essential both during oocyte maturation to generate proper ER Ca²⁺ stores and following fertilization to support persistent Ca²⁺ oscillations.

Several recent papers have contributed to our understanding of the ion channels that support Ca²⁺ influx in maturing oocytes and in eggs. Ca_V3.2 is a voltage-activated T-type Ca²⁺ channel, whose pore-forming alpha subunit is encoded by the Cacna1h gene. Ca_V3.2 supports Ca²⁺ influx during maturation and following fertilization but is not solely responsible for this function (10). Store-operated Ca²⁺ entry (SOCE), mediated by the Ca²⁺ sensor STIM proteins and plasma membrane ORAI channels, is a common mechanism for cells to trigger Ca²⁺ reuptake in response to loss of Ca²⁺ from ER stores. However, studies using knockout mouse lines demonstrated that SOCE is not required during mouse oocyte maturation or following fertilization (11). Members of a large family of cation-permeable channels known as transient receptor potential (TRP) channels are widely expressed in both excitable and nonexcitable cell types (12). One TRP family member, TRPV3, is present on the mouse egg plasma membrane and mediates Ca²⁺ influx but does not have a requisite role at fertilization (13). A channel with characteristics of transient receptor potential cation channel subfamily M, member 7 (TRPM7) is also expressed on the mouse egg surface (14). TRPM7 is a ubiquitously expressed, constitutively active channel permeable to many divalent cations, including Ca²⁺ and Mg²⁺, though it is inhibited by low millimolar concentrations of Mg²⁺ (15, 16). The TRPM7 global knockout mouse has an embryonic lethal phenotype, indicating that it has an essential function in development (17). Interestingly, TRPM7 has both ion channel and kinase activity and is therefore known as a “chanzyme.” Pharmacological modulators of TRPM7 function were used to demonstrate that a TRPM7-like channel can support Ca²⁺ influx into mouse eggs and preimplantation embryos; a role for this channel at fertilization was not tested in this study (14). However, many ion channel agonists and inhibitors have off-target effects, so there are caveats to the interpretation of these experiments. The Mg²⁺ sensitivity of TRPM7 makes it an intriguing candidate to have functional importance at fertilization because it was recently demonstrated that Ca²⁺ oscillation frequency, which depends on Ca²⁺ influx, is inversely related to Mg²⁺ concentration in the extracellular medium (18).

Here we show using a conditional knockout approach that TRPM7 has an essential role in supporting Ca²⁺ influx in mouse oocytes and eggs. We also demonstrate that Ca²⁺ influx previously attributed to SOCE is to a large extent mediated by the nearly constitutive activity of TRPM7. Additional experiments reveal that TRPM7 is involved in modulation of Ca²⁺ oscillation frequency in response to changing concentrations of extracellular divalent cations. Finally, we show using a Trpm7/Cacna1h double knockout model that these two channels are the principal mediators of Ca²⁺ influx that supports sustained Ca²⁺ oscillations following fertilization in mice.

Results

Genetic Deletion of Trpm7 Causes Loss of Classical TRPM7 Current.

Oocyte-specific deletion of Trpm7 was achieved by mating mice carrying a Trpm7 allele modified to contain LoxP sites flanking exon 17 (Trpm7^f) with a Gdf9-cre transgenic line (17, 19). Depletion of Trpm7 transcripts in cre-positive Trpm7 conditional knockout (cKO) oocytes was confirmed by real-time RT-PCR. Transcripts containing exon 17 were nearly undetectable, and levels of Trpm7 mRNA with intact 3′ exons were greatly reduced in Trpm7 cKO MII eggs compared with Trpm7^f/f controls (Fig. 1A). Whole-cell patch clamp was performed to determine whether functional TRPM7 channels could be detected in Trpm7 cKO eggs. A voltage ramp (−100 to +100 mV) elicited an outward current in Trpm7^f/f eggs; a summary of the current at +90 mV is shown (Fig. 1B). When the TRPM7-selective inhibitor NS8593 was applied to control eggs, a significant reduction in outward current was observed at 90 mV following a voltage ramp, as described previously (14, 20). NS8593 had no effect on this current in Trpm7 cKO eggs, indicating an absence of functional NS8593-sensitive TRPM7 channels. Ca²⁺ imaging using fura-2 ascertained the effect of two TRPM7 channel activators, naltriben and mibefradil (also a T-type channel inhibitor), on control and Trpm7 cKO eggs (14, 21, 22). An increase in [Ca²⁺]_i was observed when control eggs were exposed to either substance, but very little response was evident for Trpm7 cKO eggs under the same conditions (Fig. 1 C and D), further supporting the functional loss of TRPM7 channels in Trpm7 cKO oocytes.

Fig. 1.

Genetic deletion of Trpm7 causes loss of classical TRPM7 current. (A) Real-time RT-PCR of total RNA from Trpm7^f/f (control) or Trpm7^f/f;Gdf9-cre (Trpm7 cKO) eggs. Schematic shows partial genomic structure of Trpm7-floxed allele. Exons, blue boxes; loxP sites, red triangles; primer locations, gray arrows. Graph shows expression relative to that in control eggs, mean ± SEM. Eggs were collected from n = 5 individual control and n = 3 Trpm7 cKO mice. *P < 0.05 relative to control with same primer pair, t test. (B) Whole-cell patch-clamp recordings from eggs of control or Trpm7 cKO mice in response to a ramp (−100 to +100 mV). Graph shows current amplitude from individual eggs at +90 mV before (black symbols) and after (red symbols) addition of NS8593; bars indicate mean ± SEM. Control eggs, n = 8; Trpm7 cKO eggs, n = 5. *P < 0.05, paired t test. (C) Ca²⁺ response of control and Trpm7 cKO oocytes to the indicated concentrations of naltriben. Upper graphs show F340/F380 ratio in individual oocytes (gray) and the mean ratio (black) from both groups in one representative experiment. Lower graphs show mean area under the curve ± SEM for the indicated treatment. n = 3 biological replicates, a total of 18 control and 18 Trpm7 cKO oocytes were imaged. *P < 0.05, Mann–Whitney U test (40 μM naltriben); *P < 0.05, t test with Welch’s correction (80 μM naltriben). (D) Ca²⁺ response of control and Trpm7 cKO oocytes to the indicated concentrations of mibefradil (mib). Upper graphs in this panel and in all subsequent calcium traces show F340/F380 ratio in individual oocytes (gray) and the mean ratio (black) from both groups in one representative experiment. Lower graphs show mean area under the curve ± SEM for the indicated treatment. n = 2 biological replicates, a total of 23 control and 23 Trpm7 cKO oocytes were imaged. *P < 0.05, t test with Welch’s correction.

TRPM7 Is the Primary Channel Responsible for Spontaneous and Store Depletion-Induced Ca²⁺ Influx in Oocytes and Eggs.

To test whether TRPM7 supports Ca²⁺ influx in GV oocytes, we placed control or Trpm7 cKO oocytes in nominally Ca²⁺-free medium and then added Ca²⁺ to the extracellular medium while monitoring [Ca²⁺]_i. Of note, extended time in Ca²⁺-free medium was associated with an unhealthy appearance and late lysis in a significant number of both control and Trpm7 cKO oocytes; therefore, only oocytes that had a normal appearance following completion of imaging were included in this analysis. Control GV oocytes had a large increase in [Ca²⁺]_i starting ∼2–3 min following Ca²⁺ addition, whereas [Ca²⁺]_i in Trpm7 cKO oocytes changed minimally (Fig. 2A). We next tested whether spontaneous Ca²⁺ entry via TRPM7 contributed to Ca²⁺ influx that has previously been attributed to SOCE by performing a standard SOCE assay on control and Trpm7 cKO oocytes. GV oocytes were placed in nominally Ca²⁺-free medium and then thapsigargin was added to cause ER Ca²⁺ store depletion. The anticipated temporary increase in [Ca²⁺]_i was observed, and was not different between control and Trpm7 cKO GV oocytes (Fig. 2B). However, when Ca²⁺ was then added to the extracellular medium, control GV oocytes displayed a large increase in [Ca²⁺]_i, whereas Trpm7 cKO oocytes had a barely detectable increase. When a similar assay was performed on control and Trpm7 cKO eggs, [Ca²⁺]_i responses occurred in both control and Trpm7 cKO eggs, but overall response was lower for the cKO (Fig. 2C). Together, these findings indicate that thapsigargin-releasable Ca²⁺ stores in Trpm7 cKO oocytes and eggs do not differ from controls. However, Ca²⁺ influx observed after store depletion and Ca²⁺ readdition is greatly reduced in Trpm7 cKO GV oocytes and partially reduced in Trpm7 cKO eggs, indicating that TRPM7 is a major mediator of this influx.

Fig. 2.

TRPM7 is primarily responsible for spontaneous and store depletion-induced Ca²⁺ influx in oocytes and eggs. (A) Spontaneous Ca²⁺ influx into control and Trpm7 cKO GV oocytes. (Left) Representative traces; (Right) mean area under the curve ± SEM. n = 6 biological replicates, a total of 60 control and 59 Trpm7 cKO oocytes were imaged. *P < 0.05, Mann–Whitney U test. (B and C) Measurements of ER Ca²⁺ stores and Ca²⁺ influx after thapsigargin (Tg)-induced Ca²⁺ store depletion in control and Trpm7 cKO GV oocytes (B) or MII eggs (C). (Left) Representative traces; (Right) mean area under the curve ± SEM following Tg addition and following Ca²⁺ addition. Six biological replicates were performed for GV oocytes, and three for MII eggs, and the total number of cells in each group was: control GV oocytes, 46; Trpm7 cKO GV oocytes, 45; control MII eggs, 33; and Trpm7 cKO MII eggs, 35. *P < 0.05, Mann–Whitney U test (B), t test with Welch’s correction (C).

Loss of TRPM7 Alters Ca²⁺ Oscillations and Causes Abnormal Offspring Growth.

To determine whether TRPM7 mediates Ca²⁺ influx following fertilization, control and Trpm7 cKO eggs were fertilized while monitoring [Ca²⁺]_i (Fig. 3A). Consistent with the similar Ca²⁺ stores, there was no difference in the duration of the first Ca²⁺ transient. However, the oscillation frequency was significantly reduced in the Trpm7 cKO eggs, and the number of eggs that had persistent Ca²⁺ oscillations for more than 60 min was also reduced (Fig. 3 B and C). Eggs respond to increases in extracellular Ca²⁺ by increasing their Ca²⁺ oscillation frequency (8, 23). To determine whether this phenomenon was mediated by TRPM7, we fertilized eggs in the presence of 2 mM Ca²⁺ and then increased extracellular Ca²⁺ to 10 mM. In response, control eggs dramatically increased their oscillation frequency, whereas Trpm7 cKO eggs had only a subtle increase (Fig. 3 D and E). Addition of extracellular Mg²⁺ reduced oscillation frequency, consistent with the inhibitory action of Mg²⁺ on TRPM7 channel function. To further assess the role of TRPM7 in mediating the Mg²⁺ response, Mg²⁺ concentrations were varied in the presence of 2 mM Ca²⁺. As expected, increasing Mg²⁺ to 1.8 mM significantly slowed oscillation frequency, whereas diluting Mg²⁺ to 0.36 mM increased frequency in control eggs; however, these responses were substantially blunted in eggs lacking TRPM7 (Fig. 3 F and G and SI Appendix, Fig. S1). These findings indicate that TRPM7 is a key regulator of Ca²⁺ influx in response to alterations in extracellular Ca²⁺ and Mg²⁺ ion concentrations, although other channels may also participate. To test whether loss of TRPM7 impacted fertility, we performed a 6-mo breeding trial by mating wild-type males to female mice carrying either control or Trpm7 cKO oocytes. There was no difference in the number of live pups per litter (Fig. 3H); however, male pups derived from Trpm7 cKO oocytes gained less weight after weaning than control males (Fig. 3I). Furthermore, the SD of the weight measurements was significantly increased for the males from Trpm7 cKO oocytes (Fig. 3J). Because TRPM7 activity was required during a developmental period when mouse embryos are undergoing epigenetic reprogramming, we tested whether these weight differences could be related to epigenetic alterations in the offspring, in particular at imprinted genes, which have a role in regulating growth. We measured DNA methylation in liver tissue from the offspring at 3 wk of age. There was no difference in either methylation at repetitive elements measured using the luminometric methylation assay (LUMA) (24) or in methylation of imprinted genes at differentially methylated regions that control their expression (SI Appendix, Fig. S2A). These findings suggest that the Ca²⁺ oscillatory pattern immediately following fertilization has a long-term impact on the growth trajectory of male offspring, but that it is not caused by alterations in DNA methylation.

Fig. 3.

Loss of TRPM7 reduces Ca²⁺ oscillation frequency and persistence in mouse eggs and causes abnormal growth trajectory in male offspring. (A–C) Ratiometric Ca²⁺ imaging of n = 22 control and 19 Trpm7 cKO eggs was performed during IVF in three independent experiments. (A) Representative traces. (B) Oscillation frequency, mean ± SEM, *P < 0.05, t test. (C) Number of eggs with persistence of Ca²⁺ oscillations after 60 min. *P < 0.05, Fisher’s exact test. (D and E) Response of n = 18 control and 13 Trpm7 cKO eggs to the indicated changes in Ca²⁺ and Mg²⁺ concentration in three independent experiments. (D) Representative traces. (E) Graph of oscillation frequency during incubation in the indicated conditions, mean ± SEM. Different letters above columns indicate significant difference, P < 0.05, two-way ANOVA with Tukey’s multiple comparisons test. (F and G) Response of n = 29 control and 33 Trpm7 cKO eggs to the indicated changes in Mg²⁺ concentration in three independent experiments. (F) Representative traces. (G) Graph of oscillation frequency during incubation in the indicated conditions, mean ± SEM. Different letters above columns indicate significant difference, P < 0.05, two-way ANOVA with Tukey’s multiple comparisons test. (H–J) Breeding trial of wild-type males bred for 6 mo to females carrying either control or Trpm7 cKO oocytes; seven breeding pairs per group. (H) Average number of live pups per litter, mean ± SEM. (I) Weight of all male and female offspring over time, mean ± SEM, *P < 0.05, mixed-model ANOVA. ns, not significant. N = 24 female pups, 28 male pups from control dams; 32 female pups, 30 male pups from cKO dams. (J) SD of offspring weight over time. Different letters indicate statistically different SDs between groups (mixed-model ANOVA).

Combined Loss of TRPM7 and Ca_V3.2 Dramatically Alters Ca²⁺ Signals and Impairs Fertility.

We previously demonstrated that Ca_V3.2 contributes to ER Ca²⁺ stores and Ca²⁺ influx following fertilization (10), so it was possible that Ca_V3.2 was partially compensating for TRPM7 function in the Trpm7 cKO oocytes. To test this idea, we generated female mice carrying double knockout (dKO) eggs lacking both Ca_V3.2 and TRPM7 (Cacna1h^−/−;Trpm7^f/f;Gdf9-cre). For comparison, we used two sets of controls: wild-type eggs from C57BL/6J females and Ca_V3.2 knockout eggs from Cacna1h^−/−;Trpm7^f/f females (Ca_V3.2 KO), which have the same mixed genetic background (129S4/SvJae and C57BL/6J) as the dKO eggs. Thapsigargin-sensitive Ca²⁺ stores in Ca_V3.2 KO eggs averaged slightly lower than in wild-type eggs, but this decrease did not reach significance in the three-way comparison (Fig. 4A). Ca²⁺ influx following store depletion was also not different between these two groups. In contrast, dKO eggs had significantly lower Ca²⁺ stores and minimal Ca²⁺ influx following store depletion relative to that in both other egg types (Fig. 4A). The minimal Ca²⁺ influx in the dKO eggs (Fig. 4A) could be explained by the presence of functional TRPV3 channels (13). These findings indicate that in eggs, both Ca_V3.2 and TRPM7, and possibly TRPV3, support a “SOCE-like” Ca²⁺ influx following store depletion that could be misinterpreted as being mediated by STIM/ORAI interactions.

Fig. 4.

Combined loss of TRPM7 and Ca_V3.2 severely impacts ER Ca²⁺ store accrual and alters Ca²⁺ oscillations following IVF. (A) Measurements of ER Ca²⁺ stores and Ca²⁺ influx after Tg-induced Ca²⁺ store depletion in wild-type, Ca_V3.2 KO, and dKO MII eggs. (Left) Representative traces; (Right) mean area under the curve ± SEM following Tg addition and following Ca²⁺ addition. n = 3 biological replicates; the total number of cells in each group was: wild-type eggs, 27; Ca_V3.2 KO eggs, 29; and dKO eggs, 29. *P < 0.05, one-way ANOVA with Tukey’s multiple comparisons test (for Tg addition), *P < 0.05, Kruskal–Wallis test with Dunn multiple comparisons test (for Ca²⁺ addition). (B–D) Ratiometric Ca²⁺ imaging of 18 Ca_V3.2 KO and 15 dKO eggs was performed during IVF in two independent experiments. (B) Representative traces. (C) First transient duration, mean ± SEM, *P < 0.05, Mann–Whitney U test. (D) Number of eggs with indicated patterns of Ca²⁺ oscillations. Significant difference between patterns in the two groups, P < 0.05, χ² test.

We next determined Ca²⁺ oscillatory patterns following in vitro fertilization (IVF). For this experiment, we used Ca_V3.2 KO eggs as controls to avoid misinterpretation of results that could be related to mouse strain differences. Monitoring of [Ca²⁺]_i during IVF revealed that, consistent with the decrease in Ca²⁺ stores, the duration of the first Ca²⁺ transient following IVF was significantly shorter in the dKO eggs relative to eggs lacking only Ca_V3.2 (Fig. 4 B and C). However, the most dramatic difference between the Ca_V3.2 KO and dKO eggs was that there were only one or two Ca²⁺ transients in the dKO eggs before they stopped oscillating (1.71 transients ± 0.16 SEM, n = 14), which is a pattern highly reminiscent of the oscillatory behavior of fertilized wild-type eggs cultured in nominally Ca²⁺-free medium (9, 23). Interestingly, most of the dKO eggs began oscillating again almost 1 h after the first Ca²⁺ transient (52 min ± 3.7 SEM) (Fig. 4D). These findings are consistent with the idea that together, Ca_V3.2 and TRPM7 support the vast majority of Ca²⁺ influx into eggs during the first hour following fertilization, but that another Ca²⁺ channel or transporter must subsequently either appear on the plasma membrane or be activated to allow for the Ca²⁺ oscillation “restart.”

To determine whether sperm–egg membrane fusion was required for the restart oscillation feature in the dKO eggs, we next tested whether restarting occurred in dKO eggs following fertilization by intracytoplasmic sperm injection (ICSI). As a positive control for the quality of the sperm preparation, we used wild-type eggs fertilized by ICSI; these eggs displayed Ca²⁺ oscillations that had a regular frequency and persisted for over 2 h (Fig. 5A). In contrast, dKO eggs generally had only two initial Ca²⁺ transients followed by one or two transients at a very low frequency (Fig. 5B), but never displayed a restarting pattern similar to that observed following IVF. These findings suggest that there is either a unique signaling event or membrane composition change that accompanies sperm–egg fusion, but not ICSI, and can support Ca²⁺ influx into the fertilized egg in the absence of both Ca_V3.2 and TRPM7.

Fig. 5.

Combined loss of TRPM7 and Ca_V3.2 reduces Ca²⁺ oscillation frequency following ICSI, reduces litter sizes, and alters offspring growth variance. (A and B) Ratiometric Ca²⁺ imaging of 10 wild-type and 23 dKO eggs was performed immediately following ICSI. (A) Representative traces. (B) Number of Ca²⁺ oscillations during the 120-min period following ICSI, mean ± SEM, *P < 0.05, t test with Welch’s correction. (C and D) Preimplantation development of pronuclear-stage embryos derived from wild-type sperm and eggs of the indicated genotypes following in vivo fertilization and zygote collection (C) or in vitro fertilization (D); n = 6–9 females per group. 2C, two-cell stage; 4–8C, 4- to 8-cell stage; blast, blastocyst; mor, morula. (E–G) Breeding trial of wild-type males bred for 6 mo to females carrying either Ca_V3.2 KO or dKO oocytes; seven breeding pairs per group. (E) Average number of live pups per litter, mean ± SEM, *P < 0.05, Mann–Whitney U test. (F) Weight of all male and female offspring over time, mean ± SEM, mixed-model ANOVA. ns, not significant. N = 22 female pups, 20 male pups from Ca_V3.2 KO dams; 15 female pups, 16 male pups from dKO dams. (G) SD of offspring weight over time. Different letters indicate statistically different SDs between groups (mixed-model ANOVA).

To test whether the abnormal Ca²⁺ oscillatory pattern in dKO eggs affected embryo development, we collected zygotes from Ca_V3.2 KO and dKO females mated to wild-type males and then monitored preimplantation embryo development in vitro. There was no significant difference in the number of zygotes collected or in the success or timing of development to the blastocyst stage (Fig. 5C). To determine whether a combination of impaired Ca²⁺ influx and exposure to an in vitro environment during fertilization might further hamper development, eggs from Ca_V3.2 KO and dKO females were fertilized in vitro and monitored for development. Again, development to the blastocyst stage did not differ significantly between the two groups (Fig. 5D). However, a 6-mo breeding trial using wild-type males mated to females carrying either Ca_V3.2 KO or dKO oocytes revealed a significant reduction in the number of pups per litter derived from dKO oocytes (Fig. 5E). Together, these findings indicate that implantation and/or postimplantation development are impaired. Both male and female surviving pups from dKO oocytes had an average weight and growth trajectory similar to those from Ca_V3.2 KO oocytes (Fig. 5F). However, the dKO offspring had a significant increase in variability in their weight from 1 through 8 wk of age (Fig. 5G). This difference could not be explained by alterations in DNA methylation at repetitive elements or methylation of imprinted genes in liver tissue of the offspring (SI Appendix, Fig. S2B). Taken together, these findings suggest that significant abnormalities in Ca²⁺ signals immediately following fertilization in vivo can lead to postimplantation development failure and aberrations in fetal growth.

Discussion

It was shown previously that three types of ion channels that support Ca²⁺ influx are present and functional on the mouse oocyte and/or egg plasma membrane: Ca_V3.2, TRPV3, and a TRPM7-like channel (10, 13, 14). Each of these channels can participate in the accrual of ER Ca²⁺ stores during oocyte maturation, but only Ca_V3.2 has been shown to support Ca²⁺ influx required for a normal Ca²⁺ oscillatory pattern following fertilization (10). The experiments reported here build on these findings to detail the contributions of TRPM7 and Ca_V3.2 to Ca²⁺ homeostasis and Ca²⁺ signaling before and after fertilization. Using a conditional knockout approach, we demonstrate a pivotal role for TRPM7 channels in regulating Ca²⁺ oscillation frequency following fertilization. We also show that together, Ca_V3.2 and TRPM7 channels support the majority of the Ca²⁺ influx responsible for ER Ca²⁺ store accrual during oocyte maturation and are critical for establishment of an appropriate postfertilization Ca²⁺ oscillation pattern. Fertilized eggs lacking both Ca_V3.2 and TRPM7 frequently display a second wave of Ca²⁺ oscillations following a delay, suggesting activation of an alternate Ca²⁺ influx mechanism during the later stages of egg activation. Finally, we provide evidence that disruption of the initial pattern of Ca²⁺ oscillations following fertilization is associated with postimplantation development failure and alterations in postnatal growth.

TRPM7 is a key mediator of store-depletion–induced Ca²⁺ influx in GV oocytes, and together, Ca_V3.2 and TRPM7 are required for the same process in eggs. This Ca²⁺ influx was previously presumed to occur through a store-operated Ca²⁺ entry mechanism mediated by STIM/ORAI interactions. However, the data presented here and prior studies showing that Stim1, Stim2, and Orai1 are dispensable for normal Ca²⁺ homeostasis and postfertilization Ca²⁺ signaling in mouse oocytes (11, 25, 26) implicate TRPM7 as a primary mediator of this Ca²⁺ influx and Ca_V3.2 as a contributor. There are notable species differences regarding reliance of oocyte Ca²⁺ signaling on SOCE components, with evidence for STIM1 functionality in porcine oocytes (27); thus, the relative importance of TRPM7 in oocytes of other species remains to be addressed. In somatic cells, TRPM7 is also capable of mediating SOCE-like Ca²⁺ influx independently of STIM and ORAI proteins. Both macrophages and dendritic cells undergo differentiation and have completely normal function when lacking both STIM1 and STIM2 (28). It turns out that macrophages utilize TRPM7-mediated Ca²⁺ influx, rather than SOCE, to support signaling pathways downstream of toll-like receptor 4 activation by lipopolysaccharide (29). Complicating this issue, recent studies in cultured B lymphocytes indicate that while TRPM7 ion channel activity directly contributes to the maintenance of Ca²⁺ stores, TRPM7 kinase activity appears to modulate canonical SOCE by an unknown mechanism (30, 31). Despite these findings of an interaction between TRPM7 and canonical SOCE in B lymphocytes, the expendability of Stim1, Stim2, and Orai1 in mouse oocytes indicates that TRPM7 functions independently of canonical SOCE in these cells. Although our data indicate that TRPM7 contributes significantly to the Ca²⁺ influx to maintain intracellular stores in mouse oocytes, there is not sufficient evidence to claim that the channel is activated by store depletion rather than simply functioning as a Ca²⁺ conduit through its constitutive activity.

An unusual “stop–restart” pattern of Ca²⁺ oscillations was observed following IVF of dKO eggs. Importantly, this pattern was not observed following ICSI, which accomplishes fertilization without sperm–egg membrane fusion. In similar fashion, the plasma membrane block to polyspermy does not occur following ICSI or parthenogenetic activation, but instead requires some additional signal induced by sperm–egg fusion for its establishment (32–37). This result is also consistent with the recent observation that fertilization by sperm lacking PLCζ fails to initiate the typical robust pattern of Ca²⁺ oscillations but prompts a small number of Ca²⁺ transients that begin following a delay (6). This secondary Ca²⁺ signal requires sperm–egg fusion and is proposed to reflect a “primitive” intrinsic activation signal (6); it is possible that the second wave of oscillations in dKO eggs is related to this recently discovered PLCζ-independent activation mechanism. The ability of the dKO egg to restart oscillations could be a consequence of the appearance on the egg plasma membrane of sufficient Ca²⁺-permeable channels to allow refilling of the ER Ca²⁺ stores. The source of these new channels is not clear. One possibility is that an existing channel(s) at the plasma membrane that is somehow inactive in MII eggs is activated as a consequence of sperm–egg fusion. TRPV3 channels are present on the mouse egg plasma membrane (13) and could be the target of such stimulation in the dKO eggs. Another possibility is that new channels are added to the plasma membrane after sperm–egg fusion, although their source is presently unknown. It is unlikely that new channels are derived from cortical granule membranes, because cortical granule exocytosis occurs normally following ICSI (34, 35); yet ICSI eggs do not display the stop–restart Ca²⁺oscillation pattern. A possible source, however, is the sperm plasma membrane, which has several types of Ca²⁺-permeable channels that could function in the newly formed zygote; this phenomenon occurs in Caenorhabditis elegans (38). Lastly, the new channels could come from fusion of egg intracellular vesicles other than cortical granules, with the caveat that this fusion event must be dependent on a signal that originates as a result of sperm–egg fusion. It is also possible that differences in PLCζ accessibility and distribution between IVF and ICSI could influence the generation of the initial and secondary Ca²⁺ signals. These ideas will be interesting to explore in future studies.

It has been known for decades that there is a direct relationship between extracellular Ca²⁺ concentration and oscillation frequency following fertilization in mammals (8, 23). How extracellular Mg²⁺ impacts Ca²⁺ oscillation frequency was only recently described in an elegant series of experiments using mouse embryos (18). In this report, the authors found that the ratio of extracellular Mg²⁺ to Ca²⁺ was a critical factor regulating the frequency and persistence of Ca²⁺ oscillations in mouse embryos fertilized by ICSI, with a lower Mg²⁺/Ca²⁺ ratio correlating with higher oscillation frequency and persistence. Because Mg²⁺ at the levels used in this report can inhibit TRPM7 ion channel activity (15), and extracellular Ca²⁺ modulates oscillation frequency via TRPM7 (Fig. 3 D and E), this finding is consistent with our identification of TRPM7 as a critical mediator of Ca²⁺ influx following fertilization. Furthermore, it suggests that TRPM7 serves as an egg plasma membrane sensor of the extracellular Mg²⁺/Ca²⁺ ratio.

Important findings of our study were the altered growth trajectory and/or greater weight variability of offspring derived from TRPM7 cKO or dKO eggs and wild-type sperm. These findings are not surprising because previous studies have documented the importance of the Ca²⁺ oscillation pattern for promoting robust preimplantation development, postimplantation development to term, and postnatal growth rate (18, 39–41). The exact mechanisms connecting Ca²⁺ oscillations to fetal growth are not known but could be related to alterations in gene expression during preimplantation development, which have been documented in blastocyst-stage embryos generated following experimental alteration of postfertilization Ca²⁺ oscillation patterns (40). Another factor contributing to the altered growth patterns could be differences in embryo redox potential. Periodic rises in cytoplasmic Ca²⁺ levels following fertilization stimulate mitochondrial oxidative phosphorylation, which provides ATP necessary to sustain the Ca²⁺ oscillations and impacts the cellular redox state (42, 43). Importantly, experimental alterations in redox potential of pronuclear-stage mouse embryos is associated with alterations in postnatal weight variation and growth trajectories (44). These findings are consistent with our observation of an increase in weight variability and altered growth trajectories in heterozygous offspring derived from eggs that likely had an abnormal pattern of Ca²⁺ oscillations following fertilization.

Although the current studies were performed in the mouse, TRPM7, CACNA1H, and TRPV3 are all expressed in human oocytes and TRPM7 and TRPV3 are expressed in macaque oocytes as indicated by RNA sequencing (45–48). Furthermore, there are numerous missense or loss of function (LoF) mutations identified in human TRPM7 (480 missense, 27 LoF), CACNA1H (1,045 missense, 10 LoF), and TRPV3 (312 missense, 23 LoF) (see Exome Aggregation Consortium browser at exac.broadinstitute.org; ref. 49). These findings support the ideas that (i) TRPM7, Ca_V3.2, and TRPV3 could also regulate Ca²⁺ signals following fertilization in human and (ii) functional abnormalities in these channels could contribute to infertility or abnormal growth outcomes in children. Assisted reproduction procedures are responsible for the births of millions of children worldwide. Because the composition of most clinically used media is proprietary, limited information on embryo culture media composition is accessible (50–52) and Ca²⁺ and Mg²⁺ content of stand-alone fertilization media is not available at all. Embryo culture media formulations have an extremely wide range of Mg²⁺/Ca²⁺ ratios (51, 52) that could dramatically alter Ca²⁺ influx during embryo development by affecting TRPM7 function. Evidence is mounting that exposure to varying levels of Ca²⁺ and Mg²⁺ during fertilization and embryo culture can impact developmental outcomes (18, 53). Culture media formulations also have significant differences in energy substrates including those that can influence redox potential, which also impacts development (18, 44). Studies in human or nonhuman primate oocytes to determine the impact of media composition on Ca²⁺ oscillatory patterns and redox potential would provide the basis for rational design of appropriate media that optimize reproductive outcomes and offspring health.

Materials and Methods

All animal work was performed in accordance with National Institutes of Health and National Institute of Environmental Health Sciences guidelines under approved animal care and use protocols. A detailed description of mice, gamete and embryo collection and culture, fertilization, Ca²⁺ imaging, electrophysiology, real-time RT-PCR, DNA methylation analyses, and statistical analyses is provided in SI Appendix, SI Materials and Methods.