The Drosophila Trpm channel mediates calcium influx during egg activation

Significance

A rise in intracellular free calcium is a conserved feature of the egg-to-embryo transition in almost all animals. In Drosophila, as in vertebrates, the rise starts at one end of the oocyte and then travels across the egg in a wave. The Drosophila calcium rise is mediated by an influx of calcium due to the action of mechanically gated ion channels. Here, we identify the ion channel that is critical for the calcium entry as Trpm. Trpm is the ortholog of a channel recently shown to mediate the postfertilization calcium influxes needed to sustain calcium oscillations in fertilized mouse eggs, suggesting a deep homology despite species differences in the trigger for egg activation.

Abstract

Egg activation is the process in which mature oocytes are released from developmental arrest and gain competency for embryonic development. In Drosophila and other arthropods, eggs are activated by mechanical pressure in the female reproductive tract, whereas in most other species, eggs are activated by fertilization. Despite the difference in the trigger, Drosophila shares many conserved features with higher vertebrates in egg activation, including a rise of intracellular calcium in response to the trigger. In Drosophila, this calcium rise is initiated by entry of extracellular calcium due to opening of mechanosensitive ion channels and initiates a wave that passes across the egg prior to initiation of downstream activation events. Here, we combined inhibitor tests, germ-line–specific RNAi knockdown, and germ-line–specific CRISPR/Cas9 knockout to identify the Transient Receptor Potential (TRP) channel subfamily M (Trpm) as a critical channel that mediates the calcium influx and initiates the calcium wave during Drosophila egg activation. We observed a reduction in the proportion of eggs that hatched from trpm germ-line knockout mutant females, although eggs were able to complete some egg activation events including cell cycle resumption. Since a mouse ortholog of Trpm was recently reported also to be involved in calcium influx during egg activation and in further embryonic development, our results suggest that calcium uptake from the environment via TRPM channels is a deeply conserved aspect of egg activation.

In almost all animals, mature oocytes are arrested in meiosis at the end of oogenesis and require an external trigger to be activated and transition to start embryogenesis. This “egg activation” involves multiple events, including meiosis resumption and completion, maternal protein modification and/or degradation, maternal mRNA degradation or translation, and egg envelope changes (reviewed in refs. 1–5).

Triggers of egg activation vary across species. In vertebrate and some invertebrate species, fertilization triggers egg activation. However, changes in pH, ionic environment, or mechanical pressure can also trigger egg activation in other invertebrate species (reviewed in ref. 1). A conserved response to these triggers is a rise of intracellular free Ca²⁺ levels in the oocyte. This calcium rise is due to influx of external calcium and/or release from internal storage, depending on the organism (reviewed in ref. 5). The elevated Ca²⁺ concentration is thought to activate Ca²⁺-dependent kinases and/or phosphatases, which in turn change the phosphoproteome of the activated egg, initiating egg activation events (refs. 6 and 7, reviewed in ref. 3).

Drosophila eggs activate independent of fertilization, and the trigger is mechanical pressure. When mature oocytes exit the ovary and enter the lateral oviduct, they experience mechanical pressure from the reproductive tract. As the oocytes swell due to the influx of oviductal fluid, their envelopes become taut (8). Drosophila oocytes can be activated in vitro by incubation in a hypotonic buffer, although some egg activation events do not proceed completely normally in vitro (9, 10). Intracellular calcium levels rise in oocytes during egg activation, as observed with the calcium sensor GCaMP. This calcium rise takes the form of a wave that starts at the oocyte pole(s) and traverses the entire oocyte (11, 12). In vivo, the wave starts with a calcium rise at the posterior pole; in vitro it also starts predominantly at the posterior pole but sometimes (also) at the anterior pole (11, 12). Initiation of this calcium wave requires influx of external Ca²⁺, as chelating external Ca²⁺ in in vitro egg activation assays blocks the calcium wave and egg activation. Propagation of the calcium wave relies on the release of internal Ca²⁺ stores, likely through an Inositol 1,4,5-trisphosphate (IP₃)-mediated pathway as in other animals, because knocking down the endoplasmic reticulum (ER) calcium channel IP₃ receptor (IP₃R) prevents propagation of the calcium wave (11).

How mechanical forces trigger calcium entry during Drosophila egg activation was unknown. However, the lack of initiation of a calcium wave in the presence of Gd³⁺, an inhibitor of mechanosensitive ion channels (13), and N-(p-Amylcinnamoyl) anthranilic acid (ACA), an inhibitor of TRP-family ion channels (14), suggested that TRP family ion channels (reviewed in ref. 15) are likely involved (11). Further supporting this idea is the recent discovery that a TRP family channel, TRPM7, is needed for calcium influx that is necessary (but not sufficient) for calcium oscillations in activating mouse eggs (16). The Drosophila genome encodes 13 TRP family channels (reviewed in ref. 17), but according to RNA-seq data, only 3 (Painless, Trpm, and Trpml) are expressed in the ovary (reviewed in ref. 2). We used specific inhibitors, existing mutants, germ-line–specific RNAi knockdown, and knockouts that we created with CRISPR/Cas9 to screen these 3 candidates for their roles in the initiation of the calcium wave. We found that Trpm, the single Drosophila ortholog of mouse TRPM7, mediates calcium wave initiation, whereas the other 2 TRP channels are not necessary to initiate the calcium wave.

Results

Painless and Trpml Are Not Essential for the Calcium Wave Initiation or Propagation.

Three TRP channels (Painless [Pain], Trpm, and Trpml) are expressed in the Drosophila ovary and are candidates for involvement in calcium wave initiation. To examine the roles of these channels in the calcium wave during egg activation, we performed in vitro activation assays on mature oocytes dissected from WT, mutant, and/or knockdown females, and imaged the calcium waves with fluorescence microscopy (Materials and Methods). Since calcium influx is required for calcium wave initiation (11), we used the incidence of calcium waves as an indicator of calcium influx, and assessed the role of candidate channels with it.

Among the 3 candidates, Pain is reported to display mechanosensitivity (18). To examine the role of Pain, we made a CRISPR knockout line, pain^TMΔ, by deleting the region that encodes Pain’s transmembrane domains. Since pain loss-of-function mutants are viable (19), we were able to examine the calcium wave phenotype in oocytes from homozygous pain^TMΔ females that also carried a nos-GCaMP6m-attP2 transgene. The incidence of a calcium wave in vitro in mature oocytes of homozygous pain^TMΔ females did not differ significantly from that of heterozygous controls (Fig. 1A and Movies S1 and S2). The propagation rate of the calcium wave, as measured at its wave-front, was also not significantly different from control (Fig. 1B).

Fig. 1.

Pain and trpml are not essential for calcium wave initiation or propagation. (A) The incidence of a calcium wave in in vitro activation assays does not differ between control (pain^TMΔ/+, n = 33/49) and pain null mutant (pain^TMΔ/pain^TMΔ, n = 33/48), P = 1. (B) The propagation speed of the calcium wave does not differ between control (2.37 ± 0.38 μm/s) and pain null mutant (2.49 ± 0.78 μm/s), P = 0.59. (C) The incidence of a calcium wave in in vitro activation assays does not differ between control (trpml¹/+, n = 17/21) and trpml null mutant (trpml¹/trpml¹, n = 15/25), P = 0.22. (D) The propagation speed of the calcium wave does not differ between control (2.97 ± 0.74 μm/s) and trpml null mutant (2.78 ± 0.77 μm/s), P = 0.51. N.S., not significant (P > 0.05).

To examine the role of Trpml, we tested an existing null mutant, trpml¹ (20). We crossed into trpml¹ the nos-GCaMP6m-attP40 transgene and examined the calcium wave phenotype in mature oocytes from homozygous mutants or heterozygous controls. Again, we did not find a difference in the calcium wave incidence (Fig. 1C and Movies S3 and S4) or in calcium wave propagation rate between oocytes from homozygous mutant and those from heterozygous control females (Fig. 1D).

Thus, neither Pain or Trpml is essential for the initiation or propagation of the calcium wave. These channels thus are either not involved in calcium wave initiation, or they function redundantly with other channels.

Inhibitor and RNAi Perturbations of Trpm Inhibit Calcium Wave Initiation.

To test whether Trpm is needed for the calcium wave, we first examined the effect of known inhibitors of this channel. Ten millimolar Mg²⁺ is reported to inhibit several TRP channels including mouse TRPV3 (21), TRPM6, TRPM7 (22), and Drosophila Trpm (23). In the presence of 10 mM Mg²⁺ in isolation buffer (IB) and Robb’s Buffer (RB), the incidence of a calcium wave in vitro was significantly reduced (Fig. 2 A, B, and E and Movie S5). We also tested the effect of NS8593, a specific inhibitor of mouse TRPM7 channels (24). One hundred micromolar NS8593 in IB and RB completely blocked calcium wave initiation in vitro (Fig. 2 C and F and Movie S6).

Fig. 2.

Disrupting trpm function reduces calcium wave incidence in vitro. (A–C) Representative calcium waves, or lack thereof, in matα4-GAL4-VP16; UASp-GCaMP3 stage 14 oocytes in in vitro activation assays. Oocytes were incubated in: unmodified IB and RB (A); IB and RB with 10 mM of MgCl₂ (B); IB and RB with 100 μM NS8593 (C). (D) Representative pictures showing lack of calcium waves in stage 14 oocytes from trpm germ-line knockdown females (matα4-GAL4-VP16/UAS-trpm^RNAi; UASp-GCaMP3) incubated in unmodified IB and RB. An increase of calcium signal without a discernible wave is observed in some of the oocytes, possibly because of incomplete removal of Trpm channels by RNAi. (Scale bars: 50 μm.) (E–G) Quantification of calcium wave incidence in oocytes as in B–D: n = 16/18 for control (IB and RB with 30 mM sucrose to provide the same osmolarity change as 10 mM MgCl₂), n = 2/20 for 10 mM MgCl₂, P = 5.69 × 10⁻⁶ (E); n = 27/29 for control (DMSO), n = 0/29 for 100 μM NS8593, P = 2.15 × 10⁻¹¹ (F); n = 26/28 for control (matα4-GAL4-VP16; UASp-GCaMP3), n = 11/70 for trpm germ-line knockdown mutants (matα4-GAL4-VP16/UAS-trpm^RNAi; UASp-GCaMP3), P = 5.74 × 10⁻¹² (G). (H) RT-PCR quantification of trpm germ-line knockdown. Normalized expression level of trpm was calculated by normalizing trpm band intensity with that of ribosome protein gene rpl32 (C: control, KD: trpm germ-line knockdown). Transcript levels of trpm are decreased by >90% in knockdown ovaries and mature oocytes. Both sets of primers flank an exon-exon junction and will amplify a bigger band with genomic DNA as templates. This ensures that the bands we quantified were amplified from cDNA only. ***P < 0.001.

Since trpm homozygous mutants are lethal before adult stage (25), we next tested the role of trpm in calcium wave initiation using germ-line–specific RNAi. We crossed matα4-GAL4-VP16; UASp-GCaMP3 to UAS-trpm^RNAi. The female offspring expressed both GCaMP3 and trpm shRNA in the germ line. We examined the calcium waves in in vitro activation assays with mature oocytes from these females. Oocytes from trpm germ-line knockdown females displayed a significantly lower incidence of calcium waves in vitro (Fig. 2 D and G and Movie S7). To validate the efficiency of germ-line knockdown, we performed RT-PCR with ovary mRNA from these females. RT-PCR results showed that more than 90% of trpm transcripts were removed in germ-line knockdown females compared to control (Fig. 2H).

Inhibitors of Trpm and germ-line knockdown of trpm both reduced the incidence of calcium waves in in vitro egg activation assays. Thus, our data strongly suggested that Trpm is necessary for calcium influx during Drosophila egg activation.

Germ-Line–Specific CRISPR/Cas9 Knockout of trpm Inhibits Calcium Wave Initiation.

To further validate the role of Trpm in the initiation of the calcium wave, we attempted to perform germ-line–specific biallelic knockout of trpm using the CRISPR/Cas9 system. We designed 3 sets of gRNAs targeting trpm (Materials and Methods). We used the “Cas9-LEThAL” (26) method to evaluate the efficiency of our gRNA expression constructs. Among them, gRNA set 3 displayed the highest efficiency (SI Appendix, Table S1). Using this set of gRNAs, we made a dual-gRNA transgenic line based on a polycistronic gRNA design (U6:3-tRNA-gRNA1-tRNA-gRNA2, gRNA-trpm¹) and a dual transcription unit design (CR7T-gRNA1-U6:3-gRNA2, gRNA-trpm²), both ubiquitously expressing the gRNAs (SI Appendix, Fig. S1A).

We crossed each of these transgenic fly strains to nos-Cas9; matα4-GAL4-VP16; UASp-GCaMP3 to achieve germ-line–specific knockout of trpm at early oogenesis stages, via nos-Cas9, and calcium visualization at the same time (SI Appendix, Fig. S1B). By sequencing amplicons generated by single oocyte RT-PCR, we confirmed the presence of CRISPR/Cas9-generated indels (SI Appendix, Fig. S2 and Supplemental Materials and Methods). With gRNA-trpm¹, 34% of the female offspring displayed defects in ovary morphology (SI Appendix, Fig. S3 A and B). With gRNA-trpm², 96% of the female offspring displayed similar defects (SI Appendix, Fig. S3C). These data showed that trpm is required for early female germ-line development. Since Trpm is involved in maintaining cation homeostasis in cellular environments and tissue development (23, 25), we hypothesize that early knockout of trpm by nos-Cas9 interferes with normal germ-line development. The less-than-100% incidence of this phenotype is likely due to lack of 100% efficiency in generating biallelic null mutations of trpm in the germ line with less-efficient gRNA expression constructs.

Some early trpm germ-line knockout females had ovaries with grossly normal morphology, perhaps because null mutations were generated at both trpm alleles only after a critical developmental point. We tested their mature oocytes for activation in vitro. We observed that the incidence of calcium waves was significantly reduced in oocytes from these females (Fig. 3 B and D and Movie S8) compared to Cas9 control (Fig. 3A and Movie S9). Since efficient CRISPR/Cas9-mediated trpm knockout caused ovary development defects, flies with normal-looking ovaries may have less efficient trpm biallelic knockout in the germ line, which could explain the incomplete elimination of calcium waves in oocytes from these flies.

Fig. 3.

trpm germ-line specific CRISPR knockout reduces calcium wave incidence. (A–C) Representative calcium waves, or lack thereof, in in vitro activation assays. Stage 14 oocytes were dissected from: nos-Cas9; matα4-GAL4-VP16; UASp-GCaMP3 control (A); early germ-line trpm knockout (nos-Cas9; matα4-GAL4-VP16; UASp-GCaMP3/gRNA-trpm¹) (B); late germ-line trpm knockout (matα4-GAL4-VP16/UAS-GCaMP6s; gRNA-trpm¹/UAS-Cas9) (C). (Scale bars: 50 μm.) (D and E) Quantification of calcium wave incidence in oocytes as in B and C: n = 36/39 for control (nos-Cas9; matα4-GAL4-VP16; UASp-GCaMP3), n = 38/87 for early trpm germ-line knockout (nos-Cas9; matα4-GAL4-VP16; UASp-GCaMP3/gRNA-trpm¹), P = 4.64 × 10⁻⁷ (D); n = 20/27 for control (matα4-GAL4-VP16/UAS-GCaMP6s; UAS-Cas9), n = 5/62 for late trpm germ-line knockout (matα4-GAL4-VP16/UAS-GCaMP6s; gRNA-trpm¹/UAS-Cas9), P = 3.68 × 10⁻¹¹ (E). (F) Resistance to 50% bleach by mature oocytes, 1- to 5-h embryos and unfertilized eggs from control (nos-Cas9-attP2, C), trpm early germ-line knockout (nos-Cas9-attP2/gRNA-trpm¹, EKO), and late germ-line knockout (matα4-GAL4-VP16; UAS-Cas9/gRNA-trpm¹, LKO) females. All preactivation oocytes are vulnerable to bleach. One- to 5-hour embryos and unfertilized eggs from both control and trpm germ-line knockout mutants displayed resistance to bleach treatment. C oocyte n = 132, EKO oocyte n = 152, LKO oocyte n = 119, C embryo n = 76, EKO embryo n = 73, LKO embryo n = 64, C unfertilized egg n = 66, EKO unfertilized egg n = 58, LKO unfertilized egg n = 71. N.S., not significant (P > 0.05); ***P < 0.001.

To bypass the critical stage in early oogenesis that may require trpm function, we knocked out trpm at only later stages of oogenesis by crossing the UAS-GCaMP6s; gRNA-trpm¹ strain to the matα4-GAL4-VP16; UAS-Cas9 strain (SI Appendix, Fig. S1B). Since matα4-GAL4-VP16 does not drive expression of UAS constructs until mid to late oogenesis (27), female offspring of this cross will initiate germ-line knockout of trpm at later stages of oogenesis in contrast to crosses using nos-Cas9, which initiate knockout early in oogenesis. None of these germ-line knockout females displayed gross ovary morphology defects (SI Appendix, Fig.S3 D and E), suggesting that trpm knockout occurred after a critical stage in early oogenesis. We examined the calcium wave phenotype of mature oocytes from these females in in vitro egg activation assays. We observed a significant decrease in the incidence of calcium waves (Fig. 3 C and E and Movie S10).

Taken together, these results further support that Trpm is required for calcium wave initiation. In addition, our results also suggest that Trpm plays essential roles in early oogenesis.

Egg Activation Events Occur in trpm Knockout Eggs In Vivo.

We next asked whether egg activation requires normal trpm function. Almost all 1- to 5-h embryos from trpm germ-line knockout females displayed resistance to 50% bleach, an indicator of vitelline membrane cross-linking after egg activation (28) (Fig. 3F). This suggests that egg envelope hardening, one aspect of egg activation, still occurs in embryos from trpm germ-line knockout females.

Since bleach resistance does not always completely reflect the state of egg activation (10), we examined whether embryos from trpm germ-line knockout females were able to enter mitosis, implying successful completion of meiosis and fertilization. Surprisingly, anti-Tubulin and DAPI staining of 1- to 5-h-old embryos laid by trpm germ-line knockout females mated with ORP2 WT males showed they had all undergone early embryo mitoses (n = 88 for early germ-line knockout mutants, n = 36 for late germ-line knockout mutants, Fig. 4A). This indicated that cell cycles can resume in embryos from trpm germ-line knockout females mated with WT males.

Fig. 4.

Cell cycle resumption and vitelline membrane cross-linking occur normally in eggs and embryos from trpm germ-line knockout females. (A) Representative pictures of anti-Tubulin and DAPI staining of 1- to 5-h embryo laid by control and trpm germ-line CRISPR knockout females. Focal planes vary across samples. All embryos from trpm germ-line knockout females have started mitosis and progressed to multicellular stages (n = 88 for early germ-line knockout, n = 36 for late germ-line knockout). (B) Representative pictures of anti-Tubulin and DAPI staining of unfertilized eggs laid by control (nos-Cas9-attP2, n = 14) and trpm germ-line knockout (nos-Cas9-attP2/gRNA-trpm¹, n = 20) females. All eggs from both control and germ-line knockout females have completed meiosis, indicated by production of a normal polar body (Scale bars: A, 50 μm; B, 5 μm.) (C) trpm germ-line RNAi (matα4-GAL4-VP16/UAS-trpm^RNAi; UASp-GCaMP3, n = 18) does not significantly affect female egg hatchability relative to control (matα4-GAL4-VP16; UASp-GCaMP3, n = 18), P = 0.15. (D) Early trpm germ-line knockout (nos-Cas9-attP2/gRNA-trpm¹, n = 21) significantly reduces female egg hatchability relative to control (nos-Cas9-attP2, n = 16), P = 2.08 × 10⁻¹⁵. (E) Late trpm germ-line knockout (matα4-GAL4-VP16; UAS-Cas9/gRNA-trpm¹, n = 17) significantly reduces female egg hatchability relative to control (matα4-GAL4-VP16 > UAS-Cas9, n = 9), P = 5.13 × 10⁻⁵. N.S., not significant (P > 0.05); ***P < 0.001.

We then asked if egg hatchability (percent of eggs that hatched into larvae) is affected by trpm germ-line perturbations. Germ-line knockdown of trpm did not impair egg hatchability (Fig. 4C). Since this knockdown, despite being highly efficient, did not completely abolish trpm function in oocytes, we examined the hatchability of eggs from trpm germ-line knockout females. Germ-line trpm knockout females displayed significantly reduced egg hatchability compared to control, with either early (Fig. 4D) or late (Fig. 4E) germ-line specific knockout, suggesting that the lack of maternal Trpm function or a calcium wave compromised development events after initiation of the embryos’ mitotic phase. Since the fathers of these embryos were wild type, a WT trpm allele was present in the embryos. Thus, defects in hatchability must be due to the lack of maternal trpm product, because after zygotic genome activation, these embryos will express WT Trpm.

Taken together, these observations indicate that fertilized eggs are able to resume cell cycles and harden egg envelopes in the absence of Trpm function. The reduced hatchability of eggs laid by trpm germ-line knockout females suggests that lack of oocyte Trpm function or a calcium wave affects embryogenesis after egg activation.

Sperm Is Not an Alternative Trigger of Egg Activation in the Absence of Trpm Function.

Because fertilized eggs from trpm germ-line knockout females proceeded to mitotic stages, we wondered whether sperm could act as an alternative trigger for the calcium rise and egg activation (reviewed in ref. 1) in Drosophila. We mated control and trpm germ-line knockout females to spermless males (29). Eggs laid by females after the mating would experience the normal egg-activating environment in vivo but remain unfertilized. We then examined if meiosis resumption and vitelline membrane cross-linking occur in unfertilized eggs from trpm germ-line knockout females. We observed 100% resistance to 50% bleach in these unfertilized eggs (Fig. 3F). We also observed normal production of polar bodies in all unfertilized eggs from control (n = 14) and trpm germ-line knockout females (n = 20) (Fig. 4B). These results indicated that vitelline membrane cross-linking and meiosis resumption still occur in eggs from trpm germ-line knockout mutants independent of sperm, ruling out the possibility that sperm could overcome the lack of calcium influx triggered by mechanical pressure and activate eggs in the absence of Trpm function.

Discussion

Egg activation is an essential step for oocytes to transition to embryogenesis. In all species studied to date, egg activation involves a rise in intracellular calcium. In Drosophila, this calcium rise takes the form of a wave that is triggered by mechanical pressure. The initiation of this calcium wave requires influx of calcium from the environment through mechanosensitive TRP family channel(s) (11).

Here, we determined that Trpm channel is necessary for this calcium wave. Of the 3 TRP channels expressed in ovaries (Pain, Trpm, and Trpml), only impairment of Trpm affects the initiation of the calcium wave. Calcium wave phenotypes are normal in oocytes from pain or trpml null mutants. However, the incidence of the calcium wave is diminished in WT oocytes in the presence of Trpm inhibitors and in oocytes from trpm germ-line knockdown or knockout mutants. These results consistently indicated that Trpm mediates the calcium influx that initiates the calcium wave during Drosophila egg activation. Given the short time frame of calcium wave propagation and limitations of the inhibitor-test technique, we cannot determine if Trpm also plays a role in calcium wave propagation. trpm germ-line knockout females also displayed significantly decreased egg hatchability, due to defects after cell cycle resumption. The reduced egg hatchability suggested that maternal trpm function or the calcium wave is required for further embryogenesis after egg activation.

The Drosophila Trpm Channel Plays Important Reproductive Roles.

TRP family cation channels are nonselective and respond to a wide array of environmental stimuli. Drosophila Trpm has been reported to play multiple roles throughout larval development, including maintaining Mg²⁺ and Zn²⁺ homeostasis (23, 25), and sensing noxious cold in larval class III md neurons (30). However, the role of Trpm in reproduction had not been investigated because of the pupal lethality of trpm null mutants (25). Here, our germ-line–specific RNAi knockdown and CRISPR/Cas9-mediated knockout revealed 3 functions of Drosophila Trpm: supporting early oogenesis, mediating influx of environmental calcium to initiate the calcium wave during egg activation, and maternally supporting embryonic development after egg activation.

Our previous study suggested that calcium influx during Drosophila egg activation is mediated through mechanosensitive ion channels (11). Both Drosophila Trpm and mouse TRPM7 are reported to be constitutively active and permeable to a wide range of divalent cations (23, 31). Mouse TRPM7 is known to respond to mechanical pressure (32, 33), but further study will be needed to determine whether Drosophila Trpm is similarly responsive to mechanical triggers, such as those that occur during ovulation.

Requirement for Trpm during Drosophila Embryogenesis.

Germ-line knockout of trpm significantly reduced the incidence of calcium waves in in vitro egg activation assays and egg hatchability. However, this reduced egg hatchability was not due to failure of cell cycle resumption during egg activation. There are 2 possible explanations for the reduced egg hatchability of trpm germ-line knockout females.

First, it is possible that trpm plays a maternal role, independent of its role in initiating the calcium wave, such that lack of maternally deposited Trpm proteins leads to defects during embryogenesis. In mouse, TRPM7 is also required for normal early embryonic development, apart from its role in calcium oscillations. Inhibition of TRPM7 function impairs preimplantation embryo development and slows progression to the blastocyst stage (34). Drosophila trpm mutant lethality had been reported to occur during the pupal stage (25). However, those homozygous mutants were offspring of heterozygous mothers and, thus, did not lack maternal Trpm function. Our germ-line–specific depletion of trpm reveals a possible maternal role for Trpm in embryogenesis.

Alternatively, or in addition, it is possible that oocytes lacking Trpm do not pass sufficient Ca²⁺ from the environment to form a calcium wave, but that at least some events of egg activation can occur despite this. In mouse, an initial calcium rise is induced by sperm-delivered Phospholipase C ζ (PLCζ) via the IP₃ pathway (35). Yet although sperm from PLCζ null males fails to trigger normal calcium oscillations, some eggs fertilized by those sperm develop (36). Multiple oscillations following fertilization require influx of external calcium (37), mediated by TRPM7 and Ca_V3.2 (16). Although these oscillations had been reported to be needed for multiple postfertilization events (38), some TRPM7 and Ca_V3.2 double-knockout embryos still develop, albeit not completely normally (16). Together, these data suggest that egg activation can still occur in mouse with diminished intracellular calcium rises, analogous to what we see in Drosophila in the absence of maternal Trpm function.

Insufficient influx of calcium in the absence of Trpm function could disrupt later (but maternally dependent) embryogenesis. The oocyte-to-embryo transition involves multiple events. In mouse egg activation, these events take place sequentially as calcium oscillations progress, with developmental progression associated with more oscillations and more total calcium signal. Some of the events start after a certain number of oscillations but require additional oscillations to complete (38). It is possible that mechanisms critical for Drosophila embryo development also depend on reaching a precise level of calcium. A low-level calcium rise might be sufficient to trigger some egg activation mechanisms such as vitelline membrane cross-linking and cell cycle resumption, but high levels of calcium may be required for further progression.

Alternative Pathways for Calcium Influx during Egg Activation.

Given the importance of calcium in egg activation (10, 37), we were surprised that although trpm knockout eggs lacked a calcium wave in vitro, in vivo such eggs could progress in cell cycles and even, sometimes, hatch. As discussed above, there may be insufficient calcium influx in the absence of Trpm for full and efficient development, but some egg activation events may still occur. Alternatively, it is possible that redundant mechanisms permit a sufficient calcium-level increase without producing a detectable wave form. Despite being able to trigger a series of egg activation events including meiosis resumption and protein translation (9, 10), osmotic pressure during in vitro activation may have different properties from mechanical pressure exerted on mature oocytes during ovulation. The latter might allow opening of other calcium channels to initiate a normal calcium rise and complete egg activation. Two channels, TRPM7 and Ca_v3.2, are needed for the calcium oscillations following mouse fertilization, but the Drosophila ortholog of mouse Ca_v3.2, Ca-α1T, is not detectably expressed in fly ovaries (39). From RNA-seq database, another mechanosensitive, nonselective cation channel, Piezo (40), is expressed in Drosophila ovaries (39). Piezo and/or other unknown channels might play this redundant role of mediating calcium influx in vivo. In this light we note that levels of basal GCaMP fluorescence varied among oocytes incubated in IB that were inhibited from forming a wave, suggesting the possibility of a calcium increase by a redundant mechanism (SI Appendix, Fig. S4).

Conserved Role of TRPM Channels in Egg Activation.

We are intrigued that Drosophila Trpm is essential for the calcium rise at egg activation, given that mouse TRPM7 was recently reported to be required (along with Ca_V3.2) for the calcium influx needed for postfertilization calcium oscillations that are, in turn, required for egg activation events (16, 38). This apparent conservation in mechanisms in egg activation involving orthologous Trpm channels in a protostome (Drosophila) and a deuterostome (mouse) prompts us to wonder whether Trpm-mediated calcium influx is a very ancient and basal aspect of egg activation, with other more variable aspects such as sperm-triggered calcium rises being more derived, if better known, features. It is interesting in this light that a sperm-delivered TRP channel (TRP-3) has also been reported to mediate calcium influx and a calcium rise in another protostome, Caenorhabditis elegans (41).

Materials and Methods

DNA Constructs and Transgenic Flies.

Calcium waves were visualized by expressing GCaMP sensors in the female germ line using matα4-GAL4-VP16; UASp-GCaMP3 (11), matα4-GAL4-VP16; UAS-GCaMP6s, or nos-GCaMP6m. The nos-GCaMP6m strain was constructed by replacing the GCaMP3 coding sequence in the previously described nos-GCaMP3-attB construct (11) with GCaMP6m coding sequence and integrating the construct into either the attP2 or the attP40 site.

Calcium wave incidence in each experiment was compared between perturbed and control oocytes using the same calcium sensor and exposure/gain settings. Control oocytes showed lower incidence of calcium waves when expressing GCaMP6s or GCaMP6m compared to GCaMP3, likely due to weaker expression of our GCaMP6s or GCaMP6m constructs.

To create a null allele of pain, we generated pU6-chiRNA constructs following protocols described by FlyCRISPR website (42). We generated 2 constructs to express sgRNAs with the following target sequences (PAM sequences are underlined): GTCTTGCAGCTGGTTGAGTCCGG, GACGCAGACTTAAGTAGTTCGGG. These 2 constructs were coinjected by Rainbow Transgenic Flies into nos-Cas9-attP2 embryos. A strain carrying a 1,091-bp deletion (chr2R:24922130–24923220) in pain was isolated and stabilized to establish the null allele strain pain^TMΔ (SI Appendix, Supplemental Materials and Methods).

To knock out trpm, we generated the U6:3-tRNA-gRNA1-tRNA-gRNA2 (26) and CR7T-gRNA1-U6:3-gRNA2 constructs (gifts from Chun Han, Cornell University, Ithaca, NY). We designed 3 sets of gRNAs targeting trpm to express in our constructs. sgRNA target sequences in the trpm gene are as follows (PAM sequences are underlined):

• gRNA set 1: GGAACCATCGAGTTCCAGGGCGG; GATGTGGACACATGGCGAGGAGG; CTTTTGATCACCGTGCAGGGCGG; CTTGGACACGGAAATCTACGAGG

• gRNA set 2: GATGAGCGAGGAGGGCACGATGG; ACCCATAACCAAGTTCTGGGCGG

• gRNA set 3: GACTACAGGGATGAGCGAGGAGG; GAATACCACTCCTGCCACCGCGG

These constructs were injected by Rainbow Transgenic Flies into yw, nos-phiC31; PBac{attP-9A} embryos (SI Appendix, Fig. S1A).

gRNA Construct Efficiency Test.

To confirm the efficiency of constructs expressing gRNAs, we employed the “Cas9-LEThAL” method (26) to validate the efficiency of our gRNA constructs. Males carrying gRNA constructs were crossed to act-Cas9, lig4- mutant females. Our gRNA set 3 construct caused 100% lethality in lig4 mutant male offspring while the other 2 sets did not (SI Appendix, Table S1), suggesting it has the highest efficiency.

Fly Strains and Maintenance.

All Drosophila strains and crosses were maintained or performed on standard yeast-glucose-agar media at 25 °C (29 °C for all crosses involving the GAL4/UAS system to enhance its efficiency; ref. 43) on a 12/12 light/dark cycle. The following fly lines were obtained from Bloomington Drosophila Stock Center: trpml¹ (28992), UAS-trpm^RNAi (57871), matα4-GAL4-VP16 (7062), 20XUAS-IVS-GCaMP6s-attP40 (42746), UAS-Cas9-attP2 (54595), yw, nos-phiC31; PBac{attP-9A} (35569).

Buffer Reagents, Drug and Inhibitor Treatment.

IB was made as previously described (9). IB contains 55 mM NaOAc, 40 mM KOAc, 1.2 mM MgCl₂, 1 mM CaCl₂, 110 mM sucrose, 100 mM Hepes in ddH₂O. IB is adjusted to pH 7.4 with NaOH and filter sterilized. We found that modified RB functioned as well as activation buffer (AB; ref. 9) for activating eggs, so we used RB for this purpose in our experiments. RB contains 55 mM NaOAc, 8 mM KOAc, 20 mM sucrose, 0.5 mM MgCl₂, 2 mM CaCl₂, 20 mM Hepes in ddH₂O. RB is adjusted to pH 6.4 with NaOH and filter sterilized. For tests of Trpm inhibitors, stock solutions of MgCl₂ (Mallinckrodt) and NS8593 (Sigma-Aldrich) were prepared and added to IB and RB to the indicated final concentrations, on the day of the experiment. The published concentration of Mg²⁺ (10 mM) was used to inhibit Drosophila Trpm (23). For NS8593, we tried a series of concentrations starting from published results on mammalian TRPM7 (30 μM) (24). We found that 100 μM is the optimal concentration for Drosophila Trpm inhibition.

Stage 14 oocytes, dissected from virgin females that had been aged on yeast for 3 to 5 d, were incubated in inhibitor-containing IB or control IB for 30 min before experiments. Inhibitor-treated oocytes were then switched to incubation in RB that also contained the inhibitor at the same concentration, and oocytes incubated in control IB were activated in control RB lacking the inhibitor.

Imaging.

For calcium wave visualization, oocytes were imaged using a Zeiss Elyra Super Resolution Microscope with a 5× or 10× lens with Zen software. The detection wavelength was set to 493 to 556 nm for GCaMP signal.

For immunostaining, fixed and stained embryos were imaged using a Zeiss LSM880 Confocal Multiphoton Microscope with a 10× lens and the Zen software. The detection wavelength was set to 495 to 634 nm for FITC signal, and 410 to 495 nm for DAPI. Z-stack images were taken from the shallowest to the deepest visible planes with a pinhole of 100. Maximum intensity projection of captured images was performed using the Zen software.

To examine ovary morphology, ovaries were dissected from 3- to 5-d-old virgin females aged on yeast. Their images were captured using an Echo Revolve Microscope, with a 4× lens and brightfield imaging settings. Ovary images were processed with the Echo Revolve App.

All acquired images were processed with ImageJ software (44) as needed.

In Vitro Egg Activation Assay.

Oocytes were dissected from the indicated female flies and induced to activate in vitro following methods as described previously (9). For imaging, oocytes were placed in a drop of IB in a glass-bottomed Petri dish. After imaging parameters were configured, IB was replaced by RB to induce egg activation (11). Time-lapse images were taken at 1-s or 10-s intervals, for 10 to 20 min after the addition of RB. The distance traveled by the wavefront was measured using ImageJ software, and the elapsed time was recorded. The propagation rate of the calcium wave was calculated as distance traveled by the wavefront divided by time.

Egg-Laying and Egg Hatchability Assay.

Virgin females of the indicated genotypes were aged on yeasted food vials for 3 to 5 d, before mating with Oregon-R-P2 (ORP2) WT males. Matings were observed and the males were removed after a single mating had completed. Females were allowed to lay eggs in the mating vial for 24 h and were then transferred to a new vial. Females were transferred 3 times before they were discarded. The number of eggs and pupae were counted. Egg hatchability was calculated by the number of pupae divided by the number of eggs. To confirm there was no postembryonic developmental arrest in eggs laid by mutants that could have affected our hatchability score, we counted the number of unhatched eggs for 3 d after egg laying and thus determined the number of hatched eggs. This number equaled the number of pupae, indicating that our method of calculating egg hatchability was reliable.

RT-PCR.

RNA was extracted from 4 to 5 pairs of ovaries from virgin females aged in yeasted vials for 3 to 5 d using TRIzol/chloroform. Seven hundred and fifty nanograms of RNA from each sample underwent DNase (Promega) treatment and cDNA synthesis using SuperScript II Reverse Transcriptase (Invitrogen) following the manufacturer’s instructions. A 30-cycle PCR amplification was performed with GoTaq polymerase (Promega) with the following conditions: 95 °C at 2 min, 95 °C at 40 s, 54 °C at 40 s, 72 °C at 40 s, 72 °C at 10 min. PCR products were run on 1% agarose gels, and the DNA was stained with 1 μg/mL ethidium bromide. Captured gel images were processed with ImageJ software.

The expression level of ribosome protein gene rpl32 was used as a normalization control (45). The following primers were used for RT-PCR: trpm-F: TCACTGTGCTGGTGAAGATG; trpm-R: CCAGAGGTCCCAGGTATTTATTC (amplicon size with cDNA as template is 324 bp. The reverse primer spans an exon-exon junction and will amplify a 513-bp band with genomic DNA as template.). rpl32-F: CACCAGTCGGATCGATATGC; rpl32-R: CGATCCGTAACCGATGTTG (amplicon size with cDNA as template is 120 bp. The primers flank an exon-exon junction and will amplify a 182-bp band with genomic DNA as template.).

Immunostaining.

Embryos and eggs from the indicated mating were collected from grape juice/agar plates. Embryos were dechorionated in 50% commercial bleach, fixed in methanol/heptane (46), and stored at 4 °C until use. Fixed embryos were washed with phosphate buffered saline with 0.1% Tween-20 (PBST) 3 times for 5 min each and blocked with PBST containing 5% vol of normal goat serum (PBST-NGS). Embryos were then incubated overnight at 4 °C in PBST-NGS containing mouse monoclonal anti-α-Tubulin-FITC antibody (Sigma-Aldrich) at a dilution of 1:200 and RNaseA (Roche Applied Science) at a concentration of 5 μg/mL. Embryos were then washed with PBST 3 times for 5 min each. DNA was stained with 1 μg/mL DAPI (Molecular Probes) in PBST for 5 min, and embryos were mounted on glass microscope slides in antifade mounting buffer.

Bleach Resistance Assay.

Mature oocytes were dissected, in IB, from the indicated females. Eggs laid by indicated females after mating to indicated males were collected from grape juice/agar plates. Both oocytes and embryos were incubated in 50% commercial bleach for 2 min (10). Numbers of oocytes and embryos before and after incubation were counted to calculate the percentage that survived bleach treatment.

Statistics.

Pearson’s χ² test was used to compare the incidence of calcium waves and ovary morphology defects. Student’s t test was used to compare the propagation speed of calcium waves and egg hatchability.