Spindle assembly checkpoint strength is governed by cell size and PAR-mediated cell fate determination in C. elegans

The spindle assembly checkpoint (SAC) is a conserved mitotic regulator that preserves genome stability. Despite its central role in preserving the fidelity of mitosis, the strength of the SAC varies widely between cell types. How the SAC is adapted to different cellular contexts remains largely unknown. Here we show that both cell size and cell fate impact SAC strength. While smaller cells have a stronger SAC, cells with a germline fate show increased SAC activity relative to their somatic counterparts across all cell sizes. We find that enhanced SAC activity in the germline blastomere P1 requires proper specification of cell fate downstream of the conserved PAR polarity proteins, supporting a model in which checkpoint factors are distributed asymmetrically during early germ cell divisions. Our results indicate that size scaling of SAC activity is modulated by cell fate and reveal a novel interaction between asymmetric cell division and the SAC.


Introduction 1
the mitotic checkpoint complex (MCC), which acts as a diffusible "wait-anaphase" signal (11,12). 23 The MCC consists of Mad2, BubR1 and Bub3 bound to the anaphase-promoting 24 complex/cyclosome (APC/C) cofactor Cdc20 (13). The MCC binds to the APC/C (14,15), 25 blocking degradation of APC/C substrates, namely Cyclin B (16) and securin (17), thereby 26 maintaining a mitotic state. Once all kinetochores are stably bound by microtubules, 27 disassembly of the MCC (18,19), dephosphorylation of SAC proteins (20) and stripping of Mad1 28 and Mad2 from all kinetochores (21) effectively silences the SAC and permits anaphase onset 29 and mitotic exit. 30 Mitotic exit, despite failure to satisfy the SAC, occurs when activity of Cyclin B/Cdk1 falls 31 below the threshold necessary to maintain a mitotic state (5). In mammalian cells, this mitotic 32 "slippage" occurs because the active cytoplasmic pool of MCC is not sufficient to completely 33 inhibit the APC/C, and the progressive degradation of Cyclin B eventually enables mitotic exit 1 (22). The steady-state concentration of the MCC pool is influenced by the rate of new, 2 kinetochore-catalyzed MCC generation and disassembly of existing cytoplasmic MCC (23,24), 3 the former being intimately related to the severity of the spindle defect (25)(26)(27). Thus variation in 4 the strength of the SAC, as assayed by the duration of mitotic delay following spindle 5 perturbation, may be linked to differences in MCC production, activity or stability. 6 One factor that contributes to SAC strength in certain circumstances is cell size. In vitro 7 experiments using Xenopus laevis egg extracts suggested that an increased nuclear to 8 cytoplasmic ratio, as would be found in smaller cells, could increase SAC activity (28). Recent 9 work in C. elegans embryos has shown that the strength of the SAC indeed scales with cell size, 10 with smaller cells exhibiting longer SAC-dependent delays upon spindle perturbation (29). 11 However, in other organisms, the SAC remains inactive until the mid-blastula transition and 12 acquisition of SAC activity is neither accelerated by decreasing cell volume (Xenopus laevis, (8,13 9)) nor delayed by increasing cell volume (Danio rerio, (10)), indicating that SAC activity can 14 also be developmentally regulated independently of changes in cell volume. How this 15 developmental regulation of the SAC is achieved is unknown. 16 We have recently reported that the adult germline stem cells (GSCs) of C. elegans 17 exhibit a stronger SAC relative to early embryonic cells (30) providing a tractable model in which 18 to examine SAC adaptability. Here we use an inducible monopolar spindle assay to investigate 19 the developmental origins of enhanced germline SAC activity. In agreement with Galli and 20 Morgan (29), we find that the duration of SAC-dependent mitotic delays increases as cell size 21 decreases during embryogenesis. However, the relationship between cell size and SAC activity 22 is strongly influenced by cell fate, with cells in the germline lineage displaying a stronger SAC 23 relative to their cell size than cells with a somatic fate. At the 2-cell stage, we find that differential 24 SAC activity in the somatic AB versus germline P 1 blastomere requires the highly conserved 25 PAR polarity proteins, supporting a model in which asymmetric partitioning of a checkpoint-26 enhancing factor increases the strength of the SAC in germline cells. 27

Results 28
Germline blastomeres are more sensitive to spindle perturbations 29 C. elegans GSCs are derived from a single founder cell (P 4 ), which is specified during 30 embryogenesis by a series of asymmetric cell divisions (31). As the C. elegans embryonic 31 lineage is invariant and fully mapped, each cell can be identified by its position within the 32 embryo and its cell cycle characteristics ( Figure 1A, (31,32)). We were therefore able to ask 33 whether the embryonic precursors to GSCs, cells in the germline P lineage, also exhibited an 1 enhanced SAC, by comparing the duration of mitosis between each of the founding cell lineages, 2 in cells with or without spindle perturbations. As C. elegans embryos are largely refractory to 3 treatment with small molecule spindle poisons without physical or genetic manipulations to 4 permeabilize the egg shell (33,34), we opted for a genetic method to induce spindle 5 perturbations with temporal control. We utilized temperature sensitive alleles of genes involved 6 in spindle formation, combined with fluorescent markers to visualize the mitotic spindle (β-7 tubulin fused to GFP, hereafter β-tubulin::GFP) and/or the nucleus/chromatin (histone H2B 8 fused to mCherry, hereafter H2B::mCH). This allowed us to disrupt spindle formation with a 9 simple temperature shift using a standard temperature-controlled imaging chamber and to follow 10 mitotic progression by monitoring changes in nuclear and chromosome morphology. 11 We used a recessive temperature-sensitive allele of the gene encoding the polo-related 12 kinase ZYG-1, which prevents centriole duplication and produces cells with monopolar spindles 13 (zyg-1(or297), hereafter zyg-1(ts), (35,36)). Monopolar spindles can generate mitotic delays 14 that are indistinguishable from Nocodazole treatments that either partially (6,7) or completely 15 (25) inhibit microtubule formation, suggesting a robust SAC response, and are an established 16 means for SAC studies in C. elegans (37). zyg-1(ts) is fast-acting (36) and readily permitted the 17 induction of monopolar spindles in all embryonic cells up to the 16-cell stage. Following NEBD, 18 condensed chromosomes spread radially around the single spindle pole and remained 19 dispersed until the start of decondensation ( Figures 1B and 1C, Movie S1). In later stage 20 embryos, centrosome duplication was unaffected by temperature shift and all divisions were 21 bipolar (data not shown). 22 In agreement with previous reports, we found that the duration of mitosis in cells without 23 spindle perturbations was invariant across embryonic stages and cell lineages ( Figure 1D, (29, 24 38)). In cells in which spindle formation was perturbed, we noticed significant differences in the 25 duration of mitosis between cells at different embryonic stages and in different lineages. Notably, 26 cells in the germline P lineage remained in mitosis for significantly longer than both their 27 immediate somatic siblings and somatic cells at later embryonic stages ( Figures 1E). A similar 28 result was obtained using a temperature sensitive allele of the microtubule subunit β-tubulin 29 (tbb-2(or362), (39)), which alters microtubule dynamics and/or stability, to disrupt spindle 30 formation ( Figure S1 and Movie S2) 31

32
The difference in mitotic delay between germline and somatic cells is not solely due to 33 cell size 1 Specification of the germline is achieved via a series of asymmetric cell divisions, such 2 that, the germline blastomere is always smaller than its immediate somatic sibling (31). As the 3 strength of the SAC negatively correlates with cell size during C. elegans embryogenesis (29), 4 the increased duration of monopolar mitoses in germline blastomeres could be due to their 5 relatively small size. Thus, we considered cell size in two ways. We used nuclear area, which 6 scales with cell size in many organisms including C. elegans (40,41), and which we could 7 measure for the same cells in which we assayed mitotic delay, as a proxy for cell size ( Figure  8 S2A and S2B). In both the germline P and somatic AB lineages, the duration of monopolar 9 mitoses negatively correlated with nuclear area, supporting the idea that smaller cells possess a 10 stronger SAC (r = -0.73, p < 0.001 and r = -0.61, p < 0.001, respectively). However, germline 11 cells exhibited longer mitotic delays than somatic AB lineage cells with similar nuclear areas 12 ( Figure 2A), suggesting that, between comparably sized cells, the SAC is stronger in germline 13 cells. 14 Nuclear size reached an upper limit and did not reflect differences in cell volume in larger 15 cells ( Figure S2B and S2F). A similar phenomenon has been described for the size scaling of 16 both individual chromosomes and the mitotic spindle (42,43). Thus, in a separate experiment, 17 we calculated the average volume for each cell in the different lineages relative to the volume of 18 the founding P 0 blastomere ( Figure S2C-E). In both lineages, mitotic delay increased as cell 19 volume decreased ( Figure 2B, r = -0.62, p < 0.001 for the AB lineage; r = -0.72, p < 0.001 for the 20 P lineage). However, the relationship between cell volume and the duration of monopolar 21 mitoses differed significantly between the two lineages (AB versus P regression slope: p = 0.03). 22 Here again, germline cells experienced longer mitotic delays relative to their cell volume than 23 somatic AB cells, suggesting that SAC strength is also shaped by cell lineage. The increased duration of monopolar mitoses in germline blastomeres could be due to a 28 stronger SAC; alternatively, checkpoint-independent factors could contribute to a delayed mitotic 29 exit in these cells. To discriminate between these possibilities, we examined the duration of 30 monopolar mitoses in the germline P and somatic AB lineages when checkpoint activity was 31 eliminated using RNAi depletion of the C. elegans ortholog of Mad1, mdf-1(RNAi) ( Figure 2C  32 and 2D, (44)), or a null allele of the C. elegans ortholog of Mad2, mdf-2(tm2190) ( Figure 2E, (45, 33 46)). In the absence of checkpoint activity, cells with monopolar spindles rapidly exited mitosis, 1 with chromosome decondensation evident within 4 minutes of NEBD ( Figure 2C and Movie S3). 2 Plotting the duration of monopolar mitoses in the germline P and somatic AB lineages relative to 3 cell size (as approximated by nuclear area; see Figure S2 and preceding section) revealed that 4 all cells, irrespective of their size or lineage, exited mitosis with the same timing ( Figure 2D and 5 2E). The duration of checkpoint deficient monopolar mitoses was comparable to that of normal 6 bipolar divisions (3.3±0.5 versus 3.4±0.3 minutes, respectively). Thus monopolar spindle-7 induced mitotic delays are entirely checkpoint dependent and differences in the duration of 8 these delays are likely to reflect differences in the strength of the SAC. 9 10 SAC activity may also be subject to developmental regulation. 11 The strength of the checkpoint increases in both germline and somatic AB cells as cell 12 size decreases. If cell size affected SAC strength according to a universal size-scaling 13 relationship, the linear regression lines for mitotic delay versus cell volume for each lineage 14 should be parallel. Instead, SAC activity increases more rapidly with decreasing cell volume in 15 the germline P lineage ( Figure 2B and S3A), suggesting that either cell size impacts SAC 16 activity differently in the two lineages or that SAC activity is progressively modified during 17 lineage development. To distinguish between these possibilities, we modified embryo size 18 experimentally and measured the duration of monopolar mitoses at a single developmental 19 stage, in the somatic AB and germline P 1 blastomeres of 2-cell stage embryos. 20 We used RNAi depletion of ANI-2 and PTC-1 to generate small embryos, and C27D9. 1 21 to generate large embryos ( Figure 3A and data not shown, (41,(47)(48)(49)). Altering embryo size by 22 these means gave us a 5-fold range of embryo volumes, from 1.2 x 10 4 μm 3 , slightly smaller 23 than a control P 1 cell, to 6.1 x 10 4 μm 3 , about twice the size of an average control embryo, 24 without disrupting basal mitotic timing and other defining features of AB and P 1 , such as cell 25 cycle asynchrony and cell size asymmetry ( Figure S3B-D). Hereafter cell volume is expressed 26 as the radius of the equivalent sphere (R = 3 √(¾Vπ)), which we will call "cell size". 27 Cell size and the duration of monopolar mitoses were negatively correlated in both AB 28 and P 1 blastomeres ( Figure 3B, r = -0.86, p < 0.001 and r = -0.83, p < 0.001, respectively), 29 indicating that cell size can impact SAC activity irrespective of developmental stage. While the 30 rate at which the duration of monopolar mitoses changed relative to cell size was similar for AB 31 and P 1 ( Figure 3B, AB versus P 1 regression slope: p = 0.12), the average duration of monopolar 32 mitoses was always higher in P 1 cells ( Figure 3B, AB versus P 1 regression y-intercept: p = 33 0.008), indicating that P 1 cells have a stronger SAC regardless of cell size. These results 1 suggest that cell identity sets a baseline of SAC activity, which is then scaled according to cell 2 size similarly in AB and P 1 and argue against a cell fate-specific difference in SAC size scaling 3 during development of the AB and P lineages. Thus we favor the alternative hypothesis, that 4 checkpoint activity is modified within each lineage as development progresses. Altogether our 5 results demonstrate that SAC activity is determined by a combination of cell size, cell fate and 6 potentially developmental stage in the early C. elegans embryo. 7 8 PAR protein-mediated cytoplasmic asymmetries are required for differential SAC activity 9 in AB versus P 1 10 The pattern of checkpoint activity that we have observed in 2-to 16-cell stage embryos 11 mirrors the asymmetric inheritance of certain germline factors, raising the possibility that a 12 determinant of SAC activity could be similarly regulated. For example, several germline proteins 13 are initially present at low levels in the somatic siblings of germline blastomeres, but absent in 14 their progeny (50)(51)(52)(53), and germline-enriched maternal transcripts are initially symmetric during 15 cell division and subsequently lost from somatic cells and/or their progeny (54). Correspondingly, 16 the sibling of the germline blastomere in 4-and 8-cell stage embryos (EMS and C, respectively) 17 exhibited monopolar mitoses that were intermediate in duration to those in germline P and 18 somatic AB lineage cells, and, in the progeny of EMS (MS and E), these longer monopolar 19 mitoses were lost ( Figures 1E and 2B). The asymmetric inheritance of germline determinants is 20 regulated by the highly conserved PAR proteins, which lead us to ask whether differences in 21 SAC activity also require PAR proteins. 22 The polarized, cortical distribution of PAR proteins regulates both cell size and the 23 cytoplasmic distribution of cell fate determinants during the division of P 0 (55). PAR-6, PAR-3 24 and the atypical protein kinase C PKC-3 localize to the anterior cortex, while PAR-2 localizes to 25 the posterior cortex. The asymmetric cortical distribution of PAR proteins depends on mutual 26 antagonism between the anterior and posterior PARs. Consequently, in the absence of anterior 27 PARs, posterior PARs move into the anterior, and vice versa (55). The PAR-1 kinase is also 28 enriched in the posterior and is both cytoplasmic and cortical (56). Activity of PAR-1 regulates 29 the asymmetric segregation and/or degradation of cytoplasmic cell fate determinants, leading to 30 the enrichment of germline factors in P 1 and somatic factors in AB (57). PAR-1 plays a minimal 31 role in the regulation of cortical PAR protein asymmetries (58)(59)(60), but is absolutely required for 32 downstream cytoplasmic asymmetries (56,61,62). 33 To determine how much of the difference in SAC strength between AB and P 1 was due 1 to cell size, we compared the average difference in mitotic delay between AB and P 1 cells that 2 were the same size versus AB and P 1 cells where the normal cell size ratio was preserved. 3 Monopolar mitoses were 4.3±1.8 minutes longer, on average, in P 1 cells when AB was larger 4 than P 1 , irrespective of embryo volume, and 2.0±1.9 minutes longer, on average, in P 1 when AB 5 and P 1 were the same size ( Figure 3C and 3D). When we compared sibling AB/P 1 pairs from 6 control embryos versus embryos in which AB and P 1 were the same size but maintained 7 somatic versus germline fate (gpr-1/2(RNAi) (63-65)), the difference in the duration of 8 monopolar mitoses was reduced from 4.4±1.1 to 2.7±1.0 minutes ( Figure 3E and 3F and Movie 9 S4). Thus, in 2-cell stage embryos, approximately half of the difference in checkpoint strength 10 between AB and P 1 is due to their size difference. 11 We next asked whether the activity of PAR proteins was responsible for the remaining 12 difference in mitotic delay between equally sized AB and P 1 cells. While disrupting PAR proteins 13 in P 0 will change the fate of the resulting cells, for simplicity we will continue to refer to the 14 "anterior" blastomere as AB and the "posterior" blastomere as P 1 . RNAi depletion of PAR- 6 15 resulted in sibling AB and P 1 cells that were the same size, similar to gpr-1/2(RNAi) embryos 16 ( Figure 3E and 3F). However, in par-6(RNAi) embryos the difference in duration of monopolar 17 mitoses was further reduced to 1.0±0.9 minutes, with several AB/P 1 sibling pairs exiting mitosis 18 synchronously ( Figure 3G   results suggest that PAR-1-dependent cytoplasmic asymmetries, but not cortical polarity per se, 27 drive differential checkpoint activity in AB and P 1 . 28 29

Loss of either anterior or posterior PAR proteins increases SAC activity in AB 30
PAR proteins regulate the levels and localization of asymmetrically distributed cellular 31 components largely via the posterior enrichment of PAR-1. Removing anterior PAR proteins 32 permits PAR-1 activity in the anterior, resulting in two cells with germline P 1 traits. Conversely, 33 eliminating posterior PAR proteins allows somatic AB traits to emerge in both cells (55). To ask 1 whether SAC activity was subject to similar regulation, we measured the absolute duration of 2 monopolar mitoses in PAR protein-depleted AB and P 1 cells. The mean duration of monopolar 3 mitoses relative to mean cell size for par(RNAi) AB and P 1 cells (excepting par-2(RNAi)), more 4 closely resembled the predicted values for P 1 than AB control cells; whereas values for gpr-5 1/2(RNAi) embryos, in which polarity is normal, were similar to the corresponding control cell 6 ( Figure 4A). RNAi depletion of PAR proteins impacts the size of AB and P 1 , generating AB cells 7 that are slightly smaller and P 1 cells that are slightly larger than control AB and P 1 cells, 8 respectively. Thus, to confirm this trend, we utilized data from Figure 3B (hereafter referred to as 9 "control" cells), binned cells by size and compared the duration of monopolar mitoses in 10 par(RNAi) AB and P 1 cells to comparably-sized, but normally polarized and fated, control AB 11 and P 1 cells. 12 RNAi depletion of the anterior PAR-6 and PKC-3 proteins increased the duration of 13 monopolar mitoses in AB cells, such that mitotic delays in depleted AB cells closely resembled 14 those in control P 1 cells ( Figure 4B). Surprisingly, however, RNAi depletion of the posterior PAR-15 1 protein also increased the overall duration of monopolar mitoses in AB, while depletion of the 16 posterior PAR-2 protein gave an intermediate phenotype ( Figure 4B). AB and P 1 cells from gpr-17 1/2(RNAi) embryos, showed mitotic delays that were indistinguishable from comparably sized 18 control AB and P 1 cells, respectively ( Figure S4C). All par(RNAi) embryos displayed the 19 expected phenotypes with respect to cell cycle duration and asynchrony (66) ( Figure S5A) 20 indicating that our depletions were efficient. 21 We confirmed these results using a kinase-dead allele of par-1 (it51 (56)) and a near-null 22 allele of par-2 (lw32 (58, 67)), with their respective genetic controls (see Table S1 for complete 23 genotypes). AB cells in par-1(it51) embryos had monopolar mitoses similar in duration to control 24 P 1 cells and significantly longer than control AB cells ( Figure 4C). In par-2(lw32) AB and P 1 cells, 25 the mean duration of monopolar mitoses was largely unchanged. RNAi depletion of PAR-1 in 26 par-2(lw32) embryos eliminated the difference in mitotic delay between sibling AB/P 1 pairs 27 almost entirely (0.06 ± 0.9 minutes), and phenocopied par-1(it51) and par-1(RNAi) alone, with 28 unaffected P 1 cells and longer monopolar mitoses in AB cells ( Figure 4CD). par-1(it51), par-29 2(lw32) and par-2(lw32); par-1(RNAi) embryos all exhibited changes in cell cycle duration and 30 asynchrony consistent with their genotype ( Figure S5B). Under all conditions, par-1(-) AB cells 31 were significantly smaller than control AB cells and comparable in size to control P 1 cells ( Figure  32 4C and 4D); however, in light of our AB/P 1 size scaling and gpr-1/2(RNAi) results (Figure 3), 33 small AB cells should have shorter monopolar mitoses than comparably sized control P 1 cells. 1 As monopolar mitoses in par-1(-) AB cells are largely indistinguishable in duration from control 2 P 1 cells, we conclude that the impact of losing PAR-1 is not due exclusively to its effect on cell 3 size. Overall, our genetic and RNAi data are closely aligned and strongly suggest that PAR-1 4 activity is required for lower SAC activity in AB relative to P 1 . 5 6 Discussion 7 We have shown that, during C. elegans embryogenesis, both cell size and cell fate 8 impact the strength of the SAC. Spindle perturbations produce longer SAC-dependent mitotic 9 delays in smaller cells, supporting the long-standing hypothesis that the strength of the SAC can 10 be influenced by cell size (28,29), while cells in the germline P lineage delay for longer than 11 comparably-sized somatic cells, indicating that SAC strength is further subject to cell fate 12 specific regulation. Variation in the strength of the SAC could arise at many points in the 13 checkpoint signaling mechanism. Conceptually, we envision two primary possibilities: variation 14 at the level of signal generation or variation in the rate of signal degradation. The former 15 encompasses changes in the efficacy of MCC generation at the kinetochore or the number of 16 MCC-generating kinetochores. The latter is a catchall class for changes that occur away from 17 the kinetochore, including differences in the rate of cytoplasmic MCC disassembly or the 18 efficiency of APC/C inhibition. A third possibility could arise downstream of the SAC, if, for 19 example, cell size or cell fate changed the threshold of APC/C substrates at which cells exit 20 mitosis. We think this unlikely, as the timing of mitotic exit is invariant across cells of different 21 sizes and fates when SAC regulation is removed and APC/C activity is unimpeded. 22 Changes in cell size have been proposed to change the strength of the SAC via the first 23 mechanism, increased signal generation. Smaller cells have a higher ratio of DNA to cytoplasm 24 (28) and, effectively, a higher number of kinetochores per unit cytoplasm, thereby favoring MCC 25 production (29). In cells with monopolar spindles, germline cell fate could act to increase the 26 number of signal generating kinetochores. Cells with monopolar spindles fail to satisfy the SAC 27 due to the persistence of monotelic and syntelic kinetochore-microtubule attachments (68). The 28 presence of these attachments, albeit erroneous, may reduce the proportion of checkpoint-29 signaling kinetochores, as compared to Nocodazole treated cells, in which microtubule 30 formation is largely blocked (26). Accordingly, longer monopolar mitoses in germline-fated cells 31 could be achieved by further reducing kinetochore-microtubule attachment, thereby increasing 32 the proportion of MCC-producing kinetochores. Alternatively, a microtubule-dependent 33 checkpoint silencing mechanism functions at the kinetochore in C. elegans embryos (69) and 1 rendering this mechanism more or less robust could also affect the proportion of checkpoint-2 signaling kinetochores in cells with monopolar spindles. While we cannot exclude either of these 3 possibilities, we note that germline blastomeres may also exhibit longer mitotic arrests than 4 somatic cells following Nocodazole treatment, in which very few kinetochore-microtubule 5 attachments are made (29). Thus we favor the hypothesis that cell size sets the kinetochore to 6 cytoplasmic ratio and cell fate acts in parallel, either at the level of signal generation, by 7 modifying MCC production per kinetochore, or downstream of the kinetochore by modulating the 8 cytoplasmic activity of the checkpoint. 9 Our results indicate that, at the 2-cell embryonic stage, PAR protein mediated 10 cytoplasmic asymmetries underlie differential checkpoint strength between the somatic AB and Our results raise the question of why removing PAR-1, which is usually undetectable in 21 AB, increased the duration of monopolar mitoses in AB, while depleting PAR-2 had a minimal 22 effect. This result stands in contrast to a well-characterized asymmetric cellular behavior 23 regulated by PAR proteins, cell cycle asynchrony (66,70,72,73). If checkpoint activity were 24 regulated in an analogous fashion, we would expect loss of par-1 and par-2 to result in shorter 25 mitotic delays in P 1 , leaving AB largely unaffected. Instead, our results are consistent with a 26 model in which a checkpoint-promoting factor ("factor X") is enriched in P 1 via a PAR-1-27 dependent mechanism. In the absence of PAR-2, anterior PAR proteins expand into the 28 posterior, but are present in a graded fashion, and the posterior enrichment of some germline 29 factors (e.g. P granules) is often maintained (58,74). In addition, although PAR-1 is no longer 30 asymmetric in par-2(-) embryos, its activity is not lost (58,75,76). As differential mitotic delays 31 between AB and P 1 cells in par-2(lw32) embryos were entirely dependent on PAR-1, we 32 conclude that cytoplasmic asymmetries downstream of PAR-1, and partially independent of 33 PAR-2, are largely responsible for posterior enrichment of factor X. In the absence of PAR-6, 1 PKC-3 or PAR-1, this factor X is equally inherited by AB and P 1 , thereby increasing checkpoint 2 activity in AB while maintaining it in P 1 (Fig. 4e). This model requires that levels of factor X are 3 not limiting for checkpoint activity and that factor X is not degraded in the absence of PAR-1, at 4 least at the 2-cell stage. As in par-1 mutants, degradation of germline factors is delayed such 5 that germplasm proteins persist in all cells until the 4-cell stage (52), this latter condition is not 6 without precedent. 7 Altogether, our results suggest that a checkpoint-regulating factor (or factors) is 8 partitioned during the asymmetric division of germline blastomeres, downstream of PAR-9 mediated cell polarity, such that germline cells possess a stronger SAC relative to their nuclear 10 to cytoplasmic ratio than their somatic siblings. SAC activity may be further tuned within each 11 lineage as development progresses. Future work to identify the relevant molecular asymmetries 12 between somatic and germline cells driving differential SAC activity will permit further 13 investigation of this model and identification of where PAR-mediated cell polarity and checkpoint 14 regulation intersect. 15

C. elegans strains and culture 2
All strains were maintained at 15°C on NGM plates, seeded with E. coli bacteria (OP50) 3 following standard procedures (77). All genotypes are listed in Supplementary Table S1. L4 4 stage larvae were transferred to fresh OP50 plates at 15°C for 2 days, after which time embryos 5 were harvested from gravid adults for imaging. RNAi depletions were performed at 15°C by 6 feeding (78) as follows. Synchronized L1 larvae were obtained by sodium hypochlorite treatment 7 (1.2% NaOCl, 250mM KOH) and were plated onto NGM plates containing 1.5mM IPTG and 8 25μg/ml Carbenicillin, seeded with HT115 bacteria containing the empty feeding vector L4440. 9 Three days later early L4 larvae were transferred to fresh RNAi feeding plates seeded with 10 bacteria expressing doubled-stranded RNA for gene inactivation or the empty vector L4440 for 11 controls. Embryos were collected for imaging three days later. To isolate homozygous par-12 2(lw32) animals, L1 larvae on L4440 plates were shifted to 25°C for 24 hours, which allowed 13 identification of homozygous animals based on the linked, recessive, temperature-sensitive 14 allele unc-45(e286). Unc larvae were transferred to RNAi feeding plates, returned to 15°C and 15 imaged three days later. Control genotypes (UM471 or UM628, see Supplementary Table S1)  16 were similarly treated. The following clones from the Arhinger RNAi library were used: roughly 20μm thick z-stack was acquired, using either 1 or 1.5μm sectioning. 5 6 Image processing and measurements 7 Image processing and analysis was carried out in ImageJ (NIH). The duration of bipolar, 8 monopolar and tbb-2(ts) mitoses were determined manually by monitoring H2B::mCH 9 fluorescence. NEBD was defined as the first frame in which non-incorporated H2B::mCH was 10 lost from the nuclear area. DECOND was defined as the first frame in which the distribution of 11 H2B::mCH shifted from bright and compact to fainter and diffuse. Maximal z-stack projections of 12 unprocessed image files were scored. Single z-slices were examined if projected images were 13 ambiguous. For all strains in which β-tubulin::GFP was also present (all experiments excepting 14 par-1(it51) and its associated controls), the timing of NEBD was confirmed by the appearance of 15 microtubules in the nuclear space and the start of DECOND was concomitant with the start of 16 spindle disassembly and shrinking of the spindle pole. To exclude potentially confounding 17 effects of accumulated mitotic errors due to any partial conditionality of our temperature 18 sensitive alleles, only cells in which the preceding parental division was observed to be normal 19 were analyzed. AB and P 1 cell cycle duration was measured as the time from P 0 metaphase to 20 NEBD in each cell respectively, by manually monitoring H2B::mCH fluorescence. P 0 metaphase 21 was defined as the last time point prior to visible chromosome separation. NEBD was defined as 22 above. To measure nuclear area, a sum projection of the central 3 z-slices of the nucleus of 23 interest was processed and segmented. The average of the dimensions of a bounding box fit to 24 the segmented nucleus was taken as representative of nuclear diameter and nuclear area was 25 then calculated as the area of the corresponding circle ( Supplementary Fig. S2a,b). Reported 26 values reflect the average of measurements made at three time points, 1 to 2 minutes prior to 27 NEBD. To determine the relative volume of cells in each lineage from the 2 to 16-cell stage, the 28 position of the mitotic spindle midpoint along the division axis, relative to the midpoint of the 29 dividing cell was used to determine what proportion of each dividing cell was allocated to each 30 daughter ( Supplementary Fig. S2c,d). The Plot Profile tool in ImageJ was used to generate 31 signal intensity profiles along a line drawn parallel to the division axis, through the segregating 32 sister nuclei, in cells in which the cell membrane was labeled with mNeonGreen::PH (mNG::PH) 33 and nuclei were marked by H2B::mCH. Spindle displacement was defined as the offset of the 1 center point between the two membrane mNG::PH peaks versus the center point between the 2 two H2B::mCH peaks. When segregating sister nuclei were not in the same xy plane (i.e. the 3 cell division axis was tilted relative to the plane of imaging), the Stack Reslice tool in ImageJ 4 was used to construct an image of the long axis of the dividing cell from the encompassing z 5 slices. Measurements were made at three time points during anaphase, after the start of 6 membrane ingression. Nearly identical volume relationships were obtained when cell volume 7 was calculated from cross-sectional area and cell height measurements using the mNG::PH 8 membrane signal to manually outline cells (data not shown). The cell volume of AB and P 1 was 9 calculated by combining measurements of spindle displacement in P 0 , with measurements of 10 embryo volume. Spindle displacement was measured using the Plot Profile tool in ImageJ to 11 generate signal intensity profiles along the anterior to posterior axis of the embryo. bootstrapping, using a custom MATLAB script. For data presented in Fig. 3c,d, bootstrap  28 analysis was performed manually in Excel, by calculating the difference in the duration of 29 monopolar mitoses for all possible pairwise comparisons within a given size range. All possible 30 outcomes were displayed as a boxplot graph and means were compared by an Anova1 with 31 Tukey-Kramer post-hoc test. "Small" embryos are those in which AB is within ± 1 standard 32 deviation (s.d.) of the size of a Control(RNAi) P 1 cell. "Large" embryos are those in which P 1 is 33 within ± 1 s.d. of the size of a Control(RNAi) AB cell. "Normal" embryos are those in which AB is 1 within ± 1 s.d. of the size of a Control(RNAi) AB cell. "P 1 " cell size compares AB and P 1 cells 2 that are both within ± 1 s.d. of the size of a Control(RNAi) P 1 cell. "AB" cell size compares AB 3 and P 1 cells that are both within ± 1 s.d. of the size of a Control(RNAi) AB cell. For Fig. 3f,g, only  4 gpr-1/2(RNAi), par-6(RNAi) and par-1(RNAi) embryos in which the position of the mitotic spindle 5 in P 0 was roughly centered (50±2% of embryo length) were considered. For data presented in 6     relative to cell size in somatic AB (grey) and germline P 1 (red) cells following RNAi-induced 5 changes in embryo volume. Cell volumes were converted into the radius of the corresponding 6 sphere (R = 3 √(¾Vπ)), which we call "cell size". Both timing of NEBD to DECOND and cell size 7 were normalized to the mean value for Control(RNAi) AB cells. Lines represent the linear least 8 squares regression fit with 95% confidence interval (shaded regions). Linear regression models 9 are shown with statistically different coefficients in bold. Regression coefficients were compared 10 using a non-parametric bootstrap (p = 0.12 for slope and 0.008 for y-intercept). (C and D) 11 Comparison of the duration of monopolar mitoses in somatic AB and germline P 1 cells with a 12 normal cell size ratio (AB > P 1 ), in small, average and large embryos, versus AB and P 1 cells 13 that are both the same size (AB = P 1 ) as Control(RNAi) P 1 cells ("P 1 ") or Control(RNAi) AB cells  with 95% confidence interval (shaded regions) for control somatic AB (grey) and germline P 1 5 (red) cells and were generated from data in Figure 3D    (red) and somatic AB lineage (grey) in 2-to 16-cell embryos relative to mean cell size. Mean cell 3 size was calculated by converting cell volume measurements from Figure S2E into the radius of 4 the corresponding sphere (R = 3 √(¾Vπ), as in Figure 3B). Cells analyzed in (A) were treated 5 with Control(RNAi) and are independent from those analyzed in Figure 2B. The rate at which the 6 duration of monopolar mitoses increases with decreasing cell size is greater in the germline P 7 lineage than in the somatic AB lineage (p = 0.02). (B-D) Bar graphs showing the mean duration 8 of bipolar mitoses (B), cell cycle duration (C) and cell volume as a proportion of P 0 (D) in AB 9 (grey) and P 1 (red) cells following RNAi-induced changes in embryo volume. Error bars 10 represent ± 1 standard deviation of the mean. 11  with comparably sized control cells. Control cells were pulled from data represented in Figure  10 3B. Both control and gpr-2(RNAi) cells were binned by size, with bin edges set at ± 2 standard 11 deviations of the mean size for all gpr-2(RNAi) cells (both AB and P 1 ). In both Control(RNAi) and of GPR-1/2 (dark blue), PAR-6 (purple), PKC-3 (orange), PAR-1 (light blue) and PAR-2 5 (magenta). RNAi depletion of PAR-6 (purple) and PKC-3 (orange) increases cell cycle length in 6 AB, while RNAi depletion of PAR-1 (light blue) and PAR-2 (magenta) decrease cell cycle length 7 in P 1 . (B) Cell cycle lengths for par-1(it51) and par-2(lw32) AB and P 1 cells, with their respective 8 controls. Both par-1(it51) and par-2(lw32) P 1 cells have shorter cell cycle durations than control 9 P 1 cells, while par-1(it51) and par-2(lw32) AB cells are unaffected. For (A and B), data were 10 compared by an Anova1 with Tukey-Kramer post-hoc test and only statistically significant 11 differences relative to control AB or P 1 cells are shown. *** = p < 0.001, ** = p <0.01, and * = p < 12 0.05. Error bars show the 95% confidence interval for the mean. 13 1   Table S1. C. elegans strains used in this study. All strains, except UM399 (Gerhold 2015), 2 were constructed as part of this study using the following: EU782 (zyg-1(or297) II), EU858 (tbb-  All time-lapse movies were acquired on a Cell Observer SD spinning disc confocal (Zeiss; 2 Yokogawa) equipped with a stage-top incubator (Pecon) set to 26ºC, using an AxioCam 506 3