Maternal age effects on offspring life history are known in a variety of organisms, with offspring of older mothers typically having lower life expectancy (Lansing Effect). However there is no consensus on generality and mechanisms of this pattern. We tested predictions of Lansing Effect in several Daphnia magna clones and observed clone-specific magnitude and even direction of the maternal age effect on offspring longevity.  We also report ambidirectional, genotype-specific effects of maternal age on daughters' propensity to produce male offspring. Focusing on two clones with contrasting life-histories, we demonstrate that maternal age effects can be explained by lipid provisioning of embryos by mothers of different ages. Individuals from a single-generation maternal age reversal treatment showed intermediate lifespan and intermediate lipid content at birth. In the clone characterized by the “inverse Lansing Effect” neonates produced by older mothers showed higher mitochondrial membrane potential in neural tissues than their counterparts born to younger mothers. We conclude that an “inverse Lansing Effect” is possible, and hypothesize that it may be caused by age-specific maternal lipid provisioning creating a calorically restricted environment during embryonic development, which, in turn reduces fecundity and increases lifespan.

Origin and maintenance of clones

Daphnia magna clones used in this study (Supplementary Table 1) were obtained from Basel University Daphnia stock collection in Basel, Switzerland. They are a subset of clones that have been previously characterized for a number of life-history traits (Coggins et al, 2021b) and were chosen to represent a range of clone-specific life expectancies. Stocks were maintained in the lab at 20 °C in 200 mL jars with COMBO water (Kilham et al. 1998), 10 adults per jar and fed a diet of Scenedesmus acutus at the concentration of 100,000 cells per mL per day or 2x106 cells/Daphnia/day. This Daphnia and food density was the same in all experiments. Supplementary Table 1 lists clones’ IDs as recorded by Basel clone collection. For the sake of brevity clones will be thereafter referred to by the first two letters of their IDs that are indicative of the country of origin (using Internet domain two-letter codes).

Lifespan experiments

Because the interactions discussed below span three generations, we use the following conventions to avoid ambiguity. Generation 1 females’ age at the time they give birth to Generation 2 females will be referred to as “maternal age” (MA) in the analyses of its effects on their daughters (e.g. on the daughters’ longevity) and “grand-maternal age” (GMA) in the analysis of its effects on Generation 1 females’ grandchildren (e.g., grandchildren’s sex). Generation 2 females will be referred to as “daughters” relative to Generation 1 females, or “mothers” relative to Generation 3 individuals, which, in turn, are referred to as “offspring”. Males in Generation 3 are also referred to as “sons” or “grandsons” relative to Generation 2 and 1, respectively.

Details of the lifespan experiments are summarized in Table 1. We first conducted a larger experiment with five clones that were maintained in groups of 5 in 100 mL jars and in which individual life history parameters were not measured. We then chose two clones for a smaller experiment with individually maintained Daphnia and more detailed life history data recorded. The choice of the two clones was influenced by their contrasting life-histories: in four separate previous experiments (Coggins at al. 2021b; Anderson et al. 2022; L. Yampolsky and M. Ekwudo, unpublished data) they were shown to differ in early reproduction and lifespan, with the GB-EL75-69 clone (thereafter “GB”) characterized by longer lifespan than the FI-FSP1-16-2 clone (thereafter “FI”).

In both experiments, in order to obtain Generation 2 daughters of older (thereafter the “O” maternal age) and younger (thereafter the “Y” maternal age) Generation 1 females simultaneously, two grand-maternal cohorts of each clone were created by collection of neonate females born by 15-20 days old Daphnia, staggered 50-55 days apart (which corresponds to approximately the median lifespan and which implies that Daphnia in the Y treatment went through three 15-20 days long generations during the lifetime of the O treatment grandmothers, see Fig. 1). Additionally, in order to test the reversibility of any grand-maternal effects though any hypothetical “rejuvenation” effects of being born to younger mothers, a subset of Experiment 2 Generation 2 females were born to 14.6 ± 2.65 (SD) mothers, who were, in turn, daughters of 70-day old mothers (thereafter the maternal age reversal, OY, treatment, Fig. 1, gray.) This treatment was limited to only one clone, GB. Experiment 1 was conducted in 2 blocks with cohort sizes 353 individuals (in 70 independent jars) and 1095 individuals (in 220 jars).  Experiment 2 consisted of a single cohort of 153 individuals (each in an independent vial).

Sex ratios, clutch size, offspring size

Sex ratio was determined in clutches produced by females between their age at maturity and age of 40 days. Females who died before that age were excluded from the analysis to avoid bias in sex ratios, as early clutches rarely contain males. Additionally, to check for sex ratio of offspring produced by older mothers, sex ratio of offspring was measured for a subset of mothers 55-100 days old in experiment 1 and mothers of 40-85 days in experiment 2.

Nile red staining for lipids

In order to quantify maternal provisioning of storage lipids to offspring, newborn Daphnia (<24 h old) from the same clutches from which Generation 2 FI and GB females came from were stained with Nile Red dye for 2 hours with the final dye concentration 1 mg/mL, achieved by adding 5 uL of 200 mg/mL stock solution in acetone to 995 uL of combo water. Fluorescence was recorded using EVOS microscope (4x objective, aperture 0.13). Fluorescence in the entire body was measured (Supplementary Figure 1A). Because the distribution of storage lipid bubbles is patchy, the histogram of intensities was recorded and the fraction of intensities above an arbitrary chosen threshold was obtained, with the threshold chosen in such a way that it masked-in the lipid bubbles, leaving out the rest of the body. Specifically, the 8-bit image gray value of 200 was chosen as the threshold with 2.5% of pixels showing intensity above this threshold in 3 randomly chosen images. The same threshold was applied to all images. This allows the analysis of the portion of pixels located inside and the portion of fluorescence intensity emanating from the brightly fluorescent lipid vesicles, excluding background fluorescence emanating from non-storage lipids in other tissues. The portion of pixels and the portion of fluorescence from the lipid bubbles are therefore proxies for the portion of the body occupied by lipid storage and lipids density in these storage areas, respectively.

Mitochondrial potential measurements

 Mitochondrial membrane potential (DYm)is commonly used as a measure of mitochondrial quality, with the reduced DY_m values indicative of loss of membrane integrity and/or reduced function of electron transport chain (Perry et al. 2018) and correlate with apoptosis and aging (Nicholls 2004). Mitochondrial membrane potential is commonly used by means of measuring the accumulation of cationic fluorophore rhodamine-123 inside the mitochondria (Emaus et al. 1986; Huang et al. 2007; Perry et al. 2018), the rate of such accumulation is proportional to the magnitude of proton gradient across the inner mitochondrial membrane. Mitochondrial potential was measured by means of rhodamine-123 staining in neonates born to either young or old mothers treated as described above. Newborns <12 h old were placed in groups of 5 into 1.5 mL tubes containing 0 - 10 uM rhodamine-123 in COMBO water for 24 hours (Coggins et al. 2021a). The fluorescence was measured with Leica DM3000 microscope with a 10x objective (0.22 aperture) equipped with Leica DFc450C camera using the 488 nm excitation / broadband (>515 nm) emission filter. The following Regions of Interest (ROI) were selected (Supplementary Figure 1B): 2^nd antenna and heart (representing muscle tissues), brain and optical lobe (representing neural tissue), 2^nd epipodite and nuchal organ (representing excretory/osmoregulatory organs and non-neural head tissue (where the fluorescence was emitted largely from the head carapace epithelium without any organized tissue beneath). Median fluorescence (background subtracted) was recorded with exposure of 100 ms with gain 1, except the rhodamine concentration of 0, which was measured with gain 10 (and the resulting measurement divided by this factor) using ImageJ software (Rasband 2018).