Fig 2: Transcriptional changes associated with some DMR-associated genes in F1 and F2 wildtype somatic tissue.a–f RT-qPCR analysis of mRNA expression of genes located in close proximity to sperm DMRs in a, d embryos at E10.5, b, e placental trophoblast at E10.5, and c, f adult livers of wildtype (Mtrr+/+) F1 and F2 progeny. Tissue from phenotypically normal F1 wildtype individuals (orange bars) or F2 wildtype individuals (purple bars) or severely affected F2 wildtype individuals (pink bars) derived from an F0 Mtrr+/gt male was assessed. C57Bl/6 J control tissues (black bars) were assessed as controls. Pedigrees indicate specific mating schemes assessed (see also Supplementary Fig. 1a, d). Data are plotted as mean ± standard deviation, and are presented as relative expression to C57Bl/6 J levels (normalised to 1). Embryos: C57Bl/6 J, N = 4–5 embryos; F1 Mtrr+/+, N = 4–5 embryos; F2 Mtrr+/+ phenotypically normal or severely affected, N = 5 embryos. Trophoblast: C57Bl/6 J, N = 5–7 placentas; F1 or F2 Mtrr+/+, N = 8 placentas. Liver: C57Bl/6 J, N = 6–8 livers; F1 Mtrr+/+, N = 6 livers; F2 Mtrr+/+, N = 5–6 livers. Independent t tests or one-way ANOVA with Dunnett’s multiple comparisons tests were performed. d Hira **p = 0.0049, ***p = 0.0002; f Tshz3 **p = 0.0011; Cwc27 **p = 0.0095. Each gene was associated with the following sperm DMR shown in Fig. 3: Tshz3, DMR E115; Dynlt1a, DMR E52; Exoc4, DMR E112; March5, DMR C7; Cwc27, DMR D20; Hira, DMR A10. F0, parental generation; F1 first filial generation; F2, second filial generation. Pedigree legend: circle, female; square, male; blue outline, C57Bl/6 J control; black outline, Mtrrgt mouse line; white filled, Mtrr+/+; half-white/half-black filled, Mtrr+/gt. Source data are provided as a Source Data file.

Fig 3: Dysregulation of Hira RNA expression in embryos aligns with a pattern of maternal phenotypic inheritance.a–d RT-qPCR analysis of Hira mRNA (solid bars, primer set 1) and Hira lncRNA (striped bars, primer set 2) expression in embryos and placentas at E10.5. The tissue and pedigrees were assessed (see also Supplementary Fig. 1a, d–e for the breeding scheme): a F1 wildtype (Mtrr+/+) conceptuses from F0 Mtrr+/gt males (orange bars; N = 4–5 embryos, N = 8 placentas), b F2 wildtype conceptuses from F0 Mtrr+/gt males (purple bars, N = 5 embryos and N = 7–8 placentas from phenotypically normal (n) conceptuses; pink bars, N = 5 severely affected (a) embryos), c F3 wildtype conceptuses from F0 Mtrr+/gt males (teal bars, N = 3–4 phenotypically normal embryos; turquoise bars; N = 3 severely affected embryos), and d F1 wildtype embryos from F0 Mtrr+/gt females (orange bars, N = 7 phenotypically normal embryos; green bars, N = 3 severely affected embryos), F2 wildtype embryos from F0 Mtrr+/gt females (purple bars, N = 7 phenotypically normal embryos; pink bars, N = 7 severely affected embryos), and F3 wildtype embryos from F0 Mtrr+/gt females (teal bars, N = 5–7 phenotypically normal embryos; turquoise bars, N = 3 severely affected embryos). C57Bl/6 J conceptuses were controls (black bars, N = 4–8 embryos or placenta/experiment). e–h Western blot analysis showing HIRA protein expression in F1 wildtype (orange bars) and F2 wildtype (purple bars) conceptuses derived from e to f F0 Mtrr+/gt males or g–h F0 Mtrr+/gt females. C57Bl/6 J (C57) conceptuses (black bars) were assessed as controls. e, f C57Bl6/J: N = 4–5 embryos, N = 4–5 placentas; F1: N = 4 embryos, N = 3–8 placentas; F2: N = 7–9 embryos, N = 7–8 placentas. Images of western blot gels showing HIRA protein with β-actin as a loading control in f and h were quantified in e and g, respectively, using the background subtraction method. HIRA protein levels were normalised to β-actin. All RNA and protein data were plotted as mean ± standard deviation and relative to C57Bl/6 J (normalised to 1). Experiments were conducted in technical duplicates (protein) or triplicates (RNA). Statistical analyses: a–d Two-tailed independent t test or Kruskal–Wallis test with Dunn’s multiple comparison. a **p = 0.0034; b **p = 0.0049, ***p = 0.0002; c ***p = 0.0004, ****p < 0.0001; d F1: **p = 0.0056, F2 (primer set 1): #p = 0.06, *p = 0.0203, F2 (primer set 2): *p = 0.0457, F3: *p = 0.0161. e, f One-way ANOVA. p values are shown on graphs. Pedigree legend: circle, female; square, male; blue outline, C57Bl/6 J line; black outline, Mtrrgt mouse line; white fill, Mtrr+/+; half-black-half-white fill, Mtrr+/gt; black fill, Mtrrgt/gt. F0, parental generation; F1, first filial generation; F2, second filial generation; F3, third filial generation. Source data are provided as a Source Data file.

Fig 4: Proposed model of maternal grandparental phenotype inheritance in Mtrrgt model that implicates epigenetic instability in the germline.A model proposing how epigenetic instability generated by the Mtrrgt mutation might be differently inherited over multiple generations depending upon whether TEI is initiated by a an oocyte or b sperm of an F0 Mtrr+/gt individual. Trends of Hira mRNA expression are shown. Pedigree legend: circle, female; square, male; blue outline, C57Bl/6 J control; black outline, Mtrrgt mouse line; white filled, Mtrr+/+; half-white/half-black filled, Mtrr+/gt. See also Supplementary Fig. 1 for breeding schemes. Scale bar: 500 μm. F0, parental generation; F1, first filial generation; F2, second filial generation; F3, third filial generation.