**Background:** Childhood obesity is a major risk factor for adult metabolic diseases, and epidemiological evidence (e.g., the Överkalix cohort) suggests that early-life nutritional imbalances can influence metabolic health in subsequent generations through non-genetic (epigenetic) mechanisms. The authors previously developed a mouse model of early adiposity via litter size reduction (SL, 4 pups/dam vs. 8 pups/dam in controls, C). SL-F0 mice developed obesity, glucose intolerance, insulin resistance, and hepatic steatosis, with the steatosis attributed to circadian rhythm misalignment. Here, the authors investigated whether these metabolic disturbances, particularly hepatic steatosis, are transmitted to the F1 offspring via the paternal lineage and explored the roles of sperm DNA methylation and small non-coding RNAs (sncRNAs) as potential epigenetic carriers.
**Methods:** F0 male mice were reared in control (C-F0, n=8 pups/litter) or small litter (SL-F0, n=4 pups/litter) conditions. At 3 months of age, C-F0 and SL-F0 males were mated with external control females to generate F1 offspring. All F1 litters were adjusted to 8 pups/dam to normalize postnatal nutrition. Hepatic triglyceride and cholesterol content were measured in 6-month-old male mice. Global gene expression profiling was performed using Affymetrix microarrays (Mouse 430 2.0) on liver RNA from C-F1 and SL-F1 mice (3 arrays/group, each pooling RNA from 3 mice). Differential expression was analyzed using Rank Prod (q<0.1). Sperm DNA methylation was assessed using Agilent 105K Mouse CpG Island microarrays on DNA from C-F0 and SL-F0 males (n not specified). Small RNA sequencing (Illumina HiSeq2500) was performed on testis RNA from 17 male mice (C-F0 and SL-F0), and differential expression of miRNAs, tRFs, and piRNAs was analyzed using DESeq2 (controlling for batch). qPCR validation of clock and lipid-related genes was performed in livers of 4-5-month-old C-F1 and SL-F1 mice (n≥6/group).
**Key Results:** SL-F1 mice showed significantly increased hepatic triglycerides and cholesterol compared to C-F1 mice (Figure 1C, p<0.05). Microarray analysis identified 394 genes differentially expressed in SL-F1 livers. The top Gene Ontology terms were Circadian Rhythm (GO:0007623, FDR=0.11) and Lipid Metabolic Process (GO:0006629, FDR=0.08). qPCR confirmed that Per1 and Per2 (core clock genes) were significantly deregulated in SL-F1 livers (p<0.05), while Cry1, Cry2, and Per3 were not. Among lipid-related genes, rhythmic lipogenic genes (e.g., Mogat1, Dgat2, Acly) were altered, but lipid oxidation genes (Cpt1a, Cpt2) were unchanged. Sperm DNA methylation analysis revealed 763 differentially methylated CpG sites in SL-F0 vs. C-F0 mice, and 1747 differentially methylated CpG sites in SL-F1 vs. C-F1 livers. However, no CpG sites overlapped between sperm (F0) and liver (F1). Only 20 of 394 differentially expressed genes (6%) were potentially linked to DNA methylation changes. Small RNA sequencing showed that piRNAs were the most abundant sncRNA species in testes. Only two miRNAs were differentially expressed: mmu-miR-547 (adj. p=0.01) and mmu-miR-201 (adj. p=0.08). Both miRNAs map to the same chromosome, are expressed in mature spermatozoa but not oocytes, and remain detectable in morula but not blastocyst stages. The hepatic targetome analysis showed miR-547 potentially targets 339 transcripts and miR-201 targets 816 mRNAs in hepatocytes. The most significant ontologies for miR-547 included regulation of lipid biosynthetic process (GO:0051055, GO:0045834, GO:0010888), and for miR-201 included lipid metabolic process (GO:0006629). Notably, Acox1, Apob, and Scd1 (differentially expressed in the transcriptome) are potential targets of both miRNAs. Neither miRNA targeted any clock gene.
**Clinical Implications:** This study provides experimental evidence that paternal early-life overnutrition can program hepatic steatosis in the next generation through non-genetic mechanisms. The findings challenge the role of sperm DNA methylation as a primary carrier of intergenerational epigenetic information in metabolic disease, instead implicating paternally derived microRNAs (miR-201 and miR-547) as potential mediators. If confirmed in humans, these results suggest that interventions targeting paternal health during critical developmental windows (e.g., childhood and adolescence) could reduce the risk of metabolic liver disease in offspring. The study also highlights the need for further research into how paternally inherited miRNAs influence early embryonic development and adult hepatic lipid metabolism.