Seminal plasma cfDNA fragmentomics landscape delineates male infertility subtypes

While cell-free DNA (cfDNA) fragmentomics has transformed liquid biopsy applications in prenatal screening and oncology, its potential in male reproductive health remains unexplored.

Through integrated whole-genome sequencing and jagged end sequencing (Jag-Seq) coupled with non-CG(CH) methylation analysis, we established the first fragmentomic atlas of seminal plasma (SP) cfDNA from 18 healthy donors, with 20 plasma cfDNA samples. This approach was then applied to 33 patients (14 with varicocele [VC] and 19 with nonobstructive azoospermia [NOA]) to characterize disease-specific fragmentomic features. ROC analysis was employed to study the potential diagnostic ability for these two diseases.

Size distribution profiling showed SP cfDNA enrichment in short fragments (< 150 bp) with bimodal distribution (151 bp main peak/110 bp subpeak), contrasting with plasma's sharp 166-bp peak pattern ( $P<0.001$ ). Motif analysis identified SP-specific patterns: increased AAAA-end motif frequency and a strong A-base preference at positions - 2 to - 4. SP showed higher jagged end index based on Jag-Seq ( $P<0.0001$ ). In disease states, VC exhibited 7 altered frequency motifs and longer jagged end length, while NOA demonstrated higher methylation level at CH sites. Integrating these fragmentomic features, ROC analysis achieved 83% accuracy in distinguishing VC and 87% accuracy in distinguishing NOA.

The study indicates that distinct cfDNA profiles are present in certain male infertility conditions. These cfDNA metrics demonstrated disease-specific cfDNA dynamics present innovative opportunities for the development of noninvasive diagnostic tools in the field of reproductive medicine.