Burial mounds represent a challenge for microbiologists. Could ancient buried soils preserve microbiomes as they do archaeological artifacts? To investigate this question, we studied the soil microbiome under a burial mound dating from 2500 years ago in Western Kazakhstan. Two soil profile cuts were established: one under the burial mound and another adjacent to the mound surface steppe soil. Both soils represented the same dark chestnut soil type and had the same horizontal stratification (A, B, C horizons) with slight alterations. DNA samples isolated from all horizons were studied with molecular techniques including qPCR and high throughput sequencing of amplicon libraries of the 16S rRNA gene fragment. The taxonomic structure of the microbiome of the buried horizons demonstrated a deep divergence from ones of the surface, comparable to the variation between different soil types (representatives of the soil types were included in the survey). The cause of this divergence could be attributed to diagenetic processes characterized by the reduction of organic matter content and changes in its structure. Corresponding trends in the microbiome structure are obvious from the beta-diversity pattern: the A and B horizons of the buried soils form one cluster with the C horizons of both buried and surface soil. This trend could generally be designated as 'mineralization'. Statistically significant changes between the buried and surface soils microbiomes were detected in the number of phylogenetic clusters, the biology of which is in the line of diagenesis. The trend of 'mineralization' was also supported by PICRUSt2 functional prediction, demonstrating a higher occurrence of the processes of degradation in the buried microbiome. Our results show a profound shift in the buried microbiome relatively the "surface" microbiome, indicating the deep difference between the original and buried microbiomes.