103 CIRCULATING microRNAs FOR EARLY DIAGNOSIS OF BOVINE PREGNANCY

Abstract

Early diagnosis of pregnancy can shorten calving intervals, improve annual milk production and increase overall profits from modern dairy herds. At present, accurate diagnosis is only possible after the third week of pregnancy. Circulating microRNAs (miRNAs) have been proposed as diagnostic biomarkers for numerous human conditions such as cancer and diabetes. Moreover, distinct circulating miRNA profiles have been associated with different stages of human pregnancy. The objective of this study was to determine whether differential miRNA profiles occur in circulation during early pregnancy (Day 24 or earlier) in cattle that could be used for diagnostic purposes. Holstein-cross heifers were oestrous-synchronised and artificially inseminated (AI, n = 11) or sham-inseminated (control, n = 8) at first detected oestrus. Plasma samples were collected on Days 0, 8, 16 and 24 after insemination. Circulating miRNA levels were independently determined in pooled plasma samples (n = 3 pools for each of pregnant Day 24 and nonpregnant Days 0, 8, and 16) using Qiagen qPCR arrays (Qiagen, Valencia, CA, USA) and in individual samples (n = 11 samples for each pregnant Days 16 and 24, and 8 samples for each of nonpregnant Days 0, 8, and 16, respectively) using Illumina miRNA sequencing. The qPCR array data were analysed using the ΔΔCq method. The miRNA sequencing data were normalised using EdgeR. Differential expression between pregnant and nonpregnant groups was determined using 2-sample t-tests with false discovery rate (FDR) adjustment. Differences in miRNA expression were validated by RT-qPCR. Out of a total of 191 miRNAs analysed in pooled samples using qPCR arrays, 8 were differentially expressed (<3-fold, FDR <0.1) in Day 24 pregnant heifers relative to nonpregnant heifers (Days 0, 8, and 16 combined). No miRNAs were differentially expressed (FDR >0.1) between nonpregnant time-points. Changes in levels of 11 miRNAs were validated by RT-qPCR in individual plasma samples; although expression trends for these miRNAs were the same as in pooled samples, none of the changes in individual samples were significant after FDR adjustment (P > 0.1). Deep sequencing (96 million miRNA reads) identified 231 miRNAs in bovine plasma. There were no significant differences (FDR >0.1) in the expression of any miRNAs between pregnant heifers (Days 16 or 24) and nonpregnant (Days 0, 8, and 16 individually or combined). In addition, no significant differences were identified among nonpregnant time-points. In summary, we successfully performed miRNA profiling of bovine plasma using both deep sequencing and qPCR; however, we did not detect differences in miRNA expression between early pregnant (Day 16 or 24) and nonpregnant heifers. Changes in circulating miRNA levels may involve low abundance miRNAs that cannot be accurately quantified using current technology. Alternatively, changes in circulating miRNA levels may only occur later during pregnancy in cattle.