Tracing Recombinant Bovine Somatotropin Ab(Use) Through Gene Expression in Blood, Hair Follicles, and Milk Somatic Cells: A Matrix Comparison

molecules Article Tracing Recombinant Bovine Somatotropin Ab(Use) Through Gene Expression in Blood, Hair Follicles, and Milk Somatic Cells: A Matrix Comparison Alexandre Lamas * ID , Patricia Regal * ID , Beatriz Vazquez ID , JoséManuel Miranda ID , Alberto Cepeda and Carlos Manuel Franco ID Laboratorio de Higiene, Inspección y Control de Alimentos, Dpto. de Química Analítica, Nutrición y Bromatología, Universidad de Santiago de Compostela, 27002 Lugo, Spain; [email protected] (B.V.); [email protected] (J.M.M.); [email protected] (A.C.); [email protected] (C.M.F.) * Correspondence: [email protected] (A.L.); patricia.r[email protected] (P.R.); Tel.: +34-982-822-455 (A.L. & P.R.) Received: 27 June 2018; Accepted: 11 July 2018; Published: 13 July 2018   Abstract: The use of recombinant bovine somatotropin (rbST) in dairy cattle is forbidden in the European Union. Due to the very low circulating concentration of rbST in treated animals, its direct detection is still a challenge. Therefore, the use of indirect methods to detect the ab(use) of rbST in dairy cattle appears as a good alternative. In the past few years, gene expression demonstrated its utility in screening the use of illicit substances in both humans and animals. In this study, a comparison of three types of matrices (milk somatic cells, blood, and hair follicles) was carried out to evaluate their potential use for routine control of rbST using 15 gene-expression profiles. A total of six rbST-treated cows and three control cows were included in the study. A subcutaneous injection containing 500 mg of rbST was administered to the treated group. Samples of the three matrices were collected before rbST administration, and at three and nine days after treatment. The quality of RNA extracted was higher in the blood and hair-follicle samples than in the milk somatic cells. In the three matrices, there were significant differences in the expression of some genes, with milk somatic cells and blood presenting the the best matrices. On this note, the cyclin D1 (CCND1), interleukin 1 beta (IL-1 β ), tumor necrosis factor (TNF), and insulin-like growth factor 1 receptor (IGF-1R) genes showed potential as biomarkers of rbST treatment. Therefore, blood, somatic cells, and follicle hair should be considered as promising sources of RNA, and can be used in gene-expression assays to routinely control the illicit use of rbST. Keywords: rbST; nanoliter PCR; gene expression; screening; blood; hair; somatic cells 1. Introduction The growth hormone (GH) is a single-chain polypeptide secreted by the anterior pituitary gland of all vertebrates. This hormone is involved in a wide range of biological activities, such as growth, energy metabolism, sexual maturation, and immunity. These actions are initiated upon the binding of GH to membrane-bound receptors located in various tissues, including liver, muscle, bone, or mammary tissue [ 1 ]. The GH-mediated actions in mammary tissue were studied in dairy cattle for many years. Furthermore, studies carried out in Russia and England at the beginning of the 20th century found that pituitary extracts from cows increased milk yield in dairy cattle [ 2 , 3 ]. This effect was due to the presence of growth hormone in the extract. In the 1980s, recombinant DNA technology allowed the production of large quantities of recombinant bovine somatotropin (rbST) in a cost-effective and efficient manner. The galactopoietic effect of the recombinant variants was demonstrated, and the routine use of rbST in dairy farms to increase the global milk yield started being evaluated [ 4 ]. Thus, in the year 1993, the Food and Drug Administration (FDA) approved the use of rbST in the United States. Molecules 2018,23, 1708; doi:10.3390/molecules23071708 www.mdpi.com/journal/molecules Molecules 2018,23, 1708 2 of 15 Despite its capacity to enhance milk production in dairy farms, the use of rbST was controversial since the beginning. Some studies observed that the use of rbST could increase the incidence of mastitis and lameness in dairy farms, with a consequent deterioration of animal welfare [ 5 ]. Moreover, it was suggested that the increased presence of insulin-like growth factor 1 (IGF-1) in milk as a consequence of rbST use could have a detrimental effect in human health [ 6 ]. Therefore, while the use of rbST is allowed in various countries such as United States, Brazil, and Mexico, its use is banned in others such as Canada, Australia, and New Zealand. In 1999, the European Union (EU) definitively banned the use of rbST in Europe, invoking animal welfare reasons and its impact on European milk policy and consumer fears [7]. The prohibition of rbST use in the EU cannot prevent the illicit use of this substance in dairy farms. Consequently, the illegal use of rbST in dairy farms was detected in 2013 in Spain. The persons involved in this scandal introduced commercial injections from Mexico, where the use of this substance is allowed. Therefore, it is clear that rbST should be included in the European residue-control plans [ 8 ]. Analytical chemistry techniques combining chromatography and mass spectrometry are commonly used to detect the illicit use of substances in cattle (e.g., growth promoters). Furthermore, some methods were developed to detect rbST in bovine and buffalo serum, using liquid chromatography/tandem mass spectrometry [ 9 , 10 ]. However, due to different rbST variants available with varying terminal amino-acid compositions, and the low circulation levels of rbST in blood, it is difficult to detect rbST using these methods [ 11 ]. Therefore, the use of indirect methods could be an alternative for screening the use of rbST in cattle. In the past few years, transcriptomics emerged as a promising tool for evaluating the use of illicit substances in cattle [ 12 ]. Moreover, some studies used qPCR to develop a screening panel that could elucidate the use of growth promoters in cattle [ 13 , 14 ]. Some studies evaluated the gene-expression modifications caused by rbST in various tissues of dairy cattle [ 15 – 17 ]. However, none of these studies focused on the development of a screening panel to be used routinely in monitoring the illicit use of rbST in dairy farms. Routine controls of rbST ab(use) should be carried out at different time points in dairy farms to avoid the market release of milk produced using rbST. Therefore, post-mortem samples from the liver or muscle are totally discarded in this type of control. From a practical point of view, only matrices that can be collected in vivo are really interesting for routine controls. The matrix which is easiest to collect in a dairy farm is that of milk, and the somatic cells present in milk are a good alternative for transcriptomic assays [ 18 ]. The collection of other matrices requires direct contact with the animal. A good example of this is blood, which is commonly used for in vivo transcriptomics assays [ 19 ]. Another option to carry out in vivo transcriptomics assays is the hair follicle, as hair is a common matrix to detect the use of illicit substances in cattle [ 20 ]. However, until now, there are no studies focusing on the use of hair follicles for screening transcriptomics assays to control the use of banned substances in cattle. Therefore, the aim of this study was to compare the potential of three matrices (milk somatic cells, blood, and hair follicles) to detect the ab(use) of rbST in dairy cattle, using gene expression. A total of nine dairy cows were involved in the study (six rbST-treated cows and three control cows). A total of 15 target and three reference genes were selected, and their expressions in three different matrices were determined using a high-throughput real-time platform, allowing the simultaneous analysis of various genes in a wide range of samples. 2. Results 2.1. RNA Isolation and Quality The amount of RNA isolated from the samples varied across matrices (Table 1). The amount of RNA isolated from somatic cells was significantly more abundant (p< 0.001) than that in blood and hair follicles. Furthermore, the ratio of absorbances at 260 and 280 nm (A260/A280) was significantly different (p< 0.001) across matrices. The RNA extracted from hair follicles presented the highest Molecules 2018,23, 1708 3 of 15 A260/A280 ratio, with a mean value of 1.892, while the RNA from blood presented a mean A260/A280 ratio of 1.809, and the RNA from milk somatic cells presented the lowest A260/A280 ratio, with a mean value of 1.710. The RNA integrity number (RIN) values showed a similar tendency. The RIN values of RNA extracted from hair follicles and blood were significantly higher than that of RNA obtained from milk somatic cells. Table 1. The concentration of RNA, the ratio of absorbances at 260 and 280 nm (A260/A280), and the RNA integrity number (RIN) values of RNA isolated from the three matrices. Matrix RNA Concentration (ng/µL) A260/A280 Ratio RIN Value Hair follicles 53.30 ±11.97 b1.89 ±0.06 a7.76 ±0.41 a Blood 59.00 ±28.02 b1.81 ±0.04 b7.79 ±0.40 a Somatic cells 165.42 ±71.23 a1.71 ±0.08 c6.10 ±0.57 b Letters (a–c) in each column reflect significant differences. 2.2. Reference Genes Three different genes, namely ubiquitously expressed prefoldin-like chaperone (UXT), ribosomal protein S9 (RPS9), and mitochondrial glycerol 3-phosphate acyltransferase (GPAM), were evaluated as internal controls for this gene-expression study. Their stability was evaluated using the Bestkeeper ® software. With this software, the standard deviation (SD) of crossing-point values (CPs) of candidate reference genes was calculated, as well as Bestkeeper indexes using the SD values of candidate genes. The most stable genes exhibit the lowest variation, and any studied gene with an SD higher than 1 can be considered inconsistent. There were differences in the expression of reference genes. The expression of all three reference genes included in this study was detected in somatic-cell samples. However, in blood and hair-follicle samples, only the expressions of UXT and RPS9 were detected. Therefore, the stability of the three genes was calculated in somatic cells, while only the stability of UXT and RPS9 was calculated in blood and follicle samples. Table 2shows the SD ( ± CP) of the selected genes from each type of matrix. Table 2. Standard deviation of crossing-point (CP) values obtained for the three reference candidates—ubiquitously expressed prefoldin-like chaperone (UXT), ribosomal protein S9 (RPS9), and mitochondrial glycerol 3-phosphate acyltransferase (GPAM). For hair follicles and blood, the amplification of GPAM was not detected. Bestkeeper indexes were calculated based on the CP values of reference genes. For somatic cells, the Bestkeeper index was calculated based on UXT-RPS9. Matrix UXT RPS9 GPAM Bestkeeper Index Bestkeeper Index (UXT-RPS9) Hair follicles 0.90 0.91 - 0.96 0.96 Blood 0.89 0.75 - 0.71 0.71 Somatic cells 0.85 0.79 1.19 0.94 0.89 The GPAM gene was only expressed in somatic cells; in the samples analyzed for this study, the SD ( ± CP) was higher than 1, and could be considered as inconsistent. The SD ( ± CP) of the other two genes was lower than 1, and therefore, could be considered as consistent. Furthermore, in hair follicles, the SD ( ± CP) was similar for both genes (UXT and RPS9), while in the blood and in somatic cells, RPS9 showed lower SD ( ± CP) values. These results indicate a better stability of this gene in comparison with UXT. The blood samples presented the best Bestkeeper index. In the case of somatic cells, the use of only UXT and RPS9 allowed us to obtain a better Bestkeeper index. Hair follicles presented the worst Bestkeeper index. In this study, only the UXT and RPS9 genes were used to normalize the expression of target genes. Molecules 2018,23, 1708 4 of 15 2.3. Expression Pattern of Target Genes Analyzed The expressions of three genes, namely insulin-like growth factor binding protein 5 (IGFBP5), collagen type III alpha 1 chain (COL3A1), and estrogen receptor 2 (ESR2), were not detected in the matrices tested (Figure 1). In addition, the expression of IGF-1 was only detected in one cow in the milk somatic cells. The number of target genes in which expression was detected in this study varied in each matrix. In milk somatic-cell samples, the expressions of nine genes were detected, and the transcription of lactotransferrin (LTF) was only detected in this matrix. In blood samples, the expressions of eight genes were observed, and the transcription of catenin alpha-like 1 (CTNNAL1) was only detected in this matrix. Finally, in hair-follicle samples, the transcriptions of seven genes were observed, and the expression of insulin-like growth factor binding protein 3 (IGFBP3) was detected in this matrix. Molecules 2018, 23, x FOR PEER REVIEW 4 of 15 2.3. Expression Pattern of Target Genes Analyzed The expressions of three genes, namely insulin-like growth factor binding protein 5 (IGFBP5), collagen type III alpha 1 chain (COL3A1), and estrogen receptor 2 (ESR2), were not detected in the matrices tested (Figure 1). In addition, the expression of IGF-1 was only detected in one cow in the milk somatic cells. The number of target genes in which expression was detected in this study varied in each matrix. In milk somatic-cell samples, the expressions of nine genes were detected, and the transcription of lactotransferrin (LTF) was only detected in this matrix. In blood samples, the expressions of eight genes were observed, and the transcription of catenin alpha-like 1 (CTNNAL1) was only detected in this matrix. Finally, in hair-follicle samples, the transcriptions of seven genes were observed, and the expression of insulin-like growth factor binding protein 3 (IGFBP3) was detected in this matrix. Matrices Gene Blood Milk Somatic Cells Hair Follicles IGF1 IGF1-R IGFBP3 IGFBP5 IL-1β TNF LTF COL3A1 TPD52L2 ESR2 CTNNAL1 SIRT2 CCND1 MFGE8 EEF1G Figure 1. Gene-expression patterns in blood, milk somatic cells, and hair-follicle samples. Green cells represent expression of the target gene, and gray cells represent no expression of the target gene, in the samples of each matrix. For a list of the gene names evaluated, see Table 3. 2.3.1. Expression of Target Genes in Blood Samples Gene-expression assays were carried out the week before the first rbST administration, and on the third and ninth days after rbST administration. Genes differently expressed between rbST and control samples were only detected on the third day after rbST administration (Figure 2). Specifically, eukaryotic translation elongation factor 1 gamma (EEF1G), tumor necrosis factor (TNF), and interleukin 1 beta (IL-1β) were significantly upregulated in rbST samples. However, milk fat globule epidermal growth factor (EGF) 8 (MFGE8) was significantly downregulated in the rbST group on the third day. Figure 1. Gene-expression patterns in blood, milk somatic cells, and hair-follicle samples. Green cells represent expression of the target gene, and gray cells represent no expression of the target gene, in the samples of each matrix. For a list of the gene names evaluated, see Table 3. Table 3. Genes included in this study for gene-expression assays. Gene Symbol Gene Name NCBI Accession Number Assay ID IGF1 Insulin-like growth factor 1 NM_001077828.1 bt03252281_m1 IGF1-R Insulin-like growth factor 1 receptor NM_001244612.1 bt03649217_m1 IGFBP3 Insulin-like growth factor binding protein 3 NM_174556.1 bt03223809_m1 IGFBP5 Insulin-like growth factor binding protein 5 NM_001105327.2 bt03258785-g1 IL-1βInterleukin 1 beta NM_174093.1 bt03212745_m1 TNF Tumor necrosis factor NM_173966.3 bt03259156_m1 LTF Lactotransferrin NM_180998.2 bt03217382_m1 COL3A1 Collagen type III alpha 1 chain NM_001076831.1 bt03249914_m1 TPD52L2 Tumor protein D52-like 2 NM_001034615.2 bt03227133_m1 ESR2 Estrogen receptor 2 NM_174051.3 bt03259198_m1 CTNNAL1 Catenin alpha-like 1 NM_001191534.1 bt04308229_m1 SIRT2 Sirtuin 2 NM_001113531.1 bt03258971_m1 CCND1 Cyclin D1 NM_001046273.2 bt03235030_m1 MFGE8 Milk fat globule epidermal growth factor (EGF) 8 protein NM_176610.1 bt03216856_m1 EEF1G Eukaryotic translation elongation factor 1 gamma NM_001040487.2 bt03229629_g1 Reference genes UXT Ubiquitously expressed prefoldin-like chaperone NM_001037471.2 bt03229278_m1 RPS9 Ribosomal protein S9 NM_001101152.2 bt03272016_m1 GPAM Glycerol 3-phosphate acyltransferase (mitochondrial) NM_001012282.1 bt03210379_m1 Molecules 2018,23, 1708 5 of 15 2.3.1. Expression of Target Genes in Blood Samples Gene-expression assays were carried out the week before the first rbST administration, and on the third and ninth days after rbST administration. Genes differently expressed between rbST and control samples were only detected on the third day after rbST administration (Figure 2). Specifically, eukaryotic translation elongation factor 1 gamma (EEF1G), tumor necrosis factor (TNF), and interleukin 1 beta (IL-1 β ) were significantly upregulated in rbST samples. However, milk fat globule epidermal growth factor (EGF) 8 (MFGE8) was significantly downregulated in the rbST group on the third day. Molecules 2018, 23, x FOR PEER REVIEW 5 of 15 Figure 2. Relative abundance of target genes expressed in blood samples before the administration of recombinant bovine somatotropin (rbST), and on the third and ninth days after rbST administration. The bars represent the mean value of each group (rbST and control). The value used for each animal is the mean of three replicates; * p < 0.05, *** p < 0.001. 2.3.2. Expression of Target Genes in Milk Somatic-Cell Samples The IL-1β, insulin-like growth factor 1 receptor (IGF-1R), and TNF genes were significantly upregulated (Figure 3) in the rbST group on the third and ninth days after recombinant-hormone administration. The cyclin D1 (CCND1) gene was significantly upregulated in the rbST group on the ninth day. On the other hand, LTF and tumor protein D52-like 2 (TDP52L2) were significantly upregulated on the ninth day in the control group. The transcription of EEF1G was significantly Figure 2. Relative abundance of target genes expressed in blood samples before the administration of recombinant bovine somatotropin (rbST), and on the third and ninth days after rbST administration. The bars represent the mean value of each group (rbST and control). The value used for each animal is the mean of three replicates; * p < 0.05, *** p < 0.001. 2.3.2. Expression of Target Genes in Milk Somatic-Cell Samples The IL-1 β , insulin-like growth factor 1 receptor (IGF-1R), and TNF genes were significantly upregulated (Figure 3) in the rbST group on the third and ninth days after recombinant-hormone administration. The cyclin D1 (CCND1) gene was significantly upregulated in the rbST group on Molecules 2018,23, 1708 6 of 15 the ninth day. On the other hand, LTF and tumor protein D52-like 2 (TDP52L2) were significantly upregulated on the ninth day in the control group. The transcription of EEF1G was significantly different between both groups for all sample points, including those before rbST administration, and MFEG8 expression was higher in the rbST group on the sixth day prior to administration, but not after rbST administration. Molecules 2018, 23, x FOR PEER REVIEW 6 of 15 different between both groups for all sample points, including those before rbST administration, and MFEG8 expression was higher in the rbST group on the sixth day prior to administration, but not after rbST administration. Figure 3. Relative abundance of target genes expressed in milk somatic-cell samples before rbST administration, and on the third and ninth days after rbST administration. The bars represent the mean value of each group (rbST and control). The value used for each animal is the mean of three replicates. * p < 0.05, ** p < 0.01, *** p < 0.001. Molecules 2018,23, 1708 7 of 15 2.3.3. Expression of Target Genes in Hair-Follicle Samples In this matrix, significant differences between the rbST and control groups were only detected in the expression of IGF-1R on the third and ninth days, and also in CCND1 on the ninth day (Figure 4). The other genes showed no significant differences for all time points analyzed. Molecules 2018, 23, x FOR PEER REVIEW 7 of 15 Figure 3. Relative abundance of target genes expressed in milk somatic-cell samples before rbST administration, and on the third and ninth days after rbST administration. The bars represent the mean value of each group (rbST and control). The value used for each animal is the mean of three replicates. * p < 0.05, ** p < 0.01, *** p < 0.001. 2.3.3. Expression of Target Genes in Hair-Follicle Samples In this matrix, significant differences between the rbST and control groups were only detected in the expression of IGF-1R on the third and ninth days, and also in CCND1 on the ninth day (Figure 4). The other genes showed no significant differences for all time points analyzed. Molecules 2018, 23, x FOR PEER REVIEW 8 of 15 Figure 4. Relative abundance of target genes expressed in hair-follicle samples before rbST administration, and on the third and ninth days after rbST administration. The bars represent the mean value of each group (rbST and control). The value used for each animal is the mean of three replicates. * p < 0.05, *** p < 0.001. 3. Discussion On the basis of the results observed in this study, milk somatic cells may be highlighted as the best candidate for a target matrix in rbST gene-expression analysis, as it showed the highest differences in the selected genes. Transcription patterns are expected to be tissue-specific, and, on this basis, a target matrix should be selected following action-based criteria. The mammary gland is considered one of the principal target tissues of rbST, and as such, it provided the most interesting transcription pattern. Additionally, it is worth mentioning that the stability of the selected reference genes varied according to the type of matrix. Hence, it is relevant to explore new reference candidates for hair follicles, as this matrix showed the poorest stability for the housekeepers selected in this study. These results highlight the importance of evaluating various candidates to find those with better stability, as reference genes are one of the most important aspects of RT-qPCR assays [21]. In this study, various target genes were included with the aim of finding a biomarker signature of rbST administration. Samples of the three matrices evaluated were collected at three different time points (before rbST administration, and on the third and ninth days after rbST administration). The expression of insulin-like growth factor 1 (IGF1) was evaluated in this study, and its expression was only detected in one cow in the somatic cells. For the remainder of the animals, its expression was not detected in any of the evaluated matrices. It is possible that the transcription of IGF1 associated with rbST administration is focused in the liver, and IGF1 synthetized in this organ circulates to the other tissues, such as the mammary gland, where rbST exerts an indirect function through an IGF1 molecule [22]. Furthermore, the expression of IGF-1R (the receptor of IGF1) was detected in all three matrices. The expression of IGF-1R was significantly upregulated, both in somatic cells and hair follicles, in the rbST group in comparison with the control group on the third and ninth days. However, in blood, there were no differences across groups. Accordingly, Castigliego et al. [15] observed no influence of rbST on the expression of IGF-1R in the muscle of treated cows. These data highlight the tissue-specific effect of rbST, and the need for performing the selection of candidate target genes according to the matrix that is going to be analyzed. Moreover, the expression of genes coding for IGF-1-binding proteins was analyzed, and only the expression of IGFBP3 was detected in hair follicles; however, no significant differences were observed across time points. These results discredit the use of the expression of genes coding for IGF-1-binding proteins as biomarkers of rbST treatment in the matrices studied. Two genes related to the immune system (IL-1β and TNF) were included in this study. The expression of these genes was detected in milk somatic cells and in blood, but not in hair follicles. The expressions of IL-1β and TNF were clearly upregulated in milk somatic cells on the third and ninth days after rbST administration. In blood, the differences across groups for these genes were only significant on the third day, and these differences were less significant than in milk somatic cells. A meta-analysis study showed that rbST administration increases health problems in dairy cows [5]. This exogenous substance can alter the immune system of cows, and this could explain the Figure 4. Relative abundance of target genes expressed in hair-follicle samples before rbST administration, and on the third and ninth days after rbST administration. The bars represent the mean value of each group (rbST and control). The value used for each animal is the mean of three replicates. * p < 0.05, *** p < 0.001. Molecules 2018,23, 1708 8 of 15 3. Discussion On the basis of the results observed in this study, milk somatic cells may be highlighted as the best candidate for a target matrix in rbST gene-expression analysis, as it showed the highest differences in the selected genes. Transcription patterns are expected to be tissue-specific, and, on this basis, a target matrix should be selected following action-based criteria. The mammary gland is considered one of the principal target tissues of rbST, and as such, it provided the most interesting transcription pattern. Additionally, it is worth mentioning that the stability of the selected reference genes varied according to the type of matrix. Hence, it is relevant to explore new reference candidates for hair follicles, as this matrix showed the poorest stability for the housekeepers selected in this study. These results highlight the importance of evaluating various candidates to find those with better stability, as reference genes are one of the most important aspects of RT-qPCR assays [21]. In this study, various target genes were included with the aim of finding a biomarker signature of rbST administration. Samples of the three matrices evaluated were collected at three different time points (before rbST administration, and on the third and ninth days after rbST administration). The expression of insulin-like growth factor 1 (IGF1) was evaluated in this study, and its expression was only detected in one cow in the somatic cells. For the remainder of the animals, its expression was not detected in any of the evaluated matrices. It is possible that the transcription of IGF1 associated with rbST administration is focused in the liver, and IGF1 synthetized in this organ circulates to the other tissues, such as the mammary gland, where rbST exerts an indirect function through an IGF1 molecule [ 22 ]. Furthermore, the expression of IGF-1R (the receptor of IGF1) was detected in all three matrices. The expression of IGF-1R was significantly upregulated, both in somatic cells and hair follicles, in the rbST group in comparison with the control group on the third and ninth days. However, in blood, there were no differences across groups. Accordingly, Castigliego et al. [ 15 ] observed no influence of rbST on the expression of IGF-1R in the muscle of treated cows. These data highlight the tissue-specific effect of rbST, and the need for performing the selection of candidate target genes according to the matrix that is going to be analyzed. Moreover, the expression of genes coding for IGF-1-binding proteins was analyzed, and only the expression of IGFBP3 was detected in hair follicles; however, no significant differences were observed across time points. These results discredit the use of the expression of genes coding for IGF-1-binding proteins as biomarkers of rbST treatment in the matrices studied. Two genes related to the immune system (IL-1 β and TNF) were included in this study. The expression of these genes was detected in milk somatic cells and in blood, but not in hair follicles. The expressions of IL-1 β and TNF were clearly upregulated in milk somatic cells on the third and ninth days after rbST administration. In blood, the differences across groups for these genes were only significant on the third day, and these differences were less significant than in milk somatic cells. A meta-analysis study showed that rbST administration increases health problems in dairy cows [ 5 ]. This exogenous substance can alter the immune system of cows, and this could explain the upregulation of the immune-system-related genes, TNF and IL-1 β , in the treated group. Also, in previous screening transcriptomics studies, it was observed that anabolic treatments caused an upregulation of IL-1 β in blood and vaginal smear cells [ 13 , 23 ]. In addition, other processes involving infections such as subclinical mastitis can increase the expressions of IL-1 β and TNF in mammary glands [ 24 , 25 ]. Therefore, these genes cannot individually be considered as specific markers of rbST administration. Therefore, in screening transcriptomic studies, it is necessary to include a range of genes, as one gene can be upor downregulated on the basis of several factors. As the number of genes included in the panel increases, so does the discrimination power of the designed panel. Other genes with functions involving cell cycle, proliferation, differentiation, and adhesion were included in this study. It is known that rbST increases milk synthesis by increasing the turnover (proliferation/apoptosis) and activity of mammary epithelial cells, indicating that rbST strongly influences metabolic pathways that regulate cell turnover/cycle and metabolism [ 25 ]. Accordingly, the CCND1 gene plays an important role in cell physiopathology because its dysregulation is strongly Molecules 2018,23, 1708 9 of 15 related to a drive in inappropriate cell division, and the generation of genome instability, generating neoplastic growth [ 26 ]. This gene was significantly upregulated on the ninth day in the rbST group for milk somatic cells and hair follicles. These results highlight the close relationship between rbST and cell-cycle regulation. The administration of this exogenous substance causes an overexpression of CCND1 that can result in the activation of cells in the G 0 phase. Therefore, the CCND1 gene is a promising marker for detecting the use of rbST in dairy cattle, using milk somatic cells and hair follicles. However, in blood, the expression of CCND1 was not detected. On the other hand, the EEF1G and MFGE8 genes were significantly upregulated and downregulated, respectively, in the rbST group on the third day for blood samples. A previous study in the mammary tissue of rbST-treated cows also observed an upregulation of EEF1G six days after hormone administration [ 16 ]. EEF1G is involved in translation elongation via the transport of aminoacyl transfer RNAs (tRNAs) to the ribosome for protein synthesis [ 27 ]. Therefore, the upregulation of this gene is associated with a higher level of protein synthesis, potentially activated by the exogenous administration of rbST. However, in milk somatic cells and hair follicles, it was not possible to establish a relationship between the expression of this gene and rbST treatment. The MFGE8 gene is an essential factor for attenuating inflammation and inhibiting inflammasome-induced IL-1 β production [ 28 ]. As such, it is remarkable that, while TNF and IL-1 β expressions were upregulated in blood samples, MFGE8 was downregulated in treated cows on the third day. These results highlight the inverse relationship between the expressions of these genes. Due to its role in attenuating inflammation, MFGE8 could be downregulated as an inflammatory response to the external administration of rbST. In the case of sirtuin 2 (SIRT2), this gene was upregulated on the third day in the rbST group for milk somatic cells, and no differences were observed in the other two matrices. SIRT2 is closely related to cell activity, and it is involved in the cell cycle. SIRT2 is required for normal mitotic progression and the prevention of chromosomal instability. SIRT2 levels are greatly increased during mitosis, and its inhibition interferes with cell-cycle progression [ 29 ]. This result again highlights the influence of exogenous rbST on cell-cycle regulation. Finally, TDP52L2 and LTF were downregulated in the rbST group on the ninth day, and no rbST influence or expression was detected in the other two matrices tested. These data highlight the influence of matrices in the differences observed between two groups using gene expression, and the importance of selecting genes for the screening panel according the matrix that is analyzed. The transcriptional modifications after rbST administration were influenced by both time and the matrix used. Blood and milk somatic cells are good candidates to be used as routine matrices for screening purposes. In the case of hair follicles, IGF-1R and CCND1 were upregulated after rbST administration. However, two genes alone are seemingly not enough to detect rbST administration. However, the simultaneous analysis of multiple matrices can be combined. Based on the results observed across the different matrices, IL-1 β ,TNF,IGF-1R, and CCND1 are good candidates to be included in a screening panel. Although the transcription of EEF1G and MFGE8 was influenced by rbST in blood samples, they were upregulated in milk somatic cells before hormone administration, and their use in a panel should only be considered in blood matrices. It is important to note that, in the presented study, control animals were not injected with excipients of Lactotropina ® , as its exact composition is not declared by the manufacturer. On the other hand, in real-farm conditions, control animals are not treated with excipients. Nonetheless, it could be interesting to prepare excipient injections of a known composition for control animals in any future research focused on rbST analysis. OpenArray ® technology allows the simultaneous analysis of various genes in a wide number of samples [ 30 ]. Therefore, in routine controls, three matrices can be collected from the same farm and can be analyzed simultaneously. The overexpression of various selected genes in different matrices can be a good indication of rbST administration. Realistically, it is not possible to know when rbST was administered in a farm. Ideally, every day during lactation should be controlled. However, from a practical point of view, this suggestion is unrealistic. Therefore, a random collection of samples (for instance, once per week and on alternating week days) is proposed. As rbST is injected on a regular