INTRODUCTION
Mesenchymal stem cells (MSCs) have received attention in the fields of cell-based regenerative medicine and biotechnology, which is due to their advantages such as the need for ethical issues, easy accessibility, self-renewal property, multi-differentiation potentials, and immunomodulatory capacity [
1,
2]. Like numerous studies on MSCs in various species, bovine MSCs have also been widely investigated for the past few decades to elucidate their characteristics and verify their clinical potential [
3–
5]. In particular, as large animal models, including cattle, possess a greater similarity to humans than small animals such as rodents, there has been an increase in the number of studies on large animals to more reliably understand the potential of the clinical application of MSCs [
3].
Gene expression studies are indispensable in the field of cellular biology research as they enable researchers to identify the gene regulatory network in cells [
6]. In this context, quantitative real-time polymerase chain reaction (qPCR), which has the advantages of convenience, sensitivity, reproducibility, and reliability, has been most commonly used to verify the potential of MSCs and determine the change in the mRNA expression of genes of interest (GOIs) [
1,
2]. During qPCR, the GOI is normalized against a reference gene (RG), also known as a housekeeping gene, as an internal control for its relative quantification; this step corrects sample-to-sample variations in the context of different experimental conditions, sample quality, operators, and laboratories [
1,
2,
7,
8]. Therefore, RGs should be stably expressed in various samples and not be affected by various experimental conditions; in principle, RGs play a pivotal role in the vital functions of cell survival and maintenance [
7,
9].
However, till date, no single RG has been addressed to be universal and perfectly constant. It is known that the expression of RGs is also variable depending on the experimental conditions and cell types [
10]. In particular, the normalization of GOIs against inadequate or unstable RGs may result in false or contradictory conclusions [
1,
7,
11]. Therefore, validation of RGs for their stability under each experimental condition is an extremely important and prerequisite step for obtaining reliable results during qPCR assay [
2,
7,
8].
Unfortunately, information for a clear list of stable RGs in bovine MSCs is currently lacking, despite the fact that studies on cattle are being widely conducted. Therefore, the primary objective of this study was to identify the most suitable RGs in bovine MSCs, before conducting further gene expression study by qPCR. After establishing bovine bone marrow-derived MSC lines, we evaluated the stability of a set of ten traditionally used RGs (18S ribosomal RNA [18S], beta-2-microglobulin [B2M], H2A histone family, member Z [H2A], peptidylprolyl isomerase A [PPIA], ribosomal protein 4 [RPL4], succinate dehydrogenase complex, subunit A [SDHA], beta actin [ACTB], glyceraldehyde-3-phosphate dehydrogenase [GAPDH], TATA box binding protein [TBP], and hypoxanthine phosphoribosyltrasnfrase1 [HPRT1]) using the three most well-known algorithms (geNorm, Normfinder, and Bestkeeper). Thereafter, we applied the most and least stable RGs by normalization of GOIs (fatty acid binding protein 4 [FABP4], osteonectin [ON], and POU class 5 homeobox 1 (OCT4) to verify the importance of selecting suitable RGs in each study.
DISCUSSION
Although qPCR has been widely used to elucidate the characteristics of bovine MSCs and confirm their clinical potential, information for a clear list of stable RGs in bovine MSCs is currently not available. In the present study, we established bBMMSCs and investigated their Ct values using ten commonly used RGs. The Ct values were then assessed for stability using the three most well-known algorithms (geNorm, Normfinder, and Bestkeeper). Consequently,
TBP and
RPL4 were found to be the two most stable RGs in bBMMSCs, but traditional RGs such as
18S,
GAPDH, and
ACTB were determined to be less stable. These attempts to validate the suitable RGs in each experimental condition represent a prerequisite for the reliable assessment of gene expression by qPCR, as there is no equally and constantly expressed RG regardless of various experimental conditions and the usage of an inappropriate RG may lead to false or contradictory results [
7,
8,
11]. In this respect, to the best of our knowledge, the present study is the first to validate stable RGs in bovine MSCs.
Using other bovine specimens, several studies have been conducted to validate the stability of RGs in each experimental condition. Consistent with the present study results depicted in
Figure 3 to
5,
TBP was stably expressed in several bovine tissues, including the cumulus cell [
16], corpus luteum obtained from cyclic or pregnant cows [
17], and liver and thyroid [
18].
RPL15, a ribosomal protein family along with
RPL4, was also found to show stable expression in oocytes collected from cattle during winter and summer [
19]. In agreement with the results of the present study, the usage of
18S for normalization was not recommended in the bovine muscular tissue [
20] and corpus luteum [
17]. In addition, several experimental conditions with cattle specimens in terms of the mammary gland under different lactation periods [
21], polymorphonuclear leukocytes [
22], cumulus cell [
16], muscular tissue [
20], and peripheral lymphocytes [
23] were found to be unsuitable to use
ACTB and
GAPDH for the normalization step. On the other hand, using some conditions such as polymorphonuclear leukocytes [
22] and embryos produced
in vitro [
24,
25] in cattle,
18S and
GAPDH were validated as stable RGs, respectively. Altogether, the discrepancies in different stabilities of RGs in each report using cattle are believed to be caused due to the differences in experimental conditions. Therefore, validation of the stability of RGs under each experimental condition is considered as an essential step before analyzing bovine gene expression by qPCR [
2,
7,
8].
Similar studies have also been conducted in MSCs from other species. Comprehensively,
TBP was one among the three most stable RG in human [
2] and porcine [
1] MSCs regardless of cell source and differentiation induction, but
18S was the least stable RG. While
RPL13A, a ribosomal protein family along with
RPL4, was found to be stable in human MSCs derived from adipose tissue [
8], bone marrow, and fetal tissue [
9], normalization with
ACTB was not recommended due to instability.
A survey based on NCBI-PubMed data for the usage of traditional RGs reported that
GAPDH (27.24%),
ACTB (30.62%), and
18S (12.52%) were the three most widely used RGs in qPCR, semi-qPCR, and northern blotting [
26].
GAPDH is ubiquitously expressed in the cell and involved in DNA repair, tRNA export, membrane fusion, and transport, cytoskeletal dynamics, cell death, oligomerization, posttranslational modification, and subcellular localization [
27].
ACTB is an indispensable component of the cytoskeleton in the cell for cell migration, cell division, and regulation of gene expression [
28].
18S is a component of the ribosomal RNA and plays a role in the biogenesis and function of ribosome in the cell [
29]. However, its expression level is dependent on the experimental condition, in spite of its vital functions of cell survival and maintenance. In detail, the Ct values of both
GAPDH and
ACTB were decreased in
in vitro cultured blood mononuclear cells even without any treatment [
30] and altered under differentiation induction [
1] and long-term culture [
7] in MSCs. In the present study, we demonstrated the effect of the validated RGs during normalization and highlighted the possibility of false or misleading result caused by the usage of traditional RGs without validation (
Figure 6). Normalization with both the most (
TBP and
RPL4) and least (
18S) RGs could generate significant increase of lineage-specific genes in the differentiated bBMMSCs. However, there was no significant difference in
OCT4 expression, which is known to be highly expressed in MSCs than in differentiated cells as a pluripotent marker, between bBMMSCs and dermal skins when the unstable RG (
18S) was used for normalization. Similarly, although significant gradual downregulation of
OCT4 expression during long-term culture of human MSCs, indicating progressive reduction of pluripotency, after normalization against the most stable RGs was observed, the least stable RG (
GAPDH) generated no difference in
OCT4 expression [
7]. These findings indicated the importance of validation of RGs before using even though they are widely used RGs.
An ideal RG should be neither affected nor regulated by each experimental condition. However, as no single RG has till date been addressed to be universal and perfectly constant regardless of the experimental condition, the importance of validation of RGs before normalization cannot be emphasized enough to avoid generation of false or contradictory conclusions. To summarize, the present study proposes that TBP and RPL4 were suitable as stable RGs for gene expression study in bovine MSCs. These results may contribute to the experimental set-up of researchers working on animal MSCs as reference data.