A Fine-Tune Role of Mir-125a-5p on Foxn1 During Age-Associated Changes in the Thymus

Decline of transcription factor FoxN1, which predominantly regulates thymic epithelial cell (TEC) differentiation and homeostasis lifelong, is demonstrated to be casually related to age-related thymic involution. Whereas, a global role of microRNAs (miRNAs) has also been demonstrated to control and maintain TEC-constituting thymic microenvironment and to be changed in expression profile in the aged thymus. Therefore, it is urgently necessary to build knowledge regarding whether and which miRNAs regulate FoxN1 gene in the aged thymus. We primarily compared changes in miRNA expression profile between young and aged murine TECs with Mus musculus miRBase-V20 arrays (containing 1892 unique probes), and clearly identified and validated that at least one miRNA, miR-125a-5p, was increased in aged thymus. Applying miR-125a-5p mimics was able to inhibit FoxN1 3′UTR luciferase activity in a 293T cell line and to suppress FoxN1 expression in murine TEC Z210 cells. Since a single miRNA can play a fine-tuning role to regulate expression of multiple genes and a single gene can be regulated by multiple miRNAs, our result adds a single miRNA, miR-125a-5p, into the panel of FoxN1-regulating miRNAs associated with thymic aging.

findings provide a regulatory axis from miRNAs' finetuning FoxN1 to age-related disruption of TEC homeostasis in thymic involution. However, based on the principle that a single miRNA can regulate multiple genes (usually several hundreds), while a single gene can be regulated by multiple miRNAs [4,12], single miRNAs in the pool of FoxN1-regulating miRNAs associated with thymic aging still remains largely unknown.
Mature miRNAs are a large group of conserved, single-stranded, around 22 nucleotide (nt)-long, noncoding RNAs that act as post-transcriptional regulators of gene expression by binding to a target coding mRNA through imperfect complementarity at multiple sites of 3'UTR of a gene, resulting in coding mRNA cleavage, translational repression, or chromatin modification [22][23][24] to interfere a gene expression [25], which is mediated by more than one miRNAs [4,12]. It is demonstrated that although a single miRNA, miR-205, is highly and preferentially expressed in medullary TECs (mTECs), genetic knockout of it is not sufficient to induce a deficient phenotype [26].
In the current study, we primarily compared changes in miRNA expression profile between young and aged TECs with miRBase-V20 arrays (containing 1892 unique probes). We found there are 341 of them giving detectable signals in TEC from young and old mouse groups. Among the miRNAs with signal intensity over 500 units, we found 8 of them up-regulated and 4 of them downregulated significantly in TECs from aged, compared to young, murine thymuses. Among those 8 up-regulated miRNAs, we identified that miR-125a-5p negatively finetunes FoxN1 expression by validation through a real-time RT-PCR, and confirmed by a FoxN1 3'UTR luciferase assay in a 293T cell line and a miR-125a-5p mimic to FoxN1 suppressive assay in a TEC line: Z210. Since a single miRNA can play a fine-tuning role in regulation of multiple gene expression, while a single gene can be regulated by multiple miRNAs [4,12], a single miRNA, miR-125a-5p is not the sole miRNA in the panel of FoxN1-regulating miRNAs, but we indeed add a single miRNA into the panel associated with thymic aging.

Mice and animal care
Thymus samples were freshly isolated from either young (2 months) or aged (19-21 months) mice (C57BL/6 genetic background). All animal experiments were performed according to the protocols approved by the Institutional Animal Care and Use Committee of the University of North Texas Health Science Center, in accordance with guidelines from the National Institutes of Health, USA.

miRNA microarray and Real-time RT-PCR
Total RNAs for miRNA microarray were isolated from enriched TECs (based on our previously published enzymatic method [27]) of young and aged murine thymuses using QIAGEN miRNeasy Mini Kit (QIAGEN Cat#217004) according to the manufacturer's instructions. The miRNA microarray assay was performed through a service by LC Sciences, LLC (www.lcsciences.com) with miRBase-V20 arrays containing 1892 unique probes. Total RNAs for real-time RT-PCRs from enriched TECs of freshly isolated murine thymuses or from TEC line Z210 were isolated with TRIzol reagent (Invitrogen) and reverse transcribed to cDNA with the SuperScriptIII cDNA kit (Invitrogen). Real-time RT-PCRs for miRNAs were performed with TaqMan® MicroRNA Assay kits of individual miRNA primers and probes (Applied Biosystems). Real-time RT-PCR of FoxN1 primers and probe (TaqMan method) were as described previously [27]. Real-time RT-PCR was run on a Step-One-Plus thermal cycler system (Applied Biosystems). The microRNA results were internally normalized to U6snRNA control levels, while the relative expression levels of FoxN1 mRNAs from the Z210 were internally normalized to GAPDH levels. The average CT value from multiple young animals was always arbitrarily set as 1.0 in each real-time PCR reaction.

Bioinformatics prediction of the targets of miRNAs
The targets of the miRNAs were predicted using multiple databases, including the TargetScan (www.Targetscan.org), miRanda (www.microrna.org), microcosm Targets (www.ebi.ac.uk/enright-srv/ microcosm), and Segal Lab of Computational Biology (http://genie.weizmann.ac.il/index.html). The predicted hits from each algorithm were sorted as per the scores.

Aging and Disease
Aging and Disease • Volume 8, Number 3, June 2017 281

Luciferase reporter assay
Each well of 293T cells cultured on a 12-well plate was transfected with 50ng of wild-type FoxN1 3'UTR luciferase reporter plasmids or m(364-387) mutated reporter plasmids, and miR-125a-5p mimics (with a various of concentrations ). Aliquots of medium from the transfected wells were collected at 24h, 48h, and 72h post transfection to measure luciferase activity. 50μl of the medium (diluted if necessary) was mixed with 100μl of substrate solution containing 0.5μg/ml of coelenterazine (CTZ), 200mM NaCl, 50mM Tris-HCl and 0.01% Triton X-100, at pH 8.7. The light emission was measured at a wavelength of 480 nm and normalized with the SEAP expression.

Statistics
Data are presented as mean ± SEM. Statistical significance was analyzed by unpaired Student's t-test and two-way ANOVA test. Differences were considered statistically significant at values of p < 0.05.

Different expression profiles of miRNAs in TECs from young and aged murine thymuses
To screen age-related miRNAs in murine thymic epithelial cells (TECs), we assessed the miRNA expression profiles of total RNAs from enriched TECs isolated from the thymuses of 2-month-old and 21months-old mice, respectively, with miRNA microarray (miRBase-V20 arrays containing 1892 unique probes, performed by LC Sciences). The results show that 341 miRNAs were detectably expressed in TECs with signal intensity over 500 in any of each (young or aged) or both (young and aged) groups (Red dots in Figure 1A). After future analysis by focusing on those with signaling strength ≥ 500, we found there are 4 miRNAs decreased (fold change ≤ -2.0) and 8 miRNAs increased (fold change ≥ 2.0) in aged TECs compared to young TECs. A hierarchical clustering/ dendrogram analysis of these 12 miRNAs between 2 old (Old 1 & 2) and young (Young1 & 2) mice is shown in Figure 1B, and their detailed information is shown in Figure 1C, suggesting that TECs from young and old mice indeed have a certain of miRNA expression distinctly, and these twelve differently expressed miRNAs are potentially related to TECconducted thymic aging, associated with FoxN1 expression. Methods for details). (C) One hundred nM of miR-CON (mimic negative control) or miR-125a-5p was co-transfected with 50ng of the wild type (WT) or mutant FoxN1 3'UTR luciferase reporter plasmids, and the SEAP-expressing plasmid. GLuc activity was measured 24h post transfection and normalized as described above. Data are presented as mean ± SEM (n = 9, i.e. nine independent experiments). *p < 0.05, **p < 0.01, and ***p < 0.001, versus control (two-way ANOVA test).

Validation of expression of miRNAs in TECs potentially associated with thymic aging
To validate the results obtained from the microarray assay (Figure 1), we chose those screened 12 miRNAs to perform a more in-depth assay on their expression in TECs with more numbers of 2-month-old and 19-monthold mice by a real-time RT-PCR approach. Results are shown in Figure 2. In those 12 miRNAs, there were 6 miRNAs with statistically different expression, among which two (miR-18a-5p, miR-466f) were down-regulated (< -2.0 folds) in aged TECs which were consistent the result with ones in the microarray (Fig.2 top panels), and four (miR-342-3p, miR-6931-5p, miR-125a-5p, miR320-5p) were up-regulated (> 2.0-fold), confirmed the trend in the microarray data ( Fig. 2 middle and bottom panels). The real-time RT-PCR utilized more numbers of animals, which provide reliable biological standard deviation, confirm that microarray clues are informational.

Predicting conserved target sites of potential miRNAs in the FoxN1 3'UTR
Since we focused on miRNAs with directly-inhibiting FoxN1 role, we particularly aimed those up-regulated 4 miRNAs and predicted their potential target sites on FoxN1 3'UTR using the online target prediction programs [28] of TargetScan Microcosm Targets, and miRanda. We did not find targeting sites of miR-6931-5p and miR-342-3p in FoxN1 3'UTR, while we found that miR-320-3p and miR-125a-5p have 3 and 5 targeting sites in FoxN1 3'UTR, respectively (Fig. 3). Because miR-125a-5p has 5 targeting sites in FoxN1 3'UTR, in which nucleotide positions 364-387 were aligned with each other (as shown in Figure 3A top panel), miR-125a-5p attracts us. This clue suggests that miR-125a-5p has a higher potential to target FoxN1 to play a negative fine-tune for inhibition of FoxN1 expression in aged thymus. To further explore the potential role of miR-125a-5p in repressing FoxN1, we constructed a FoxN1 3'UTR luciferase reporter plasmid. A ~1.25Kb FoxN1 3'UTR DNA fragment [containing the last 50bps of FoxN1 translated region, via PCR from a mouse bacterial artificial chromosome (BAC) clone RP24-137A13] was sub-cloned into downstream of Gaussia luciferase (Gluc) that is driven by PGK promoter (kindly provided by Dr. Sui) [29] (Figure 4). Meanwhile, we also made a mutant 3'UTR of FoxN1 [a 7bp mutation on the first target site for miR-125a-5p, termed m(364 -387)], as a control vector (Fig. 4A bottom panel), along with wide-type (WT) 3'UTR of FoxN1 vector (Fig. 4A top penal). NIH293T cells were transfected with an either WT FoxN1 3'UTR luciferase reporter plasmid followed by adding synthetic mimics of miR-125a-5p with a final concentration of 0 -200nM (Fig. 4B). A significant dose-dependent repression in the WT FoxN1 3'UTR luciferase reporter activity was observed at 24h, 48h and 72h post co-culture (Fig. 4B). In addition, a mutant FoxN1 3'UTR luciferase reporter plasmid m (364-387) (at 100nM) was also transfected into 293T cells. However, the suppression, seen in WT FoxN1 3'UTR transfection (Fig. 4B), could not be observed in the 24hr post co-culture compared with miR-CON (a negative control miRNA), which is a random sequence of a miRNA molecule that has been extensively tested in human cell lines and tissues not to produce identifiable effects on known miRNA function. (Fig. 4C). Taking together, the result that miR-125a-5p mimic was able to suppress post-transcription activity of WT 3'UTR of FoxN1 at a dose-dependent manner, but could not suppress this activity of mutant 3'UTR of FoxN1, which indicates that miR-125a-5p specifically targets 3'UTR of FoxN1.

In vitro effects of miR-125a-5p on FoxN1 expression in a TEC line
Since FoxN1 is endogenously expressed in epithelial cells, to determine whether miR-125a-5p is able to fine-tune endogenous FoxN1 expression, we transfected the synthetic miR-125a-5p mimics into murine TEC line Z210 (a murine thymic stromal cell line with a medullary phenotype, generated by Dr. Farr) [30] and observed inhibiting effects of miR-125a-5p on the TEC endogenous FoxN1 expression with Real-time RT-PCR. As shown in Figure 5, increasing concentration of miR-125a-5p mimics (25, 50, 100nM) negatively regulates the expression of endogenous FoxN1 in TECs at 48hrs post transfection (Fig. 5A). We further analysis the timecourse inhibiting effect at a highest concentration of miR-125a-5p mimics (100 nM) and found that there is not a time-course change, but the inhibiting effect cannot be reversed even by 72hrs (Fig. 5B).

DISCUSSION
The thymus is a unique organ to be responsible for T cell development and central tolerance establishment. Agerelated thymic involution has a severe clinical outcome, which results not only in reduced output of naï ve T cells but also in increased output of self-reactive T cells in the elderly [31]. The thymus consists of two major cellular components: thymocytes from hematopoietic origin and stromal cells, mostly thymic epithelial cells (TECs), from non-hematopoietic origin. TECs mainly constitute thymic microenvironment to support thymocyte development. However, TEC development and homeostasis are mainly regulated by an epithelial cell-autonomous transcription factor, FoxN1 gene, through the p63-FoxN1 regulatory axis [32]. Age-related thymic involution [33][34][35] has been demonstrated to be causally associated with decline of FoxN1 [1][2][3]. However, decline of FoxN1 with age could be mostly regulated by epigenetic alteration, since mounting evidence indicates that changes in many transcription factors are epigenetically regulated in the context of the aged microenvironment [36][37][38][39][40], where epigenetic modifications are associated with remodeling of chromatin structure to play a central role in controlling molecular expression [37][38][39][40] via alteration of transcriptional activity. Epigenetic regulation includes three major mechanisms at three layers: DNA methylation (DNA layer), histone modifications (chromatin layer), and non-coding microRNA expression (post-transcription layer). However, limited information is available on how changes in organism affect FoxN1 expression in the aged thymus. Our study suggests a new target at posttranscription level in the aged thymic context. We identified a potential FoxN1 negative regulator, miR-125a-5p, which is able to fine-tune FoxN1 expression based on (1) increased miR-125a-5p expression in the aged thymuses, (2) an inhibiting effect on FoxN1 3'UTR luciferase reporter by transferring miR-125a-5p mimics, and (3) a dose-dependent decrease of endogenous FoxN1 expression in a TEC line by transferring miR-125a-5p mimics.
Regulation of TEC biological processes by miRNAs has been recognized in many context, particularly in TEC development [7,8,11]. FoxN1 gene is critical in the regulation and maintenance of TEC biological function and it also controls postnatal TEC homeostasis. Therefore, miRNAs appear to be involved in the epigenetic regulation of TEC aging through repressing FoxN1 posttranscription in the aged thymus. However, the regulation of thymic aging by miRNAs' suppressive function to FoxN1 gene is the lack of evidence, while studying on miRNA profiles between young and aged thymus just began [9,10]. It is well-known that a single gene can be regulated by multiple miRNAs. However, specific miRNAs on regulation of age-related decline of FoxN1 has been less well investigated. Although some miRNA candidates, such as miR-18b and miR-518b, were reported to lead to downregulation of FOXN1 in stem cells [11], herein, we add one more candidate miR-125a-5p leading to downregulation of FoxN1 in the aged thymus. A steady-state up-regulation of miR-125a-5p was also observed in 10-and 19-month-old murine thymuses by other group [9], which the possibility that miR-125a-5p is indeed involved in thymic aging associated with a decline of FoxN1. Because a single miRNA may only play a role for a fine-tune, we do not expect that totally abolishing FoxN1 expression can be achieved by knocking out or inhibiting single miR-125a-5p.
MiR-125 is a group of miRNA family, whose members include miR-125a-5p and miR-125b etc. MiR-125 play important roles in development and cell differentiation [41]. A significant down-regulation in miR-125a-5p and miR-125b was observed in malignant and aggressive breast cancer tissues [42]. An important role for miR-125a-5p in angiogenesis regulation via RTEF-1 was elucidated in aging mice [43]. Evidence also shows that miR-125a-5p may be as a potential tumorsuppressor in gastric cancer [44,45], and acts as an inhibitor to epidermal growth factor receptor signaling to block migration and invasion of lung cancer cells [46]. These findings of mir-125a-5p as a suppressor of proliferation and migration in cancer cells may provide a clue that an increase of mir-125a-5p in the aged thymus can suppress thymic epithelial cell homeostasis during aging. Additionally, in non-cancer cells, miR-125a-5p also has a role in suppressing classical activation of macrophages while promoting alternative activation by targeting Kruppel-like transcription factor 13 (KLF13) [41], which negatively regulates inflammation associated with T cell activation. An important role for miR-125a-5p in angiogenesis regulation via RTEF-1 was elucidated in aging mice [43]. In a recent study, miR-125a-5p appears to cause inflammatory activation of THP-1 cells [47], and activates NF-κB activity in diffuse large B-cell lymphoma cells [48]. These findings may be reminiscent of increased inflammation in the aged thymus.
In conclusion, our finding suggests that miR-125a-5p plays a role as a negative regulator of FoxN1 in thymic involution during aging. Targeting miR-125a-5p in aged thymus should be one of therapeutic strategies to delay age-related thymic involution and/or improve thymic function in the elderly.