Long Noncoding RNA, MicroRNA, Zn Transporter Zip14 (Slc39a14) and Inflammation in Mice

Integration of non-coding RNAs and miRNAs with physiological processes in animals, including nutrient metabolism, is an important new focus. Twenty-three transporter proteins control cellular zinc homeostasis. The transporter Zip14 (Slc39a14) responds to proinflammatory stimuli. Using enterocyte-specific Zip14 knockout mice and RNA-sequencing and quantitative polymerase chain reaction (qPCR), we conducted transcriptome profiling of proximal small intestine, where Zip14 is highly expressed, using RNA from whole intestine tissue, isolated intestinal epithelial cells (IECs) and intestinal organoids. H19, U90926, Meg3, Bvht, Pvt1, Neat1 and miR-7027 were among the most highly expressed genes. Enterocyte-specific deletion of Zip14 demonstrated tissue specific expression, as such these changes were not observed with skeletal muscle. Chromatin immunoprecipitation (ChIP) assays of chromatin from isolated intestinal epithelial cells showed that enterocyte-specific Zip14 deletion enhanced binding of proinflammatory transcription factors (TFs) signal transducer and activator of transcription 3 (STAT3) and nuclear factor kappa beta (NF-ĸβ) to promoters of H19, Meg3 and U90926. We conclude enterocyte-specific ablation of Zip14 restricts changes in those RNAs to the intestine. Binding of proinflammatory TFs, NF-ĸβ and STAT3 to the H19, Meg3 and U90926 promoters is consistent with a model where Zip14 ablation, leads to increased TF occupancy, allowing epigenetic regulation of specific lncRNA genes.


Introduction
There is robust literature on long non-protein-coding RNA (lncRNA) and small noncoding RNA or microRNA (miRNA) associated with an array of biological processes. The abundance of current information on lncRNAs and miRNAs stems to a considerable degree on data sets from RNA-sequencing and other transcriptome screening platforms. These efforts illustrate the roles of non-coding RNAs associated with necessary cellular functions. In most cases, their functions influence regulatory pathways and gene expression necessary for normal cell function. Atypical or excessive abundance of lncRNAs and miRNAs appear to be associated with disease, either extensions of normal functions or abnormal departures to disrupt normal homeostatic mechanisms [1][2][3][4].
An important focus of non-coding RNAs is their association with necessary cellular functions, which have developmental and physiological outcomes in animals. Expectedly, this includes nutrient metabolism. Early examples of such roles are the influence of miR-122 in iron homeostasis [5] and miR-33 for cholesterol transport [6]. Zinc, an essential micronutrient, has documented roles in biology that are classified as structural, catalytic and regulatory [7]. Frequently they overlap, e.g., the zinc finger motifs of proteins maintain structure, but also influence gene expression. Similarly, most histone deacetylase enzymes (HDACs) are zinc-dependent and require zinc for catalytic activity, yet they function to influence gene expression [8,9]. Our interest in zinc metabolism and status assessment stimulated our initial research on identification of zinc-responsive miRNAs. Using serum episomes from human subjects that were fed a zinc-deficient diet under controlled conditions, we found that a select sub-set of miRNAs were influenced by zinc intake [10]. As detected by quantitative polymerase chain reaction (qPCR), this sub-set of miRNAs decreased with dietary zinc restriction and returned to normal abundance levels within days with acute zinc repletion. These experiments were never followed up, despite the fact that they respond to dietary zinc restriction in humans and there is a need for a biomarker of zinc status [7]. Other investigators have described studies associating miRNAs/lncRNAs with zinc, including zinc effects on prostate cancer cell lines [11], transporter Zip9 and radiation-induced skin damage [12], transporter Zip7 and gastric cancer [13], ZnT4 and acute cerebral ischemia [14], human deficiency and esophageal cancer [15], and Zip8 and osteoarthritis [16].
To broaden our understanding of the influence of cellular zinc trafficking in skeletal muscle, we conducted experiments using a murine global knockout for the metal transporter Zip14 (Slc39a14). Transcript profiling with arrays resulted in the relative expression of the lncRNA, H19, and a derivative miRNA, miR-675, prominently upregulated in the whole body Zip14 knockout mice (WB-KO) [17]. Skeletal muscle from these genotypes shows muscle wasting, accompanied by decreased zinc uptake. The mice with the Zip14 ablation had a proinflammatory phenotype. High relative expression of H19/miR-675 was found in muscle of the WB-KO mice. Moreover, Zip14 and H19/miR-675 were induced in muscle with Lipopolysaccharide (LPS) treatment. Differential expression of specific transcripts was positively related to expression changes of specific lncRNAs/miRNAs, as a result of deletion of Zip14 and a reduction in zinc transport activity.
Here we show comparative effects of whole body Zip14 deletion on lncRNA/miRNA expression in skeletal muscle and proximal small intestine, as well as experiments with lncRNA/miRNA expression differences in intestinal epithelial cells (IECs) and intestinal organoids from mice with an enterocyte specific Zip14 (Slc39a14) deletion.

Animals
Whole Body Zip14−/− (WB-KO) mice were developed, characterized, and backcrossed as reported previously [18]. Wild-type (WT) mice of the same strain were used as controls. Heterozygous mutant (−/+) × heterozygous mutant (−/+) mating scheme was used to generate both genotypes for experiments. Enterocyte-specific Zip14 knockout mice (Zip14∆IEC) were generated by targeted deletion of introns 4-8 [19]. Briefly, founder Zip14 flox/+ mice were bred to obtain a floxed Zip14 (Zip14 F/F ) mouse line. They were crossed with Villin-Cre (VC-Zip14+/+) mice of the C57BL/6 line from The Jackson Laboratory (JAX # 004586; Bar Harbor, ME, USA) to create enterocyte-specific Zip14 conditional knockout (Zip14 ∆IEC ) mice. Zip14 F/F mice were used as controls for experiments with this strain. Breeding and genotyping were conducted at the University of Florida (Gainesville, FL, USA). Starting at weaning, the mice were fed ad libitum a commercial diet (Teklad LM-485 Mouse/Rat Sterilizable Diet; Harlan 7012, ENVIGO, Madison, WI, USA) and tap water. A 12 h light/dark cycle was used, and the mice were maintained in plastic barrier caging. Young adult male and female mice (8-16 weeks of age) were used for the experiments. Mice were killed by exsanguination, using cardiac puncture under isoflurane anesthesia, between 11:00 am and 2:00 pm. Protocols were approved by the University of Florida Institutional Animal Care and Use Committee (protocol # 202007015).

Hematoxylin and Eosin Staining (H&E)
Mice intestinal tissues were fixed in 4% paraformaldehyde for 24 h, embedded in paraffin, cut into 5-µm thick sections, dewaxed in xylene, and rehydrated with ethanol. Sections were then stained with hematoxylin for 2 min, followed by eosin for 2 min. Light microscopy was used for histological and pathological analysis at the UF molecular pathology core. Histopathological scores of intestine staining were assigned in a blinded manner.

Tissue Preparation and RNA Isolation
Tissues (liver, lung, pancreas, colon, gastrocnemius muscle, intestine, spleen, kidney) from enterocyte-specific and WB-KO mice were excised and samples were flash-frozen in liquid nitrogen (N 2 ) and stored at −80 • C prior to RNA extraction. Homogenization with TRI Reagent®(Molecular Research Center, Cincinnati, OH, USA) was performed with a Bullet Blender®(Next Advance, Raymertown, NY, USA). IECs were isolated from the first 20 cm of the intestine, after perfusion with ice-cold buffer (10 mM EDTA, 10 mM HEPES, 0.9% NaCl), using a chelation method [20]. For RNA-seq analysis, RNA from sections of proximal small intestine of WB-KO and WT mice was extracted as above. All RNA was isolated using RNeasy ® reagents and spin columns (Qiagen #217604, Hilden, Germany). RNA samples were treated with deoxyribonuclease (DNase). Quantity and quality of the RNA were assessed spectrophotometrically and further analyzed for integrity using a Bioanalyzer (Agilent, Santa Clara, CA, USA). Only RNA with an integrity number of at least 6.8 were used for RNA-sequencing (RNA-seq).

RNA-Sequencing
RNA-seq libraries, prepared from samples from individual mice, were sequenced on an Illumina Nova Seq 6000 Platform at Novogene Corp. (Sacramento, CA, USA). Genes with a Log2 fold change (FC) of >2, and lncRNAs and miRNAs with an FC of >1 were used for differential expression analysis [21,22]. RNA Sequencing data files were deposited into Gene Expression Omnibus-NCBI (accession number GSE210160).

Chromatin Immunoprecipitation (ChIP) Assays
ChIP assays were performed using intestinal epithelial cells from enterocyte-specific Zip14 knockout mice [19]. Briefly, tissue or cells were placed in PBS containing 1% formaldehyde for crosslinking. The crosslinking reaction was stopped by incubating in 20X glycine (0.2 M Tris, 1.5 M glycine (pH 8.0)) for 5 min. After washing, the chromatin was resuspended in sonication cell lysis buffer and fragmented by sonication (Bioruptor ® sonicator (Diagenode, Denville, NJ, USA)). The chromatin nuclear extract was immunoprecipitated following the manufacturer's protocol (Cell Signaling Technology, Danvers, MA, USA). The antibodies used were nuclear factor kappa beta (NF-kβ) (#8242) and signal transducer and activator of transcription 3 (STAT3) (#4904) from Cell Signaling Technology (Danvers, MA, USA). The primers used for H19, Meg3 and U90926 promoters are described in the supplemental materials (Supplementary Materials Table S3).

Preparation and Treatment of Intestinal Organoids (Enteroids)
A 20 cm segment of proximal small intestine from both Zip14 ∆IEC and Zip14 F/F mice were excised and perfused with phosphate-buffered saline (PBS), opened longitudinally, cut into 5 mm lengths and placed in Cell Dissociation Reagent (Stemcell Technologies, Vancouver, BC, Canada). The domes produced, were seeded into Matrigel ® (Corning, Corning, NY, USA) and incubated for 20 min at 37 • C, with InstestiCult™ (Stemcell Tech-Nutrients 2022, 14, 5114 4 of 13 nologies, Vancouver, Canada) added thereafter. The enteroids were cultured for 10 days. These media were replaced every three days. Enteroids in the Matrigel domes were placed in Organoid Harvesting Solution (Cultrex, R&D Systems, Inc., Minneapolis, MN, USA), washed with PBS, and RNA was isolated using RNeasy ® and TRI Reagent ® as above.

Statistics
Data are presented as means ± standard error of the mean (SEM) of biological replicates. Prism 6.0 software (Graph Pad Software Inc., La Jolla, CA, USA) was used for these analyses. Differences between genotypes were evaluated using Student's t-test and multiple comparisons used ANOVA, followed by Tukey's test. Statistical significance was indicated as * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; ns = not significant.

Influence of WB-Zip14 Ablation on Differential Expression of Specific lncRNA Genes in Proximal Small Intestine
Transcriptome profiling was carried out to examine the influence of Zip14 ablation in small intestine. A WB-Zip14 knockout model was used, with profiling by DNA array and RNA-sequencing, respectively [17,23]. Focusing on the small intestine of mice, RNA-seq was used to profile changes in lncRNA and miRNA transcripts. The volcano plot ( Figure 1A) shows that many lncRNAs had expression differentials (Fold Change) of >1.0 or <−1.0. A total of 53 and 55 lncRNAs met those criteria, respectively. Only five miRNAs had a FC of >1.0 and none were <−1.0. The full list, including cDNA sequences, predicted genes and uncharacterized cDNAs is presented as Supplementary Materials Tables S1 and S2. Read densities for H19, U90926 and Meg3 (Table 1) display the marked differences in lncRNA sequences produced by WB-Zip14 ablation ( Figure 1B). Note read densities for Malat1, used as a control, are uniform compared to those of the differentially expressed lncRNAs. A number of transcripts identified as upregulated were selected and confirmed as upregulated ( Figure 1C). To assess whether Zip14 ablation is associated with a specific mucosal pattern of inflammation, H&E-stained histological sections from intestinal tissue (duodenum) were examined. Histological examination of the small intestine showed no difference in the number of crypts, villus and goblet cell between both genotypes (Supplementary Materials Figure S1).

Tissue Specific Expression of lncRNAs in Mice with WB-Zip14 Deletion
To compare tissue-specificity of lncRNA and miRNA expression changes produced by Zip14 ablation, we used RNA from liver, lung, pancreas, colon, muscle, intestine, isolated IECs and enteroids from mice of the WB-KO and WT genotypes. U90926, H19 and Meg3 were selected as representing a range of FC differences. U90926 was differentially expressed in intestinal tissue, colon and IECs, but was undetectable in the other tissues examined, except WT lung (Figure 2A). H19 mRNA was detected in liver, muscle, enteroids, tissues of the gastrointestinal (GI) tract, and WT lung ( Figure 2B). Similarly, Meg3 is highly upregulated in the GI tract ( Figure 2C). Of note is that neither Meg3 nor U90926 are detectable with enteroid cultures (Figure 2A,C). Overall, these lncRNAs were found to be the most differentially expressed in the GI tract compared to the other tissues in our WB-KO mice.
Meg3 were selected as representing a range of FC differences. U90926 was differentially expressed in intestinal tissue, colon and IECs, but was undetectable in the other tissues examined, except WT lung (Figure 2A). H19 mRNA was detected in liver, muscle, enteroids, tissues of the gastrointestinal (GI) tract, and WT lung ( Figure 2B). Similarly, Meg3 is highly upregulated in the GI tract ( Figure 2C). Of note is that neither Meg3 nor U90926 are detectable with enteroid cultures (Figure 2A, C). Overall, these lncRNAs were found to be the most differentially expressed in the GI tract compared to the other tissues in our WB-KO mice.

Tissue Specific Expression of lncRNAs in Mice with Zip14 ΔIEC Deletion
qPCR assays of proximal small intestine and muscle RNA from Zip14 ΔIEC (ΔIEC) and Zip14 F/F (F/F) control mice were compared to visualize the changes in lncRNA expression. H19, Meg3, U90926, Bvht, and Pvt1 levels were significantly (p < 0.05, 0.01) increased in intestine of ΔIEC mice ( Figure 4A). In contrast, none of these lncRNAs were affected in

Tissue Specific Expression of lncRNAs in Mice with Zip14 ∆IEC Deletion
qPCR assays of proximal small intestine and muscle RNA from Zip14 ∆IEC (∆IEC) and Zip14 F/F (F/F) control mice were compared to visualize the changes in lncRNA expression. H19, Meg3, U90926, Bvht, and Pvt1 levels were significantly (p < 0.05, 0.01) increased in intestine of ∆IEC mice ( Figure 4A). In contrast, none of these lncRNAs were affected in skeletal muscle with this enterocyte specific Zip14 deletion ( Figure 4B). Normalization of 1.0 indicates a detectable signal rather than non-detectable qPCR signals.

Promoter Occupancy May Explain Differential Expression of H19, Meg3 and U90926 with Zip14 Ablation
Bioinformatic searches revealed promoters of Meg3, H19 and U90926 all have consensus binding sites for the transcription factors STAT3 and NF-ĸβ. ChIP assays using chromatin derived from IECs, isolated from ΔIEC and F/F mice, were conducted. No difference in binding of STAT3 were noted for the H19 promoter ( Figure 5A). Increased binding of this TF in one of the identified STAT3 sites in both the Meg3 and U90926 promoters was detected. There was increased binding of NF-ĸβ at two sites of the H19 promoter and

Promoter Occupancy May Explain Differential Expression of H19, Meg3 and U90926 with Zip14 Ablation
Bioinformatic searches revealed promoters of Meg3, H19 and U90926 all have consensus binding sites for the transcription factors STAT3 and NF-kβ. ChIP assays using chromatin derived from IECs, isolated from ∆IEC and F/F mice, were conducted. No difference in binding of STAT3 were noted for the H19 promoter ( Figure 5A). Increased binding of this TF in one of the identified STAT3 sites in both the Meg3 and U90926 promoters was detected. There was increased binding of NF-kβ at two sites of the H19 promoter and one site of the Meg3 promoter ( Figure 5B). Decreased NF-kβ binding was detected on Nutrients 2022, 14, 5114 9 of 13 the U90926 promoter in the ∆IEC mice ( Figure 5B). These differences suggest that the upregulation of these genes in the ∆IEC mice could be a proinflammatory response produced by these TFs through loss of Zip14.

Discussion
Ablation of Zip14 in mice leads to diminished transport of zinc into enterocy [17,24], and produces interesting phenotypes including decreased intestinal barrier fun tion, adipose hyperplasia, muscle wasting, glucose processing defects, and skeletal defe with aging [25]. Mutations of Zip14 in humans leads to neurological defects and skele abnormalities [15,16].
Our approach to examine a biochemical basis of Zip14 function has been throu transcriptome comparisons of mice with deletions of the Zip14 gene. Whole body deleti of Zip14 yields upregulation and down-regulation of specific genes in muscle [17]. S prisingly, among the most upregulated genes were H19, and miR-675, which is deriv from H19. These findings confirmed those of others; H19 is not silenced at birth in mus and is expressed in mice into adulthood [16,26]. We also showed that LPS administrati Figure 5. ChIP assays of STAT3 (A) and NF-κβ (B) binding to promoters for H19, Meg3 and U90926 using chromatin from enterocyte-specific Zip14 knockout mice. Values are reported as the mean ± SEM. n = 3 (each dot indicates data measurement for one mouse). * p < 0.05; ** p < 0.01; *** p < 0.001; ns = non-significant. Data were analyzed using Student's t test for F/F and ∆IEC comparison. F/F = floxed/floxed mouse; ∆IEC = Enterocyte-specific Zip14 knockout mice; ChIP = chromatin immunoprecipitation; STAT3 = signal transducer and activator of transcription 3; NF-kβ = nuclear factor kappa beta.

Discussion
Ablation of Zip14 in mice leads to diminished transport of zinc into enterocytes [17,24], and produces interesting phenotypes including decreased intestinal barrier function, adipose hyperplasia, muscle wasting, glucose processing defects, and skeletal defects with aging [25]. Mutations of Zip14 in humans leads to neurological defects and skeletal abnormalities [15,16].
Our approach to examine a biochemical basis of Zip14 function has been through transcriptome comparisons of mice with deletions of the Zip14 gene. Whole body deletion of Zip14 yields upregulation and down-regulation of specific genes in muscle [17]. Surprisingly, among the most upregulated genes were H19, and miR-675, which is derived from H19. These findings confirmed those of others; H19 is not silenced at birth in muscle and is expressed in mice into adulthood [16,26]. We also showed that LPS administration induced miR-675 in muscle, a response that was markedly enhanced in whole body Zip14 KO mice [17]. LPS administration also induces Zip14 and increases zinc uptake into muscle of WT mice, but not in WB-KO mice. Since Zip14 ablation leads to enhanced binding of inflammation-related TFs in muscle, we concluded the elevated miR-675 is a reflection of loss of a mechanism associated with muscle wasting [17]. We propose that the ablation of Zip14 function leads to upregulated expression of lncRNAs and miRNAs that in turn regulate specific inflammation-related genes.
The influence of nutrition on epigenetic changes and its relationship with lncRNAs or miRNAs modification has not been widely studied [27]. Minerals such as zinc are known to have the ability to modify epigenetic markers and alter cell signaling during inflammation or intestinal homeostasis [28,29]. The in-silico integrative analysis of the lncRNAs and miRNAs (RNA-seq analysis data) shows that the possible target genes of these miRNAs are related to the signaling pathways of transcriptional factors and inflammation pathway mediated by cytokines and chemokines. Zinc, as a signaling molecule, is critical to maintain proliferative cellular capacity that allows self-renewal of intestinal cells to maintain homeostasis of the intestinal epithelium [30]. The cytokine-and chemokinemediated inflammation pathway includes key protein/lncRNAs/miRNAs in intercellular communication that help maintain intestinal homeostasis [31], as well as driving intestinal inflammation in various pathological processes [32,33].
We have focused in this experimental series on the small intestine, as Zip14 is most highly expressed in the proximal region and decreases along the gastrointestinal axis [24]. Furthermore, Zip14 deletion in the intestine leads to decreased barrier function and mild endotoxemia [36]. There is considerable interest in lncRNAs in intestinal physiology [37]. Intestinal miRNAs are important factors in the function and regeneration of the intestinal epithelium. They have an important role in gene regulation under pathological and physiological conditions, including inflammatory response. RNA-seq differential expression analysis showed three highly expressed miRNAs including miR-99agh and -7027 (Supplementary Materials Table S2). A role for miR-7027 in the regulation of intestinal inflammation and gene regulation has not previously been demonstrated. After in-silico analysis, a relevant function has been proposed for miR-7027 in the intestinal mouse epithelium where this miRNA directly regulates inflammation by binding to 3 untranslated region (3 UTR) of STAT3 and interleukin-21 (IL-21) genes. Upregulation of these genes has been shown by RNA-seq analysis (GSE210160). miR-7027 is differentially expressed in both crypts and villus and was reported to be associated with pathological disorders [38]. It also has a compensatory function under the ablation or downregulation of other miRNAs [39].
As has been shown numerous times, and again here, H19 is expressed in the intestine. Functions related to intestinal permeability and inflammation have been shown where this lncRNA acts to regulate specific mRNAs [40,41]. Meg3 has been implicated in LPS-induced intestinal injury [42] and is believed to be upregulated by STAT3 in cardiomyocytes [43]. This latter finding agrees with results reported here. Expression of U90926 in intestine has been reported [44] and elevated intestinal expression has been demonstrated as a response to Giardia infection [45]. It has been proposed that lncRNAs in intestine are a reflection of resident consensual microbiota [46]. As we have found, Zip14 deletion increases the abundance of microbes of the Akkermansia phylum (In-press). Hence deletion of a metal transporter that supplies zinc to enterocytes may alter intestinal microflora that provide signals regulating lncRNA production. In addition, the bulk of the literature on enteric lncRNAs relate to cancer, thus the loss of zinc transport through Zip14 deletion could establish precancerous conditions.

Conclusions
The consistent and precise finding of the genomic driver of intestinal homeostasis and remodeling may have a critical implication for intestinal diseases and nutrition. The experiments reported here demonstrate deletion of the metal transporter Zip14 gene initiates the upregulation of specific lncRNAs and miRNAs, as detected by both RNA-Seq and specific qPCR assays. Enterocyte-specific ablation of Zip14 restricts changes in those RNAs to the intestine. The binding of proinflammatory TFs, NF-kβ and STAT3, to the H19, Meg3, and U90926 promoters is consistent with earlier work where Zip14-induced epigenetic factors influence TF occupancy of specific regulatory sites. These findings provide supporting evidence on the role of the Zip14 transporter, as its ablation initiates an inflammatory cascade of responses by activating noncoding regulatory elements in the intestinal epithelium. The Zip14 transcriptome study of comparative patterns in the intestinal epithelium is valuable to better understand the function and regulation of human orthologous genes during intestinal diseases. Nevertheless, future studies are needed to perform further validation of the functions of lncRNAs and microRNAs in response to intestinal microbiota, nutrition, and disease.