Abstract

Objectives

   Thyroid hormone (T3) and high glucose concentrations are critical
   components of β-cell maturation and function. In the present study, we
   asked whether T3 and glucose signaling pathways coordinately regulate
   transcription of genes important for β-cell function and proliferation.

Methods

   RNA-seq analysis was performed on cadaveric human islets from five
   different donors in response to low and high glucose concentrations and
   in the presence or absence of T3. Gene expression was also studies in
   sorted human β-cells, mouse islets and Ins-1 cells by RT-qPCR.
   Silencing of the thyroid hormone receptors (THR) was conducted using
   lentiviruses. Proliferation was assessed by ki67 immunostaining in
   primary human/mouse islets. Chromatin immunoprecipitation and proximity
   ligation assay were preformed to validate interactions of ChREBP and
   THR.

Results

   We found glucose-mediated expression of carbohydrate response element
   binding protein alpha and beta (ChREBPα and ChREBPβ) mRNAs and their
   target genes are highly dependent on T3 concentrations in rodent and
   human β-cells. In β-cells, T3 and glucose coordinately regulate the
   expression of ChREBPβ and PCK1 (phosphoenolpyruvate carboxykinase-1)
   among other important genes for β-cell maturation. Additionally, we
   show the thyroid hormone receptor (THR) and ChREBP interact, and their
   relative response elements are located near to each other on mutually
   responsive genes. In FACS-sorted adult human β-cells, we found that
   high concentrations of glucose and T3 induced the expression of PCK1.
   Next, we show that overexpression of Pck1 together with dimethyl malate
   (DMM), a substrate precursor, significantly increased β-cell
   proliferation in human islets. Finally, using a Cre-Lox approach, we
   demonstrated that ChREBPβ contributes to Pck1-dependent β-cell
   proliferation in mouse β-cells.

Conclusions

   We conclude that T3 and glucose act together to regulate ChREBPβ,
   leading to increased expression and activity of Pck1, and ultimately
   increased β-cell proliferation.

   Keywords: ChREBP, Diabetes, Pancreatic β-cell, Glucose, Thyroid
   hormone, Proliferation

Graphical abstract

   Image 1
   [29]Open in a new tab

   Created with BioRender.com

Highlights

     * •
       T3 is necessary for and glucose-mediated transcription on rodent
       and human β-cells.
     * •
       T3 and glucose coordinately regulate the expression of ChREBPβ and
       PCK1 among other important genes for β-cell maturation.
     * •
       Fine-tuning of glucose and T3 signals can regulate gene expression
       and proliferation of rodent and human pancreatic β-cells.
     * •
       THR and ChREBP are interacting have common target genes and
       interact in close proximity in β-cells.
     * •
       PCK1 activity is sufficient to drive β-cell proliferation.

1. Introduction

   The association between thyroid dysfunction and diabetes has long been
   recognized, and both hypothyroidism and hyperthyroidism are associated
   with diabetes [[30][1], [31][2], [32][3], [33][4], [34][5], [35][6],
   [36][7], [37][8], [38][9], [39][10]]. Thyroid hormones act to promote
   or antagonize insulin's actions depending on the context as well as the
   cell type they are acting upon. Thus, thyroid hormones participate in a
   fine balance that promotes normal glucose metabolism and any deviation
   of thyroid hormone abundance can perturb glucose homeostasis [[40]4].

   One way that T3 affects glucose homeostasis is through its influence on
   β-cell mass. Thyroid hormone (T3) is required for islet development and
   function [[41][11], [42][12], [43][13], [44][14], [45][15]]. T3
   promotes β-cell proliferation in human and rodent cell lines and in the
   embryonic murine pancreas in explant culture [[46]13,[47][16],
   [48][17], [49][18]]. Glucose is also a known β-cell mitogen, implicated
   in adaptive β-cell expansion [[50][19], [51][20], [52][21], [53][22]].
   One transcription factor known to mediate this effect is Carbohydrate
   Response Element Binding Protein (ChREBP) [[54]23,[55]24]. ChREBP is a
   glucose responsive transcription factor that has two splice isoforms.
   One is ChREBPα which is mostly cytoplasmic and repressed in low
   glucose. The protein consists of an N-terminal low glucose inhibitory
   domain, containing a nuclear export signal that folds over and
   represses the activation domain. The C-terminal contains a
   beta-helix-loop-helix Zip DNA-binding domain. The other major isoform
   is ChREBPβ, which is a product of alternative splicing Where the low
   glucose inhibitory domain and nuclear export signals are removed but is
   otherwise identical to ChREPBα [[56]25]. Consequently, ChREBPβ is
   mostly nuclear, and is constitutively and potently active [[57]25].
   Notably, both T3 and high glucose concentrations are critical
   components of protocols that drive differentiation of stem cells to
   β-cells [[58]14,[59][26], [60][27], [61][28]].

   In mouse brown adipose tissue (BAT) we demonstrated that T3 and glucose
   synergistically regulate ChREBP, which in turnupregulates Ucp1, Glut4
   and Fasn, resulting in increased thermogenesis, decreased body weight,
   and improved glycemic levels. Recently, T3 was shown to promote
   lipogenesis in hepatocytes [[62]30]. Similarly, T3 and glucose were
   shown to coordinately interact to activate ChREBPβ transcription, which
   in turn activates lipogenesis and fatty acid oxidation in hepatocytes
   [[63]31]. In islets, both ChREBP splice isoforms- α & β [[64]25], are
   expressed [[65]29]. The expression of the β isoform is induced in
   response to increased glucose concentrations and is mostly nuclear,
   while ChREBPα is mostly cytoplasmic [[66]25,[67]32]. In β-cells,
   ChREBPβ (but not ChREBPα) expression is upregulated in response to
   glucose, leading to increased expression of known ChREBP target genes
   and increased β-cell proliferation [[68]29]. Furthermore, this
   upregulation of ChREBPβ is required for glucose-stimulated β-cell
   proliferation and adaptive expansion of β-cell mass [[69]29,[70]32]. In
   pancreatic β-cells, ChREBP is a known regulator of liver-type pyruvate
   kinase (Pklr), which encodes an enzyme that catalyzes the conversion of
   phosphoenolpyruvate to pyruvate, the last step of glycolysis [[71]33].
   ChREBP also regulates expression of thioredoxin-interacting protein
   (Txnip) [[72]34] which is involved in oxidative stress and is
   implicated in the regulation of β-cell death [[73]35,[74]36]. Other
   target genes of ChREBP include lipogenic genes, and hence ChREBP is
   thought to play a role in mediating glucolipotoxicity in β-cells
   [[75]32,[76][37], [77][38], [78][39]]

   Since ChREBP was shown to play a key role in glucose stimulated β-cell
   proliferation [[79]29,[80]40], we tested the hypothesis that glucose
   and T3 have a synergistic effect on ChREBP transcription and thus
   β-cell proliferation. We found that T3 and glucose act together to
   regulate expansion of β-cells in response to glucose. We identified a
   novel pathway that controls proliferation in pancreatic β-cells, the
   activation of phosphoenolpyruvate Carboxykinase (PEPCK-C) activity.
   PEPCK-C (gene name PCK1) is a main control point for the regulation of
   gluconeogenesis. PEPCK-C converts oxaloacetate and GTP into
   phosphoenolpyruvate, GDP and CO[2]. PEPCK promotes cancer cell
   proliferation in vitro and in vivo by increasing glucose and glutamine
   utilization toward anabolic metabolism. This effect is mediated at
   least partially by mTORC1 [[81]41,[82]42]. PCK1 was demonstrated by
   Shalev et al. to be the second most glucose responsive gene in
   pancreatic human islets after Txnip [[83]43]. In the liver, ChREBP is
   regulated by glucose levels [[84]25,[85]44], and also by T3
   [[86]45,[87]46]. However, crosstalk or cooperative signaling effects
   between glucose and T3 in β-cells have not been studied.

   While it is now established that human and murine α-cells express PCK1
   [[88]47], it is widely thought that mature β-cells do not express PCK1
   [[89]48]. In this study and by examining various available data sets
   for β-cell and human and rodent pancreatic progenitor cell
   differentiation, we found that PCK1 is expressed during maturation and
   development of β-cells [[90][49], [91][50], [92][51], [93][52],
   [94][53]], at a time when the proliferative capacity of β-cells is the
   highest [[95]54,[96]55]. We hence suggest a mechanism whereby T3 and
   glucose signaling pathways coordinately regulates transcription of
   genes important for β-cell function and mass, a novel concept in islet
   biology.

2. Materials and methods

2.1. Cell culture

   INS-1–derived 832/13 rat insulinoma cells were maintained in RPMI 1640
   medium with 10% FBS, 10 mM HEPES, 2 mM l-glutamine, 1 mM sodium
   pyruvate, and 50 mM β-mercaptoethanol, 100 U/mL penicillin, 100 mg/mL
   streptomycin and further supplemented with 11 mM glucose, at 37 °C in a
   5% CO[2] incubator. To specifically study the effect T3, 10%
   resin-stripped FCS, was used to deplete thyroid hormones as described
   in Cao et al. [[97]56].

2.2. RNA-seq analysis

   Total RNA from ∼100 islets per condition, from five different human
   donors was isolated using the RNAeasy micro kit (Qiagen) according to
   the manufacturer's protocol. RNA integrity was assessed using Ribogreen
   to determine total mass and Fragment Analyzer. All samples passed QC.
   The RQN (RNA quality) scores ranged from 7.7 to 10. Samples were
   submitted to the New York Genome Center and RNA was amplified via the
   NuGEN Ovation RNA-Seq System V2 prior to RNA sequencing. 35–40 million
   2 × 50 bp paired-end reads were sequenced per sample on the HiSeq2500
   instrument (Illumina). Raw count data was pre-filtered to keep genes
   with CPM >1.0 for at least 60% of the samples. After filtering, count
   data was normalized via the weighted trimmed mean of M-values [[98]57]
   and normalized counts were further transformed into normally
   distributed expression values via the voom-transformation using a model
   that included technical and demographic covariates (gender, age, body
   mass index, intronic rate). We estimated the correlation between
   measurements made on the same subject using the limma function,
   duplicate Correlation and the intersubject correlation was input into
   the linear model fit using the limma block design [[99]58]. The
   voom-transformed, adjusted expression data was the final input for
   statistical modeling. Statistical analysis was carried out using R
   language version 3.0.3 and its available packages [[100]59]. Volcano
   plots were generated using ggplot2 function in R [[101]60]. Data is
   available in GEO ([102]GSE218334).

   Comparisons between groups (log-fold-changes) were obtained as
   contrasts of the fitted linear modes generated using weighted least
   squares (lmFit) and empirical Bayes method [[103]58,[104]61]. A
   factorial design was also used to determine if genes respond
   differently to thyroid stimulation in low glucose versus high glucose
   concentrations (interaction term).

2.3. Identification of ChoREs

   Carbohydrate response elements (ChoREs) binding motifs were downloaded
   from the Schmidt et al. paper [[105]62], which aimed at determining
   such motifs by ChIP-seq in rat. By using the “seq2profile.pl” function
   of HOMER version 4.11 displayed in over the ChREBP chromatin peaks, we
   regenerated the ChoRE motif matrix used to build the top logo of
   [106]Figure 3F from Schmidt et al. We then further “trained” the motif
   matrix by adding the ChoRE binding sites described by Poungvarin et al.
   [[107]63] for mouse exons 1a and 1b. The final matrix
   ([108]Supplementary Figure 9) was fed to the “findMotifs.pl” HOMER
   function by using the human GRCh38/hg38 and the GRCm38/mm10 mouse
   genomes. The coordinates of the ChoRE sites mapping within each of the
   genes (±5,000 bp) of [109]Figure 5A and [110]Supplementary Figure 8
   were determined by using the “genome_join” function of the “fuzzyjoin”
   version 0.1.6 package of r 4.2.0.

Figure 3.

   [111]Figure 3
   [112]Open in a new tab

   T3 and glucose enhance β cell proliferation. Human islets (A), Mouse
   islets (B) or Ins-1 cells (C) were dispersed and incubated at the
   indicated glucose or T3 concentrations in RPMI containing 10%
   resin-stripped serum. After 48 h, cells were fixed and immunolabeled
   for Ki67 and insulin. Presented are the percent of Ki67-positive and
   Insulin-positive cells. Data are the means ± SEM of at least three
   independent experiments. ∗P < 0.05; ∗∗P < 001; ∗∗∗P < 005; ∗∗∗∗P < 001
   by two-way ANOVA.

Figure 5.

   [113]Figure 5
   [114]Open in a new tab

   Promoters of key regulatory genes for islet development contain THR and
   ChREBP binding sites. A. ChREBP and THRB binding sites in human
   selected genes. Each panel is arranged as follows. The ideogram of the
   gene with its chromosomal location from the UCSC genome browser is
   shown. The representation displays exons (dark blue boxes) and introns
   (dark blue lines with arrowheads pointing to the direction of
   transcription). The promoter region (TSS ± 2,500 bp) is shown as a
   transparent red arrow. For the ChREBP gene, the position of the
   additional exon 1b is marked with a purple box and the intron between
   exons 1b and 1a is marked with a purple line with arrowheads oriented
   as for the rest of the gene. Blue and red upward arrowheads identify
   the center of ChREBP and THRB binding sites. ChREBP binding sites have
   been scored with the HOMER package (see material and methods) by using
   the frequency matrix of [115]Supplementary Fig. 9, except for three
   sites that have been experimentally validated and are marked with
   asterisks near the respective arrowheads. The two ChREBP binding sites
   experimentally validated within exon 1b of the ChREBP gene have been
   tested by our lab. The single ChREBP binding site upstream of the PCK1
   promoter has been tested by Jeong et al. [[116]76]. THRB sites have
   been extracted from the ReMap2022 database. [117]Supplementary Table 3
   provides the coordinates of both ChREBP and THRB sites displayed. B.
   Chromatin Immunoprecipitation in Ins-1 cells grown in RPMI (11 mM
   glucose) supplemented with regular FCS. ChIP was performed using ChREBP
   and THR (alpha and beta, [[118]30,[119]67]) antibodies to detect
   binding on ChREBP promoter area and actin coding area (C). D. Proximity
   ligation assay for ChREBP and THR in Ins-1 cells was performed as
   described in materials and methods. Cells were growing low and high
   glucose, in the presence or absence of T3. Bottom panel-quantification
   of cells showing positive proximity ligation signal. ∗P < 0.05;
   ∗∗∗P < 0.01; ∗∗∗∗P < 0.005 using one-way ANOVA. (For interpretation of
   the references to color in this figure legend, the reader is referred