Peroxisome proliferator-activated receptor gamma (PPARγ) regulates lactase expression and activity in the gut

Abstract

Lactase (LCT) deficiency affects approximately 75% of the world's adult population and may lead to lactose malabsorption and intolerance. Currently, the regulation of LCT gene expression remains poorly known. Peroxisome proliferator activator receptorγ (PPARγ) is a key player in carbohydrate metabolism. While the intestine is essential for carbohydrate digestion and absorption, the role of PPARγ in enterocyte metabolic functions has been poorly investigated. This study aims at characterizing PPARγ target genes involved in intestinal metabolic functions. In microarray analysis, the LCT gene was the most upregulated by PPARγ agonists in Caco-2 cells. We confirmed that PPARγ agonists were able to increase the expression and activity of LCT both in vitro and in vivo in the proximal small bowel of rodents. The functional response element activated by PPARγ was identified in the promoter of the human LCT gene. PPARγ modulation was able to improve symptoms induced by lactose-enriched diet in weaned rats. Our results demonstrate that PPARγ regulates LCT expression, and suggest that modulating intestinal PPARγ activity might constitute a new therapeutic strategy for lactose malabsorption.

Keywords: PPARgamma; hypolactasia; intestinal epithelial cells; lactase; lactose intolerance.

Figure 1. PPARγ agonists specifically induce LCT expression and activity in Caco‐2 cells

1. Quantitative PCR (qPCR) analysis of LCT gene expression in stimulated Caco‐2 cells. Cells were stimulated for 24 h with each agonist. Results represent the fold change of LCT gene expression normalized to GAPDH level. The expression level measured in control cells was used as reference and defined as 1.

2. Immunofluorescence staining of Caco‐2 cells for LCT protein (green). Cells were stimulated for 24 h with each agonist. Nuclei are stained with DAPI (blue). Non‐relevant IgG was used as control (“IgG control”). Scale bar, 100 μm. Magnification ×20.

3. Western blot analysis of LCT protein expression from stimulated Caco‐2 cells. Densitometric analysis was used to quantify LCT protein.

4. LCT activity in Caco‐2 cells stimulated for 24 h. Results represent the fold change of LCT activity with respect to the activity measured in control cells arbitrarily defined as 1.

Data information: (A, D) Data are expressed as mean ± SEM (two to four independent experiments). Statistical analysis: two‐tailed nonparametric Mann–Whitney U‐test. ***P < 0.0001. Source data are available online for this figure.

Figure EV1. Dose effect of GED, CLA, pioglitazone, and rosiglitazone on LCT expression in Caco‐2 cells

Caco‐2 cells were stimulated with various concentrations of each PPARγ agonist for 24 h as indicated. LCT gene expression was determined by qRT–PCR of corresponding reverse‐transcribed mRNA. Results represent the mean ± SEM (two to three independent experiments in triplicate or sixplicate) of the fold change of LCT gene expression. The expression level measured in control cells, arbitrarily defined as one, was used as reference. NS, not significant. Statistical analysis: two‐tailed nonparametric Mann–Whitney U‐test.

Figure EV2. GED, pioglitazone, and rosiglitazone do not induce upregulation of other disaccharidases expressed by Caco‐2

Relative expression of genes encoding sucrase‐isomaltase (SIM) and maltase‐glucoamylase (MGAM) compared to LCT in Caco‐2 cells as determined by qPCR analysis of reverse‐transcribed mRNA. qPCR analysis of SIM and MGAM mRNA expression in Caco‐2 cells stimulated with 1 mM GED, 1 μM Pio, and 1 μM Rosi. Results represent the mean ± SEM (two independent experiments, 8 < n < 9) of the fold change of expression of SIM and MGAM genes normalized to GAPDH level. The expression level measured in control cells, arbitrarily defined as one, was used as reference. NS, not significant. Statistical analysis: two‐tailed nonparametric Mann–Whitney U‐test.

Figure EV3. LCT genotyping of polymorphisms C/T₁₃₉₁₀ and G/A₂₂₀₁₈ for Caco‐2 cells

DNA genomic fragments encompassing C/T₁₃₉₁₀ and G/A₂₂₀₁₈ nucleotides were amplified and digested as described in Matthews et al (2005). The digested PCR products were analyzed on agarose gel. Lanes 1 and 8 show molecular weight markers. Lanes 2 and 9 show non‐digested (ND) PCR products. Lanes 3 and 10 show the CC₁₃₉₁₀ and GG₂₂₀₁₈ homozygous lactose intolerance genotypes, respectively. Lanes 4 and 11 show the TT₁₃₉₁₀ and AA₂₂₀₁₈ homozygous lactase persistent genotypes, respectively. Lanes 5 and 12 show the CT₁₃₉₁₀ and GA₂₂₀₁₈ heterozygous lactase persistent genotypes, respectively. Lanes 6 and 13 show non‐digested PCR products from Caco‐2 cells. Lanes 7 and 14 show digested PCR products displaying LCT genotyping polymorphisms C/T₁₃₉₁₀ and G/A₂₂₀₁₈ for Caco‐2 cells. Migration profile reveals that Caco‐2 cells possess the CC₁₃₉₁₀ (lane 7) and GG₂₂₀₁₈ (lane 14) homozygous lactose intolerance genotypes.

Figure 2. PPARγ is a transcriptional regulator of the LCT gene

1. Chromatin immunoprecipitation (ChIP) assay. The top diagram depicts the PPRE predicted by in silico analysis. The picture shows PCR amplification of the 8a–8b fragment in ChIP assay from Caco‐2 cells. Graph bars represent quantification of the 8a–8b fragment by qPCR. Results are expressed as fold enrichment with amplification from control cells defined as 1.

2. Luciferase gene reporter assay in Caco‐2 cells transfected with pGL4Luc‐PromLCT, pGL4Luc‐PromLCT MUT, and pGL4Luc‐PromLCT DEL reporter constructs. Cells transfected with empty pGL4Luc were used as control. Results represent the fold change luciferase activity normalized for protein content.

3. LCT gene expression measured by qPCR and LCT activity in PPARγ knockdown Caco‐2 cells (ShPPAR) compared to control cells (ShLuc). LCT activity in ShPPAR cells compared to ShLuc cells stimulated by GED and CLA. The activity levels measured in control cells were arbitrarily defined as one.

4. Effect of GW9662 on GED‐dependent induction of LCT gene expression in Caco‐2 cells. LCT gene expression was determined by qPCR. Results represent the fold change of LCT gene expression. The expression level measured in control cells (w/o GED and GW9662) was used as reference and defined as 1.

5. LCT mRNA expression in the proximal small intestine of PPARγ^ΔIEC mice. Results represent the mean ± SD.

6. LCT mRNA expression and activity in Caco‐2 cells stimulated with fenofibrate compared to control cells (DMSO).

7. PPARα mRNA expression in the small intestine of PPARγ^ΔIEC mice (left) and in Caco‐2 ShPPARγ/ShLuc cells (right). For mice results, data represent the mean ± SD.

8. Effect of GW6471 on GED‐, CLA‐, Pio‐, and Rosi‐dependent induction of LCT gene expression in Caco‐2 cells. Results represent the fold change of LCT gene expression. The expression level measured in control cells was used as reference and defined as 1.

Data information: Data are expressed as mean ± SEM (two to four independent experiments) (except for panels A, E and mouse data shown in G). Statistical analysis: two‐tailed nonparametric Mann–Whitney U‐test. ***P < 0.0001; NS, not significant.

Figure EV4. LCT and PPARγ genes expression correlate in the rat duodenum and jejunum

1. A

   Comparison of the LCT and PPARγ mRNA levels along the gut of “not weaned” and “weaned” rats. Gene expression level was determined by qPCR of corresponding mRNA. Results represent the mean ± SD of the relative expression normalized to GAPDH level (for each group n = 6). *P < 0.05, **P < 0.01. Statistical analysis: two‐tailed nonparametric Mann–Whitney U‐test.

2. B, C

   Correlation between the LCT mRNA and PPARγ mRNA levels in the duodenum and jejunum of rats (not weaned and weaned).

Figure 3. LCT expression and activity in vivo in rodents following GED administration

1. A, B

   LCT gene expression (qPCR) and LCT activity were assessed in the proximal small intestine of weaned C57BL/6 mice (A) and Sprague Dawley rats (B) treated with oral GED (30 mg/kg) for 7 days. Results represent the sum of three independent experiments. Horizontal bars represent mean values. LCT activities in rats are expressed as the percentage of activity compared to that measured in control animals (arbitrarily defined as 100%).

2. C

   Stool consistency scores over time in weaned rats fed with lactose‐enriched diets. P‐values between lactose groups and control diet are indicated.

3. D

   Stool consistency score in rats with and without lactose‐enriched diet treated or not with GED. Results represent the sum of two independent experiments (n = 20 for each group).

4. E

   Cecum dilatation induced by lactose diet was improved by GED treatment. Photographs show representative pictures of cecum size and morphology in the various groups. Results of cecum weight for one experiment (n = 10 for each group) are represented in the dot plot graph. Horizontal bars represent mean values.

5. F

   Total SCFA concentration (mmol/l) in the cecal contents (n = 10 for each group). Horizontal bars represent mean values.

Data information: (C, D) Data are expressed as mean ± SEM. Statistical analysis: two‐tailed nonparametric Mann–Whitney U‐test.

Figure EV5. LCT expression and activity in rats following CLA administration

LCT gene expression (qPCR) and LCT activity were assessed in the proximal small intestine of weaned Sprague Dawley rats treated with oral CLA (200 mg/kg) for 5 days. Horizontal bars represent mean values (n = 10). Statistical analysis: two‐tailed nonparametric Mann–Whitney U‐test.