THE MECHANISM OF PPARN3 MEDIATED DOWN REGULATION OF SODIUM HYDROGEN EXCHANGER 1 (NHE1) GENE EPXRESSION AND ITS INHIBITION BY ESTROGEN RECEPTOR n1 2

Intrigued by the presence of putative PPRE on NHE1 promoter, we next assessed the effect of PPARγ ligands on NHE1 expression both at protein level and mRNA level.. if these synthetic PPA

3 RESULTS

3A PPARγ-MEDIATED REGULATION OF NHE1

3A.1 PPARγ AND THE EXPRESSION OF NHE1

It has been shown that activation of peroxisome proliferator-activated receptor γ

(PPARγ) inhibits proliferation of multiple cancer cells both in vitro and in vivo;

however, the downstream targets responsible for this anti-tumorigenic effect remain unidentified Our group previously identified a putative peroxisome proliferator response element (PPRE) in the promoter region of the Na+/H+

transporter gene NHE1 The first aim of this study was to investigate the role of

NHE1 in PPARγ-mediated anti-proliferative effect in breast cancer cells and the mechanism by which PPARγ regulates NHE1 expression

3A1.1 Identification of putative PPRE on NHE1 promoter

As a classical nuclear receptor, PPARγ has been shown to be involved in transcription regulation of various target genes The binding of the PPARγ to Peroxisome Proliferator Response Element (PPRE) either represses (Ricote et al., 1999) or activates (Barak et al., 1999) the target gene, depending on the cofactors recruited

To establish NHE1 as a bona fide target gene of PPARγ, we first examined the

5’-proximal promoter region of human NHE1 for any presence of potential Peroxisome Proliferator Response Element (PPRE) The DNA sequence of NHE1 promoter 5 kb upstream of the transcription start site was extracted from NCBI

http://www.ncbi.nlm.nih.gov (Accession number: L25272) Consensus PPRE typically consists of Direct Repeat (DR) of AGGTCA spaced by one nucleotide (DR1) However, binding of PPAR to direct repeat of AGGTCA spaced by two nucleotides (DR2) has also been found in the promoter of myeloperoxidase (MPO) gene (Kumar et al., 2004) DNA sequence analysis for potential DR1 and DR2 revealed a putative PPRE in the human NHE1 promoter region between –977 to –

990 upstream of TATA box (Figure 1A) The putative PPRE is located in a primate-specific Alu reporter response element (AluRRE), which is reported to be recognized by different nuclear receptors (Vansant and Reynolds, 1995)

Sequence alignment of AluRRE in NHE1 with that found in MPO revealed high sequence similarity (Figure 1B) The putative PPRE in NHE1 contains only one mismatch in nucleotide sequence compared to PPRE of MPO Though not optimal, binding of PPARγ to DR2 was previously demonstrated on MPO promoter (Kumar et al., 2004) High sequence similarity between MPO and NHE1 PPRE predicts possibility of PPARγ binding to DR2 in NHE1 promoter, and subsequent regulation of NHE1 gene expressions in a similar manner as MPO gene Besides high sequence similarity to DR2 in MPO, the hexamer sequence on NHE1 is identical to the first half of the consensus PPRE Alignment with consensus PPRE (Figure 1B) highlights the exact match in the first hexamer sequence of identified NHE1 PPRE to that in classical PPRE

A list of known PPAR target genes containing the sequence of their corresponding PPRE was compiled in Table 1 As shown, majority of PPRE from literature composes of general DR1 consensus of AGGTCA N AGGTCA with a

few mismatches in nucleotides This general pattern of 6-N-6 is relatively conserved in all PPREs, and the identified NHE1 PPRE bears high sequence similarities to these PPREs

In this section, we identified a putative PPRE –977 to –990 upstream of TATA box, after searching through NHE1 promoter for potential motifs of DR1 or DR2

(A)

-1344

GGAATCGCATATCAAGCTTTCCAGTGATTCCATTGTACAGCCATGATCCCTTGAACCTCACCAA -1280

TTTCAACCAAACTATAGGTTCAAATTTAAGTTCCACTACTTAAAGCATGCCACTGTTGTGGGTT -1216

GAATTGTGTCCCGGCAAAAGAAGTTGAAGTCCTAATGCCCAGTGCCTATGAAGATGGACTAATT -1152

AGGATGCAGTCTTTGAAGATGTTCAGGTTAAGATGAGGTAATTACGTTGGATTTCTAATCCAAT -1088

CCATGGAATACCAGATAGTCCTAACAAACCACTGGAAGGTAGGAGAAAGGCATGGGACAGATTC -0640

TCCCTCATAGCTCTCAGCTGAAACCAACCCTGCCAACACCTAGATCCGACCTCCAGCCTCCAGA -0576

ACTGTGAGACAATCAATTTCTGTTGTTGCAGCCACCCAGTTTGGGGTGATACTTTGTTACGGCA -0512

GCCCTAGTAAGCAATACAACTACTTGCATAGTAGCCAGGGGACTCTCTTCACCTGTTTCCTCAT -0448

CTGTAAAAGTGGAATTGTAATAATGTGCCAGGGTGCATTCCAAATAGTTTACACGGATTGTCTC -0384

AGTCATTACATCATCCCTCTGACATAGTCACTATTACTGTCTCTACTTAACAGATGAGAAAGTT -0320

GTGAAACAGGTTAAGTAACTTGCTCAAGGTCACACGGTAACTAAATACATAAACTAATAATACA -0256

TTCTTCACAGGATTATTCGAAAGCCCTTATGAGACTGCAGATGTGGACGTGAAATCGTTTTGTA -0192

AGTAGTCGGCATTTTACTCGCGTTAGTGAGGTTCTCTGTATATTCAGGACTTTTTTTTTTTTTT

Figure 1: Sequence Analysis of NHE1 Promoter

(A) Sequence of 5’ proximal promoter region 1344bp upstream of human NHE1

gene was retrieved from NCBI (Accession # L25272) Bold denotes Alu element

(Alusq) whereas TATA signal denotes start of gene The underlined sequences are

four hexamer repeats that are present in AluRRE (B) The AluRRE of NHE1

promoter is aligned with AluRRE of myeloperoxidase (MPO) (Kumar et al.,

2004).The putative PPRE in 5’ proximal promoter region of NHE1 is aligned with

PPRE of PTEN (Patel et al., 2001) and the consensus PPRE

AGGTCA

Peroxisomal enoyl-CoA

hydratase/3-hydroxyacyl-CoA dehydrogenase

AGGTCC T AGTTCA

Phosphoenolpyruvate

carboxykinse(PEPCK1)

CGGCCA A AGGTCA

3A1.2 Down-regulation of NHE1 by PPARγ ligands

Intrigued by the presence of putative PPRE on NHE1 promoter, we next assessed the effect of PPARγ ligands on NHE1 expression both at protein level and mRNA level To this end, three human breast cancer cell lines, MCF-7, MDA-MB-231 and T47D were selected to be exposed to various PPARγ ligands at different doses The three cell lines express different levels of endogenous PPARγ receptor (Figure 2D)

Prostaglandin J2 and its derivatives are reported to be activators of peroxisome proliferator-activated receptors α and γ (Kliewer et al., 1995) Among them, the PGJ2 metabolite 15-deoxy-12,14-PGJ2 has been identified to be the most potent endogenous ligand for PPARγ It is found to bind directly to PPARγ receptor and elicit efficient differentiation of C3H10T1/2 fibroblasts to adipocytes (Lehmann

et al., 1995) The effect of 15d-PGJ2 on NHE1 protein expression was analyzed using Western blot The result showed that after 24h of exposure to 1µM, 3µM and 5µM of 15d-PGJ2, the NHE1 protein decreased in a dose-dependent manner

in MCF-7 cells (Figure 2A)

Besides the endogenous PPARγ ligand, thiazolidinedione (TZD) is a class of synthetic PPARγ ligands which are clinically available for treatment of type 2 diabetes Members of TZDs include troglitazone, rosiglitazone, and pioglitazone, marketed as Rezulin, Avandia and Actos respectively Ciglitazone though not available in market, is a prototypical compound for the TZD class To investigate

if these synthetic PPARγ ligands produce similar effect of repressing NHE1 protein expression, MCF-7 cells were treated with troglitazone at increasing doses for 24h Immunoblotting of the NHE1 protein showed similar dose-dependent decrease in NHE1 protein level (Figure 2B) These data suggest that the effect of PPARγ ligands on NHE1 protein is not drug specific, but is conserved in both the endogenous and synthetic PPARγ ligands.The difference in extent of NHE1 repression by 15d-PGJ2 and synthetic PPARγ ligands may be attributed to the different affinities of these ligands for PPARγ receptor (Boitier et al., 2003)

To further demonstrate that the down-regulation of NHE1 protein expression was not cell type specific, MDA-MB-231 and T47D were treated with 3µM and 5µM

of 15d-PGJ2 for 24h In agreement with the results obtained for MCF-7, Western blot analysis of NHE1 protein showed similar dose-dependent down-regulation of NHE1 protein level in both cell lines (Figure 2E) Interestingly, the levels of PPARγ protein in these three cell lines are ranked in the order of MDA-MB-231> MCF-7> T47D (Figure 2D) This difference in PPARγ expression correlated well with the repressive efficacy of 15d-PGJ2 on NHE1 In T47D cells that contain lowest level of PPARγ receptor, 15d-PGJ2 showed the least pronounced effect on NHE1 repression, while in MDA-MB-231 cells which express the highest amount

of PPARγ, the same drug induced the most significant down-regulation of NHE1

To further verify that the reduction in NHE1 protein level was a result of transcriptional repression of NHE1 gene, we quantified the mRNA expression of NHE1 using real-time PCR Cells were treated with various PPARγ ligands for

Consistent with the results obtained for NHE1 protein expression, mRNA expression of NHE1 decreased in a concentration-dependent manner in MCF-7 cells treated with various PPARγ ligands It is noteworthy that 15d-PGJ2, as in the case of NHE1 protein expression, remained to be the most efficacious in down-regulating NHE1 mRNA compared to troglitzone and ciglitazone: 3µM of 15d-PGJ2 induced 55% decrease from the vehicle control (Figure 3A), whereas the similar extent of reduction was only achieved at 15µM of troglitazone and at 10µM of ciglitazone respectively (Figure 3B, C) In agreement with the result of NHE1 protein, NHE1 mRNA in MDA-MB-231 cells was also more sensitive to repression by 15d-PGJ2-induced PPARγ activation compared to MCF-7 cells In MCF-7 cells treated with 5µM of 15d-PGJ2, NHE1 mRNA level decreased to 40%

of the vehicle control, while the same concentration reduced NHE1 mRNA to 20%

of the vehicle control in MDA-MB-231 cells (Figure 3D) These data again demonstrated that the repressive efficacy of PPARγ ligand on NHE1 expression

in different cell lines mirror-imaged the endogenous level of PPARγ receptor present

Together, these results confirm that PPARγ agonists down-regulate NHE1 protein

as well as mRNA in three breast cancer cell lines

Figure 2: PPARγ ligands down-regulate NHE1 protein levels in human

breast cancer cells

MCF-7 (3 X105 cells/6-well dishes) were exposed to different PPARγ ligands: (A) 15d-PGJ2, (B) troglitazone for 24h, and the protein expression of NHE1 was analyzed by Western blot (C) MDA-MB-231 (2X105 cells/6-well dishes), and T47D cells (3 X105 cells/6-well dishes) were treated with increasing doses of 15d-PGJ2 NHE1 protein expression was then determined by Western blot, NHE1 band intensity was normalized to β-actin (D) MCF-7 (3 X105 cells/6-well dishes), MDA-MB-231 (2X105 cells/6-well dishes) and T47D cells (3 X105 cells/6-well dishes) were subjected to nuclear-cytosol fractionation as described in Materials and Methods The PPARγ levels in nuclear lysates were analyzed by Western blot, using PARP as a loading control

Figure 3: PPARγ ligands down-regulate NHE1 mRNA levels in human breast cancer cells

MCF-7 (3 X105 cells/6-well dishes) cells were exposed to increasing doses of PPARγ ligands: (A) 15d-PGJ2, (B) ciglitazone, (C) troglitazone for 16h, and the fold change of NHE1 mRNA expression was determined by Taqman real-time PCR, normalized to the endogenous control: human 18s Relative NHE1 mRNA expression is expressed as percentage of control Results denote means +/-SD

computed from two experiments done in duplicate (D) MCF-7 (3 X10 well dishes), MDA-MB-231 (2X105 cells/6-well dishes) and T47D cells (3 X105cells/6-well dishes) were exposed to increasing doses of 15d-PGJ2 for 24h, and the mRNA expression of NHE1 was analyzed by real-time PCR, normalized to the endogenous control: human 18s Relative NHE1 mRNA expression is expressed as percentage of control Results denote means +/-SD computed from two experiments done in duplicate *, p<0.05, **, p<0.01, treated versus untreated control

cells/6-3A1.3 Down-regulation of NHE1 by PPARγ ligands is PPARγ-dependent

The data so far demonstrated that cells that have higher level of PPARγ are more sensitive to PPARγ ligands both at protein and mRNA level As shown previously

in figure 2, In T47D cells, where PPARγ receptor was present at low level, the inhibitory effect of 15d-PGJ2 on NHE1 expression was significantly weaker as compared to MDA-MB-231 cells which express higher level of PPARγ receptor These results suggest that PPARγ could be responsible for ligand-induced down-regulation of NHE1 expression

PPARγ ligands were reported to have both PPARγ dependent and independent effects For instance, it was shown that PPARγ was required for positive effects of its ligands on lipid metabolism in macrophage, but not for its repressive effects on cytokine production and inflammation (Chawla et al., 2001) To further confirm that the inhibitory effect of PPARγ ligands on NHE1 was indeed due to the function of PPARγ receptor, we over-expressed murine PPARγ in T47D cells which expressed low endogenous level of PPARγ The effect of 15d-PGJ2 on NHE1 expression in T47D transfected with plasmid encoding murine PPAR

(mPPARγ) or with empty parent vector was assessed both at protein and mRNA level As expected, overexpression of PPARγ in T47D induced a more significant reduction in NHE1 protein expression compared to cells transfected with the vector control (Figure 4A of published paper) (Kumar et al., 2009) Consistent with the result obtained from Western blot of NHE1 protein, real-time PCR assay

of NHE1 mRNA also demonstrated a more pronounced decrease in NHE1 mRNA

in cells overexpressing PPARγ compared to cells transfected with the parent vector (Figure 4A) Interestingly, the NHE1 protein and mRNA level decreased upon PPARγ overexpression even in the absence of exogenous ligand treatment This could be explained by the presence of endogenous PPARγ ligands activating exogenous PPARγ in transfected cells, hence repressing the NHE1 transcription even in the absence of exogenous ligands

Our previous results demonstrated that both synthetic and endogenous ligands of PPARγ were capable of inducing down-regulation of PPARγ NHE1 and mRNA Among them, the inhibitory effect was the most pronounced in cell lines expressing higher level of PPARγ receptor These observations lead to the speculation of PPARγ receptor’s involvement in PPARγ ligand-induced down-regulation of NHE1 The data obtained from PPARγ- overexpressing T47D further confirmed the important role of functional PPARγ receptor in PPARγ ligand-induced repression of NHE1 protein and mRNA

Figure 4: Overexpression of PPARγ enhances the inhibition of 15d-PGJ2 on NHE1 expression

(A) T47D (2 X105 cells/6-well dishes) were transfected with 5µg plasmid encoding mouse PPARγ (mPPARγ) or empty pCMX (vector) as described in Materials and Methods 48h after transfection, cells were treated with increasing doses of 15d-PGJ2 Overexpression of PPARγ after 48h of transfection is shown at top panel After 16h of drug treatment, the mRNA expression of NHE1 was analyzed by real-time PCR, normalized to the endogenous control: human 18s Relative NHE1 mRNA expression is expressed as percentage of control Results denote means +/-SD computed from two experiments done in duplicate *, p<0.05

3A1.4 Silencing PPARγ abrogates the effect of PPARγ ligand on NHE1

Another means to assess whether the observed down-regulation of NHE1 expression was mediated by PPARγ receptor was by silencing PPARγ protein PPARγ silencing was performed in MDA-MB-231 cells as described in Materials and Methods with scrambled si RNA as a negative control or with PPARγ si RNA These two cell lines were chosen because they express higher level of endogenous PPARγ and showed greater sensitivity to PPARγ ligands on NHE1 repression After 48h of silencing, the cells were assayed using Western blot to verify PPARγ level In both cell lines, there was effective reduction in PPARγ receptor level (Figure 5) Transfected cells were then exposed to increasing dose of 15d-PGJ2

for 24h before they were harvested and checked for NHE1 expression using Western blot As expected, the cells silenced with negative si RNA showed down-regulation of NHE1 in both cell lines This observation was consistent with previous findings in figure 1 that PPARγ ligands repressed NHE1 in a concentration-dependent manner Interestingly, silencing PPARγ receptor drastically abrogated the inhibitory effect of 15d-PGJ2 on NHE1 expression in both MDA-MB-231 cells (Figure 5) However, the reversal of the down-regulation of NHE1 by silencing was more significant at 3µM than 5µM of 15d-PGJ2 In silenced cells, exposure to 3µM of 15d-PGJ2 showed no significant reduction in NHE1 protein level from the untreated control On the other hand, 3µM of 15d-PGJ2 drastically down-regulated NHE1 protein in cells transfected with negative si RNA It should also be noted that silencing PPARγ could not completely block the effect of 15d-PGJ2 on NHE1 expression at higher concentration of 5µM This phenomenon could be explained by incomplete

knocking down of PPARγ receptor and its stronger activation at higher concentration of the ligand, leading to the observed reduction in NHE1 protein from the untreated control The inhibition on NHE1 protein in PPARγ-silenced cells at higher concentration of 15d-PGJ2 could also be attributed to PPARγ-independent effect of the ligand The PPARγ-indepedent mechanism of PPARγ ligand-mediated inhibition on NHE1 will be further discussed in later sections

Figure 5: Silencing PPARγ attenuates the inhibition of 15d-PGJ2 on NHE1 expression

MDA-MB-231 (1.5 X 105 cells/6-well dishes) cells were transfected with 200ng

of PPARγ Si RNA or control Si RNA as described in Materials and Methods 48h after transfection, cells were treated with increasing doses of 15d-PGJ2 Reduced PPARγ expression after 48h of transfection is shown at top panel After 24h of drug treatment, NHE1 protein expression was then determined by Western blot NHE1 band intensity was normalized to β-actin

3A1.5 Pharmacologcial PPARγ antagonist abrogates the effect of PPARγ

To further test whether PPARγ activity is required for the ligand-induced regulation of NHE1 gene, PPARγ agonist, 2-chloro-5-nitrobenzanilide (GW9662) was used GW9662 functions as an antagonist of PPARγ by covalently modifying cysteine286 of the ligand binding domain, and hence preventing ligand-induced activation of PPARγ receptor Because the modification is irreversible, GW9662

down-is considered to be a non-competitive antagondown-ist of PPARγ Although GW9662 was reported to bind to all three PPAR subtypes, its inhibition on PPARγ is 100-

1000 times more than on PPARα and β (Leesnitzer et al., 2002) Henceforth, GW9662 is considered as a specific PPARγ antagonist

In order to confirm GW9662’s ability to inhibit PPARγ activation, MCF-7 cells were transfected with 3X PPRE-Luc or pTA-luc (control plasmid) together with

Renilla plasmid as a control for transfection efficiency Luciferase reporter assay

was then performed on transfected MCF-7 cells which were exposed to different doses of 15d-PGJ2 with or without 2h preincubation with GW9662 As expected, 15d-PGJ2 significantly up-regulated PPRE reporter activity in cells that were not exposed to GW9662 (Figure 6A), showing that the ligand was able to transactivate PPARγ in MCF-7 However, the presence of GW9662 prior to treatment of 15d-PGJ2 successfully attenuated PPARγ activity in MCF-7 cells Although 15d-PGJ2 was still able to activate PPRE reporter in the presence of GW9662, the magnitude of activation was significantly lower than that without the PPARγ antagonist (Figure 6A) This result confirms GW9662’s ability to inhibit PPARγ-mediated luciferase reporter activity, and its role as a pharmacological PPARγ antagonist Furthermore, the specificity of GW9662 as a

PPARγ antagonist also implies that PPRE reporter activity is largely driven by PPARγ but not other PPAR subtypes

After establishing the function of GW9662 as a PPARγ antagonist, we tested its ability to rescue the inhibitory effect of PPARγ ligand on NHE1 expression MDA-MB-231 cells were pre-incubated with GW9662 for 2h, before they were exposed to increasing doses of 15d-PGJ2 Western blot and real-time PCR assays were then performed to assess the NHE1 protein and mRNA level respectively In accordance with our previous findings, treatment with 15d-PGJ2 resulted in a dose-dependent down-regulation of NHE1 protein and mRNA However, this inhibitory effect by PPARγ ligand was abrogated when cells were pre-incubated with 10µM of GW9662 (Figure 6B, C) In cells treated with 15d-PGJ2 alone, 5µM

of the drug substantially reduced NHE1 mRNA expression The same concentration albeit was not able to produce a significant decrease from the control in the presence of the PPARγ antagonist (Figure 6C)

Taken together, the above data conclusively demonstrated that GW9662 blocked PPARγ activation and transcriptional activity by functioning as a PPARγ antagonist Furthermore, the suppressive effect of PPARγ ligand on NHE1 expression could be rescued by the presence of the functional PPARγ antagonist GW9662

Figure 6: PPARγ inhibitor abrogates the effects of 15d-PGJ2 on PPARγ activity and on NHE1 expression

(A) MCF7 cells (7.5 X 104 cells/12-well dishes) were co-transfected with 3µg of reporter plasmid 3XPPRE-luc and 0.3µg of renilla as described in Materials and Methods 48h after transfection, cells were exposed to increasing doses of 15d-PGJ2 for 5h with or without 2h preincubation with 15µM GW9662 The activity

of PPARγ was then determined using luciferase assay and the result was calculated as luciferase RLU/renilla/µg total Data represents the average +/- SD

of three experiments (B) MDA-MB-231 (2X105 cells/6-well dishes) cells were treated with increasing doses of 15d-PGJ2 for 24h with or without 2h preincubation with 10µM GW9662 NHE1 protein expression was then determined by Western blot NHE1 band intensity was normalized to β-actin (C) MCF-7 (3X105 cells/6-well dishes) cells were treated with increasing doses of 15d-PGJ2 for 16h with or without 2h preincubation of 15µM GW9662 The mRNA expression of NHE1 was analyzed by real-time PCR, normalized to the endogenous control: human 18s Relative NHE1 mRNA expression is expressed

as percentage of control Results denote means +/-SD computed from two experiments done in duplicate *, p<0.05, **, p<0.01, treated versus untreated control

3A.2 THE MECHANISM OF PPARγ-MEDIATED

DOWN-REGULATION OF NHE1

After establishing the crucial role of activated PPARγ receptor in down-regulating NHE1 expression, we set out to investigate the mechanism involved in the repression of NHE1 by PPARγ

3A2.1 Transcription-defective PPARγ abrogates the effect of PPARγ ligand

on NHE1 gene expression

From previous sections, we used PPARγ overexpression, PPARγ silencing and PPARγ antagonist to confirm the role of PPARγ receptor in PPARγ ligand-