Research in Veterinary Science 93 (2012) 770–775

Contents lists available at SciVerse ScienceDirect

Research in Veterinary Science

journal homepage: www.elsevier .com/locate / rvsc

cDNA cloning and mRNA expression of canine pancreatic and duodenumhomeobox 1 (Pdx-1) q

Hiroshi Takemitsu, Ichiro Yamamoto, Peter Lee, Taizo Ohta, Nobuko Mori, Toshiro Arai ⇑Department of Veterinary Science, School of Veterinary Medicine, Nippon Veterinary and Life Science University, 1-7-1 Kyonancho, Musashino, Tokyo 180-8602, Japan

a r t i c l e i n f o a b s t r a c t

Article history:Received 9 November 2010Accepted 2 November 2011

Keywords:CanineInsulinPdx1BETA2MafAqRT-PCR

0034-5288/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.rvsc.2011.11.003

q This sequence data have been submitted to the DDunder Accession No. HQ454506.⇑ Corresponding author. Tel.: +81 422 31 4151; fax

E-mail address: tarai@nvlu.ac.jp (T. Arai).

Pancreatic and duodenal homeobox 1 (Pdx-1) is a critical insulin transcription factor expressed by pan-creatic b-cells, and is crucial in the early stage of pancreas development. Unfortunately, nothing concern-ing Pdx-1 in canine has been elucidated yet. In this study, full length canine Pdx-1 cDNA was cloned and itwas 1498 bp in length, consisting of a 99 bp 50-untranslated region (UTR), a 849 bp coding region, and a550 bp 30-UTR region. A deduced 282 amino acid sequence of canine PDX-1 displayed high overallsequence identity with human, bovine, and mouse PDX-1. qRT-PCR analysis revealed that a high levelof Pdx1 mRNA expression is exists in the duodenum and pancreas of canines. In addition, functionalcanine insulin promoter–luciferase reporter constructs with various canine insulin promoter region frag-ments revealed that our Pdx-1 isolated cDNA sequence encodes for a functionally active PDX-1 protein.Significant promoter activity was observed within the �583 bp 50-upstream region of canine insulin genewith Chinese hamster ovary cells. In addition, Pdx-1 appears to have a synergistic effect with beta celltransactivator 2 (BETA2) and V-maf avian musculoaponeurotic fibrosarcoma oncogene homolog A(MafA), which also have important roles in the activation of the insulin gene promoter. Our results con-firm that similar to humans, Pdx1 plays an important role in expression of insulin gene in canines.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Type 1 diabetes mellitus (T1DM) is an autoimmune diseasecharacterized by the development of autoantibodies and destruc-tive T-cell infiltration of insulin-producing islet b-cells (Gillespie,2006). Insulin deficiency due to reduced pancreatic islet b-cellnumber underlies the progression of both T1 and T2DM, promptingefforts to develop b cell replacement therapies. Recently, it hasbeen reported in humans that pancreatic and duodenal homeobox1 (Pdx-1) is a novel pancreatic b-cell-specific autoantigen recog-nized by both B- and T-cells in NOD mice (Li et al., 2010).

PDX-1 (also known as insulin promoter factor 1 (IPF1)/islet duo-denal homeobox 1 (IDX-1)/somatostatin transcription factor 1(STF1)) (Kaneto et al., 1997) is a transcription factor initially ex-pressed in both exocrine and endocrine cells (Kim and Hebrok,2001), and is important and crucial in the early stage of pancreasdevelopment. At an early stage of embryonic development, Pdx-1is initially expressed in the gut region when the foregut endodermbecomes committed to common pancreatic precursor cells. This ex-plains why Pdx-1 is also expressed in the regions of developing

ll rights reserved.

BJ/EMBL/GenBank databases

: +81 422 31 7841.

stomach and duodenum. However, in mature pancreas, Pdx-1expression is restricted primarily to insulin-producing islet b-cells,with expression in some somatostatin-producing d-cells, and low-level expression in subpopulations of acinar and duct cells (Guzet al., 1995). PDX-1 is necessary for normal pancreatic developmentand islet b-cell differentiation and maturation. In fact, Pdx1�/�

knockout mice have demonstrated a loss of pancreas formation inaddition to a loss of Brunner’s glands and deficiency of enteroendo-crine differentiation in the stomach and duodenum (Miller et al.,1994). Whereas, heterozygous Pdx1+/� mice exhibit an age-depen-dent worsening of glucose tolerance, reduced glucose-stimulatedinsulin release and higher susceptibility to apoptosis (Ahlgrenet al., 1998).

Coincidentally, T1DM is the most common type of DM encoun-tered in canines (Rand et al., 2004). However, very little has beenelucidated concerning PDX-1 in canines. Hence, we sought todetermine the existence of a canine PDX-1 ortholog by cloningthe full length canine Pdx-1 cDNA and deducing its amino acid se-quence for comparison against other known species. We alsosought to determine where Pdx-1 is normally expressed in theadult canine by performing qRT-PCR on various canine tissues.Lastly, we wanted to determine and characterize canine PDX-1’smolecular mechanism of action and association with the insulinpromoter. In order to do this, we performed a luciferase expressionpromoter assay using various sized portions of the canine insulin

H. Takemitsu et al. / Research in Veterinary Science 93 (2012) 770–775 771

gene promoter region, allowing PDX-1 to interact with two otherimportant insulin transcription factors, beta cell transactivator 2(BETA2) and V-maf avian musculoaponeurotic fibrosarcoma onco-gene homolog A (MafA).

2. Materials and methods

2.1. Cloning of canine Pdx1 cDNA

Canine pancreas tissue sample was purchased from Zyagen (SanDiego, California) 2 years-old male beagle canine. Total RNA wasextracted using TRIzol (Invitrogen, Carlsbad, CA) and quantifiedusing a spectrophotometer at 260 nm. Poly A + RNA was preparedfrom total RNA using Oligo-Tex dT30 Super (Takara, Shiga, Japan)according to the manufacturer’s instructions, and used to preparea cDNA library using the SMARTer RACE cDNA Amplification Kit(Clontech, Mountain View, CA). A full length fragment of caninePdx-1 cDNA was amplified using primers pdx1-cDNA-F and pdx1-cDNA-R (Table 1), designed from the predicted sequence for caninePdx-1 mRNA (GenBank Accession No. XM_543155) and an adapterprimer supplied by the manufacturer. PCR was performed at 30 cy-cles of 98 �C for 10 s, 60 �C for 30 s, and 72 �C for 60 s with Ex TaqDNA polymerase (Takara), and 0.2 lM each of primers. The ampli-fied fragment was cloned into a T-Vector pMD19 (Takara) andsequencing was carried out with the BigDye Terminator v3.1 CycleSequencing kit (Applied Biosystems, Foster City, CA) and ABI PRISM310 Genetic Analyzer (Applied Biosystems).

2.2. Quantitative real-time PCR (qRT-PCR) analysis of Pdx-1 in variouscanine tissues

Total RNA was obtained from Zyagen (San Diego, California). To-tal RNA (1 lg) was reverse-transcribed at 42 �C for 15 min in 20 ll ofQuantitact (Qiagen Hilden Germany). After inactivating the reversetranscription reaction by heating at 95 �C for 3 min, the cDNA prod-uct was used for qRT-PCR. Reactions were carried out with PerfectReal Time SYBR Premix Ex Taq II (Takara) using an ABI 7300 RealTime PCR system Sequence Detection System (Applied Biosystems)using the following Shuttle PCR protocol: 95 �C for 30 s; followed by40 cycles of 95 �C for 5 s and 60 �C for 34 s, in 20 ll reaction volumescontaining: 2 ll template cDNA, 0.8 ll (0.4 ll) of RT-pdx-F (sense)and RT-pdx1-R (antisense) primers designed from cloned Pdx-1 se-quence (GenBank Accession No. HQ454506), 10 ll of SYBR PremixEx taq (Takara) and 0.4 ll ROX reference Dye II (Takara) and 6.0 llof distilled water. After qRT-PCR amplification, absolute quantifica-tion was performed according to the method of Whelan et al. (2003),by establishing a linear amplification curve from 10-fold serial dilu-tions of cloned and sequenced plasmid DNA containing canine Pdx-1

Table 1Primers used for gene cloning.

Name 50–30

beta2-1 CACCATGACCAAATCGTACAGCGAGAbeta2-2 CTAGTCGTGGAAGATGGCATTmafa-1 CACCATGGCCGCGGAGCTGGCGATmafa-2 GCGCTCACAGGAAGAAGTCGpdx1-cDNA-F CGACGCAGCTCTACAAGGACpdx1-cDNA-R GTATTTGTTGAACAGGAACTCCTTCpdx1-Prot-F CACCATGAACAGCGAGGAGCAGTTpdx1-Prot-R CCATCTCTCACCGGGGCTCCTRT-pdx1-F TCCCGTGGATGAAGTCTACCRT-pdx1-R CGTGGCCTCGAGATGTATTTins-1 AAGCTTGGGACAGGCAGGGTTGAGins-2 AGATCTCCCAGCACTGGGGAAATGATCCAins-3 AGATCTTTGGCCCCAGCTGTTAGTTGGins-4 AGATCTCCTCATGGTCTTGCACCATCCT

cDNA. Each value of mRNA expression was calculated and expressedas copies (copies/ng of cDNA input).

2.3. Construction of the expression vector for canine Pdx1, BETA2, andMafA

Using a pcDNA3.2/V5/GW/D-TOPO expression kits (Invitrogen),mammalian cell based expression vectors for PDX-1, BETA2 andMafA were generated by subcloning fragments generated withgene specific primers, pdx1-Prot-F, pdx1-Prot-R, beta2-1, beta2-2,mafa-1 and mafa-2, respectively (Table 1). Canine Pdx-1 was de-signed our cloning result (GenBank Accession No. HQ454506). Ca-nine BETA2 (GenBank Accession No. XM_545553) and MafA(GenBank Accession No. AC197467) sequences were found in thecanine genome database. Cloned canine Pdx-1, BETA2 and MafAexpression vectors were sequenced and confirmed the mRNAexpression in cultured cell line.

2.4. Canine insulin gene promoter fragment construction for use in aluciferase reporter gene assay

Different canine insulin promoter fragment sequence lengthswere generated by PCR using ins1–4 primers (Table 1) and ExTaq PCR Hot Start Version DNA polymerase (Takara). The ins-1 pri-mer has a Hind III restriction site whereas ins2–4 primers have aBgl II restriction site; hence canine insulin promoter PCR fragmentswere cloned into the pGL3-Basic luciferase reporter vector be-tween Hind III and Bgl II restriction sites (Promega, Madison, WI,USA) and sequenced. The �318 bp length fragment contains E1and C1 elements, the �411 bp length fragment contains E1, C1and A1 elements, and the �583 bp length fragment contains E1,C1, A1 and A3 elements.

2.5. Luciferase reporter assay in Chinese hamster ovary cells

Chinese hamster ovary (CHO-K1) cells were maintained in F-12Ham’s medium supplemented with 10% fetal bovine serum,100 units/ml penicillin G, 100 lg/ml streptomycin, and 0.25 lg/mlamphotericin B at 37 �C under 5% CO2. CHO-K1 cells were plated at1 � 104/well in 96 well plates and transfected with 200 ng of DNA(pGL3-Basic + 318 bp or pGL3-Basic + 411 bp or pGL3-Basic +583 bp) using Lipofectamine 2000 (Invitrogen) according to themanufacturer’s instructions. The pRL-TK vector was included in alltransfections to allow normalization for transfection efficiency.Luciferase activity was measured and defined as the ratio of fireflyto Renilla luciferase luminescence and was further normalizedagainst luciferase activity from cells transfected with the pGL3-Basicconstruct (Promega Dual-Luciferase Reporter Assay System).

Direction Position Gen Bank No.

Sense 1 XM_545553Anti-sense 1071 XM_545553Sense 1 AC197467Anti-sense 1041 AC197467Sense 128 NW_876278Anti-sense 606 NW_876278Sense 1 HQ454506Anti-sense 849 HQ454506Sense 451 HQ454506Anti-sense 620 HQ454506Anti-sense �5 NW_876266.1Sense �318 NW_876266.1Sense �411 NW_876266.1Sense �583 NW_876266.1

772 H. Takemitsu et al. / Research in Veterinary Science 93 (2012) 770–775

2.6. Statistical analysis

Values were expressed as mean ± SD values. Holm-Sidak OneWay ANOVA was utilized to perform multiple group comparison.Statistical significance was set at P < 0.05 for both the U-test andOne Way ANOVA. All tests were conducted using Sigmaplot analy-sis software (Sigmaplot 11.0, Build 11.0.077; Systat Software Inc.,San Jose, CA).

3. Results

3.1. cDNA and genomic structure of canine PDX-1

Canine Pdx-1 cDNA was successfully cloned from a canine pan-creas cDNA library (Fig. 1). The cDNA consisted of a 99 bp 50-untranslated region (UTR), an 849 bp coding region, and a 550 bp30-UTR, which translated into a 282 amino acid protein sequence.The deduced canine PDX-1 protein sequence displayed a high over-all sequence identity with PDX-1 orthologs from human (92.9%),bovine (90.0%), mouse (87.2%), chicken (60.6%), and zebrafish(44.7%). The homeodomain (amino acid residues 143–209) wasconserved in canine Pdx-1, and it exhibited high homology withthe corresponding region in other vertebrates. In addition, similarto that the other vertebrates, the canine PDX-1 homeodomainhad helix 1, 2 and 3 motifs.

3.2. Pdx-1 mRNA expression profile in canine tissues

Canine Pdx-1 mRNA expression levels in various tissues (namely,tissue of cerebral cortex, duodenum, heart, kidney, liver, pancreas,skeletal muscle, spleen, and stomach) derived from a 2 year oldmale beagle were profiled by qRT-PCR (Fig. 2). A high expressionlevel was observed in four tissues in particular, with expression inthe order of highest to lowest being as follows: duodenum >

Fig. 1. Multiple sequence alignment of PDX-1 amino acid sequences from dog, human (Nzebrafish (NP_571518). Black shaded background indicates 100% of homology betweenspecies. The homeodomain is underlined and the helix motifs are indicated with a box,

pancreas > stomach > liver. In fact, Pdx-1 mRNA expression in theduodenum (25.3 ± 0.8 copies/ng total RNA) was three-fold higherthan that in the pancreas (8.3 ± 1.5 copies/ng total RNA), four-foldhigher than that in the stomach (5.9 ± 0.8 copies/ng total RNA),and six-fold higher than that the liver (4.3 ± 2.0 copies/ng totalRNA). A low level of Pdx-1 mRNA expression was observed in, heart,kidney, spleen, cerebral cortex, and skeletal muscle tissues.

3.3. PDX-1 binding with canine insulin gene promoter – luciferasereporter assay

In order to test the ability of recombinant PDX-1 to induce insu-lin expression, we fused different length fragments of the canineinsulin promoter region, which contained various transcriptionalbinding elements, to a luciferase reporter. Subsequently, a combi-nation of three known insulin TFs (BETA2, PDX-1, and MafA) wasallowed to interact with the luciferase gene construct (Fig. 3).The highest level of luciferase expression was observed with the�583 bp promoter fragment whereas significant (Holm-SidakOne Way ANOVA, P < 0.05) 25% and 80% reductions were observedin the �411 and �317 bp fragments, respectively. The �317 bpinsulin promoter fragment contained binding elements for BETA2and MafA. The �411 bp fragment was similar to the �317 bp frag-ment with the addition of 1 of the 2 known for canine PDX-1 bind-ing elements. Lastly, the �583 bp fragment was similar to the�317 bp fragment with the addition of both the canine PDX-1binding elements.

Our results demonstrate that recombinant canine PDX-1 exhib-its transcriptional activity with the canine insulin gene promoter.Moreover, the A1 and A3 elements found in the canine insulin pro-moter region are both important and critical for enhanced insulintranscription by PDX-1. The highest level of PDX-1 induced expres-sion occurred when both A1 and A3 elements were presentwhereas it was reduced when only the A1 element was present.

P_000200), bovine (XP_583722), mouse (NP_032840), chicken (XP_001234636) andall species, whereas gray shaded backgrounds indicates >50% homology between

respectively.

Fig. 2. Measurement of Pdx-1 mRNA expression levels in various dog tissues asdetermined by quantitative RT-PCR. High levels of Pdx-1 mRNA expression weredetected in duodenum, pancreas, stomach and liver.

H. Takemitsu et al. / Research in Veterinary Science 93 (2012) 770–775 773

3.4. Association of PDX-1 with BETA2 and MafA transcription factorbinding with the canine insulin gene promoter – luciferase reporterassay

For a better understanding transcriptional activity of PDX-1’s,we compared and contrasted different combinations of TFs (BETA2,MafA, and PDX-1) against various insulin gene promoter regionbackgrounds in order to determine whether there was any synergybetween the TFs (Fig. 4). With the �318 bp insulin promoter frag-ment, MafA as alone resulted in significantly higher luciferaseexpression than BETA2 alone. Moreover, the combination of MafAand BETA2 resulted in higher luciferase expression compared witheither of the TFs alone. Pdx-1 had no effect on luciferase expres-sion, nor did it exhibits any synergistic effect on either BETA2 orMafA. No difference in luciferase expression was observed whenall three TFs were combined (5.67 ± 00.93) as compared with towhen BETA2 and MafA were used together (6.26 ± 0.78).

With the �411 bp insulin promoter fragment, PDX-1 alone didnot appear to have any significant effect because it only marginally

Fig. 3. Promoter activity of the 50-flanking region of the dog insulin gene. CHO-K1 cellsfragments of the 50-flanking region of the dog insulin gene (�317 or �411 or �583 bp)TATA designates TATA box, E1 designates BETA2 binding site, C1 designates MafA bindinindependent experiments. Non-parametric statistical analysis was performed by KruskallP < 0.05. Single asterisk designates significance as compared to control luciferase expressusing �317 bp insulin promoter fragment. Triple asterisk designates significance when

increased luciferase expression as compared to baseline value (noTFs) (3.77 ± 0.45 versus 2.63 ± 0.29). However, PDX-1 had a signif-icant synergistic effect in conjunction with BETA2, but not withMafA, on luciferase expression compared with BETA2 and MafAalone. Surprisingly, the degree of luciferase expression inducedby MafA alone (15.03 ± 1.28) was equal in range to the various dualtranscription activator combinations tested (PDX-1 + BETA2, PDX-1 + MafA, and BETA2 + MafA). The highest luciferase expression le-vel was observed with the �411 bp insulin promoter fragment(26.09 ± 2.12) when all three TFs were used together.

Finally, compared to that with the �411 bp fragment, the�583 bp insulin promoter fragment did not alter the luciferaseexpression induced by BETA2 or MafA alone. However, the effectof PDX-1 alone as well its synergistic effect appeared to be strongerwhen it was used in combination with either BETA2 or MafA. Sur-prisingly, the luciferase expression level induced by PDX-1 + BETA2was comparable to that of BETA2 + MafA (20.01 ± 0.27 versus19.00 ± 1.75). The highest luciferase expression level was observedwith the �538 bp insulin promoter fragment (35.08 ± 2.71) whenall three TFs were used together.

Taken together, these data demonstrate that when the three TFswere used individually. MafA had the strongest induction of lucif-erase activity. In addition, MafA exerted an additive effect onBETA2. However, PDX-1 appeared to have a strong synergistic ef-fect with BETA2, but only an additive effect with MafA. The stron-gest level of luciferase expression occurred with all three TFs intandem on a �583 bp insulin gene promoter region background.

4. Discussion

Canine Pdx-1 cDNA was successfully cloned from a canine pan-creas cDNA library. The full length cDNA sequence consisted of a99 bp 50-untranslated region (UTR), an 849 bp coding region, anda 550 bp 30-UTR, which translated into a 282 amino acid protein se-quence. The deduced canine PDX-1 protein sequence displayed ahigh overall sequence identity with PDX-1 orthologs from otherspecies such as human, bovine, and mouse. The PDX-1 homeodo-main consists of three a-helical regions separated by loop and turn(Singh et al., 1991). The most highly conserved region of thehomeodomain is amino acid motif ‘‘KIWFQN’’ within the DNArecognition helix 3 (Scott et al., 1989). A computer-assisted searchof our canine Pdx-1 cDNA sequence within the canine genome

were transfected with luciferase reporter gene constructs containing various lengthand three insulin transcription factor expression vectors (PDX-1, BETA2 and MafA).g site, and A1 and A3 designates PDX-1 binding sites. Values are means ± SD of four–Wallis One Way ANOVA (Student–Newman–Keuls Method) with significance set ation. Double asterisk designates significance when compared to luciferase expressioncompared to luciferase expression using �411 bp insulin promoter fragment.

Fig. 4. Comparison of insulin promoter activity according to individual versusvarious combinations of transcription factors (TFs). CHO-K1 cells were transfectedwith luciferase reporter gene constructs containing various length fragments of the50-flanking region of the dog insulin gene (�317 or �411 or �583 bp) and singleinsulin transcription factor expression vectors (PDX-1 or BETA2 or MafA) orcombinations thereof. Non-parametric statistical analysis was performed byKruskall–Wallis One Way ANOVA (Student–Newman–Keuls Method) with signif-icance set at P < 0.05. Single asterisk designates significance as compared to anysingle transcription factor based luciferase expression. Double asterisk designatessignificance when compared to any dual transcription factor combination basedluciferase expression.

774 H. Takemitsu et al. / Research in Veterinary Science 93 (2012) 770–775

database (www.ncbi.nlm.nih.gov/BLAST) showed the presence ofan approximately 3.5 kb intron located between positions 498and 499 bp in the Pdx-1 cDNA, indicating that the canine Pdx-1consists of two exons, which is similar to humans (Inoue et al.,1996).

The predicted canine PDX-1 amino acid sequence included 75amino acid residues in the homeodomain which is highly con-served among mammals and birds (100%) (Singh et al., 1991). Inthe homeodomain, the so-called helix 1–3, and in particular helix3 interacts and binds with a PDX-1 binding site ‘‘TAAT motif’’ foundon insulin or somatostatin promoter regions (Peshavaria et al.,1994). This finding indicates that the canine Pdx1 has importantrole as in other animals.

Pdx-1 mRNA expression level was highest in the duodenum andwas moderate in the pancreas, stomach, and liver. High mRNAexpression level in the duodenum is considered to be related toCalbindin-D9k expression. Calbindin-D9k is an intestinal vitaminD-dependent Ca2+-binding protein with a PDX-1 binding site (Bar-ley et al., 1999). Pancreatic expression of PDX-1 is considered to beindicative of pancreatic b-cell activity. Pdx-1 expression in thestomach and liver is considered to be related to the formation ofBrunner’s glands in the stomach and transcriptional activation ofheparin-binding EGF-like growth factor in the liver (Campbelland Macfarlane, 2002; Fujitani et al., 2006).

Pancreatic b-cell-specific insulin gene expression is regulated bya variety of pancreatic TFs. Indeed, PDX-1 binding to the ‘‘TAAT mo-tif’’ (A element in humans, and A1 and A3 elements in canines) (Ohn-eda et al., 2000) and BETA2 binding to the GCCATCTGC motif (E1element) (Ray et al., 2003) are crucial for insulin gene transcription(Babu et al., 2008). Moreover, recently AAATTGCAGCCTCAGCCTCmotif (C1 element) binding TF was identified as MafA, which is abasic-leucine zipper TF that functions as a potent transactivatorfor the insulin gene (Matsuoka et al., 2003). Within the human insu-lin promoter region, the E1, C1, A1, and A3 elements are located atpositions �364/�355, �350/�341, �320/�317 and �455/�452,respectively (Matsuoka et al., 2003). On the other hand, E1, C1, A1,

and A3 elements within the canine insulin promoter are located atpositions �275/�266, �296/�277, �370/�367 and �475/�472,respectively. These conserved elements of A3, C1 and E1 in the insu-lin promoter are essential for the activation of the insulin gene.

In our study, we used several canine insulin promoter regionfragments of different lengths for a luciferase reporter assay withrecombinant PDX-1, BETA2, and MafA. Recombinant canine PDX-1 demonstrated strong transcriptional activity with our 411 and583 bp insulin promoter fragment – luciferase constructs. Lucifer-ase expression was considerably higher in the �538 bp constructsince it contained both A1 and A3 elements whereas the �411 bpconstruct only contained the A1 element. In comparison to the�318 bp construct (contains BETA2 and MafA but no PDX-1 bind-ing elements), luciferase expression in the 411 and 538 bp con-structs were 4.5 and 6 times higher, respectively, thusillustrating the critical importance of PDX-1 for insulin production.In addition, our luciferase reporter construct assays also revealedthat PDX-1 associations with other TFs such as BETA2 and MafAare important and different. For example, the effect of PDX-1 withBETA2 was stronger than with MafA. In fact, the association be-tween PDX-1 and BETA2 appeared to be quite synergistic whereasthe PDX-1 appeared to have an additive effect on MafA. This con-tradicts what has been previously reported in a mouse study citingstrong cooperation between PDX-1 and MafA TFs (Zhao et al.,2005). We can speculate that this contradiction may be due to dif-ferences inherent in the canine insulin promoter sequence from ro-dent or human counterparts. In addition, BETA2 can synergize witheither PDX-1 or MafA resulting in similar levels of luciferaseexpression. However, the level of luciferase expression signifi-cantly dropped when PDX-1 and MafA were allowed to interactwith each other suggesting that PDX-1 and MafA may share a sim-ilar synergistic mechanism on BETA2.

5. Conclusions

Just as is the case with humans, canine PDX-1 appears to play avery important role in pancreatic b-cell insulin production. Its roleas a TF and its associations with other important insulin TFs suchas BETA2 and MafA appear to be critical for successful insulin pro-duction. In future studies, we hope to clone and examine PDX-1from Type 1 DM suffering dogs more in depth in an attempt todetermine if mutations exist which prevent or inhibit successfulbinding to A1 and A3 binding elements in the insulin promoter re-gion. We would also like to examine mRNA expression of PAX-1 inType 1 DM animals also since it has been reported that under Type2 diabetic conditions, chronic hyperglycemia gradually deterio-rates pancreatic b-cell function, which is accompanied by de-creased expression and/or DNA binding activities of MafA andPDX-1 (Zhang et al., 2005). Interestingly though, MafA overexpres-sion with PDX-1 and BETA2 markedly induces insulin biosynthesisin various non-b-cells and thereby might be a useful tool to effi-ciently induce insulin-producing surrogate b-cells. In addition,treatment of streptozotocin-induced mouse diabetes (Type 1) withrecombinant PDX-1 can reverse diabetes by stimulating b-cellregeneration and liver cell reprogramming into insulin-producingcells.

Acknowledgements

This work was supported in part by the Strategic Research BaseDevelopment Program for Private Universities from the Ministry ofEducation, Culture, Sports, Science and Technology of Japan(MEXT), 2008–2012, and Grant-in-Aid for Scientific Research (No.21380195 to T. Arai) from the MEXT.

H. Takemitsu et al. / Research in Veterinary Science 93 (2012) 770–775 775

References

Ahlgren, U., Jonsson, J., Jonsson, L., Simu, K., Edlund, H., 1998. Beta-cell-specificinactivation of the mouse Ipf1/Pdx1 gene results in loss of the beta-cellphenotype and maturity onset diabetes. Genes and Development 12, 1763–1768.

Babu, D.A., Chakrabarti, S.K., Garmey, J.C., Mirmira, R.G., 2008. Pdx1 and BETA2/NeuroD1 participate in a transcriptional complex that mediates short-rangeDNA looping at the insulin gene. Journal of Biological Chemistry 283, 8164–8172.

Barley, N.F., Prathalingam, S.R., Zhi, P., Legon, S., Howard, A., Walters, J.R., 1999.Factors involved in the duodenal expression of the human calbindin-D9k gene.Biochemical Journal 341, 491–500.

Campbell, S.C., Macfarlane, W.M., 2002. Regulation of the pdx1 gene promoter inpancreatic beta-cells. Biochemical and Biophysical Research Communications299, 277–284.

Fujitani, Y., Fujitani, S., Boyer, D.F., Gannon, M., Kawaguchi, Y., Ray, M., Shiota, M.,Stein, R.W., Magnuson, M.A., Wright, C.V., 2006. Targeted deletion of acid-regulatory region reveals differential gene dosage requirements for Pdx1 inforegut organ differentiation and pancreas formation. Genes and Development20, 253–266.

Gillespie, K.M., 2006. Type 1 diabetes: pathogenesis and prevention. CanadianMedical Association Journal 175, 165–170.

Guz, Y., Montminy, M.R., Stein, R., Leonard, J., Gamer, L.W., Wright, C.V., Teitelman,G., 1995. Expression of murine STF-1, a putative insulin gene transcriptionfactor, in beta cells of pancreas, duodenal epithelium and pancreatic exocrineand endocrine progenitors during ontogeny. Development 121, 11–18.

Inoue, H., Riggs, A.C., Tanizawa, Y., Ueda, K., Kuwano, A., Liu, L., Donis-Keller, H.,Permutt, M.A., 1996. Isolation, characterization, and chromosomal mappingof the human insulin promoter factor 1 (IPF-1) gene. Diabetes 45, 789–794.

Kaneto, H., Miyagawa, J., Kajimoto, Y., Yamamoto, K., Watada, H., Umayahara, Y.,Hanafusa, T., Matsuzawa, Y., Yamasaki, Y., Higashiyama, S., Taniguchi, N., 1997.Expression of heparin-binding epidermal growth factor-like growth factorduring pancreas development. A potential role of PDX-1 in transcriptionalactivation. Journal of Biological Chemistry 272, 29137–29143.

Kim, S.K., Hebrok, M., 2001. Intercellular signals regulating pancreas developmentand function. Genes and Development 15, 111–127.

Li, S.W., Koya, V., Li, Y., Donelan, W., Lin, P., Reeves, W.H., Yang, L.J., 2010. Pancreaticduodenal homeobox 1 protein is a novel b-cell-specific autoantigen for type Idiabetes. Laboratory Investigation 90, 31–39.

Matsuoka, T.A., Zhao, L., Artner, I., Jarrett, H.W., Friedman, D., Means, A., Stein, R.,2003. Members of the large Maf transcription family regulate insulin genetranscription in islet beta cells. Molecular and Cellular Biology 23, 6049–6062.

Miller, C.P., McGehee Jr., R.E., Habener, J.F., 1994. IDX-1: a new homeodomaintranscription factor expressed in rat pancreatic islets and duodenum thattransactivates the somatostatin gene. EMBO Journal 13, 1145–1156.

Ohneda, K., Ee, H., German, M., 2000. Regulation of insulin gene transcription.Seminars in Cell and Developmental Biology 111, 227–233.

Peshavaria, M., Gamer, L., Henderson, E., Teitelman, G., Wright, C.V., Stein, R., 1994.XIHbox 8, an endoderm-specific Xenopus homeodomain protein, is closelyrelated to a mammalian insulin gene transcription factor. MolecularEndocrinology 8, 806–816.

Rand, J.S., Fleeman, L.M., Farrow, H.A., Appleton, D.J., Lederer, R., 2004. Canine andfeline diabetes mellitus: nature or nurture? Journal of Nutrition 134, 2072S–2080S.

Ray, S.K., Nishitani, J., Petry, M.W., Fessing, M.Y., Leiter, A.B., 2003. Noveltranscriptional potentiation of BETA2/NeuroD on the secretin gene promoterby the DNA-binding protein Finb/RREB-1. Molecular and Cellular Biology 23,259–271.

Scott, M.P., Tamkun, J.W., Hartzell III, G.W., 1989. The structure and function of thehomeodomain. Biochimica et Biophysica Acta 989, 25–48.

Singh, G., Kaur, S., Stock, J.L., Jenkins, N.A., Gilbert, D.J., Copeland, N.G., Potter, S.S.,1991. Identification of 10 murine homeobox genes. The Proceedings of theNational Academy of Science of the United States of America 88, 10706–10710.

Whelan, J.A., Russell, N.B., Whelan, M.A., 2003. A method for the absolutequantification of cDNA using real-time PCR. Journal of ImmunologicalMethods 278, 261–269.

Zhang, C., Moriguchi, T., Kajihara, M., Esaki, R., Harada, A., Shimohata, H., Oishi, H.,Hamada, M., Morito, N., Hasegawa, K., Kudo, T., Engel, J.D., Yamamoto, M.,Takahashi, S., 2005. MafA is a key regulator of glucose-stimulated insulinsecretion. Molecular and Cellular Biology 25, 4969–4976.

Zhao, L., Guo, M., Matsuoka, T.A., Hagman, D.K., Parazzoli, S.D., Poitout, V., Stein, R.,2005. The islet beta cell-enriched MafA activator is a key regulator of insulingene transcription. Journal of Biological Chemistry 280, 11887–11894.