A humanized osteopontin mouse model and its application in ...

3 downloads 0 Views 3MB Size Report
Osteopontin (OPN) is a multifunctional protein involved in several inflammatory pro- cesses and pathogeneses including obesity-related disorders and cancer.
A humanized osteopontin mouse model and its application in immunometabolic obesity studies € KARINA ZEYDA, VERONICA MORENO-VIEDMA, KARIN STROHMEIER, NICOLE G. GRUN, € GUNTHER STAFFLER, MAXIMILIAN ZEYDA, and THOMAS M. STULNIG VIENNA, AUSTRIA

Osteopontin (OPN) is a multifunctional protein involved in several inflammatory processes and pathogeneses including obesity-related disorders and cancer. OPN binds to a variety of integrin receptors and CD44 resulting in a proinflammatory stimulus. Therefore, OPN constitutes a novel interesting target to develop new therapeutic strategies, which counteract OPN’s proinflammatory properties. We established a humanized SPP1 (hSPP1) mouse model and evaluated its suitability as a model for obesity and insulin resistance. Unchallenged hSPP1 animals did not significantly differ in body weight and gross behavioral properties compared to wild-type (WT) animals. High-fat diet-challenged hSPP1 similarly developed obesity and inflammation, whereas insulin resistance was markedly changed. However, OPN expression profile in tissues was significantly altered in hSPP1 compared to WT depending on the diet. In conclusion, we developed a versatile humanized model to study the action of OPN in vivo and to develop strategies that target human OPN in a variety of pathologies. (Translational Research 2016;178:63–73) Abbreviations: OPN ¼ osteopontin; hSPP1 ¼ humanized SPP1; ECM ¼ extracellular matrix; AT ¼ adipose tissue; GWAT ¼ gonadal white AT; SWAT ¼ subcutaneous white AT; mRNA ¼ messenger RNA; HF ¼ high-fat; LF ¼ low-fat; DIO ¼ diet-induced obesity; FRT ¼ flippase recognition target; BAC ¼ bacterial artificial chromosome; WT ¼ wild-type; AUC ¼ area under the curve; ITT ¼ insulin tolerance test; PCR ¼ polymerase chain reaction

INTRODUCTION

From the Christian Doppler Laboratory for Cardio-Metabolic Immunotherapy and Clinical Division of Endocrinology and Metabolism, Department of Medicine III, Medical University of Vienna, Vienna, Austria; FH Campus Wien, University of Applied Sciences, Department Health, Section Biomedical Science, Vienna, Austria; AFFiRiS AG, Vienna, Austria; Department of Pediatrics and Adolescent Medicine, Clinical Division of Pediatric Pulmonology, Allergology and Endocrinology, Medical University of Vienna, Vienna Austria. Submitted for publication June 3, 2016; revision submitted July 4, 2016; accepted for publication July 11, 2016. Reprint requests: Thomas M. Stulnig, Christian Doppler Laboratory for Cardio-Metabolic Immunotherapy and Clinical Division of Endocrinology and Metabolism, Department of Medicine III, Medical University of Vienna, W€ahringer G€urtel 18-20, 1090 Vienna, Austria; e-mail: [email protected]. 1931-5244/$ - see front matter Ó 2016 Published by Elsevier Inc. http://dx.doi.org/10.1016/j.trsl.2016.07.009

O

steopontin (OPN) is a multifunctional protein, first identified as a T helper type 1 cytokine, which is mainly expressed in various tissues and cells including macrophages, T-cells, dendritic cells, hepatocytes, smooth muscle cells, endothelial, and epithelial cells.1-3 OPN exhibits properties not only of a soluble cytokine that acts in an autocrine or paracrine manner on liver cells but also of an extracellular matrix (ECM)-bound protein.1 OPN plays a prominent role in monocyte migration, adhesion, and differentiation4-6 as well as in the pathophysiology of chronic inflammatory diseases, such as rheumatoid arthritis,7 Crohn’s disease,8 cancer,9 and cardiovascular disease.10,11 OPN is involved in the pathogenesis of obesity-associated complications, including adipose tissue (AT) inflammation, insulin resistance, type 2 diabetes, nonalcoholic fatty liver disease, and atherosclerosis.12-16 In vitro studies with adipocytes identified OPN as a modulator of insulin resistance and impaired 63

64

Translational Research December 2016

€ n et al Gru

AT A GLANCE COMMENTARY € n NG, et al. Gru Background

Osteopontin (OPN) is involved in a number of inflammatory processes, including obesity-linked complications, asthma, rheumatoid arthritis as well as cancer progression and metastasis. Proteolytic cleavage of OPN increases OPN’s proinflammatory properties and therefore constitutes an interesting target to develop new strategies to counteract OPN’s actions. To enable research on human OPN, we generated a humanized SPP1 (hSPP1) mouse and evaluated it as a diet-induced obesity model. Translational Significance

Our findings propose the hSPP1 animal model to be applied for a variety of pathologies in which OPN is a key player, including obesity-driven, cardio-metabolic as well as other diseases that can be translated to humans.

glucose uptake.17 Obese mice and humans revealed markedly increased messenger RNA (mRNA) levels for OPN in visceral and subcutaneous fat.18 Also, plasma OPN concentrations were markedly elevated in obese human subjects compared with lean controls.19,20 Importantly, high-fat (HF) diet treatment of mice lacking OPN, improved AT inflammation and insulin sensitivity12 as well as reduced hyperleptinemia and adipocyte hypertrophy.21 OPN acts via integrin binding by different domains. Although the arg-gly-asp (RGD)-containing domain mainly binds integrins avb1, avb3, avb5, and a5b122 on the cell surface, the ser-val-val-tyr-glu-leu-arg (SVVYGLR) motif interacts with integrins a4b1, a4b7, and a9b1.23,24 Importantly, the RGD region is fully conserved between species, whereas the SVVYGLR motif differs between humans (SVVYGLR) and rodents (SLAYGLR).23 The third integrin-binding domain ELVTDFTDLPAT has also been described to bind a4b1.25 Overall, human and murine OPN show only 63% amino acid sequence identity (National Center of Biotechnology Information). OPN’s actions are affected by protease cleavage. The cryptic region SVVYGLR, which is usually hidden in intact OPN, becomes accessible on thrombin and matrix metalloproteinase cleavage. Interestingly, although enzymatic thrombin and matrix metalloproteinase cleavage produces a proinflammatory

N-terminal OPN fragment, the less described C-terminal OPN form may even attenuate inflammatory processes.26 Moreover, bonding with the hyaluronic acid receptor CD44 leads to the induction of macrophage chemotaxis and the b3-integrin receptors engagement.27 To develop therapeutic strategies against OPN actions, small molecules are not a primary option but sequence-specific approaches are needed such as specific antibodies, RNA aptamers, and antisense oligonucleotides or interfering RNAs. Such approaches targeting OPN have been successfully applied on human tissues transplanted into immunodeficient mice, for example, in cancer studies.28-30 However, human cell transplantation is not feasible for most disease models, particularly inflammatory diseases. To facilitate in vivo studies of blocking human OPN, we developed a humanized mouse model by replacing the mouse gene with human SPP1 sequence. Owing to OPN’s crucial involvement in a variety of inflammatory pathologies and particularly its association with cardio-metabolic complications of obesity, we evaluated the humanized SPP1 (hSPP1) mouse model as a diet-induced obesity (DIO) model. We show the functionality of hSPP1 mice as a DIO model by generation of obesity-associated insulin resistance and inflammation following HF diet feeding. Hence, hSPP1 mice are a highly useful model to investigate OPN’s role not only to study prevention and treatment of obesity and its complications but also to facilitate preclinical studies on neutralization of human OPN in a large variety of other inflammatory diseases and cancer. MATERIAL AND METHODS

Heterozygous C57BL/6-Spp1tm2737 (hSPP1 ) mice were developed with Taconic (K€oln, Germany). In short, the chimeric SPP1 protein sequence was chosen as depicted in Fig 1, A. Accordingly, exon 2 contains the translation initiation codon and the sequence, which encodes the cleavable signal peptide. The mouse genomic sequence from amino acid 21 (valine) in exon 3 up to the termination codon in exon 7 has been replaced by its human counterpart to leave a single N-terminal mouse-specific amino acid in the mature protein (p.Ile19Leu; Fig 1). Positive selection markers, including neomycin resistance flanked by flippase recognition target (FRT) and puromycin resistance flanked with F3, were inserted into human intron 3 as well as downstream of the mouse 30 untranslated region, respectively. Bacterial artificial chromosome (BAC) clones from the mouse C57BL/6J were used for the targeting vector that was further Animals and diets.

(SPP1) Arte

1/2

Translational Research Volume 178

€ n et al Gru

65

Fig 1. Generation of the humanized SPP1 gene for homologous recombination. Description of the constructed chimeric SPP1 in comparison to the murine and human form (A). Stepwise generation of the humanized allele is given in (B). NeoR, neomycin resistance; PuroR, puromycin resistance.

Fig 2. Blood glucose levels are significantly lower in hSPP1 compared to WT animals with similar body weight. Male hSPP1 and WT mice (n 5 20 per group) were kept on normal chow until 9 weeks of age. Body weight (A) and blood glucose levels (B) were determined before LF and HF administration. Data is expressed as mean 6 standard error of the mean. ***P , 0.001. WT, wild-type; LF, low-fat diet; HF, high-fat diet; hSPP1, humanized SPP1.

transfected using electroporation into a Taconic C57BL/ 6N embryonic stem cell line. After homologous recombination, clones were selected with double positive neomycin resistance–puromycin resistance selection, and the humanized allele was obtained after Flp-mediated

removal of these selection markers. Expression of the human SPP1 gene is controlled by the mouse Spp1 promoter and results in the loss of expression of the mouse Spp1 gene (Fig 1, B). Generation of the transgenic mouse model was according to well-described

66

€ n et al Gru

Translational Research December 2016

Fig 3. Dietary challenge alters body weight and insulin resistance in hSPP1 mice. Male hSPP1 and WT mice (n 5 8–10 per group) were fed a HF and LF diet for 22 weeks. Body weight was determined weekly over 22 weeks (A,B). After 15 weeks on diet, an ITT was performed. Glucose levels were measured over 120 minutes after intraperitoneal insulin application of 0.75-IU/kg body weight for HF and 0.25-IU/kg body weight for LF, respectively, (C) and AUC (D) was calculated. Fasting glucose and insulin levels (E) were measured after 22 weeks, and insulin resistance is given by the HOMA-IR (F). Data is expressed as mean 6 standard error of the mean. *,#,§P , 0.05, **,##,§§P , 0.01, ***,###,§§§P , 0.001. (*) indicates the comparison between WT LF and HF, (#) comparison between hSPP1 LF and WT LF; (§) comparison between hSPP1 LF and HF. AUC, area under the curve; ITT, insulin tolerance test; HOMA-IR, homeostasis model assessment of insulin resistance; WT, wild-type; LF, low-fat diet; HF, high-fat diet; hSPP1, humanized SPP1.

instructions.31 Heterozygous hSPP11/2 animals were crossbred to homozygosity in house. At 9 weeks of age, 10 male hSPP1 and 10 C57BL/6JRj (WT) mice (Janvier Labs, Saint Berthevin, France) were set on high-fat diet (HF, 60 kcal%, D12492; Research

Diets, New Brunswisk, NJ) for 22 weeks to induce obesity, whereas 10 male hSPP1 and 10 WT mice were fed a low-fat diet (LF, 10 kcal%, D12450 B; Research Diets) to serve as negative controls for DIO. All mice were kept on normal chow until 9 weeks of age.

Translational Research Volume 178

€ n et al Gru

67

Fig 4. Obesity-induced adipose tissue inflammation in hSPP1 mice is similar to WT animals. Male hSPP1 and WT mice (n 5 8–10 per group) were set on indicated diets for 22 weeks. GWAT and SWAT weight (A) was determined after sacrification; RNA was isolated and transcribed to complementary DNA. Gene expression of insulin sensitivity markers Adipoq (B) and Pparg (C), macrophage marker Emr1 (F4/80) (D) as well as of cytokines Ccl2 (E) and Tnf-a (F) was assessed by quantitative real-time PCR. Ubc was used as house-keeping gene, and data are given after normalization to GWAT WT LF as mean 6 standard error of the mean. *P , 0.05, **P , 0.01, ***P , 0.001. WT, wild-type; PCR, polymerase chain reaction; Ubc, ubiquitin C; LF, low-fat diet; HF, high-fat diet; mRNA, messenger RNA; hSPP1, humanized SPP1; GWAT, gonadal white AT; SWAT, Subcutaneous white AT.

All mice were housed in a specific pathogen-free facility, maintaining a 12-hour light/dark cycle and had free access to water and food. Food intake was measured 3 times per week. Weight was determined weekly, blood was drawn every third week and immediately before mice were sacrificed. Subcutaneous white AT (SWAT)

and gonadal white AT (GWAT) as well as liver were collected, weighed, and stored for further experiments. The protocol was approved by the local ethics committee for animal studies and followed the guidelines on accommodation and care of animals formulated by the European Convention for the Protection of Vertebrate

68

€ n et al Gru

Translational Research December 2016

Fig 5. Dietary challenge induces adipocyte hypertrophy and crown-like structure formation in both genotypes. Male hSPP1 and WT mice (n 5 8–10 per group) were fed a HF and LF diet for 22 weeks. Paraffin-embedded GWAT sections were stained for Mac-2 and visualized with light microscopy (A). Formation of crown-like structures (CLS) normalized to total number of adipocytes (B) and adipocyte size (C) was assessed with ImageJ software. Representative pictures are shown after 10-fold magnification. Data are expressed as mean 6 standard error of the mean. *P , 0.05, ***P , 0.001. WT, wild-type; LF, low-fat diet; HF, high-fat diet; hSPP1, humanized SPP1; GWAT, gonadal white AT; SWAT, Subcutaneous white AT.

Animals Used for Experimental and Other Scientific Purposes. Polymerase chain reaction (PCR) for genotyping analysis. Genomic DNA was isolated from ear tissue af-

ter clipping to identify homozygous hSPP1 mice using 2xKAPPA 2G Fast HS Kit. PCR was performed according to manufacturer’s instructions (KAPA Biosystems, Wilmington, Mass). To amplify the constitutive humanized and WT allele, 2 different primer pairs were used: for the humanized SPP1 allele forward primer

50 -CTGTTTGGCATTGCCTCC-30 and reverse primer 30 -TATACTGTGCTTGGTACTGGCC-50 , for the murine WT Spp1 allele forward primer 50 -TAG CAACCTTTGCGACGC-30 and reverse primer 30 -TATACTGTGCTTGGTACTGGCC-5’. In addition, forward primer 50 -CGAATGTAAGGAGTCTGGTGG-30 and reverse primer 30 -CAAGTTTCAAGCT 0 TTTGCTCG-5 were used to detect both, the humanized and WT allele. To assess DNA size, 100 base pair (bp) DNA ladder was used (Invitrogen, Carlsbad, Calif).

Translational Research Volume 178

The humanized allele was displayed at 362 bp, whereas the WT allele showed a 254 bp band. Genomic DNA isolated from WT and heterozygous mice was employed as negative controls. Serum metabolic parameters. Serum insulin (Mercodia, Uppsala, Sweden) and OPN (R&D Systems, Minneapolis, USA) were analyzed with commercially available enzyme-linked immunosorbent assays. After a 5-hour fasting period, insulin sensitivity was assessed after 15 weeks on indicated diets by measuring glucose levels before, 30, 60, 90, and 120 minutes after an intraperitoneal injection of recombinant human insulin (0.75 IU/kg body weight NovoRapid for HF and 0.25 IU/kg body weight for LF, respectively; Novo Nordisk, Bagsværd, Denmark). Blood glucose levels in mice on normal chow diet were measured without fasting using OneTouch Ultra test strips (LifeScan, Milpitas, USA). Homeostasis model assessment of insulin resistance was calculated as an index for insulin resistance. Reverse transcription and gene expression. One part of SWAT, GWAT, and liver was stored in RNAlater (Qiagen, Hilden, Germany) overnight at 4 C. RNAlater was removed, tissues were snap frozen and stored at 280 C. RNA isolation was performed with peqGOLD TriFastTM (Peqlab, Erlangen, Germany) by using an automated homogenizer (Precellys 24, Erlangen, Germany), and one microgram RNA was transcribed to complementary DNA with M-MLV Reverse Transcriptase (Promega, Madison, USA). Gene expression of F4/80 (Emr1, Mm00802529_m1), Mcp1 (Ccl2, Mm00441242_m1), tumor necrosis factor-a (Tnf, Mm00443258_m1), adiponectin (Adipoq, Mm004564 25_m1), peroxisome proliferator-activated receptor gamma (Pparg, Mm01184322_m1), integrin alpha x (Itgax; Cd11c, Mm00498698_m1), murine OPN (mSpp1; Spp1, Mm00436767_m1), human OPN (hSPP1; SPP1, Hs00959010_m1), and reference gene ubiquitin C (Ubc; Ubc, Mm01201237_m1) was analyzed by quantitative real-time (qRT)-PCR on an ABI Prism 7000 cycler with commercially available assays-on demand kits (TaqMan Gene Expression Assay, Applied Biosystems, Foster City, Calif) and normalized to Ubc. The comparative threshold cycle method was used to calculate relative expression.32 Immunohistochemistry. Murine GWAT samples were fixed with neutral buffered 4% paraformaldehyde at 4 C and were paraffin-embedded. After dewaxation and rehydration, sections were immunoh istochemically stained for Mac-2 (Cedarlane, Ontario, Canada) using the ABC kit (Vector Laboratories) according to manufacturer’s protocol. Counterstaining was performed by using hematoxylin (Merck,

€ n et al Gru

69

Darmstadt, Germany), and sections were analyzed with standard light microscopy using a 10-fold magnification. Crown-like structures and density of adipocytes were quantified with ImageJ software, thereby assessing 5 independent fields per sample. Crown-like structures were normalized to total adipocyte number. Statistical analysis. Data are given as mean and standard error of the mean (mean 6 standard error of the mean). Comparisons between hSPP1 and WT were assessed by 1-way analysis of variance using Tukey’s multiple comparisons test. For body weight gain and insulin tolerance test results, a 2-way analysis of variance with Tukey’s multiple comparisons test was performed. Impact of dietary treatment on OPN expression within a genotype was analyzed by 2-tailed independent Student’s t-test. A P value of #0.05 was considered statistically significant. RESULTS Generation and characterization of the humanized 1/2 hSPP1 mouse. Heterozygous hSPP1 animals were

generated, bred to homozygosity, and subsequently evaluated with respect to its functionality as a DIO model. All hSPP1 mice used were genotyped at 3 weeks of age and displayed a distinct 362 bp DNA band indicating the humanized allele. Moreover, heterozygous hSPP11/2 animals showed a double-band at 362 bp and 254 bp, the latter identifying the mouse allele, whereas genotyping analysis in WT mice revealed a single band at 254 bp (17 hSPP1 animals of 20 are given in Fig S1). A pilot study using 3 unchallenged homozygous hSPP1 mice and WT control animals confirmed the selective expression of human OPN and the complete loss of murine OPN in our humanized mouse model (data not shown). In addition, serum human OPN levels (96.41 6 18.16 ng/ml) were comparable to serum murine OPN concentrations (71.30 6 9.74 ng/ml; data not shown) in lean hSPP1 and WT animals. Breeding efficacy, litter size, and sex distribution of littermates of hSPP1 mice was comparable to WT animals. Unchallenged homozygous hSPP1 animals did not differ in size or gross behavior. Moreover, no apparent abnormalities were observed in hSPP1 mice. Body weight was measured in unchallenged hSPP1 and WT mice at 9 weeks of age giving comparable results (Fig 2, A). However, blood glucose levels were significantly lower in nonfasting hSPP1 compared with WT mice (Fig 2, B). Evaluation of the hSPP1 mouse as a DIO model. To evaluate whether dietary treatment leads to the development of obesity in homozygous hSPP1 comparable to that of

70

€ n et al Gru

WT mice, 9-week-old male hSPP1 and WT mice were subdivided into 2 groups, involving 8–10 animals per group and were challenged by HF and LF diet for 22 weeks, respectively. Body weight was measured weekly (Fig 3, A) and was significantly increased in hSPP1 and WT mice on HF diet compared with respective lean controls after 22 weeks. Importantly, there was no difference between obese hSPP1 and WT mice (Fig 3, B). Interestingly, hSPP1 on LF diet significantly increased the body weight compared with WT mice starting at 6 weeks of LF treatment (Fig 3, A and B), whereas food intake was comparable between all groups (Fig S2). Metabolic changes on HF diet in hSPP1 mice. To determine insulin resistance, an insulin tolerance test was performed in diet-challenged mice, and area under the curve was calculated. Owing to a strong insulin response in WT mice on LF diet, we discontinued glucose measurements in WT mice after 30 minutes and rescued with oral glucose. hSPP1 were markedly less insulin sensitive compared to WT on LF (Fig 3, C and D). In addition, we observed significantly higher serum insulin levels in both genotypes after 22 weeks on HF diet compared with corresponding lean controls (Fig 3, E), which were mirrored by elevated homeostasis model assessment of insulin resistance (Fig 3, F). Hence, HF diet challenge induced obesity and insulin resistance in hSPP1 mice comparable to WT mice. However, hSPP1 mice on LF diet had significantly higher body weight and diminished insulin sensitivity compared with lean WT mice. HF diet induces adipose tissue hypertrophy as well as adipose tissue and liver inflammation similar to WT. Next,

we investigated the impact of HF diet on adipose tissues (GWAT and SWAT) of hSPP1 and WT mice. GWAT weight was not increased in hSPP1 animals on HF vs those on LF diet, whereas it was elevated in the latter compared to WT mice on LF diet. SWAT weight was significantly higher in hSPP1 after HF diet compared with lean controls, whereas SWAT weight was increased in hSPP1 on LF compared to WT (Fig 4, A). Gene expression of Adipoq and Pparg was significantly lower in hSPP1 on LF diet compared with WT animals, whereas mRNA levels were markedly higher in GWAT of WT mice fed a LF diet compared to obese WT mice (Fig 4, B and C). In both mouse models, gene expression levels of macrophage marker Emr1 and inflammation markers Ccl2 and Tnf-a were significantly elevated after HF diet in both adipose tissues than in respective control mice, indicating comparable inflammatory gene alterations of hSPP1 mice in adipose tissue under HF diet (Fig 4, D–F).

Translational Research December 2016

To extend our findings regarding AT inflammation, we investigated macrophage infiltration and adipocyte hypertrophy in GWAT. Mac-2 staining of macrophages showed significantly more crown-like structures after HF diet in both mouse genotypes compared to LF controls (Fig 5, A and B). GWAT adipocyte size, which is often related to insulin resistance,33 was increased after HF diet in WT animals but not significantly changed by HF diet in hSPP1. Notably, adipocyte size of hSPP1 mice on LF diet was significantly increased compared with lean WT mice (Fig 5, C). Because it is well-recognized that obesity-linked inflammation also occurs in the liver, we measured liver weight as well as hepatic macrophage and inflammation markers. All mice on HF diet markedly increased liver weight compared to corresponding lean controls (Fig S3, A). As expected, mRNA levels of macrophage markers Emr1 and Itgax as well as of proinflammatory markers Ccl2 and Tnf-a were significantly upregulated to a similar extent in both mouse genotypes after HF diet compared to respective lean controls (Fig S3, B–E). In conclusion, although several differences in fat distribution and GWAT adipocytes could be found under LF diet, adipose tissue properties were very similar in hSPP1 mice compared to WT mice in HF diet-induced obesity. Furthermore, dietary treatment resulted in hepatomegaly and obesityassociated liver inflammation in hSPP1 comparable to WT mice. Regulation of OPN expression in hSPP1 mice. To analyze the regulation of OPN gene expression after insertion of the human SPP1 gene under LF and HF diet, tissue-specific gene expression and serum levels of murine and human OPN were determined. As expected, homozygous hSPP1 mice did not express mSpp1, neither in AT depots nor in the liver and vice versa (Fig 6, A–D). WT mice showed significantly higher mSpp1 mRNA levels in GWAT, SWAT, and liver after HF diet compared to respective lean controls (Fig 6, A and C). In hSPP1 mice, however, SPP1 mRNA expression was not altered by HF in GWAT but significantly upregulated in SWAT of obese hSPP1 mice (Fig 6, B). Assessment of SPP1 mRNA expression in the liver revealed significantly lower SPP1 mRNA levels in obese hSPP1 mice compared to lean controls, whereas serum human OPN levels were not affected after dietary administration in hSPP1 mice (Fig 6, E). Thus, although OPN expression was significantly upregulated in SWAT in hSPP1 mice on HF challenge, no upregulation was found in GWAT and liver due to high expression already on LF diet.

Translational Research Volume 178

€ n et al Gru

71

Fig 6. Insertion of human OPN results in differential OPN expression patterns. Male hSPP1 and WT mice (n 5 8– 10 per group) were fed a HF and LF diet for 22 weeks. Gene expression of mSpp1 and hSPP1 was assessed in GWAT, SWAT (A, B), and liver (C, D). Fasting serum hOPN was determined in hSPP1 mice (E). Data are expressed as mean 6 standard error of the mean. *P , 0.05, **P , 0.01, ***P , 0.001. OPN, osteopontin; WT, wild-type; LF, low-fat diet; HF, high-fat diet; hSPP1, humanized SPP1; GWAT, gonadal white AT; SWAT, Subcutaneous white AT.

DISCUSSION

Humanized animal models were developed for various research fields, including immunology or cancer research.34 Here, we successfully inserted the human SPP1 gene into C57BL/6 mice and demonstrated similarities in body weight and glucose levels of nonfasting hSPP1 and WT animals, indicating no apparent abnormalities after transgenic modification, although the expression of OPN was high in GWAT and liver already

under LF diet. Moreover, phenotypical parameters including litter size, sex distribution, and breeding efficacy of hSPP1 mice did not differ when compared to WT animals. Notably, research data on Spp12/2 animals revealed improved insulin sensitivity, tissue inflammation, and hepatic steatosis when challenged with HF diet,12 whereas our hSPP1 animal model developed obesity-induced complications similar to a well-known DIO model. Therefore, our results indicate the

72

Translational Research December 2016

€ n et al Gru

functionality of this classical knock-out-knock-in model. Taken together, these characteristics make the developed hSPP1 model a versatile tool for a variety of preclinical in vivo and ex vivo studies to target human OPN. Diet-induced obesity strikingly resembles the human metabolic syndrome including inflammatory and metabolic characteristics. HF diet administration tremendously increased body weight as well as SWAT and liver weight in the hSPP1 mouse model. Moreover, tissue-specific inflammation and macrophage infiltration was significantly upregulated after HF diet. Importantly, infiltrating macrophages phagocyte dying adipocytes, thereby forming crown-like structures, which have been reported to play a functional role in obesity. Here, we demonstrate that obese hSPP1 mice showed significantly more crown-like structures than lean controls and together with changes in inflammatory gene expression this indicates changes in obesityinduced inflammation in hSPP1 similar to WT animals. In addition, obesity-associated insulin resistance was induced by HF diet in both animal models to a comparable extent. Although mice on HF diet were strikingly similar irrespective of the genotype, hSPP1 mice on LF diet in parallel to high OPN expression, revealed altered body weight, GWAT weight, adipocyte hypertrophy, insulin resistance, and SPP1 expression levels. The increased body weight in hSPP1 compared to WT mice on LF diet as reflected by a marked increase in GWAT weight due larger adipocytes is regarded as fundamental to the other observed alterations including hepatic alterations and insulin resistance.33 Also decreased Adipoq and Pparg gene expression levels in GWAT match the diminished insulin sensitivity and increased GWAT weight. Because food intake was similar in hSPP1 and WT mice on LF, we presume lower energy expenditure in lean hSPP1 mice as a basis for increased body weight in hSPP1 compared to WT animals. Further studies including multiparameter monitoring in metabolic cages are warranted for a more detailed characterization of this model for its use in metabolic studies. Moreover, the hSPP1 model in LF diet constitutes an interesting model for obesityindependent inflammatory and infectious diseases. Although changes in GWAT physiology in hSPP1 mice on LF diet could be a basis for the failure to upregulate hSPP1 gene expression in GWAT and liver by HF diet, intrinsic changes in SPP1 gene regulation by the transgenic modification cannot be ruled out. Although murine 50 regulatory sequences including the signal peptide were conserved, intronic regulatory components that were replaced by the genomic human sequence may play a role for the observed alterations in GWAT and liver. More data in other and more pro-

nounced inflammatory disease models have to be studied to evaluate the comparability of changes in hSPP1 compared to WT mice. Because there is a slight difference between the genotypes used in this study, comparisons were primarily made within a genotype, and similarities or important differences were highlighted between hSPP1 and WT animals. To date, there is a high prevalence of obesityassociated disorders including AT inflammation, insulin resistance, and cardio-metabolic disease. In 2014, the World Health Organization stated a number of approximately 1.9 billion overweight adults, of which over 600 million were recognized of being obese.35 According to this background and the crucial role of OPN in obesity-associated diseases, hSPP1 mice represent a new interesting model to evolve novel strategies to combat OPN-induced inflammation and metabolic complications, for example, by immunotherapy. Moreover, this model is now available to target human SPP1/OPN in a variety of other inflammatory, infectious, and neoplastic disease models in which OPN plays a key role. ACKNOWLEDGMENTS

Conflicts of Interest: The authors have read the journal’s authorship statement and policy on conflicts of interest and declare no commercial or financial conflict of interest. This work was supported by the Federal Ministry of Economy, Family and Youth and the National Foundation for Research, Technology and Development (to T. M. Stulnig). The authors gratefully thank Melina Amor for technical support. The article has been reviewed and approved by all authors. REFERENCES

1. Lund SA, Giachelli CM, Scatena M. The role of osteopontin in inflammatory processes. J Cell Commun Signal 2009;3:311–22. 2. Tokairin T, Nishikawa Y, Watanabe H, et al. Osteopontin expression in the liver with severe perisinusoidal fibrosis: autopsy case of Down syndrome with transient myeloproliferative disorder. Pathol Int 2008;58:64–8. 3. Mazzali M, Kipari T, Ophascharoensuk V, Wesson JA, Johnson R, Hughes J. Osteopontin–a molecule for all seasons. QJM 2002;95: 3–13. 4. Giachelli CM, Lombardi D, Johnson RJ, Murry CE, Almeida M. Evidence for a role of osteopontin in macrophage infiltration in response to pathological stimuli in vivo. Am J Pathol 1998;152: 353–8. 5. Weber GF, Zawaideh S, Hikita S, Kumar VA, Cantor H, Ashkar S. Phosphorylation-dependent interaction of osteopontin with its receptors regulates macrophage migration and activation. J Leukoc Biol 2002;72:752–61. 6. Crawford HC, Matrisian LM, Liaw L. Distinct roles of osteopontin in host defense activity and tumor survival during squamous cell carcinoma progression in vivo. Cancer Res 1998;58:5206–15.

Translational Research Volume 178

7. Yamamoto N, Sakai F, Kon S, et al. Essential role of the cryptic epitope SLAYGLR within osteopontin in a murine model of rheumatoid arthritis. J Clin Invest 2003;112:181–8. 8. Agnholt J, Kelsen J, Schack L, Hvas CL, Dahlerup JF, Sørensen ES. Osteopontin, a protein with cytokine-like properties, is associated with inflammation in Crohn’s disease. Scand J Immunol 2007;65:453–60. 9. Leitner L, J€ urets A, Itariu BK, et al. Osteopontin promotes aromatase expression and estradiol production in human adipocytes. Breast Cancer Res Treat 2015;154:63–9. 10. Waller AH, Sanchez-Ross M, Kaluski E, Klapholz M. Osteopontin in cardiovascular disease: a potential therapeutic target. Cardiol Rev 2010;18:125–31. 11. Bruemmer D, Collins AR, Noh G, et al. Angiotensin IIaccelerated atherosclerosis and aneurysm formation is attenuated in osteopontin-deficient mice. J Clin Invest 2003;112:1318–31. 12. Kiefer FW, Neschen S, Pfau B, et al. Osteopontin deficiency protects against obesity-induced hepatic steatosis and attenuates glucose production in mice. Diabetologia 2011;54:2132–42. 13. Takemoto M, Yokote K, Nishimura M, et al. Enhanced expression of osteopontin in human diabetic artery and analysis of its functional role in accelerated atherogenesis. Arterioscler Thromb Vasc Biol 2000;20:624–8. 14. Sahai A, Malladi P, Melin-Aldana H, Green RM, Whitington PF. Upregulation of osteopontin expression is involved in the development of nonalcoholic steatohepatitis in a dietary murine model. Am J Physiol Gastrointest Liver Physiol 2004;287: G264–73. 15. Isoda K, Kamezawa Y, Ayaori M, Kusuhara M, Tada N, Ohsuzu F. Osteopontin transgenic mice fed a high-cholesterol diet develop early fatty-streak lesions. Circulation 2003;107:679–81. 16. Bertola A, Deveaux V, Bonnafous S, et al. Elevated expression of osteopontin may be related to adipose tissue macrophage accumulation and liver steatosis in morbid obesity. Diabetes 2009;58: 125–33. 17. Zeyda M, Gollinger K, Todoric J, et al. Osteopontin is an activator of human adipose tissue macrophages and directly affects adipocyte function. Endocrinology 2011;152:2219–27. 18. Nomiyama T, Perez-Tilve D, Ogawa D, et al. Osteopontin mediates obesity-induced adipose tissue macrophage infiltration and insulin resistance in mice. J Clin Invest 2007;117:2877–88. 19. Kiefer FW, Zeyda M, Todoric J, et al. Osteopontin expression in human and murine obesity: extensive local up-regulation in adipose tissue but minimal systemic alterations. Endocrinology 2008;149:1350–7. 20. G omez-Ambrosi J, Catalan V, Ramırez B, et al. Plasma osteopontin levels and expression in adipose tissue are increased in obesity. J Clin Endocrinol Metab 2007;92:3719–27.

€ n et al Gru

73

21. Chapman J, Miles PD, Ofrecio JM, et al. Osteopontin is required for the early onset of high fat diet-induced insulin resistance in mice. PLoS One 2010;5:e13959. 22. Yokosaki Y, Tanaka K, Higashikawa F, Yamashita K, Eboshida A. Distinct structural requirements for binding of the integrins alphavbeta6, alphavbeta3, alphavbeta5, alpha5beta1 and alpha9beta1 to osteopontin. Matrix Biol J Int Soc Matrix Biol 2005;24:418–27. 23. Scatena M, Liaw L, Giachelli CM. Osteopontin: a multifunctional molecule regulating chronic inflammation and vascular disease. Arterioscler Thromb Vasc Biol 2007;27:2302–9. 24. Ito K, Kon S, Nakayama Y, et al. The differential amino acid requirement within osteopontin in alpha4 and alpha9 integrinmediated cell binding and migration. Matrix Biol J Int Soc Matrix Biol 2009;28:11–9. 25. Bayless KJ, Davis GE. Identification of dual a4b1 integrin binding sites within a 38 amino acid domain in the N-terminal thrombin fragment of human osteopontin. J Biol Chem 2001;276:13483–9. 26. Takahashi K, Takahashi F, Tanabe KK, Takahashi H, Fukuchi Y. The carboxyl-terminal fragment of osteopontin suppresses arginine-glycine-asparatic acid-dependent cell adhesion. Biochem Mol Biol Int 1998;46:1081–92. 27. Okamoto H. Osteopontin and cardiovascular system. Mol Cell Biochem 2006;300:1–7. 28. Ye QH, Qin LX, Forgues M, et al. Predicting hepatitis B viruspositive metastatic hepatocellular carcinomas using gene expression profiling and supervised machine learning. Nat Med 2003;9: 416–23. 29. Zhao J, Dong L, Lu B, et al. Down-regulation of osteopontin suppresses growth and metastasis of hepatocellular carcinoma via induction of apoptosis. Gastroenterology 2008;135:956–68. 30. Bhattacharya SD, Mi Z, Kim VM, Guo H, Talbot LJ, Kuo PC. Osteopontin regulates epithelial mesenchymal transition-associated growth of hepatocellular cancer in a mouse xenograft model. Ann Surg 2012;255:319–25. 31. Doyle A, McGarry MP, Lee NA, Lee JJ. The construction of transgenic and gene knockout/knockin mouse models of human disease. Transgenic Res 2012;21:327–49. 32. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods San Diego Calif 2001;25:402–8. 33. Kim JI, Huh JY, Sohn JH, et al. Lipid-overloaded enlarged adipocytes provoke insulin resistance independent of inflammation. Mol Cell Biol 2015;35:1686–99. 34. Shultz LD, Ishikawa F, Greiner DL. Humanized mice in translational biomedical research. Nat Rev Immunol 2007;7:118–30. 35. WHO. Obesity and overweight [Internet]. WHO. Available at: http://www.who.int/mediacentre/factsheets/fs311/en/. Accessed August 20, 2015.

73.e1

€ n et al Gru

Translational Research December 2016

APPENDIX

Fig S1. PCR analysis confirms homozygosity of hSPP1 mice. Three-week-old male hSPP1 mice (n 5 20; genotyping of 17 animals are given as an example) were ear-clipped, and ear tissue was used for DNA isolation. Genotyping was performed using the 2xKAPPA 2G Fast kit and indicated primer pairs. Electrophoresis revealed the humanized allele at 362 bp and the WT allele at 254 bp. PCR, polymerase chain reaction; WT, wild-type; hSPP1, humanized SPP1.

Fig S2. Food intake of hSPP1 and WT mice. Male hSPP1 and WT mice (n 5 8–10 per group) were fed a HF and LF diet for 22 weeks. Food intake was measured over 3 weeks and calculated as grams per mouse per day. Data is expressed as mean 6 standard error of the mean. WT, wild-type; LF, low fat; HF, high fat; hSPP1, humanized SPP1.

Translational Research Volume 178

€ n et al Gru

Fig S3. Hepatic gene expression displays inflammation after HF diet. Male hSPP1 and WT mice (n 5 8–10 per group) were set on indicated diets for 22 weeks. Liver weight was determined after sacrification (A). Hepatic gene expression levels of macrophage marker Emr1 (B) and Itgax (C) as well as of inflammation markers Ccl2 (D) and Tnf-a (E) were assessed by quantitative real-time PCR. Gene expression was normalized to WT LF. Data is given as mean 6 standard error of the mean. *P , 0.05, **P , 0.01, and ***P , 0.001. mRNA, messenger RNA; WT, wild-type; LF, low fat; HF, high fat; PCR, polymerase chain reaction; hSPP1, humanized SPP1.

73.e2