High glucose-induced human cellular immune ...

YBCMD-01904; No. of pages: 6; 4C: Blood Cells, Molecules and Diseases xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Blood Cells, Molecules and Diseases journal homepage: www.elsevier.com/locate/bcmd

High glucose-induced human cellular immune response is governed by miR-2909 RNomics Deepak Kaul ⁎, Sugandha Sharma Department of Experimental Medicine, Post-graduate Institute of Medical Education & Research, Chandigarh 160012, India Department of Biotechnology, Post-graduate Institute of Medical Education & Research, Chandigarh 160012, India

a r t i c l e

i n f o

Article history: Submitted 18 November 2014 Accepted 11 January 2015 Available online xxxx (Communicated by M. Narla, DSc, 11 January 2015) Keywords: CCL5 IL-17 IFN-γ miR-2909 RNomics Aerobic-glycolysis

a b s t r a c t Regulation of NFkB family member RelA translocation by tumour suppressor genes encoding p53 and KLF4, has been widely recognized as the critical for human peripheral blood mononuclear cells (PBMCs) to meet their energy requirement for tailoring their immune response against any perceived threat. Our study was addressed to understand as to how human PBMCs respond to high glucose threat in terms of their genomics-directed immune response. The results of such a study revealed for the first time that NFkB induced miR-2909 RNomics is crucial for the regulation of RelA translocation within human PBMCs exposed to high glucose thereby enabling these epigenetically programmed cells to tailor immune response involving genes coding for CCL5; IFN-γ and IL-17. Based upon these results an attempt was also made to propose a mechanistic pathway that links high glucose induced cellular miR-2909 RNomics with the genes involved in energy metabolism and immune response. © 2015 Elsevier Inc. All rights reserved.

Introduction Aerobic-glycolysis is not only the unique feature of cancer cells but also of the normal immune/stem cells that use this phenomenon to produce intermediary metabolites of glycolysis as substrates for growth [1,2]. The cellular cyclic-transition from aerobic respiration to glycolysis would have been impossible without the existence of masterregulatory genes coding for p53 and NFkB family member RelA because p53-dependent genomics promotes aerobic-respiration leading to cellular differentiation [3,4] whereas RelA-dependent genomics initiates aerobic-glycolysis reverting fully differentiated cells into more primitive de-differentiated form [5]. Recently, a novel emerging role of p53 in curtailing the stem cell expansion has unfolded the mechanisms through which p53 restricts self-renewing divisions and reprogramming of somatic cells into stem cells [6]. Importantly, KLF4 was shown to physically interact with p53, resulting in the synergistic activation of p21 promoter thereby acting as a critical mediator of p53-induced growth arrest [7]. It is pertinent to note that the polycomb group protein Bmi-1 has been shown to support stem cells renewal by ensuring reduced p53 protein half-life [8] whereas KLF4 restricts the

⁎ Corresponding author at: Department of Experimental Medicine, Post-graduate Institute of Medical Education & Research, Chandigarh 160012, India. Fax: +91 172 2744401. E-mail address: [email protected] (D. Kaul).

transcriptional expression of Bmi-1 [9]. Further the stimulating effect of Bmi-1 upon NFkB activation [10] is reversed by both p53 as well as KLF4 through the inhibition of IKK [3,11]. The tumour suppressor p53 not only restricts cancer cell addiction for sugar through repression of glucose transporters 1 and 4 (GLUT 1 and GLUT 4) but also inhibits glucose-induced cell proliferation through the activation of NFkB [3]. Interestingly, PPAR-γ also has the capacity to inhibit NFkB activation [12] as well as increase in the expression of genes coding for KLF4 [13] and GLUT1 but not GLUT4 [14]. Cellular PPAR-γ mediated transcriptional activation has been shown to be suppressed by PER-3 [15]. Apart from the role of these genes, coding for KLF4, p53, NFkB and PPAR-γ, in the cellular metabolism and differentiation, they also play crucial role in programming the cellular immune response. KLF-4 has been found to play crucial role in the development of IL-17 producing CD4+ T-cells independently of RORγt [16] whereas p53 was shown to induce type-1 interferons and regulatory T-cells [17,18] as well as to inhibit translational expression of IFN-γ [19,20]. PPAR-γ was shown to restrict CD4+ T-cell differentiation to Th1 and Th17 cell subsets and promote Th2 differentiation [21] as well as favour phenotype-switch from Th17 into Treg CD4+ T cells [22]. At this stage, it is pertinent to note that NFkB was shown to regulate the expression of a novel microRNA (designated as miR-2909) encoded by human cellular AATF-genome [11, 23]. This miR-2909 was shown to regulate a large number of genes involved in the immunity and cancer (Fig. 1) through its capacity to suppress KLF4 translational expression [11,23,24]. Keeping in view the fact that orchestrated cross-talk between coding RNAs and regulatory

http://dx.doi.org/10.1016/j.bcmd.2015.01.009 1079-9796/© 2015 Elsevier Inc. All rights reserved.

Please cite this article as: D. Kaul, S. Sharma, High glucose-induced human cellular immune response is governed by miR-2909 RNomics, Blood Cells Mol. Diseases (2015), http://dx.doi.org/10.1016/j.bcmd.2015.01.009

2

D. Kaul, S. Sharma / Blood Cells, Molecules and Diseases xxx (2015) xxx–xxx

Fig. 1. Functional miR-2909 RNomics in HeLa cells: miR-2909 RNomics governed by regulatory-link between miR-2909 and KLF4-dependent genes involved in immunity, cancer, metabolism etc. (NCBI: Geo Accession No. GSE54949).

miRNAs (within human genome) has provided compelling evidence for the existence of flexible programming of immune cells depending upon time and space of their activation, the present study was addressed to understand as to how human peripheral blood mononuclear cells (PBMCs) exposed in vitro to high-glucose respond to this threat in terms of influencing their miR-2909 RNomics involving genes that link energy metabolism with immune response. Methods Cellular model and invitro culture design Human PBMCs were isolated from healthy volunteers, who were fasting for 12 h and had abstained from any medication for 2 weeks before blood donation, using Ficoll-Hypaque density gradient method as reported earlier [11]. Subsequently these cells were exposed to RPMI-1640 culture medium enriched with 25 mM glucose and incubated up to 72 h at 37 °C in 5% CO2 atmosphere. High glucose-induced cellular RelA translocation The cells from each culture well, at different incubation periods ranging from 0–72 h, were processed to prepare cytosolic and nuclear-protein extracts reported by us earlier [11]. These protein samples were subsequently subjected to SDS page followed by Western blotting and immunodetection using specific antibodies against phospho IkBα, RelA and Histone H3 (used as invariant control). High glucose programmed cellular genomic expression At the end of specified incubation periods, the cells from each well were processed for total cellular as well as small non-coding RNA extraction using miReasy mini kit (Qiagen). The extracted RNA was reverse transcribed using miScript reverse transcription kit (Qiagen). Differential expression of genes coding for miR-2909; PER3; NANOG;

GLUT1; PPAR-γ; Bmi-1; IFNγ; Cyclin ‘E’; C-myc; p53 was studied with real time PCR using gene specific primers (Table 1). β2M and U6 were used as invariant controls for expression studies of various genes and miR-2909 respectively. Total cellular protein was isolated using standard Lammelli's method [11,24]. The isolated proteins from each culture well were subjected to Western-blotting followed by immunodetection using specific antibodies against Cyclin ‘D’, KLF4, IL-17, CCL5, p53 and Histone H3 (used as an invariant control). Each band on the immunoblot was scanned densitometrically using Scion Image Analysis software, and the results were expressed as intensity ratio of target protein to histone protein taken as AU (arbitrary unit). KLF4-dependent genomics Normal human PBMCs were transfected with either KLF4 expression plasmid (Addgene plasmid 17967) or control plasmid with scrambledinsert using escort transfection reagent (Sigma) and incubated for 48 h in nutrient medium at 37 °C in 5% CO2 atmosphere. At the end of incubation period cells from each well were processed for RNA isolation and cDNA preparation as described earlier. The expression of various

Table 1 Primer sequences. Gene

Forward primer (5′ → 3′)

Reverse primer (5′ → 3′)

NANOG Bmi-1 GLUT1 PPAR-γ IFN-γ CCL5 p53 Cyclin E C-myc β2M

5′GTCTTCTGCTGAGATGCCTCACA3′ 5′TCTGCAGCTCGCTTCAAGAT3′ 5′TGAGCATCGTGGCCATCTTT3′ 5′GCATTATGAGACATCCCC3′ 5′GTTTGGGTTCTCTTGGCTGTT3′ 5′CGTGCCCACATCAAGGAGTA3′ 5′GAAGACCCAGGTCCAGATGA3′ 5′GACATACTTAAGGGATCAGC3′ 5′CGGAACTCTTGTGCGTAAGG3′ 5′GAATTGCTATGTGTCTGGGT3′

5′CTTCTGCGTCACACCATTGCTAT3′ 5′AGTGGTCTGGTCTTGTGAAC3′ 5′CCGGAAGCGATCTCATCGAA3′ 5′GCGATTCCTTCACTGATAAC3′ 5′CTCCTTTTTCGCTTCCCTGTTTT3′ 5′CTTCTCTGGGTTGGCACACA3′ 5′CTGCCCTGGTAGGTTTTCTG3′ 5′GGGGACTTAAACGCCACTTA3′ 5′GGATTGAAATTCTGTGTAACTGC3′ 5′CATCTTCAAACCTCCATGATG3′



genes coding for Bmi-1, NANOG, CCL5, IL-17, C-myc, PER3, PPAR-γ was analysed at transcriptional level with real time PCR. Bioinformatics analysis Promoter sequence for NANOG gene was obtained from UCSC genome browser (https://genome.ucsc.edu/cgi-bin/hgTables) and putative binding sites for various transcription factors on this promoter were retrieved from JASPAR database (http://jaspar.cgb.ki.se) and TFSEARCH (http://www.cbrc.jp/research/db/TFSEARCH.html) at a default threshold score of 85.0. Statistical analysis Statistical analyses were performed by SPSS Windows version 19. Statistical comparisons were made between two groups by student t-test and Man Whitney and between multiple groups by one way analysis of variance (ANOVA) followed by appropriate post hoc test. Differences were considered significant at p b 0.05. Results Exposure of high glucose to human PBMCs up to 72 h resulted in oscillatory expressional pattern of genes coding for miR-2909 and PER3 depicting an inverse relationship between their genomic expression at 48 h and 72 h (Fig. 2a). This phenomenon was accompanied by significantly higher NFkB family-member RelA nuclear translocation at 48 h exposure time as compared that either at 0 h or 72 h exposure

3

time (Fig. 2b) indicating a direct correlation between NFkB nuclear translocation and transcriptional expression of miR-2909 which is in conformity with our earlier observations [11]. Interestingly, human PBMCs exposed to high glucose at 48 h time interval exhibited significantly higher expression of genes coding for NANOG; Bmi-1; PPAR-γ; GLUT1 (Fig. 2c–f) as compared to that observed in either at 0 h or 72 h time interval (Fig. 2c–f). Similar expressional pattern could be observed at 48 h exposure of PBMCs to high glucose as far as genes coding for C-myc, Cyclin ‘E’ and Cyclin ‘D’ are concerned (Fig. 3b,c). However, translational expression of KLF4 was significantly reduced at 48 h interval followed by significant increase at 72 h exposureinterval with respect to that observed in 0 h control cells (Fig. 3d) whereas p53 translational expression was significantly increased at 72 h time interval as compared to either 0 h or 48 h exposure interval (Fig. 3d). Translational expression of CCL5 was significantly increased at 48 h exposure time and reduced at 72 h exposure time as compared to cells at 0 h exposure (Fig. 3d). Both IFN-γ and IL-17 showed time dependent increase in the transcriptional and translational expression respectively up to 72 h exposure time (Fig. 3a,d). Since in our earlier studies we have shown that NFkB nuclear translocation is required for miR-2909 expression and subsequent regulation of genes through repression of KLF4 gene accompanied by up-regulation of Sp1 gene [11], the experiments were directed to understand as to how ectopic expression of KLF4 within normal human PBMCs does regulate the genes coding for PER3; Bmi-1; CCL5; NANOG; PPAR-γ; C-myc and IL-17. The results of such a study showed that the increased expression of KLF4 results in the increased expression of genes coding for PER3; NANOG; IL-17 as well as decreased expression of genes coding for Bmi-1; CCL5

Fig. 2. Epigenomic programming of human PBMCs induced by high glucose as a function of time: 2a) Reciprocal oscillatory expression pattern of genes coding for miR-2909 and PER3 as a function of incubation time up to 72 h; 2b) cellular status of NFkB activation coupled with RelA nuclear translocation as a function of time; 2c–f) transcriptional expression of genes coding for NANOG, Bmi-1, PPAR-γ and GLUT1 as a function of incubation time. For translational expression Histone H3 was used as an invariant control in our experiments because this gene did not show any changes in its expression due to glucose exposure. Each bar represents mean ± SD of experiments done in triplicate. Statistical significance is shown as ** p b 0.05.


4


Fig. 3. Epigenomic programming of human PBMCs induced by high glucose: 3a,b) Time-dependent transcriptional expression of genes coding for IFN-γ, C-myc,Cyclin ‘E’ and p53; 3c, d) translational expression of genes coding for Cyclin ‘D’; KLF4; p53; CCL5 and IL-17 as a function of high-glucose exposure time. Histone H3 was used as an invariant control in our experiments because this gene did not show any changes in its expression due to glucose exposure. Each bar represents mean ± SD of experiments done in triplicate. Statistical significance is shown as ** p b 0.05.

and C-myc (Fig. 4a,b) whereas there was no significant effect of KLF4 gene ectopic-expression upon the PPAR-γ gene expression (Fig. 4b). Discussion Cyclic regulation of aerobic-respiration and aerobic-glycolysis provides metabolic energy to human blood mononuclear cells for tailoring their immune-response in form of helpers or warriors or regulators. This cyclic regulation of these two major metabolic processes would have been impossible without mechanism responsible for the regulation of the NFkB family member RelA translocation to mitochondria by tumour suppressor genes coding for p53 and KLF4 [5,25]. Existence of functional p53 or KLF4 restricts immune cells to employ aerobic glycolysis for shaping their immune response to either endogenous molecular threat or microbial infection [5,11]. It is in this context, that the results reported here assume importance because NFkB induced miR-2909 expression ensures suppression of KLF4 [26] (Fig. 3d) resulting in the up-regulation of Bmi-1 gene expression (Fig. 2d) which, in turn, ensures degradation of p53 protein as well as availability of NFkB family member RelA for initiating aerobic-glycolysis [5]. This phenomenon observed in PBMCs exposed to high glucose up to 48 h incubation, was accompanied by the over-expression of genes coding for miR-2909; Bmi-1; NANOG; GLUT1; PPAR-γ; C-myc; cyclin ‘E’; Cyclin ‘D’; CCL5 and IFN-γ as well as down-regulation of genes coding for PER3 and p53 at the transcriptional and translational level respectively (Figs. 2 and 3). However such observed genomic-expression pattern in human PBMCs, exposed to high glucose, was abolished at 72 h

incubation period resulting in the down-regulation of genes coding for miR-2909; Bmi-1; C-myc; Cyclin ‘D’; CCL5; PPAR-γ; GLUT1 as well as upregulation of genes coding for KLF4; p53; IL-17; and IFN-γ (Figs. 2 and 3). At 48 hour time interval, p53 protein is degraded because of high expression of Bmi-1 (widely known to facilitate the degradation of p53 protein) even if at this incubation time-point the transcription expression of p53 is very high. This phenomenon was further strengthened by the fact that at 72 h incubation period, the reduction in Bmi-1 gene expression is accompanied by the high p53 protein expression. It is not unlikely that the biphasic effect of high glucose upon human PBMCs at 48 h and 72 h time intervals may be as a result of increase in the expression of genes coding for PPAR-γ; IFN-γ and C-myc observed at 48 h interval keeping in view the fact that these genes have the inherent capacity to increase the expression of KLF4 and p53 observed at 72 h interval [13,27,28] resulting in the inactivation of NFkB [3,11]. However more data is required before such a possibility becomes acceptable. Further, KLF4 is known to induce IL-17 gene expression and p53 has been shown to inhibit IFN-γ translational expression [16,20]. Based upon our results, human PBMCs respond to the threat of high glucose by tailoring immune-response involving genes coding for CCL5; IFN-γ and IL-17 (Fig. 3d). CCL5 has been shown to induce the proliferation of Th-1 cells that secrete IFN-γ as well as help in the recruitment of NK cells and regulatory T-cells [25, 29] whereas IL-17 secreting Th17 cells has been shown not only to play crucial protective role in mobilizing host immunity to extracellular and intracellular microbial pathogens but also contribute to autoimmune response [27]. However, it is not clear as to how human PBMCs



5

capacity to inhibit NFkB activation [3,11,12]. Hence based upon our results reported here, we have attempted to propose a pathway (Fig. 5) that for the first time links NFkB activated miR-2909 RNomics with energy-metabolic state that favours aerobic-glycolysis resulting in the immune response involving CCL5; IFN-γ and IL-17 within human PBMCs. References

Fig. 4. Gene-regulation by KLF4: 4a) Bioinformatic analysis of NANOG gene promoter indicating both KLF4 and SP1 binding sites; 4b) differential transcriptional regulation of genes coding for KLF4, PER3, Bmi-1, CCL5, NANOG, PPAR-γ, Cmyc and IL-17 within human PBMCs having high ectopic expression of KLF4 gene as compared to that in the control cells. Each bar represents mean ± SD of experiments done in triplicate. Statistical significance is shown as ** p b 0.05.

fight high glucose exposure through such an immune response involving CCL5, IFN-γ and IL-17. Interestingly, PPAR-γ is shown to upregulate GLUT1 gene expression [14] whereas p53 has been shown to repress the glucose transporter genes coding for GLUT1 and GLUT4 [3]. Further, KLF4, p53 and PPAR-γ have been shown to have inherent

[1] P.P. Hsu, D.M. Sabatini, Cancer cell metabolism: Warburg and beyond, Cell 134 (2008) 703–707. [2] M.M. Delmastro-Greenwood, J.D. Piganelli, Changing the energy of an immune response, Am. J. Clin. Exp. Immunol. 2 (2013) 30–54. [3] A.M. Puzio-Kuter, The role of p53 in metabolic regulation, Genes Cancer 2 (2011) 385–391. [4] C.R.H. Tanupriya Contractor, p53 negatively regulates transcription of the pyruvate dehydrogenase kinase Pdk2, Cancer Res. 72 (2011) 560–567. [5] R.F. Johnson, I.-I. Witzel, N.D. Perkins, p53-dependent regulation of mitochondrial energy production by the RelA subunit of NF-κB, Cancer Res. 71 (2011) 5588–5597. [6] G. Bonizzi, A. Cicalese, A. Insinga, P.G. Pelicci, The emerging role of p53 in stem cells, Trends Mol. Med. 18 (2012) 6–12. [7] H.S. Yoon, X. Chen, V.W. Yang, Kruppel-like factor 4 mediates p53-dependent G1/S cell cycle arrest in response to DNA damage, J. Biol. Chem. 278 (2003) 2101–2105. [8] M. Calao, E.O. Sekyere, H.J. Cui, B.B. Cheung, W.D. Thomas, J. Keating, et al., Direct effects of Bmi1 on p53 protein stability inactivates oncoprotein stress responses in embryonal cancer precursor cells at tumor initiation, Oncogene 32 (2013) 3616–3626. [9] T. Yu, X. Chen, W. Zhang, D. Colon, J. Shi, D. Napier, et al., Regulation of the potential marker for intestinal cells, Bmi1, by β-catenin and the zinc finger protein KLF4: implications for colon cancer, J. Biol. Chem. 287 (2012) 3760–3768. [10] H.R. Siddique, M. Saleem, Role of BMI1, a stem cell factor, in cancer recurrence and chemoresistance: preclinical and clinical evidences, Stem Cells Dayt. Ohio. 30 (2012) 372–378. [11] M. Arora, D. Kaul, Y.P. Sharma, Human coronary heart disease: importance of blood cellular miR-2909 RNomics, Mol. Cell. Biochem. 392 (2014) 49–63. [12] A.H.V. Remels, R.C.J. Langen, H.R. Gosker, A.P. Russell, F. Spaapen, J.W. Voncken, et al., PPARgamma inhibits NF-kappaB-dependent transcriptional activation in skeletal muscle, Am. J. Physiol. Endocrinol. Metab. 297 (2009) E174–E183. [13] M. Cekanova, S.-H. Lee, M.F. McEntee, S.J. Baek, MCC-555-induced NAG-1 expression is mediated in part by KLF4, Eur. J. Pharmacol. 637 (2010) 30–37. [14] C. Nugent, J.B. Prins, J.P. Whitehead, D. Savage, J.M. Wentworth, V.K. Chatterjee, et al., Potentiation of glucose uptake in 3T3-L1 adipocytes by PPAR gamma agonists is maintained in cells expressing a PPAR gamma dominant-negative mutant: evidence for selectivity in the downstream responses to PPAR gamma activation, Mol. Endocrinol. Baltim. Md. 15 (2001) 1729–1738. [15] M.J. Costa, A.Y.-L. So, K. Kaasik, K.C. Krueger, M.L. Pillsbury, Y.-H. Fu, et al., Circadian rhythm gene period 3 is an inhibitor of the adipocyte cell fate, J. Biol. Chem. 286 (2011) 9063–9070. [16] L. Lebson, A. Gocke, J. Rosenzweig, J. Alder, C. Civin, P.A. Calabresi, et al., Cutting edge: the transcription factor Kruppel-like factor 4 regulates the differentiation of

Fig. 5. Proposed high glucose induced miR-2909 RNomics pathway within human PBMCs depicting an epigenomic link between their energy metabolic state and immune response involving CCL5; IFN-γ and IL-17.


6


[17] [18]

[19]

[20]

[21] [22]

Th17 cells independently of RORγt, J. Immunol. Baltim. MD 1950 (185) (2010) 7161–7164. S. Pestka, A dance between interferon-alpha/beta and p53 demonstrates collaborations in tumor suppression and antiviral activities, Cancer Cell 4 (2003) 85–87. H. Kawashima, H. Takatori, K. Suzuki, A. Iwata, M. Yokota, A. Suto, et al., Tumor suppressor p53 inhibits systemic autoimmune diseases by inducing regulatory T cells, J. Immunol. Baltim. MD 1950 (191) (2013) 3614–3623. R.-W. Chen, P.A. Saunders, H. Wei, Z. Li, P. Seth, D.-M. Chuang, Involvement of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and p53 in neuronal apoptosis: evidence that GAPDH is upregulated by p53, J. Neurosci. 19 (1999) 9654–9662. C.-H. Chang, J.D. Curtis, L.B. Maggi, B. Faubert, A.V. Villarino, D. O'Sullivan, et al., Posttranscriptional control of T cell effector function by aerobic glycolysis, Cell 153 (2013) 1239–1251. Y. Yang, A.E. Lovett-Racke, M.K. Racke, Regulation of immune responses and autoimmune encephalomyelitis by PPARs, PPAR Res. 2010 (2010). A. Carbo, R. Hontecillas, S. Hoops, B. Kronsteiner-Dobramysl, P. Lu, K. Wendelsdorf, et al., PPAR{gamma} activation drives Th17 cells into a Treg phenotype, J. Immunol. 188 (2012) 163.7.

[23] D. Kaul, S. Sharma, M. Sharma, M. Arora, M. Arora, Arsenic programmes cellular genomic-immunity through miR-2909 RNomics, Gene 536 (2014) 326–331. [24] M. Sharma, S. Sharma, M. Arora, D. Kaul, Regulation of cellular Cyclin D1 gene by arsenic is mediated through miR-2909, Gene 522 (2013) 60–64. [25] X. Qian, J. Zhang, J. Liu, Tumor-secreted PGE2 inhibits CCL5 production in activated macrophages through cAMP/PKA signaling pathway, J. Biol. Chem. 286 (2011) 2111–2120. [26] D. Malik, D. Kaul, N. Chauhan, R.K. Marwaha, miR-2909-mediated regulation of KLF4: a novel molecular mechanism for differentiating between B-cell and T-cell pediatric acute lymphoblastic leukemias, Mol. Cancer 13 (2014) 175. [27] Z.Y. Chen, J.L. Shie, C.C. Tseng, STAT1 is required for IFN-gamma-mediated gutenriched Krüppel-like factor expression, Exp. Cell Res. 281 (2002) 19–27. [28] M.S. Lindström, K.G. Wiman, Myc and E2F1 induce p53 through p14ARFindependent mechanisms in human fibroblasts, Oncogene 22 (2003) 4993–5005. [29] N. Dilek, N. Poirier, C. Usal, B. Martinet, G. Blancho, B. Vanhove, Control of transplant tolerance and intragraft regulatory T cell localization by myeloid-derived suppressor cells and CCL5, J. Immunol. Baltim. MD 1950 (188) (2012) 4209–4216.
