Stimulatory Influences of Far Infrared Therapy on the Transcriptome and Genetic Networks of Endothelial Progenitor Cells Receiving High Glucose Treatment

Abstract

Background

Endothelial progenitor cells (EPCs) play a fundamental role in vascular repair and angiogenesis- related diseases. It is well-known that the process of angiogenesis is faulty in patients with diabetes. Long-term exposure of peripheral blood EPCs to high glucose (HG-EPCs) has been shown to impair cell proliferation and other functional competencies. Far infrared (FIR) therapy can promote ischemia-induced angiogenesis in diabetic mice and restore high glucose-suppressed endothelial progenitor cell functions both in vitro and in vivo. However, the detail mechanisms and global transcriptome alternations are still unclear.

Methods

In this study, we investigated the influences of FIR upon HG-EPC gene expressions. EPCs were obtained from the peripheral blood and treated with high glucose. These cells were then subjected to FIR irradiation and functional assays.

Results

Those genes responsible for fibroblast growth factors, Mitogen-activated protein kinases (MAPK), Janus kinase/signal transducer and activator of transcription and prostaglandin signaling pathways were significantly induced in HG-EPCs after FIR treatment. On the other hand, mouse double minute 2 homolog, genes involved in glycogen metabolic process, and genes involved in cardiac fibrosis were down-regulated. We also observed complex genetic networks functioning in FIR-treated HG-EPCs, in which several genes, such as GATA binding protein 3, hairy and enhancer of split-1, Sprouty Homolog 2, MAPK and Sirtuin 1, acted as hubs to maintain the stability and connectivity of the whole genetic network.

Conclusions

Deciphering FIR-affected genes will not only provide us with new knowledge regarding angiogenesis, but also help to develop new biomarkers for evaluating the effects of FIR therapy. Our findings may also be adapted to develop new methods to increase EPC activities for treating diabetes-related ischemia and metabolic syndrome-associated cardiovascular disorders.

INTRODUCTION

Deregulated angiogenesis has been proven to be involved in major complications in diabetes, leading to a diminished ability for collateral vessels to be formed in response to ischemia in the heart and peripheral tissues.1 The best way to maintain vascular access thereby becomes a critical medical issue to address in diabetes patients. Mounting evidence indicates that the administration of angiogenic growth factors increases nutrient perfusion through neovascularization, and also reduces ischemia-related organ damage.2 Nevertheless, clinical evidence has also shown that certain patients appear to be refractory to administration of exogenous growth factors and fail to develop collateral circulation after tissue ischemia.3 It is therefore necessary to develop alternative therapeutic approaches to restore endogenous angiogenesis and vasculogenesis more efficiently in patients.

Far infrared (FIR) therapy, a noninvasive and convenient therapeutic modality, is known to improve both blood flow and endothelial function.4-6 FIR rays are defined as invisible electromagnetic waves with a wavelength between 5.6 to 1000 μm that can be perceived as heat by thermoreceptors in the surrounding skin.7,8 Currently, FIR irradiation is implicated in the treatment of ischemic lesions and necrosis in skin tissue as a result of trauma, diabetes, and peripheral arterial occlusive disease. FIR treatment reduces the frequency of cardiovascular diseases (CVDs) as well as improves arteriovenous fistula (AVF) access flow in hemodialysis patients through both its thermal and non-thermal (endothelial-improving, anti-inflammatory, antiproliferative, antioxidative) effects.4 FIR radiation improves ventricular arrhythmias and endothelial function in patients with heart disease, and enhances access flow and patency of AVF in hemodialysis patients.5,9,10 In diabetic mice, FIR therapy also promotes ischemia-induced angiogenesis.6

Enhancement of the regenerative capacity of the injured endothelium appears to be one way to reduce the incidence of CVDs or AVF malfunction lesions. According to the traditional view, endothelium integrity is maintained by neighboring mature endothelial cells which migrate and proliferate to restore the injured endothelial cells. However, a series of clinical and basic studies prompted by the discovery of bone marrow-derived endothelial progenitor cells (EPCs) have demonstrated that the injured endothelial monolayer may be regenerated partly by circulating EPCs.11 These circulating EPCs are mobilized endogenously and triggered by tissue ischemia, or exogenously by cytokine stimulation such as vascular endothelial growth factor, matrix metalloproteinase-9 (MMP-9), and stromal cell-derived factor-1 (SDF-1).11-13 Reduced levels of circulating EPCs independently predict atherosclerotic disease progression and development of cardiovascular events.14 Clinical studies have demonstrated that levels of circulating EPCs are associated with vascular endothelial function and cardiovascular risk factors, and help to identify patients at increased cardiovascular risk.15,16 There are several ways to increase levels of circulating EPCs and improve their function by pharmacological strategies and lifestyle modification.13,17 Many studies have evaluated the possibility of EPC transplantation for improving disease recovery; for example, one investigation described the long-term stroke outcome in a mouse model of transient middle cerebral artery occlusion (MCAO).18,19

Recent evidence has indicated that levels of circulating EPCs and their associated arterial stiffness were closely related to glycemic control in patients with type 2 diabetes mellitus (DM).6 Long-term exposure of EPCs to elevated levels of glucose enhances cellular senescence and decreases cell numbers and functional competencies via nitric oxide (NO)-related mechanisms.20 These findings provide a rationale for developing potential therapeutic approaches regarding hyperglycemia-related EPC dysfunction. Far infra-red therapy, on the other hand, can promote ischemia-induced angiogenesis in diabetic mice and restores high glucose-suppressed endothelial progenitor cell functions both in vitro and in vivo.6 Pilot mechanistic studies showed that FIR radiation upregulates high glucose-impaired endothelial nitric oxide synthase (eNOS) production, as well as the phosphorylation and activation of protein kinase B (AKT), extracellular regulated protein kinases (ERK) and p38 Mitogen-activated protein kinases (MAPK).6 However, the detail mechanisms and global transcriptome alternations of this process have not yet been addressed. Deciphering FIR-affected genes will not only inform us of new knowledge relating to angiogenesis, but also help to develop new biomarkers for evaluating FIR therapy effects in vitro and in vivo. So, in this study, the main objective was to investigate the influences of FIR upon EPC gene expression profiles, working with the hypothesis that the levels of specific genes must be influenced in activated EPCs after FIR stimulation.

MATERIALS AND METHODS

Human EPC isolation and cultivation

All patients gave informed consent, and the study was approved by the local research ethics committee. The protocols of this study are consistent with the ethical guidelines of the 1975 Helsinki Declaration. Peripheral blood EPC isolation and characterization were done as described previously with minor modifications.20,21 Cell and motility assays were performed by Transwell assays as described.21 In brief, peripheral blood samples were obtained from healthy donors, and total mononuclear cells (MNCs) were isolated by density gradient centrifugation with Histopaque-1077 (1.077 g/ml, Sigma, St. Louis, MO, USA). MNCs (5 × 106) were plated in 2 ml endothelial growth medium (EGM-2 MV Cambrex Corp., East Rutherford, NJ, USA) with supplements on fibronectin-coated 6-well plates. After 4 days of culturing, the medium was changed, and EPC (or ECFCs, endothelial colonies forming cells) emerged 2-4 weeks after the start of the MNCs culture. These EPCs exhibited a cobblestone morphology and monolayer growth pattern typical of mature endothelial cells at confluence.11,21 For high glucose culture, EPCs/ECFCs were treated with 25 mM of glucose for 4 days before being subjected to functional assays or FIR irradiation as described.11

ECFC characterization

Tube formation assays were performed on EPCs/ECFCs to assess their capacity for EPC vasculogenesis, which is believed to be important in new vessel formation. In brief, the in vitro tube formation assay was performed by thawing Matrigel at 4 °C overnight, and then by placing it into a 96-well plate at 37 °C for 1 hour to allow the matrix solution to solidify. EPCs were harvested with trypsin/ethylenediaminetetraacetic acid, and 1 × 10⁴ EPCs/ECFCs or human umbilical vein endothelial cells (HUVECs) were placed on Matrigel with EGM-2 medium or serum-free Dulbecco’s modified eagle medium and incubated at 37 °C for 3 hours. Tubule formation was inspected under an inverted light microscope (100x), and four representative fields were taken. For 3D angiogenesis assay, collagen type I acidic solution was mixed with 1/2 volume of basic conditioned medium with 0.2 ug/ml SDF-1α (R&D Systems, Minneapolis, MN USA) and solidify for 30 minutes in a 96-well plate at 37 °C in a 5% CO₂ incubator. 10⁵ cells per well were seeded and assayed.

FIR treatment on EPC

Treatment of EPCs with FIR was performed as described.5,6 In brief, human peripheral blood EPCs were exposed to FIR radiation using a WS TY101 FIR emitter (WS Far Infrared Medical Technology Co, Ltd.). This device generates electromagnetic waves with wavelengths between 3 and 25 μm (peak 5 to 6 μm). The radiator was set at a height of 25 cm above the bottom of the tissue culture plates and cells were exposed to FIR radiation for 30 minutes according to the parameters determined previously.6

RNA isolation and quantitative reverse transcription polymerase chain reaction (RT-PCR)

Total mRNA were extracted by the RNeasy mini kit (Qiagen, Germany), and 100 ng to 1 μg of total RNA were subjected to reverse transcription using a First cDNA Synthesis kit (Fermentas, USA). For quantitative real-time PCR analysis, human pre-messenger RNA sequence was obtained from the NCBI (National Center for Biotechnology Information) AceView program (www.ncbi.nlm.nih.gov/AceView/). All primers were designed to cross introns using the Primer3 website (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi/) or Primer Express software (Applied Biosystems, CA, USA). Thermodynamics and primer specificity analysis were performed by the Vector NTI suite (Invitrogen, CA, USA) and the NCBI reverse e-PCR program (http://www.ncbi.nlm.nih.gov/sutils/e-pcr/reverse.cgi/). Real-time PCR reactions were performed using Maxima^TM SYBR Green qPCR Master Mix (Fermentas, USA), and the specific products of the PCRs were detected and analyzed using a StepOne^TM sequence detector (Applied Biosystems, USA). The expression level of each gene was normalized against the expression level of glyceraldehyde 3-phosphate dehydrogenase (GAPDH).

Gene expression microarrays and data analysis

Array data for FIR-treated human peripheral blood EPCs were deposited in the Gene Expression Omnibus database with an accession number of GSE37044. Total RNA collection, cRNA probe preparation, array hybridization and data analysis were done as previously described.22 Return merchandise authorization (RMA) log expression units were calculated from Affymetrix^TM HG-U133 Plus 2.0 whole genome array data using the ‘affy’ package included in the Bioconductor (http://www.bioconductor.org/) suite of software for the R statistical programming language (http://www.r-project. org/). The default RMA settings were used to background correct, normalize and summarize all expression values. Significant differences between the sample groups were identified using the ‘limma’ (Linear Models for Microarray Analysis) package of the Bioconductor suite, and an empirical Bayesian moderated t-statistic hypothesis test between the two specified phenotypic groups was performed. To control for multiple testing errors, we then applied a false discovery rate algorithm to these p values in order to calculate a set of q values, thresholds of the expected proportion of false positives, or false rejections of the null hypothesis.

Gene enrichment analysis was performed using the Gene Ontology database with the WebGestalt (http://bioinfo.vanderbilt.edu/webgestalt/).23 Differential gene expression profiles were imported into the Ingenuity Pathways Analysis (IPA) software (Ingenuity Systems, Redwood City, CA; http://www.ingenuity.com/) to obtain canonical pathways, biological functions, and genetic networks. The knowledge base behind IPA was built upon properly accumulated scientific evidence, manually curated from thousands of journal articles, textbooks, and other data sources. After a list of signature genes was uploaded, interaction among focus genes and interaction among interacting genes and molecules from the knowledge base were used to combine genes into networks according to their probability of having more focus genes than expected by chance. Networks are scored on the basis of the number of uploaded signature genes they contain. The network score is based on the hypergeometric distribution and is calculated with the right-tailed Fisher’s exact test. The score is the negative log of this p value. The higher the score, the lower the probability of finding the observed number of uploaded signature genes in a given network by random chance.

RESULTS

FIR radiation stimulates angiogenesis-related activities in human EPCs

EPCs (late EPCs, ECFCs) were obtained from the peripheral blood of healthy subjects using an established protocol.20,21 These cells have a cobblestone-like morphology, which is similar to mature endothelial cells (Figure 1A). Isolated EPCs/ECFCs, just like HUVECs, formed capillary-like structures in a 3-dimensional (3D) angiogenesis assay (Figure 1B).

Figure 1.

Effects of FIR therapy on eNOS, EPC migration, and tube formation in vitro. (A) Cultivation of EPCs/ECFCs. Blood mononuclear cells were isolated and plated on fibronectin-coated culture dishes. Twenty-one days after plating, late EPCs (EPCs, ECFCs) with a cobblestone-like morphology were selected, reseeded, and grown to confluence (left). (B) Both EPCs and matured HUVECs formed vessel structures in 3D-angiogenesis assays. Arrows: tubular structures; arrowheads: tip cell-like structures. (C) Induction of eNOS mRNA production in HG-EPCs. EPCs cultured in 25 mM glucose for 4 days were subjected into FIR irradiation for 30 min before RNA extraction. The enhanced expression of eNOS in HG-EPCs after FIR irradiation was verified by RT-qPCR. *: p < 0.05 by 2-way ANOVA. Data are means ± SEM; n = 3 in each experiment. (D) Modified Boyden chamber assay was used to assess the effects of FIR radiation on the motility of HG-EPCs. VEGF was used as a chemoattracting factor. (E) A MatrixGel tube formation assay to monitor the in vitro angiogenesis ability of FIR-treated and control HG-EPCs. (F) Validation of microarray data. RT-qPCR assays were performed either on the same batch of RNAs used in array hybridization (left) or an independent batch of HG-EPCs RNAs were used (right). *: p < 0.05 by 2-way ANOVA. Data are means ± SEM. ECFCs, endothelial colonies forming cells; eNOS, endothelial nitric oxide synthase; EPCs, endothelial progenitor cells; FIR, far infrared; HUVECs, human umbilical vein endothelial cells; SEM, standard error of mean.

Here, peripheral blood EPCs were treated with high glucose (25 mM) for 4 days (HG-EPCs) before 30 minute of FIR treatment was administered. Figure 1C shows that FIR radiation administrated on HG-EPCs upregulated significantly the level of eNOS transcripts. We also used the modified Boyden chamber assay (Figure 1D) and the MatriGel tube formation assay (Figure 1E) to verify the increased migration and microvasculature formation functions of FIR-irradiated HG-EPCs. Compared with the control group, the migratory and microtubule formation abilities in HG-EPCs receiving FIR irradiation were significantly increased.

Global gene expression profile alternations after FIR radiation on EPCs

To reveal the underlying mechanisms of FIR-mediated phenotypes, we identified signature genes for FIR-treated HG-EPCs as well as control un-irradiated cells using whole genome gene expression microarrays. In total, 627 probe sets were specifically enriched in HG-EPCs after FIR treatment, while another 623 ones were repressed by FIR. Up- and down-regulated transcripts with more than 2-fold changes are shown in Table 1 and Table 2, respectively. Results of RT-qPCR validation performed on 2 independent batches of HG-EPCs were shown in Figure 1F. Among the up-regulated genes, RDX (radixin) and ARHGEF7 (Rho guanine nucleotide exchange factor (GEF) 7) are involved in the enhanced metastatic potential of cancer cells.24,25 VIP (vasoactive intestinal peptide) enhances angiogenesis in rats with focal cerebral ischemia.26 and increases VEGF expression.27 It has been reported that FIR treatment on high glucose-repressed EPCs induced p38 MAPK phosphorylation,6 and MAP3K8 (mitogen-activated protein kinase kinase kinase 8) is also in our gene list (Table 1, genes discussed here are indicated by asterisks). FGF18 (fibroblast growth factor 18) is wildly expressed in the cardiovascular tissues and can induce endothelial cells migration, maybe via a short transient activation of the MAPK pathway.28 Another member of the FGF signaling pathway, FRS2 (fibroblast growth factor receptor substrate 2), was also induced by FIR irradiation (Table 1). Another angiogenesis pathway that may be involved in FIR functions is the famous prostaglandin signaling pathway: PTGS2 (prostaglandin-endoperoxide synthase 2, also known as cyclooxygenase-2 (COX2)), PTGER3 and PTGER4 (prostaglandin E receptor 3 and 4, respectively). All were increased after FIR irradiation. There has been no published report to date that shows the prostaglandin pathway is involved in FIR functions.

Table 1. Known genes stimulated by FIR irradiation in PB-EPC (> 2 folds) .

Probe set ID	Uni gene ID	Gene title	Gene symbol	Location
238825_at	Hs.135167	Acidic repeat contauning	ACRC	chrXq13.1
235412_at	Hs.508738	Rho guanine nucleotide exchange factor (GEF) 7	*ARHGEF7	chr13q34
204194_at	Hs.154276	BTB and CNC homology1, basic leucine zipper transcription factor 1	BACH1	chr21q22.11
225551_at	Hs.368353	Consortin, connexin sorting protein	CNST	chr1q44
239203_at	Hs.396189	Chromosome 7 open reading frame 53	C7 or f53	chr7q31.1
206727_at	Hs.654443	Complement component 9	C9	chr5p14-p12
1555411_a_at	Hs.4859	Cyclin L1	CCNL1	chr3q25.32
204995_at	Hs.500015	Cyclin-dependent kinase 5, regulatory subunit 1(p35)	CDK5R1	chr17q11.2
215318_at	Hs.687692	N4BP2L2 intronic transcript 1 (non-protein coding)	N4BP2L2-IT1	chr13q12-q13
214683_s_at	Hs.433732	CDC-like kinase 1	CLK1	chr2q33
210346_s_at	Hs.406557	CDC-like kinase 4	CLK4	chr5q35
224831_at	Hs.127126	Cytoplasmic polyadenylation element binding protein 4	CPEB4	chr5q21
206424_at	Hs.150595	Cytochrome P450, family26, subfamily A, polypeptide 1	CYP26A1	chr10q23-q24
227335_at	Hs.517172	Death inducer-obliterator 1	DIDO1	chr20q13.33
208891_at	Hs.298654	Dual specificity phosphatase 6	DUSP6	chr12q22-q23
224973_at	Hs.10784	Family with sequence similarity 46, member A	FAM46A	chr6q14
206987_x_at	Hs.87191	Fibroblast growth factor 18	*FGF18	chr5q34
228250_at	Hs.483329	Rap guanine nucleotide exchange factor (GEF) 6///folliculin interacting protein 1	FNIP1///RAPGEF6	chr5q23.3///chr5q31.1
226045_at	Hs.593446	Fibroblast growth factor receptor substrate 2	*FRS2	chr12q15
204472_at	Hs.654463	GTP binding protein overexpressed in skeletal muscle	GEM	chr8q13-q21
203159_at	Hs.116448	Glutaminase	GLS	chr2q32-q34
211998_at	Hs.533624	H3 histone, family 3B (H3.3B)	H3F3B	chr1q41
209315_at	Hs.378532	HBS1-like (S. cerevisiae)	HBS1L	chr6q23-q24
210387_at	Hs.591809	Histone cluster 1, H2bg	HIST1H2BG	chr6p21.3
214438_at	Hs.74870	H2.0-like homeobox	HLX	chr1q41-q42.1
201566_x_at	Hs.180919	Inhibitor of DNA binding 2, dominant negative helix-loop-helix protein	ID2	chr2p25
221763_at	Hs.413416	Jumonji domain containing 1C	JMJD1C	chr10q21.2-q21.3
206765_at	Hs.1547	Potassium inwardly-rectifying channel, subfamily J, member 2	KCNJ2	chr17q23.1-q24.2
225582_at	Hs.523252	Inositol 1, 4, 5-trisphosphate receptor interacting protein	ITPRIP	chr10q25.1
226370_at	Hs.495854	Kelch-like 15 (Drosophila)	KLHL15	chrXp22.1-p21
207247_s_at	Hs.522845	Zinc finger protein, X-linked///zinc finger protein	ZFX///ZFY	chrXp21.3///chrYp11.3
226682_at	Hs.560343	RAR-related orphan receptor A	RORA	chr15q21.3
205027_s_at	Hs.432453	Mitogen-activated protein kinase kinase kinase 8	*MAP3K8	chr10p11.23
1556382_a_at	Hs.555985	NMDA receptor regulated 1	NARG1	chr4q31.1
223218_s_at	Hs.319171	Nuclear factor of kappaa light polypeptide gene enhancer in B-cells inhibitor, zeta	NFKBIZ	chr3p12-q12
224958_at	Hs.462598	Nuclear fragile X mental retardation protein interacting protein 2	NUFIP2	chr17q11.2
231838_at	Hs.641481	Poly(A) binding protein, cytoplasmic 1-like	PABPC1L	chr20
1569431_at	Hs.696131	Platelet-activating factor acetylhydrolase, isoform lb, beta subunit 30 kDa	PAFAH1B2	chr11q23
214415_at	Hs.652174	Plasminogen-like B2///plasminogen-like B1	PLGLB1///PLGLB2	chr2p11-q11///chr2p11.2
242804_at	Hs.368454	Polymerase (DNA directed) nu	POLN	chr4p16.3
201702_s_at	Hs.106019	Protein phosphatase 1, regulatory (inhibitor) subunit 10	PPP1R10	chr6p21.3
202014_at	Hs.631593	Protein phosphatase 1, regulatory (inhibitor) subunit 15A	PPP1R15A	chr19q13.2
235987_at	Hs.12250	Protein kinase, X-linked, pseudogene 1	PRKXP1	chr15q26.3
1554997_a_at	Hs.196384	Prostaglandin-endoperoxide synthase 2	*PTGS2	chr1q25.2-q25.3
205178_s_at	Hs.188553	Retinoblastoma binding protein 6	RBBP6	chr16p12.2
244739_at	Hs.263671	Radxin	*RDX	chr11q23
223707_at	Hs.523463	Ribosomal protein L27a	RPL27A	chr11p15
218041_x_at	Hs.221847	Solute carrier family 38, member 2	SLC38A2	chr12q
213164_at	Hs.302742	Solute carrier family 5 (inositol transporters), member 3	SLC5A3	chr21q22.12
219820_at	Hs.130949	Solute carrier family 6, member 16	SLC6A16	chr19q13.1-q13.4
208127_s_at	Hs.468426	Suppressor of cytokine signaling 5	SOCS5	chr2p21
243570_at	Hs.282700	Signal peptidase complex subunit 2 homolog (S. cerevisiae)	SPCS2	chr11q13.4
236123_at	Hs.201921	Suppression of tumorigenicity 7 like	ST7L	chr1p13.2
1554250_s_at	Hs.661254	Tripartite motif-containing 73	TRIM73	chr7q11.23
1558356_at	Hs.108049	Uveal autoantigen with coiled-coil domains and ankyrin repeats	UACA	chr15q22-q24
206577_at	Hs.53973	Vasoactive intestinal peptide	*VIP	chr6q25
205092_x_at	Hs.655536	Zinc finger and BTB domain containing 1	ZBTB1	chr14q23.3

*, Discussed in the text.

Table 2. Known genes down-regulated in PB-EPC after FIR treatment (> 2 folds) .

Probe set ID	Uni gene ID	Gene title	Gene symbol	Location
220518_at	Hs.477015	ABI gene family, member 3 (NESH) binding protein	ABI3BP	chr3q12
215483_at	Hs.651221	A kinase (PRKA) anchor protein (yotiao) 9	AKAP9	chr7q21-q22
224797_at	Hs.24684	Arrestin domain containing 3	ARRDC3	chr5q14.3
229694_at	Hs.144447	Bromodomain and WD repeat domain containing 2	BRWD2	chr10q26
233106_at	Hs.645410	FRMD6 antisense RAN 1 (non-protein coding)	FRMD6-AS1	chr14q22.1
205525_at	Hs.490203	Caldesmon 1	CALD1	chr7q33
1570571_at	Hs.653125	Coiled-coil domain containing 91	CCDC91	chr12p11.22
239442_at	Hs.709257	Centrosomal protein 68 kDa	CEP68	chr2p14
215629_s_at	Hs.659291	Deleted in lymphocytic leukemia, 2	DLEU2	chr13q14.3
230229_at	Hs.292549	Discs, large homolog 1 (Drosophila)	DLG1	chr3q29
236649_at	Hs.127432	DTW domain containing 1	DTWD1	chr15q21.2
235405_at	Hs.485557	Glutathione S-transferase A4	GSTA4	chr6p12.1
223412_at	Hs.63841	Kelch repeat and BTB (POZ) domain containing 7	KBTBD7	chr13q14.11
215268_at	Hs.658760	Hypothetical LOC643314	KIAA0754	chr1p34.2
215599_at	Hs.529793	Glucuronidase, beta pseudogene///similar to Beta-glucuronidase-like protein SMA3	LOC653188///SMA4	chr5q13///chr5q13.2
1554155_at	Hs.709634	Microcephalin 1	MCPH1	chr8p23.1
217373_x_at	Hs.567303	Mdm2 p53 binding protein homolog (mouse)	*MDM2	chr12q14.3-q15
213684_s_at	Hs.480311	PDZ and LIM domain 5	PDLIM5	chr4q22
242248_at	Hs.78060	Phosphorylase kinase, beta	PHKB	chr16q12-q13
214981_at	Hs.664318	Periostin, osteoblast specific factor	POSTN	chr13q13.3
204284_at	Hs.303090	Protein phosphatase 1, regulatory (inhibitor) subunit 3C	PPP1R3C	chr10q23-q24
201427_s_at	Hs.709405	Selenoprotein P, plasm, 1	SEPP1	chr5q31
220983_s_at	Hs.323308	Sprouty homolog 4 (Drosophila)	SPRY4	chr5q31.3
236561_at	Hs.494622	Transforming growth factor, beta receptor I	TGFBR1	chr9q22
224565_at	Hs.523789	Nuclear paraspeckle assembly transcript 1 (non-protein coding)	NEAT1	chr11q13.1
244132_x_at	Hs.657337	Zinc finger protein 518A	ZNF518A	chr10q23.33

*, Discussed in the text.

Among the genes down-regulated by FIR in HG-EPCs, the p53-binding E3 ubiquitin ligase protein mouse double minute 2 homolog (MDM2) is present. The reduced levels of MDM2 and another gene PTK2 in 2 different batches of HG-EPCs by FIR were verified by RT-qPCR (Figure 1F).

Functional module analysis as a framework for the interpretation of FIR effects

The gene lists outlined above gave us preliminary insights into the functional consequences of detected differential gene expression. FGF, MAPK and prostaglandin signaling pathways were found to be associated with the effects of FIR. To understand more about how the gene expression profiles might be correlated with FIR functions as well as to provide quantitative evidence, the signature mRNAs were subjected to a Gene Ontology (GO) database search in order to find statistically overrepresented functional groups within the gene lists. The WebGestalt web tool23 was applied to provide statistical analysis and visual presentation of the results. The GO categories of biological processes that were statistically overrepresented (p < 0.05) among up or down genes are shown in Figure 2A and B, respectively. Consistent with the altered expression profile after FIR treatment, genes involved in cellular metabolism were significantly induced by FIR irradiation (118 genes, p = 1.4e-03; Figure 2A). Another significant biological process associated with this group is related to the protein phosphorylation processes (45 genes, p = 2.76e-6; Figure 2A), correlating with the status that FIR radiation could promote ERK and p38 MAPK activation in EPCs in response to high glucose stimulation. 6 Other predominant processes in the GO group include genes response to external stimulus (22 genes, p = 0.0014; gene names are shown in Figure 2A). Genes involved in cytokine (incl. IL6ST and IL20RB) and prostaglandin pathways (incl. PTGS2, PTGER3 and PTGER4) were induced and enriched by FIR, corresponding to the more angiogenic and activated status of treated EPCs.

Figure 2.

Significantly affected biological processes in HG-EPCs after FIR irradiation. Gene set enrichment analysis was performed according to the Gene Ontology (GO) classification. Up- (A) and down-regulated (B) probe sets were subjected to the GO database search via the WebGestalt interface. The number of genes, gene symbols, their percentages and the p values for each category that show significance (p < 0.05) are listed.

In contrast, the principal functions of the down-regulated genes include those related to the regulation of cellular macromolecule biosynthesis (75 genes, p = 7.3e-3; Figure 2B). The glycogen metabolic process, in particular, was repressed by FIR (7 genes, p = 8.5e-5; genes listed in Figure 2B), which may reflect the energy consumption requirement during the active migration and microtubular formation in FIR-treated EPCs.

Coordinated canonical pathway changes after FIR irradiation in human EPCs

It was shown that administration of FIR radiation promotes AKT, ERK and p38 MAPK activation and phosphorylation in EPCs in response to high glucose stimulation.6 However, additional signaling pathways may also be influenced by FIR. When the FIR-stimulated genes were subjected to a canonical pathway database using the IPA web tool, a total of 14 pathways were significantly enriched (Figure 3A). The most significant canonical pathway mapped is the “caveolar-mediated endocytosis signaling” pathway, and other top-ranked canonical pathways including G-protein coupled receptor signaling and JAK/Stat signaling pathways (Figure 3A). Four genes, including son of sevenless homolog 1 (SOS1), suppressors of cytokine signaling 3 (SOCS3), suppressors of cytokine signaling 5 (SOCS5), and protein inhibitors of activated stats 2 (PIAS2), are in the JAK/Stat signaling pathway, whose details are illustrated in Figure 3B.

Figure 3.

Induced canonical pathways and functional modules in HG-EPCs after FIR irradiation. FIR-induced probe sets were subjected to canonical pathway (A) and functional module (C) search via the IPA web tool. (B) Details of the JAK/Stat signal pathway. Genes highlighted: FIR-induced genes in this pathway.

To obtain more insights into these FIR-stimulated genes, a similar module analysis was performed according to the IPA toxicity database. As shown in Figure 3C, the top ranked functional groups included “Cardiac Hypertrophy”, “Liver Proliferation/Necrosis/Cell Death”, and “Cardiac Necrosis/Cell Death“. Genes regulating cell proliferation and apoptosis were therefore enriched among our gene list. Five genes involved in the “Gene regulation by peroxisome proliferators via PPARα” category were enriched (incl. inhibitor of kappa light polypeptide gene enhancer in B-cells, kinase beta, SOS1, fatty acid binding protein 1, prostaglandin-endoperoxide synthase 2, and nuclear receptor interacting protein 1; Figure 3C), indicated that the PPARα pathway is crucial for high glucose-incubated EPCs to respond to FIR.

Among genes down-regulated by FIR irradiation, the protein kinase A signaling canonical pathway was the most significantly affected one (Figure 4A). Other predominant pathways were Cardiac β-adrenergic signaling, Role of BRCA1 in DNA Damage Response, CDK5 signaling and others (Figure 4A). Genes involved in IGF1 signaling and Insulin receptor signaling, consistent with the phenotypes that Far infra-red therapy promotes ischemia-induced angiogenesis in diabetic mice and restores high glucose-suppressed endothelial progenitor cell functions.6 On the other hand, genes involved in hepatic (incl.collagen, type I, alpha 1, fibronectin 1, collagen, type IV, alpha 3, insulin-like growth factor binding protein 3, insulin-like growth factor binding protein 5, transmembrane protein 67, and chemokine (C-X-C motif) ligand 2) or cardiac (incl. protein tyrosine kinase 2, adenosine A3 receptor, fibronectin 1, adrenergic, beta-1-, receptor, periostin, osteoblast specific factor, immunoglobulin heavy constant mu, and dicer 1, ribonuclease type III) fibrosis were significantly repressed by FIR (Figure 4B).

Figure 4.

Repressed canonical pathways and functional modules in HG-EPCs after FIR irradiation. FIR-suppressed probe sets were subjected to canonical pathway (A) and functional module (B) search via the IPA web tool.

Genetic networks of FIR-treated EPCs

Increasing evidence has shown that genes do not act as individuals but collaborate in genetic networks. To better understand how FIR-affected genes form functional modules to regulate EPC biology, we performed systems biology genetic network analysis on signature genes. Identified up- or down-regulated genes were again inputted into the IPA tool to construct network modules. The knowledge base behind IPA summarizes known as molecular interactions are noted in published references (described in Materials and Methods). The term “network” in IPA is not the same as a biological or canonical pathway with a distinct function (i.e., angiogenesis) but a reflection of all interactions of a given protein as defined in the literature.

A major network consisting of 96 genes (96/627 = 15.31%) was identified (Figure 5A). This network was enriched with genes involved in hematological system development and function (20 genes incl. CASP8 and FADD-like apoptosis regulator, dual specificity phosphatase 4, eukaryotic translation initiation factor 4E, F-box protein 32, interleukin 6 signal transducer, integrin, beta 6, mitogen-activated protein kinase, Niemann-Pick disease, type C1, protein phosphatase, Mg2+/Mn2+dependent, 1D, PTGS2, prostaglandin E receptor 3, prostaglandin E receptor 4, Ras association (RalGDS/AF-6) domain family member 5, sirtuin (silent mating type information regulation 2 homolog) 1, SOCS3, thrombospondin 1, toll-like receptor 4, vasoactive intestinal peptide, X-box binding protein 1, and zinc finger CCCH-type containing 12A). MAPK was found to regulate most of these genes and functions as a “hub” gene, which had higher connectivity to others or resided in a position among submodules in the major network (Figure 5A). This observation was also consistent with the finding that FIR induces p38 MAPK phosphorylation and activation.6 XBP1, GATA binding protein 3, PTGS2, SIRT1, Histone H3, hairy and enhancer of split 1 and sprouty homolog 2 were also hubs connecting different submodules in the major network component (Figure 5A).

Figure 5A.

Interaction network analysis as a framework for the interpretation of FIR effects. FIR-induced probe sets were subjected to genetic network analysis via the Ingenuity Pathway Analysis (IPA) web tool. This network is displayed graphically as nodes (gene products) and edges (biological relationships between nodes) mapped by IPA. The intensity of the node color indicates the degree of differential expression. Hub genes in this genetic network are highlighted.

When a similar function network analysis was conducted on FIR-repressed genes, a major network consisting of 80 genes (80/623 = 12.84%) was also found (Figure 5B). Central to the network, there were significant hubs, including nuclear receptor subfamily 3 group C member 1, Mdm2 p53 binding protein homolog, GATA binding protein 4, transforming growth factor, beta receptor 1, FN1, COL1A1, brain-derived neurotrophic factor, and CXCL2 (Figure 5B, cellular compartments of these genes are also shown). MDM2 down-regulation has been verified by RT-qPCR in Figure 1. Top toxicity genes associated with this genetic network according to the IPA database are those involved in cardiac hypertrophy (incl. DICER1, GATA4, MDM2, POSTN, PTK2 and S100 calcium binding protein A10; indicated in Figure 5B), reflecting the healthy circumstance of these irradiated EPCs.

Figure 5B.

Interaction network analysis as a framework for the interpretation of FIR effects. A major functional genetic network composed of multiple genes down-regulated in HG-EPCs after FIR irradiation. Genes were arranged according to not only their relationships (indicated by lines) but also their cellular compartments. Hub genes are highlighted.

DISCUSSION

EPC research these years is quite active, partly due to the clinical application potential of these cells. These cells are potential sources for cell therapy that aim to enhance the neovascularization of tissue engineered constructs or ischemic tissues.29 It has been shown that both short-term and long-term FIR can increase access flow, in part due to the activation of both endothelial cells and EPCs.6 Infrared radiation is an invisible electromagnetic wave with a longer wavelength than that of visible light. According to the difference in wavelength, infrared radiation can be divided into 3 categories: near-infrared radiation (0.8-1.5 μ m); middle-infrared radiation (1.5-5.6 μ m); and FIR radiation (5.6-1,000 μ m). Many efforts have been undertaken to elucidate the mechanisms of FIR therapy: Akasaki et al. found that repeated FIR therapy upregulates eNOS expression and augments angiogenesis in an apolipoprotein E (ApoE)-deficient mouse hind limb ischemia model.4 Ikeda et al. reported that 4 weeks of sauna therapy significantly increased arterial eNOS expression and nitric oxide production in cardiomyopathic hamster.30 Huang et al. reported that FIR therapy enhanced blood flow recovery and new vessel formation in ischemic hindlimbs of diabetic mice, as well as the restoration of high glucose-suppressed functions of peripheral blood EPCs, presumably via activation of AKT, ERK and p38 MAPK pathways.6 Here we showed again a favorable effect of FIR treatment on EPCs derived from peripheral blood and provided for the first time the global transcriptomic alternations caused by FIR irradiation. Given that FIR therapy may counteract the detrimental effect of a diabetic environment on improvement of EPC functions, our results may provide some novel rationales for the potential clinical applications, such as the development of new biomarkers to monitor FIR effects in vivo or in vitro.

In this study we have disclosed genes and functional modules affected by FIR, and hence maybe in angiogenesis. Besides reproducing what was already known for FIR functions and EPC biology, novel genes that may play crucial roles in EPC activities were also revealed. Among the many genes affected by FIR irradiation, we spotted some genes that have more interaction/connectivity to other genes. These “hub” genes are theoretically more crucial than other in terms of FIR functions. It is clear that dysregulation of hubs may eventually lead to the disruption of the genetic network and the malfunction of cells.31 Of these hub genes, 2 GATA family transcription factors, GATA3 and GATA4, were either induced or repressed by FIR (Figure 5A & Figure 5B). Both GATA factors play an important role in angiogenesis: GATA3 is known to mediate Tie2 expression and functions in large vessel endothelial cells,32 and GATA4 expression in endothelial-derived cells is crucial in the development of cardiac valves.33,34 The repression of GATA4 in glucose-treated EPCs may help to enhance EPC stemness and thereby motility. Other transcription factors including HES1, SPRY2 and SIRT1 are also hubs in the FIR-affected genetic network (Figure 5A). The Notch-HES1 signaling axis controls hemato-endothelial fate decisions of human embryonic and induced pluripotent cells,35 and SIRT1, together with nitric oxide, plays a critical role in vasodilation and atherosclerosis prevention.36 The FGF antagonist Sprouty2 (SPRY2) expression controls endothelial monolayer integrity and quiescence, which is fundamental to the formation of mature blood vessels.37 However, no report yet links these genes to EPC biology or FIR effects.

Infrared radiation transfers energy that is perceived as heat by thermoreceptors in the skin.38 According to the report by Hartel et al., the temperature can be increased up to 4 °C in 10 mm depth of tissue.39 In addition, infrared therapy may allow multiple energy transfers as deep as 2-3 cm into subcutaneous tissue without irritating or overheating the skin like unfiltered heat radiation.39 The skin temperature steadily increases to a plateau of around 38-39 °C during FIR treatment as long as the distance between the ceramic plate and the skin was more than 20 cm.40 On the other hand, cellular photosensors may also be involved in FIR functions. It has been shown that green light-emitting diode (LED) irradiation can activate directional motility in human orbital fat stem cells through activation of the ERK/MAP kinase/p38 signaling pathway.41 Two non-visual opsins, encephalopsin (OPN3) and short-wave-sensitive opsin 1 (OPN1SW), serve as photoreceptor responses to green LED irradiation.41 Elucidation of the levels of various opsins in healthy and high glucose-treated EPCs, as well as roles of opsins in FIR-induced phenotypes and gene expression alternations, will be a significant study to pursue in the future.

CONCLUSIONS

In summary, our research combined systems biology tools to decipher transcriptome expression alternations in high glucose-treated EPCs. With these findings, it will be possible to discover novel molecular mechanisms that are crucial to ECFC/late EPC functioning. In this context, our research will also lead to the development of new diagnosis/prognosis approaches for cardiovascular disorders as well as other EPC-related diseases such as diabetes. Novel diagnostic modalities that are able to trace FIR effects could also be developed accordingly.

CONFLICT-OF-INTEREST DISCLOSURE

The authors declare no competing financial interests.