Global Remodeling of the Vascular Stem Cell Niche in Bone Marrow of Diabetic Patients: Implication...

Beltrami, Constanza Emanueli and Paolo MadedduFranco Carnelli, Francesco Rosa, Stefano Riboldi, Fausto Sessa, Elisa Avolio, Antonio Paolo Gaia Spinetti, Daniela Cordella, Orazio Fortunato, Elena Sangalli, Sergio Losa, Ambra Gotti,

Implication of the microRNA-155/FOXO3a Signaling PathwayGlobal Remodeling of the Vascular Stem Cell Niche in Bone Marrow of Diabetic Patients :

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 2012 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/CIRCRESAHA.112.3005982013;112:510-522; originally published online December 18, 2012;Circ Res.

http://circres.ahajournals.org/content/112/3/510World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circres.ahajournals.org/content/suppl/2012/12/18/CIRCRESAHA.112.300598.DC1.htmlData Supplement (unedited) at:

http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on January 31, 2013http://circres.ahajournals.org/Downloaded from

http://circres.ahajournals.org/content/112/3/510

http://circres.ahajournals.org/content/suppl/2012/12/18/CIRCRESAHA.112.300598.DC1.html

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circres.ahajournals.org//subscriptions/

http://circres.ahajournals.org/

510

Clinical/Translational Research

An imbalance between circulating proinflammatory and proangiogenic cells leading to counterproductive cell re-

cruitment to sites of endothelial injury contributes to vascular complications in patients with diabetes mellitus.1–6 Studies in animal models suggest that the altered spectrum of circulating cells is consequent to a deregulated control of cell mobilization from bone marrow (BM).7–10 In line, clinical data show that the BM of diabetic patients has an impaired capacity to release he-matopoietic stem cells (HSCs) after stimulation with granulocyte

colony-stimulating factor,11 a defect for which the term diabetic stem cell (SC) mobilopathy was recently proposed.12 Moreover, diabetes mellitus might impinge on the integrity of SCs/progeni-tor cells (PCs) by altering the marrow microenvironment, which consists of stromal, endosteal, and microvascular cells.10,13,14

In This Issue, see p 407Besides providing oxygen and nutrients, the mar-

row microvasculature plays a key role in the regulation of

© 2012 American Heart Association, Inc.

Circulation Research is available at http://circres.ahajournals.org DOI: 10.1161/CIRCRESAHA.112.300598

RES

Circulation Research

0009-7330

10.1161/CIRCRESAHA.112.300598

201540

Spinetti et al Remodeling of Diabetic Bone Marrow

Circulation Research Month, XXXX

1

February

2013

112

3

510

522

11122012

17122012

© 2012 American Heart Association, Inc.

Rationale: The impact of diabetes mellitus on bone marrow (BM) structure is incompletely understood.Objective: Investigate the effect of type-2 diabetes mellitus (T2DM) on BM microvascular and hematopoietic cell

composition in patients without vascular complications.Methods and Results: Bone samples were obtained from T2DM patients and nondiabetic controls (C) during

hip replacement surgery and from T2DM patients undergoing amputation for critical limb ischemia. BM composition was assessed by histomorphometry, immunostaining, and flow cytometry. Expressional studies were performed on CD34pos immunosorted BM progenitor cells (PCs). Diabetes mellitus causes a reduction of hematopoietic tissue, fat deposition, and microvascular rarefaction, especially when associated with critical limb ischemia. Immunohistochemistry documented increased apoptosis and reduced abundance of CD34pos-PCs in diabetic groups. Likewise, flow cytometry showed scarcity of BM PCs in T2DM and T2DM+critical limb ischemia compared with C, but similar levels of mature hematopoietic cells. Activation of apoptosis in CD34pos-PCs was associated with upregulation and nuclear localization of the proapoptotic factor FOXO3a and induction of FOXO3a targets, p21 and p27kip1. Moreover, microRNA-155, which regulates cell survival through inhibition of FOXO3a, was downregulated in diabetic CD34pos-PCs and inversely correlated with FOXO3a levels. The effect of diabetes mellitus on anatomic and molecular end points was confirmed when considering background covariates. Furthermore, exposure of healthy CD34pos-PCs to high glucose reproduced the transcriptional changes induced by diabetes mellitus, with this effect being reversed by forced expression of microRNA-155.

Conclusions: We provide new anatomic and molecular evidence for the damaging effect of diabetes mellitus on human BM, comprising microvascular rarefaction and shortage of PCs attributable to activation of proapoptotic pathway. (Circ Res. 2013;112:510-522.)

Key Words: bone marrow ■ diabetes mellitus type 2 ■ macroangiopathy ■ microangiopathy ■ stem cells

lww

190,97

Circ Res

Global Remodeling of the Vascular Stem Cell Niche in Bone Marrow of Diabetic Patients

Implication of the microRNA-155/FOXO3a Signaling Pathway

Gaia Spinetti, Daniela Cordella, Orazio Fortunato, Elena Sangalli, Sergio Losa, Ambra Gotti, Franco Carnelli, Francesco Rosa, Stefano Riboldi, Fausto Sessa, Elisa Avolio, Antonio Paolo Beltrami,

Constanza Emanueli, Paolo Madeddu

Original received August 6, 2012; revision received December 11, 2012; accepted December 17, 2012. In November 2012, the average time from submission to first decision for all original research papers submitted to Circulation Research was 15.8 days.

From the Laboratories of Experimental Cardiovascular Medicine (P.M., E.A.) and Vascular Pathology and Regeneration (C.E.), University of Bristol, Bristol, United Kingdom; IRCCS MultiMedica, Milan, Italy (G.S., D.C., O.F., E.S., S.L., A.G., F.C., S.R.); School of Medicine, University of L’Aquila, L’Aquila, Italy (F.R.); Department of Surgical and Morphological Science, University of Insubria (F.S.), Varese, Italy; and Interdepartmental Centre for Regenerative Medicine, University of Udine, Udine, Italy (E.A., A.P.B.).

The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA. 112.300598/-/DC1

Correspondence to Paolo Madeddu, Chair of Experimental Cardiovascular Medicine, Regenerative Medicine Section, Bristol Heart Institute, School of Clinical Sciences, University of Bristol, Level 7, Bristol Royal Infirmary, Upper Maudlin St, Bristol BS2 8HW, United Kingdom. E-mail [email protected]


http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.112.300598/-/DC1

http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.112.300598/-/DC1

mailto:[email protected]


Spinetti et al Remodeling of Diabetic Bone Marrow 511

hematopoiesis.15,16 Furthermore, vascular sinusoids constitute a dedicated interface for cell exchanges with the peripheral circulation. In a mouse model, we showed that type 1 diabetes mellitus causes microvascular rarefaction, resulting in critical hypoperfusion, depletion of SCs at the level of the endosteal niche, and altered transendothelial cell trafficking.17 It would be of paramount importance to understand whether similar microvascular pathology occurs in BM of patients with diabe-tes mellitus. Current knowledge is limited to flow cytometry analysis of aspirates, which showed reduction of CD34pos-PCs in BM of diabetic patients.5 However, this sampling procedure is not suited to define the anatomic structure of the marrow and may provide inaccurate SC counts if the marrow is remodeled. Moreover, the molecular mechanisms underlying diabetes mellitus-induced SC depletion are incompletely understood. We showed that experimental diabetes mellitus causes an ele-vation in reactive oxygen species, infringes on DNA integrity, and induces apoptosis of Sca-1posc-Kitpos cells.17 In addition, signaling mechanisms that maintain the self-renewal capacity and prevent senescence of SCs, like the polycomb group gene Bmi-1,18 are downregulated in BM cells from diabetic mice.13

Recent evidence indicates that microRNAs (miRs) regulate the maturation of different hematopoietic lineages.19–21 In par-ticular, a restricted subset of miRs expressed in CD34pos HSCs is hierarchically organized in a circuitry that controls prolifera-tion, viability, and differentiation.21,22 Among HSC-associated miRs, miR-221 reportedly regulates terminal sta ges of erythro-poiesis via repression of c-Kit, whereas miR-155 acts upstream of miR-221 holding HSCs at an early stem-progenitor stage through inhibition of differentiation-associated molecules, like CCAAT/enhancer-binding protein-β, cAMP response element- binding protein, JUN and FOS.22,23 In addition, miR-155 inhibits the forkhead transcription factor FOXO3a.24 In hema-topoietic cells with incurred DNA damage, FOXO3a induces cell cycle arrest and apoptosis, via transcriptional regulation of the cyclin-dependent kinase inhibitor p27Kip1 and proapoptotic Bcl-2 family member Bim.25–27 By inhibiting FOXO3a, miR-155 exerts prosurvival effects in HSCs.

The present study examines the damaging action of diabetes mellitus on human BM using femoral bone specimens collect-ed from orthopedic surgery. To determine whether macrovas-cular disease can further disrupt BM integrity, we also studied

patients with type 2 diabetes mellitus (T2DM) undergoing amputation for critical limb ischemia (CLI). Furthermore, we investigated the influence of diabetes mellitus and high glu-cose (HG) on miR expression in CD34pos immunosorted BM PCs and the ability of miR-155 to reverse FOXO3a upregula-tion in HG-challenged CD34pos cells.

MethodsA Supplemental Methods section is available in the Online Data Supplement.

Patients and Study ProtocolThe study complied with the principles stated in the Declaration of Helsinki and was covered by institutional ethical approval (protocol number 20/2010). Eligible subjects were screened from a consecutive series referring to MultiMedica Hospital for hip replacement surgery (nondiabetic controls and T2DM patients) or limb amputation (T2DM patients with CLI) from August 2010 to November 2012. T2DM and CLI were defined according to the American Diabetes Association and the Inter-Society Consensus for the Management of Peripheral Arterial Disease (TASC II), respectively.28 Exclusion criteria were acute infection, immune diseases, current or past hematologic disor-ders or malignancy, drug-induced diabetes mellitus, unstable angina, recent (within 6 months) myocardial infarction or stroke, liver failure, dialysis because of renal failure, pregnancy, and lack of consent to participate in the study. Online Table I illustrates main clinical data of the 82 subjects who entered the study.

Immunohistochemistry, flow cytometry, and molecular biology analyses were performed on femoral head leftovers from orthopedic surgery and proximal part of amputated femurs. In addition, a 30 mL peripheral blood (PB) sample was obtained by venipuncture the day before interventional procedures.

ResultsCharacteristics of the Study PopulationAs shown in Online Table I, groups were similar with regard to age and sex distribution. T2DM patients showed a higher body mass index than controls and T2DM+CLI, whereas the T2DM+CLI group had the highest percentage of smokers and a longer duration of diabetes mellitus. Hypertension was more frequent in the 2 diabetic groups than in controls, and cardiovascular complications were common in T2DM+CLI patients. Insulin was the preferred antidiabetic medication in T2DM+CLI, whereas T2DM patients were mainly treated with oral antidiabetic drugs including glitazones. Between 40% and 52% of diabetic patients were taking statins. Fasting glucose was higher in the 2 diabetic groups, especially in T2DM+CLI.

Diabetes Mellitus Alters BM CompositionHistomorphometry demonstrates a remarkable remodeling of BM from diabetic patients, consisting of decreased hemato-poietic tissue, fat deposition, and bone rarefaction (Figure 1A and 1B). The effect of diabetes mellitus was significant, incre-mental in T2DM+CLI patients and independent of background factors (Online Table III). Multiple regression analysis showed that the dependent variables, hematopoietic and fat fractions, can be predicted by grouping factor, fasting glucose and dura-tion of diabetes mellitus, with no effect of other independent variables except for body mass index, which in association with duration of diabetes mellitus predicts the abundance of fat in BM (Online Table V).

Nonstandard Abbreviations and Acronyms

BM bone marrow

CLI critical limb ischemia

EC endothelial cell

HG high glucose

HSC hematopoietic stem cell

KDR kinase insert domain receptor

miR microRNA

NK natural killer lymphocytes

PB peripheral blood

PC progenitor cell

SC stem cell

T2DM type 2 diabetes mellitus



Diabetes Mellitus and CLI Additively Reduce Vascular Density in Human BMWe next assessed the marrow microvasculature by immuno-histochemistry. Capillaries were recognized as CD31-positive structures whose lumen-size does not exceed the diameter of an erythrocyte, whereas venous sinusoids were identified as larg-er irregular structures containing several erythrocytes (Figure 1C). Furthermore, microvessels and arterioles were identified by confocal and fluorescence microscopy using the endothelial marker von Willebrand factor and the vascular smooth muscle cell marker α-smooth muscle actin (Figure 1D).

Analysis of microvascular density identified a significant difference among groups for all sets of vascular structures (P=0.005; Figure 2E). Pairwise comparison indicates a large decrease of capillary density in BM of T2DM patients (P<0.05 versus controls) and a further reduction of all 3 vascular frac-tions in T2DM+CLI patients (P<0.01). The effect of diabetes mellitus on BM microvascular density was confirmed when considering background covariates (Online Table III). In mul-tiple regression analysis, the dependent microvascular variables were predicted by grouping factor, duration of diabetes mellitus, and fasting glucose, in linear combination with hypertension as

Figure 1. Diabetes mellitus induces bone marrow (BM) remodeling and vascular rarefaction. A and B, Histomorphometric analysis shows replacement of marrow with fat and bone rarefaction. A, Representative microphotograph of hematoxylin and eosin-stained BM sections: (i) Control, (ii) type-2 diabetes mellitus (T2DM) patient, and (iii) T2DM patient with critical limb ischemia (CLI). B, Bar graph showing average data of marrow fractions. C, Representative microphotograph of human BM showing CD31-positive vascular structures. Arrows indicate the type of vessel. D, Confocal microscopy photographs of human BM showing vascular structures stained with the endothelial marker von Willebrand factor (VWF) and the vascular smooth muscle marker α-smooth muscle actin (αSMA). Nuclei are stained blue by 4',6-diamidino-2-phenylindole (DAPI). E, Bar graph showing average data of microvascular density. *P<0.05 and **P<0.01 versus Controls, §P<0.05 versus T2DM. Controls, n=10; T2DM, n=7; T2DM+CLI, n=10.

512 Circulation Research February 1, 2013




far as capillaries and sinusoids are concerned (Online Table V). Furthermore, we found that vascular density is directly correlat-ed with the hematopoietic fraction and inversely correlated with fat abundance in BM (P<0.01 for both comparisons).

Altogether, these data indicate that diabetes mellitus causes microangiopathy in human BM, with vascular rarefaction being aggravated by CLI. Among associated risk factors,

hypertension interacts with diabetes mellitus in influencing capillary and sinusoid density.

Diabetes Mellitus Reduces the Abundance of Hematopoietic Progenitors in BMBM-derived CD34pos cells are a well-characterized popula-tion that have been used for hematopoietic reconstitution and

Figure 2. Reduced abundance of CD34pos cells in bone marrow (BM) of type-2 diabetes mellitus (T2DM) patients. A, Representative confocal microscopy photographs of human BM showing the presence of CD45pos (green arrow), CD45posCD34pos (pink arrowhead) (i), and CD45negCD34pos cells (red arrow) (ii). Nuclei are stained blue with 4',6-diamidino-2-phenylindole (DAPI). Bar graphs showing the average density of CD45posCD34pos cells (iii, median and 5%–95% distribution) and CD34posCD45neg cells (iv, mean±SEM). B, Increased abundance of apoptotic mononuclear cells and CD34pos cells in BM from diabetic patients. Representative microphotographs of fluorescent terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) positive cells (i) and bar graphs showing average data (ii and iii). *P<0.05, **P<0.01 and ***P<0.001 versus controls, §P<0.05 and §§P<0.01 versus T2DM. Controls, n=10; T2DM, n=7; T2D+critical limb ischemia (CLI), n=8.




more recently for the treatment of myocardial and peripheral ischemia.29 We investigated the abundance of CD45posCD34pos and CD45negCD34pos cells in BM by confocal microsco-py (Figure 2Ai and 2Aii). ANOVA detected a difference among groups with regard to the CD45posCD34pos fraction (P=0.008) and pairwise comparison showed a large reduc-tion of this cell population in T2DM+CLI compared with controls (P<0.05) (Figure 2Aiii). Likewise, CD45negCD34pos cells were different among groups (P=0.001) because of a large reduction in T2DM (P=0.008 versus controls) and T2DM+CLI patients (P=0.0001 versus controls) (Figure 2Aiv). The difference among groups was confirmed after consideration of background factors by ANCOVA (Online Table III). Grouping factor was a strong predictor of the dependent variables, CD45negCD34pos and CD45posCD34pos cells. Furthermore, duration of diabetes mellitus and fasting glucose predicted the abundance of CD45negCD34pos cells (Online Table V). Moreover, in situ detection of DNA frag-mentation by terminal deoxynucleotidyl transferase dUTP nick end labeling assay indicates activation of apoptosis

in BM-mononuclear cells and CD34pos-PCs of T2DM and T2DM+CLI patients compared with controls (Figure 2B and Online Figure I).

We next used multicolor flow cytometry for quantifica-tion of various cell populations in BM and PB (Figure 3A).30 CD45dimCD34pos cells were reduced in BM of T2DM (P=0.02 versus controls) and T2DM+CLI patients (P=0.004), but did not differ between the diabetic groups (Figure 3Bi), with these data being confirmed by ANCOVA (Online Table III). In contrast, PB CD45dimCD34pos cells were similar in control and diabetic patients (Figure 3Bii). The relative abundance of CD45dimCD34pos cells in BM could be predicted independently by grouping factor, duration of diabetes mellitus, and fasting glucose in the multiple regression model (Online Table V).

The surface antigen CD133, a prominin 5 transmembrane gly-coprotein 1, has been used to identify primitive hematopoietic and nonhematopoietic PCs31 endowed with proangiogenic and healing activities in animal models32 and patients with acute myo-cardial infarction.33–35 We found that CD45dimCD133posCD34pos cells are remarkably reduced in BM of T2DM (P=0.02 versus

Figure 3. Flow cytometry characterization of hematopoietic cells in bone marrow (BM) and peripheral blood (PB) of type-2 diabetes mellitus (T2DM) patients. A, Gating strategy of multicolor flow cytometry. B to D, Bar graphs showing the abundance of CD45dimCD34pos (Bi and Bii), CD45dimCD133posCD34pos cells (Ci and Cii) and CD45posCD133posCD34neg cells (Di and Dii). *P<0.05, **P<0.01 and ***P<0.001 versus controls. Controls, n=8 to 14; T2DM, n=7 to 9; T2DM+critical limb ischemia (CLI), n=7 to 11.




Figure 4. Flow cytometry characterization of bone marrow (BM) and peripheral blood (PB) endothelial progenitors. A, Gating strategy of CD45dimCD34posKDRpos mononuclear cells (MNCs) (i), and bar graphs showing the abundance of this population in BM (ii) and PB (iii). B, Gating strategy of CD34posCD14posCD45dimKDRposCXCR4pos cells (i), and bar graphs showing the abundance of this population in BM (ii) and PB (iii). *P<0.05 versus controls. Controls, n=8 to 11; type-2 diabetes mellitus (T2DM), n=6 to 9; T2DM+critical limb ischemia (CLI), n=7 to 10.




controls) and T2DM+CLI (P=0.01), with no difference between the 2 diabetic groups (P=0.83) (Figure 3Ci and Online Table III). The levels of CD45dimCD133posCD34pos cells were pre-dicted by grouping, duration of diabetes mellitus, and fasting glucose (Online Table V). In contrast, no group difference was found in PB CD45dimCD133posCD34pos cells (P=0.35) (Figure 3Cii). Moreover, there was no difference between diabetic and control subjects with regard to BM and PB CD133posCD34neg cells (Figure 3Di and 3Dii), which are considered very early hematopoietic precursors.36

Differential Effects of Diabetes Mellitus on Endothelial PCs and Lineage Committed Hematopoietic CellsSeveral BM cell subpopulations, including hematopoietic and nonhematopoietic SCs, have the capacity to differentiate into endothelial cells (ECs) or promote EC growth by paracrine mechanisms. For instance, BM CD34pos kinase insert domain receptor (KDR)pos cells have been postulated to comprise the hemangioblast, that is, a bipotent cell type able to form both ECs and blood cells.37–39 Moreover, CD34posKDRpos cells are re-portedly decreased in PB of patients with traditional cardiovas-cular risk factors including diabetes mellitus.3,40,41 As shown in Figure 4Ai, CD34posKDRpos cells were mainly enriched in the CD45dim mononuclear cell fraction. Moreover, we could con-firm a large reduction of BM and PB CD45dimCD34posKDRpos cells in both subgroups of diabetic subjects as compared with controls (P=0.006 for both comparisons), with no dif-ference between T2DM and T2DM+CLI (Figure 4Aii and 4Aiii and Online Table III). The dependent variable, BM CD45dimCD34posKDRpos cells, was predicted by grouping fac-tor, duration of diabetes mellitus, and fasting glucose, whereas PB CD45dimCD34posKDRpos cells were predicted by grouping factor and duration of diabetes mellitus (Online Table V).

We also verified the abundance of proangiogenic PCs de-fined as CD34posCD14posCD45dimKDRposCXCR4pos mononucle-ar cells.42,43 This cellular subpopulation was reduced in BM and PB of both diabetic groups, with no difference between T2DM and T2DM+CLI (Figure 4Bi–4Biii). However, this effect was attenuated when considering other covariates (Online Table III). Nonetheless, the reduction of the PB cell fraction was predicted by grouping factor (Online Table V). Furthermore, a recent study showed that BM CD31pos cells are enriched of highly angiogenic cells.44 However, we found no difference among groups regarding this cell population (data not shown).

We next verified the influence of diabetes mellitus on lin-eage committed hematopoietic cells (Figure 5A–5D) and mature ECs (Figure 5E). No difference was found in the BM and PB levels of CD45posCD19pos B-lymphocytes (Figure 5Bi and 5Bii) and CD45posCD3pos T lymphocytes (Figure 5Ci and 5Cii). In contrast, CD3negCD56posCD16pos natural killer (NK) lymphocytes were increased in BM of T2DM patients, with this effect being confirmed after correction for other covari-ates. In multiple regression analysis, fasting glucose was a predictor of NK lymphocytes (P=0.001). Moreover, PB NK lymphocytes did not differ between diabetic and control sub-jects (Figure 5Di and 5Dii). Finally, CD45negCD31posCD144pos ECs were reduced in BM of T2DM+CLI (Figure 5E), thus, confirming the in situ data from immunohistochemistry.

Altogether, these data indicate that diabetes mellitus causes a selective reduction of endothelial PCs in BM and PB, with-out altering the abundance of lineage committed hemato-poietic cells with exception of NK lymphocytes which were augmented in BM.

Effect of Diabetes Mellitus on Non-HSCsThis fraction represents a heterogeneous popula-tion that comprises CD45negCXCR4posCD34pos SCs, CD34negCD45negCD14negCD90posCD73posCD105pos mesenchy-mal SCs and cKitpos PCs.45–47 We could not find any differ-ence between diabetic patients and controls with regard to BM CD45negCXCR4posCD34pos SCs and mesenchymal SCs (data not shown). Surprisingly, cKitpos PCs were increased in BM and PB of T2DM patients (P<0.05 versus controls) but not in T2DM+CLI patients (Online Figure II). However, in multiple regression analysis, diabetes mellitus was not a predictor of cKitpos PCs.

Activation of the Proapoptotic FOXO3a Signaling Pathway in BM CD34pos CellsWe investigated the expression levels of miR-155 and its vali-dated target FOXO3a24 in CD34pos immunosorted BM PCs. Results indicate a downregulation of miR-155 in CD34pos-PCs from diabetic patients compared with controls (P=0.019) (Figure 6A). In contrast, miR-221 did not differ among groups (data not shown). Moreover, we found FOXO3a mRNA lev-els to be increased in BM CD34pos-PCs of diabetic patients (P=0.004) (Figure 6B) and inversely correlated to miR-155 levels (R2=−0.41; P=0.002) (Figure 6C). Of note, in situ analy-sis of FOXO3a expression showed a remarkable increase of the transcription factor in BM cells from both diabetic groups (P<0.0007) (Figure 6Di and 6Dii), with evidence of nuclear translocation which is required for transcriptional activation of target genes (Figure 6Diii and 6Div). In line, we found that the FOXO3a targets p21 and p27kip1 are upregulated in CD34pos cells from diabetic patients (P=0.001 and 0.006, respectively) (Figure 6E and 6F). The effect of diabetes mellitus on miR-115, FOXO3a, and p21 was confirmed when considering back-ground covariates (Online Table VI). P21 and p27kip1 inhibit cell cycle progression by binding to, and inactivating, cyclin- dependent kinase complexes. Analysis of cell cycle by flow cytometry confirmed that CD34pos cells from diabetic BM are stalled at the G1 checkpoint and undergo apoptosis with high frequency (Online Figure III).

We next investigated whether in vitro exposure to HG mimics diabetes mellitus in altering the miR-155/FOXO3a/p21/p27kip1 signaling pathway in CD34pos-PCs. We found that HG decreases cell counts and miR-155 levels and increases FOXO3a (Online Figure IVA–IVC). Moreover, HG increases the levels of p21 and p27kip1 (Online Figure IVD and IVE). The osmotic control mannitol did not alter the expression of miR-155 and related target genes (Online Figure V).

To verify whether miR-155 overexpression can rescue the effect of HG on target genes, we transfected HG-challenged CD34pos cells with pre–miR-155 and confirmed that transfec-tion results in upregulation of related mature miR (Online Figure VI). Of note, forced expression of miR-155 contrast-ed the diminishing effect of HG on CD34pos cell counts and abrogates the HG-induced upregulation of FOXO3a, p21,




and p27kip1 (Figure 7A and 7B). Finally, in inductive colony forming unit assays on HG-challenged CD34pos cells, miR-155-transduced cells generated fewer myeloid and erythroid colonies. Furthermore, the colonies generated by miR-155–transduced cells were smaller than controls (Figure 7C).

DiscussionTo the best of our knowledge, this is the first study showing the damaging effect of diabetes mellitus on BM SCs and their mi-crovascular environment in human subjects. We demonstrate for the first time the presence of microangiopathy in BM of diabetic patients. Furthermore, we screened a large spectrum of SCs/PCs and mature hematopoietic cells in BM and PB by flow cytometry. Results newly document that the BM of dia-betic patients is depleted of hematopoietic and proangiogenic

PCs but not of differentiated hematopoietic cells. Finally, we show that both diabetes mellitus and HG alter the miR-155/FOXO3a/p21/p27kip1 signaling pathway, which represents a crucial mechanism controlling human CD34pos-PC renewal and differentiation. Importantly, forced expression of miR-155 inhibits the inductive effect of HG on the FOXO3a/p21/p27kip1 trio and reduces differentiation of CD34pos-PCs in a colony forming unit assay.

Diabetic Microangiopathy in Human BMA prominent feature of diabetic BM consists of substitu-tion of the hematopoietic component with adipose tissue. Interestingly, the effect of diabetes mellitus on marrow com-position was independent of background factors, with ex-ception of fat accumulation which was also predicted by the

Figure 5. Flow cytometry characterization of lineage committed hematopoietic cells and endothelial cells (ECs). A, Gating strategy for identification of B-lymphocytes, T-lymphocytes, and natural killer (NK) cells. B to D, Bar graphs showing the abundance of B-lymphocytes (B), T-lymphocytes (C) and NK cells (D) in bone marrow (BM) (i) and peripheral blood (PB) (ii). E, Gating strategy for identification of BM ECs (i) and bar graph showing average values (ii) *P<0.05 and **P<0.01 versus controls, §P<0.05 versus type-2 diabetes mellitus (T2DM). Controls, n=9 to 16; T2DM, n=7 to 9; T2DM+critical limb ischemia (CLI), n=7 to 8.




combination of duration of diabetes mellitus and body mass index. Although additional investigation is needed to interpret the relationship between fatty infiltration of the marrow and obesity, evidence indicates that BM adipocytes could contrib-ute to inhibit hematopoiesis in an obese model of T2DM.48

We found a direct correlation between vascular rarefaction and marrow remodeling, in line with the concept that a reduc-tion of nurturing vasculature is detrimental for SC homeostasis. Of note, regression analysis indicates that, besides grouping factor, both duration of diabetes mellitus and hypertension are good predictors of capillary and sinusoid rarefaction, whereas

arteriole density could be predicted only by variables inherent to diabetes mellitus, that is, disease duration and fasting glu-cose. The relationship between glycemic control and hyper-tension in the development of microangiopathy in peripheral organs is well documented. In CLI patients, who showed the most striking microvascular remodeling, analyses were per-formed on the proximal part of the amputated femoral bone which contains healthy tissue. We cannot exclude that the dif-ferent anatomic source, in addition to ischemia, might have an impact on the incremental reduction of vascularity observed in this category.

Figure 6. Diabetes mellitus-induced expressional changes in bone marrow (BM) CD34pos cells. A and B, Bar graphs showing mRNA levels of microRNA (miR)-155 (A) and its target FOXO3a (B). Controls, n=10; type-2 diabetes mellitus (T2DM), n=7; T2DM+critical limb ischemia (CLI), n=6. C, Graph showing the inverse correlation between miR-155 and FOXO3a mRNA levels in CD34pos cells. D, Representative microphotographs (i) and bar graph (ii) showing the in situ expression of FOXO3a in BM cells. Confocal microphotographs showing FOXO3a (red) localization in the cytoplasm (iii) and nucleus (iv) of CD34pos cells (green). Nuclei are stained blue with 4',6-diamidino-2-phenylindole (DAPI). n=5 per group. E and F, Bar graphs showing mRNA levels of CDKN1A/p21 (E) and CDKN1B/p27kip1 (F). Controls, n=10; T2DM, n=7; T2DM+CLI, n=6. *P<0.05 and **P<0.01 versus controls.




Figure 7. MicroRNA (miR)-155 overexpression in bone marrow (BM) CD34pos cells reverts high glucose (HG)-induced expressional changes and reduces the generation of myeloid and erythroid colonies. A, Bar graphs showing the effect of HG and pre–miR-155 or control scramble (SCR) transfection on CD34pos cell number. *P<0.05 versus SCR HG. Cells were cultured either in normal glucose (NG) (5 mmol/L D-glucose, NG) or HG (25 mmol/L D-glucose, HG) for 48 hours. B, Bar graphs showing mRNA levels of (i) FOXO3a, (ii) CDKN1A/p21, and (iii) CDKN1B/p27kip1 after pre–miR-155 or control SCR transfection. N=5 healthy donors per group assayed in duplicate. C, Colony forming unit (CFU) assay of CD34pos cells. (i) Representative images of CFU-granulocytes, erythroid, macrophage, megakaryocyte (CFU-GEMM), CFU-granulocyte, macrophage (CFU-GM), and CFU-erythroid (CFU-E) colonies. ii, Immunocytochemical characterization of CFU-isolated cells by staining for myeloperoxidase (MPO), a marker for granulocytes, CD68, a marker of macrophages, and glyphorin, a marker of erythrocytes. Arrows in CFU-GEMM point at MPOpos granulocytes. Nuclei are stained blue with hematoxylin. iii, Bar graph showing the effect of miR-155 overexpression on number (top) and area (bottom) covered by colonies. N=5 healthy donors per group. *P<0.05 and **P<0.01 versus SCR HG, §P<0.05 and §§P<0.01 versus SCR NG. Data represent means±SEM.




Impact of Diabetes Mellitus on PC AbundanceComparing marrow aspirates from 10 nondiabetic and 10 diabetic subjects, Fadini et al5 reported a significant reduc-tion of CD34pos cells in the latter group. Extending those data, we observed a decrease in both CD34posCD133pos and CD34posKDRpos PCs in BM of diabetic patients. In contrast, lineage committed hematopoietic cells, including B and T lymphocytes, were similarly abundant or even increased, as in the case of NK subpopulation. Altogether, these findings newly indicate a disparity in the depleting effect of diabe-tes mellitus on different hematopoietic cell lineages, which might contribute to the altered spectrum of circulating cells. Importantly, the dominant influence of diabetes mellitus on these cellular endpoints was confirmed by analysis of associ-ated risk factors and confounding variables.

Diabetes Mellitus Impinges on Master Molecular Regulators of HematopoiesisThe present study highlights new molecular mechanisms that may account for HSC depletion in diabetes mellitus. Results of cell cycle analysis show that freshly sorted CD34pos-PCs from BM of diabetic patients are held in G

1 phase, the typi-

cal restriction checkpoint where HSCs are cycle-arrested to prevent accrual of DNA damage. Moreover, diabetic BM cells show increased apoptosis, as assessed by immunohistochem-istry and cell cycle analysis.

FOXO transcription factors are typically involved in enforc-ing cell cycle checkpoints in hematopoietic cells with DNA damage.27 A potential mediator of FOXO-induced cell cycle arrest as well as of irreversible progression toward cell apop-tosis is p27kip1, a cyclin-dependent kinase inhibitor that report-edly reduces proliferation and survival of HSCs.49 We found that FOXO3a and downstream mediators, p21 and p27kip1, are remarkably upregulated in CD34pos-PCs from BM of diabetic patients, with these transcriptional changes being associated with reduction of cell viability. In vitro experiments challeng-ing CD34pos cells with HG confirm that glucose is sufficient to activate the proapoptotic signaling pathway in healthy HSCs.

MiRs regulate major cellular processes, including metabo-lism, apoptosis, and differentiation and also participate in the pathogenesis of human diseases.50–52 Seminal studies using conditional deletion of Dicer, which disrupts miR processing, revealed critical roles of miRs during development of hemato-poietic cell lineages.53,54 In particular, miR-155 regulates early stages of hematopoiesis through inhibition of multiple genes implicated in HSC survival and differentiation, such as the hu-man FOXO3a gene.22 Current knowledge, however, is mainly centered on the implication of miR-155 in inflammation and cancer.55 Intriguingly, a recent study showed that circulating levels of miR-155 are lower in patients with coronary artery disease compared with healthy controls, with an additive di-minishing effect of diabetes mellitus.56 It remains unknown what is the source of circulating miR-155 and whether a re-duction of miR-155 expression in BM-derived cells can con-tribute to this defect.

Here, we report for the first time that miR-155 expres-sion is reduced in BM CD34pos cells from diabetic patients and inversely correlated with levels of its validated target FOXO3a. Importantly, the inhibitory effect on miR-155 and

the increase of FOXO3a, p21 and p27kip1 by diabetes mellitus were independent of other background factors and all repli-cated by challenging healthy CD34pos cells with HG. To estab-lish causality, we next asked whether miR-155 would prevent the effects of HG. Results indicate that forced expression of miR-155 is able to reverse the HG-induced upregulation of FOXO3a, p21, and p27kip1. Furthermore, miR-155–transduced BM CD34pos cells formed fewer myeloid and erythroid colo-nies compared with scramble-transfected controls. The latter result replicates data from a previous study showing the abil-ity of miR-155 in blocking differentiation in models of human hematopoiesis.22 It remains unknown whether the described miR-dependent mechanism is implicated in other deficien-cies of diabetic CD34pos cells, including unresponsiveness to granulocyte colony-stimulating factor. Intriguingly, a recent study in healthy primates showed that granulocyte colony-stimulating factor–mobilized CD34pos cells express higher miR-155 levels compared with nonstimulated or Plerixafor-mobilized cells.57

Altogether, these findings suggest that deregulation of the miR-155/FOXO3a/p27 signaling pathway might contribute to BM CD34pos cell depletion in diabetes mellitus. More investi-gation is warranted to establish whether other HSC-associated miRs participate in determining an imbalance between endo-thelial progenitors and mature hematopoietic cells.

Clinical ImplicationsIn conclusion, this study draws attention to the BM as a pri-mary target of diabetes mellitus-induced damage. Our data suggest that the severity of systemic vascular disease has an impact on BM remodeling. Conversely, more severe BM pathologies can cause (or contribute to) macroangiopathy, through shortage of vascular regenerative cells. Moreover, it should be acknowledged that this study was conducted on aged subjects and that inference to a younger population is uncertain. Further research is warranted to find specific treat-ments able to preserve BM integrity in patients with diabetes mellitus.

Sources of FundingThis study was supported by a grant from the British Heart Foundation (BHF) entitled Bone marrow dysfunction alters vascular homeosta-sis in diabetes. C. Emanueli is a BHF senior basic research fellow. In addition, the study was supported by a grant from the Cariplo Foundation (Ref. 2011-0566).

DisclosuresNone.

References 1. Tepper OM, Galiano RD, Capla JM, Kalka C, Gagne PJ, Jacobowitz GR,

Levine JP, Gurtner GC. Human endothelial progenitor cells from type II diabetics exhibit impaired proliferation, adhesion, and incorporation into vascular structures. Circulation. 2002;106:2781–2786.

2. Loomans CJ, de Koning EJ, Staal FJ, Rookmaaker MB, Verseyden C, de Boer HC, Verhaar MC, Braam B, Rabelink TJ, van Zonneveld AJ. Endothelial progenitor cell dysfunction: a novel concept in the pathogenesis of vascular complications of type 1 diabetes. Diabetes. 2004;53:195–199.

3. Fadini GP, Miorin M, Facco M, Bonamico S, Baesso I, Grego F, Menegolo M, de Kreutzenberg SV, Tiengo A, Agostini C, Avogaro A. Circulating endothelial progenitor cells are reduced in peripheral




vascular complications of type 2 diabetes mellitus. J Am Coll Cardiol. 2005;45:1449–1457.

4. Fadini GP, Maruyama S, Ozaki T, Taguchi A, Meigs J, Dimmeler S, Zeiher AM, de Kreutzenberg S, Avogaro A, Nickenig G, Schmidt-Lucke C, Werner N. Circulating progenitor cell count for cardiovascular risk stratification: a pooled analysis. PLoS ONE. 2010;5:e11488.

5. Fadini GP, Boscaro E, de Kreutzenberg S, Agostini C, Seeger F, Dimmeler S, Zeiher A, Tiengo A, Avogaro A. Time course and mecha-nisms of circulating progenitor cell reduction in the natural history of type 2 diabetes. Diabetes Care. 2010;33:1097–1102.

6. Fadini GP, Albiero M, Seeger F, Poncina N, Menegazzo L, Angelini A, Castellani C, Thiene G, Agostini C, Cappellari R, Boscaro E, Zeiher A, Dimmeler S, Avogaro A. Stem cell compartmentalization in diabetes and high cardiovascular risk reveals the role of DPP-4 in diabetic stem cell mobilopathy. Basic Res Cardiol. 2013;108:313.

7. Segal MS, Shah R, Afzal A, Perrault CM, Chang K, Schuler A, Beem E, Shaw LC, Li Calzi S, Harrison JK, Tran-Son-Tay R, Grant MB. Nitric ox-ide cytoskeletal-induced alterations reverse the endothelial progenitor cell migratory defect associated with diabetes. Diabetes. 2006;55:102–109.

8. Kränkel N, Katare RG, Siragusa M, et al. Role of kinin B2 receptor sig-naling in the recruitment of circulating progenitor cells with neovascu-larization potential. Circ Res. 2008;103:1335–1343.

9. Busik JV, Tikhonenko M, Bhatwadekar A, et al. Diabetic retinopathy is associated with bone marrow neuropathy and a depressed peripheral clock. J Exp Med. 2009;206:2897–2906.

10. Ferraro F, Lymperi S, Méndez-Ferrer S, et al. Diabetes impairs hemato-poietic stem cell mobilization by altering niche function. Sci Transl Med. 2011;3:104ra101.

11. Fadini GP, Albiero M, Vigili de Kreutzenberg S, Boscaro E, Cappellari R, Marescotti M, Poncina N, Agostini C, Avogaro A. Diabetes impairs stem cell and proangiogenic cell mobilization in humans. Diabetes Care. October 30, 2012. doi: 10.2337/dc12-1084. http://care.diabetesjournals.org/content/early/2012/10/28/dc12-1084.long. Accessed January 7, 2013.

12. DiPersio JF. Diabetic stem-cell “mobilopathy”. N Engl J Med. 2011;365:2536–2538.

13. Orlandi A, Chavakis E, Seeger F, Tjwa M, Zeiher AM, Dimmeler S. Long-term diabetes impairs repopulation of hematopoietic progenitor cells and dysregulates the cytokine expression in the bone marrow mi-croenvironment in mice. Basic Res Cardiol. 2010;105:703–712.

14. Saito H, Yamamoto Y, Yamamoto H. Diabetes alters subsets of endothe-lial progenitor cells that reside in blood, bone marrow, and spleen. Am J Physiol, Cell Physiol. 2012;302:C892–C901.

15. Avecilla ST, Hattori K, Heissig B, et al. Chemokine-mediated interaction of hematopoietic progenitors with the bone marrow vascular niche is required for thrombopoiesis. Nat Med. 2004;10:64–71.

16. Butler JM, Nolan DJ, Vertes EL, et al. Endothelial cells are essential for the self-renewal and repopulation of Notch-dependent hematopoietic stem cells. Cell Stem Cell. 2010;6:251–264.

17. Oikawa A, Siragusa M, Quaini F, Mangialardi G, Katare RG, Caporali A, van Buul JD, van Alphen FP, Graiani G, Spinetti G, Kraenkel N, Prezioso L, Emanueli C, Madeddu P. Diabetes mellitus induces bone marrow mi-croangiopathy. Arterioscler Thromb Vasc Biol. 2010;30:498–508.

18. Park IK, Morrison SJ, Clarke MF. Bmi1, stem cells, and senescence regulation. J Clin Invest. 2004;113:175–179.

19. Chen CZ, Li L, Lodish HF, Bartel DP. MicroRNAs modulate hematopoi-etic lineage differentiation. Science. 2004;303:83–86.

20. Garzon R, Volinia S, Liu CG, et al. MicroRNA signatures associated with cytogenetics and prognosis in acute myeloid leukemia. Blood. 2008;111:3183–3189.

21. O’Connell RM, Chaudhuri AA, Rao DS, Gibson WS, Balazs AB, Baltimore D. MicroRNAs enriched in hematopoietic stem cells differen-tially regulate long-term hematopoietic output. Proc Natl Acad Sci USA. 2010;107:14235–14240.

22. Georgantas RW 3rd, Hildreth R, Morisot S, Alder J, Liu CG, Heimfeld S, Calin GA, Croce CM, Civin CI. CD34+ hematopoietic stem-progenitor cell microRNA expression and function: a circuit diagram of differentia-tion control. Proc Natl Acad Sci USA. 2007;104:2750–2755.

23. Masaki S, Ohtsuka R, Abe Y, Muta K, Umemura T. Expression pat-terns of microRNAs 155 and 451 during normal human erythropoiesis. Biochem Biophys Res Commun. 2007;364:509–514.

24. Yamamoto M, Kondo E, Takeuchi M, Harashima A, Otani T, Tsuji-Takayama K, Yamasaki F, Kumon H, Kibata M, Nakamura S. miR-155, a modulator of FOXO3a protein expression, is underexpressed and can-not be upregulated by stimulation of HOZOT, a line of multifunctional treg. PLoS ONE. 2011;6:e16841.

25. Stahl M, Dijkers PF, Kops GJ, Lens SM, Coffer PJ, Burgering BM, Medema RH. The forkhead transcription factor FoxO regulates transcription of p27Kip1 and Bim in response to IL-2. J Immunol. 2002;168:5024–5031.

26. Kops GJ, Dansen TB, Polderman PE, Saarloos I, Wirtz KW, Coffer PJ, Huang TT, Bos JL, Medema RH, Burgering BM. Forkhead transcription factor FOXO3a protects quiescent cells from oxidative stress. Nature. 2002;419:316–321.

27. Lei H, Quelle FW. FOXO transcription factors enforce cell cycle check-points and promote survival of hematopoietic cells after DNA damage. Mol Cancer Res. 2009;7:1294–1303.

28. Norgren L, Hiatt WR, Dormandy JA, Nehler MR, Harris KA, Fowkes FG; TASC II Working Group. Inter-Society Consensus for the Management of Peripheral Arterial Disease (TASC II). J Vasc Surg. 2007;45(suppl S):S5–67.

29. Mackie AR, Losordo DW. CD34-positive stem cells: in the treat-ment of heart and vascular disease in human beings. Tex Heart Inst J. 2011;38:474–485.

30. Sutherland DR, Anderson L, Keeney M, Nayar R, Chin-Yee I. The ISHAGE guidelines for CD34+ cell determination by flow cytom-etry. International Society of Hematotherapy and Graft Engineering. J Hematother. 1996;5:213–226.

31. Gallacher L, Murdoch B, Wu DM, Karanu FN, Keeney M, Bhatia M. Isolation and characterization of human CD34(-)Lin(-) and CD34(+)Lin(-) hematopoietic stem cells using cell surface markers AC133 and CD7. Blood. 2000;95:2813–2820.

32. Barcelos LS, Duplaa C, Kränkel N, et al. Human CD133+ progeni-tor cells promote the healing of diabetic ischemic ulcers by paracrine stimulation of angiogenesis and activation of Wnt signaling. Circ Res. 2009;104:1095–1102.

33. Stamm C, Kleine HD, Choi YH, Dunkelmann S, Lauffs JA, Lorenzen B, David A, Liebold A, Nienaber C, Zurakowski D, Freund M, Steinhoff G. Intramyocardial delivery of CD133+ bone marrow cells and coronary artery bypass grafting for chronic ischemic heart disease: safety and ef-ficacy studies. J Thorac Cardiovasc Surg. 2007;133:717–725.

34. Bartunek J, Vanderheyden M, Vandekerckhove B, Mansour S, De Bruyne B, De Bondt P, Van Haute I, Lootens N, Heyndrickx G, Wijns W. Intracoronary injection of CD133-positive enriched bone marrow pro-genitor cells promotes cardiac recovery after recent myocardial infarc-tion: feasibility and safety. Circulation. 2005;112:I178–I183.

35. Mansour S, Roy DC, Bouchard V, Stevens LM, Gobeil F, Rivard A, Leclerc G, Reeves F, Noiseux N. One-year safety analysis of the COMPARE-AMI trial: comparison of intracoronary injection of CD133 bone marrow stem cells to placebo in patients after acute myo-cardial infarction and left ventricular dysfunction. Bone Marrow Res. 2011;2011:385124.

36. Friedrich EB, Walenta K, Scharlau J, Nickenig G, Werner N. CD34-/CD133+/VEGFR-2+ endothelial progenitor cell subpopulation with po-tent vasoregenerative capacities. Circ Res. 2006;98:e20–e25.

37. Pelosi E, Valtieri M, Coppola S, Botta R, Gabbianelli M, Lulli V, Marziali G, Masella B, Müller R, Sgadari C, Testa U, Bonanno G, Peschle C. Identification of the hemangioblast in postnatal life. Blood. 2002;100:3203–3208.

38. Botta R, Gao E, Stassi G, Bonci D, Pelosi E, Zwas D, Patti M, Colonna L, Baiocchi M, Coppola S, Ma X, Condorelli G, Peschle C. Heart in-farct in NOD-SCID mice: therapeutic vasculogenesis by transplanta-tion of human CD34+ cells and low dose CD34+KDR+ cells. FASEB J. 2004;18:1392–1394.

39. Madeddu P, Emanueli C, Pelosi E, Salis MB, Cerio AM, Bonanno G, Patti M, Stassi G, Condorelli G, Peschle C. Transplantation of low dose CD34+KDR+ cells promotes vascular and muscular regeneration in ischemic limbs. FASEB J. 2004;18:1737–1739.

40. Vasa M, Fichtlscherer S, Aicher A, Adler K, Urbich C, Martin H, Zeiher AM, Dimmeler S. Number and migratory activity of circulating endo-thelial progenitor cells inversely correlate with risk factors for coronary artery disease. Circ Res. 2001;89:E1–E7.

41. Pirro M, Schillaci G, Menecali C, Bagaglia F, Paltriccia R, Vaudo G, Mannarino MR, Mannarino E. Reduced number of circulating endothe-lial progenitors and HOXA9 expression in CD34+ cells of hypertensive patients. J Hypertens. 2007;25:2093–2099.

42. Hirschi KK, Ingram DA, Yoder MC. Assessing identity, phenotype, and fate of endothelial progenitor cells. Arterioscler Thromb Vasc Biol. 2008;28:1584–1595.

43. Spinetti G, Fortunato O, Cordella D, Portararo P, Kränkel N, Katare R, Sala-Newby GB, Richer C, Vincent MP, Alhenc-Gelas F, Tonolo


http://care.diabetesjournals.org/content/early/2012/10/28/dc12-1084.long

http://care.diabetesjournals.org/content/early/2012/10/28/dc12-1084.long



G, Cherchi S, Emanueli C, Madeddu P. Tissue kallikrein is essen-tial for invasive capacity of circulating proangiogenic cells. Circ Res. 2011;108:284–293.

44. Kim H, Cho HJ, Kim SW, Liu B, Choi YJ, Lee J, Sohn YD, Lee MY, Houge MA, Yoon YS. CD31+ cells represent highly angiogenic and vas-culogenic cells in bone marrow: novel role of nonendothelial CD31+ cells in neovascularization and their therapeutic effects on ischemic vas-cular disease. Circ Res. 2010;107:602–614.

45. Kucia M, Reca R, Jala VR, Dawn B, Ratajczak J, Ratajczak MZ. Bone marrow as a home of heterogenous populations of nonhematopoietic stem cells. Leukemia. 2005;19:1118–1127.

46. Orlic D, Kajstura J, Chimenti S, Jakoniuk I, Anderson SM, Li B, Pickel J, McKay R, Nadal-Ginard B, Bodine DM, Leri A, Anversa P. Bone mar-row cells regenerate infarcted myocardium. Nature. 2001;410:701–705.

47. Dominici M, Le Blanc K, Mueller I, Slaper-Cortenbach I, Marini F, Krause D, Deans R, Keating A, Prockop Dj, Horwitz E. Minimal criteria for de-fining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy. 2006;8:315–317.

48. Naveiras O, Nardi V, Wenzel PL, Hauschka PV, Fahey F, Daley GQ. Bone-marrow adipocytes as negative regulators of the haematopoietic microenvironment. Nature. 2009;460:259–263.

49. Medema RH, Kops GJ, Bos JL, Burgering BM. AFX-like Forkhead tran-scription factors mediate cell-cycle regulation by Ras and PKB through p27kip1. Nature. 2000;404:782–787.

50. Caporali A, Emanueli C. MicroRNA regulation in angiogenesis. Vascul Pharmacol. 2011;55:79–86.

51. Shantikumar S, Caporali A, Emanueli C. Role of microRNAs in diabetes and its cardiovascular complications. Cardiovasc Res. 2012;93:583–593.

52. Fasanaro P, Greco S, Ivan M, Capogrossi MC, Martelli F. microRNA: emerging therapeutic targets in acute ischemic diseases. Pharmacol Ther. 2010;125:92–104.

53. Cobb BS, Nesterova TB, Thompson E, Hertweck A, O’Connor E, Godwin J, Wilson CB, Brockdorff N, Fisher AG, Smale ST, Merkenschlager M. T cell lineage choice and differentiation in the absence of the RNase III enzyme Dicer. J Exp Med. 2005;201:1367–1373.

54. Koralov SB, Muljo SA, Galler GR, Krek A, Chakraborty T, Kanellopoulou C, Jensen K, Cobb BS, Merkenschlager M, Rajewsky N, Rajewsky K. Dicer ablation affects antibody diversity and cell survival in the B lymphocyte lineage. Cell. 2008;132:860–874.

55. Tili E, Michaille JJ, Wernicke D, Alder H, Costinean S, Volinia S, Croce CM. Mutator activity induced by microRNA-155 (miR-155) links inflammation and cancer. Proc Natl Acad Sci USA. 2011;108: 4908–4913.

56. Fichtlscherer S, De Rosa S, Fox H, Schwietz T, Fischer A, Liebetrau C, Weber M, Hamm CW, Röxe T, Müller-Ardogan M, Bonauer A, Zeiher AM, Dimmeler S. Circulating microRNAs in patients with coronary ar-tery disease. Circ Res. 2010;107:677–684.

57. Donahue RE, Jin P, Bonifacino AC, Metzger ME, Ren J, Wang E, Stroncek DF. Plerixafor (AMD3100) and granulocyte colony-stim-ulating factor (G-CSF) mobilize different CD34+ cell populations based on global gene and microRNA expression signatures. Blood. 2009;114:2530–2541.

What Is Known?

• Diabetes mellitus is associated with reduced levels of circulating progenitor cells.

• Studies in diabetic animal models indicate the presence of microangi-opathy in bone marrow endangering resident stem cell integrity.

• A restricted spectrum of microRNAs (miRs) that regulate self-renewal is expressed in hematopoietic stem cells.

What New Information Does This Article Contribute?

• In patients with diabetes mellitus, there is microangiopathy in the bone marrow, which is associated with incremental vascular damage in the presence of critical limb ischemia.

• The bone marrow of diabetic patients is depleted of hematopoietic and proangiogenic progenitor cells but not differentiated hematopoi-etic cells.

• Hyperglycemia inhibits miR-155, thereby releasing the activity of proapoptotic pathway involving FOXO3a/p21.

Although remodeling of blood vessels in peripheral organs has been extensively studied, little is known about the impact

of common diseases on the marrow vascular niche, which is crucial for stem cell homeostasis and mobilization. Therefore, we investigated the effect of diabetes mellitus on bone marrow stem cells and their nurturing vasculature in humans. Results show a profound remodeling of the vascular niche, which is mainly replaced by fat, especially in patients with critical isch-emia. Stem cell depletion did not preclude a regular abundance of mature hematopoietic cells, suggesting a defect in stem cell self-renewal. Investigation of underpinning mechanisms re-vealed that hyperglycemia inhibits a master regulator of hema-topoietic cell fate, miR-155, thereby resulting in an unbalance between proliferation and differentiation. This was corrected by forced expression of miR-155. Our findings advance current understanding of pathological mechanisms leading to collapse of the vascular niche and reduced availability of proangiogenic cells. These data provide a key for interpretation of diabetes mellitus-associated defect in stem cell mobilization. In addition, our study reveals a negative circuitry, normally contrasted by a hematopoietic miR- responsible for pauperization of the marrow regenerative potential.

Novelty and Significance



Supplemental Material Detailed Methods Study protocol and statistics Primary endpoints of the study were BM capillary density and abundance of BM CD34pos cells. Secondary endpoints comprise sinusoid and arteriole density and antigenic and transcriptional characteristics of CD34pos cells. In pilot studies, we determined that a sample size of 7 in each of the 3 groups would have an 80% power to detect a difference between means of 13.6 capillaries per mm2 with an expected average standard deviation of 8.3 and a significance level (alpha) of 0.05 (two-tailed). The same group size would have a 90% power to detect a difference between means of 1.1 CD34pos cells/100 BM mononuclear cells (BM-MNCs) with an expected average standard deviation of 0.58. Recruitment of patients was continued until reaching the requested sample size as by power calculation for primary endpoints. Since patients were screened consecutively, the actual sample size eventually exceeded the minimum requested in 2 of the 3 groups. Moreover, depending on the dimensions of bone leftovers, we attempted to perform all the analyses on the same sample. When this was not possible, priority was given to histological analyses until reaching the requested group size and then to flow cytometry. Online Table II summarizes the samples attribution to different assays. In order to verify the direct effect of hyperglycemia on molecular mechanisms, CD34pos-PCs were immunosorted from the BM of 5 non-diabetic patients undergoing hip replacement surgery and used for expressional studies and colony forming unit (CFU) assays, following exposure to HG with or without miR-155 forced expression. Results are presented as means±SEM. Normal distribution of the variables of interest was verified with D’agostino and Pearson test. data failed to pass normality or equal variance tests, non-parametric analysis was applied and results were expressed as median with 5-95 percentile distribution. For gene expression studies, logarithm transformation of the data was used followed by parametric tests. Multiple groups were compared by parametric analysis of variance (ANOVA), followed by Bonferroni t test, or non-parametric analysis of variance on ranks, followed by Tukey pairwise comparison or Dunnett’s test for multiple comparisons against a single control group. Comparison of 2 groups was carried out by paired or unpaired Student t test or Mann-Whitney rank sum test. Comparison of categorical variables was made by chi-square test or Fisher's exact test. In addition, the Cohen’s d index was calculated to determine if a statistically significant difference was of biological importance. The results of Cohen’s d are considered to be small (< 0.5), medium (0.5 – 0.8), or large (>0.8). In order to control the effects of background factors, means of dependent variables were compared and adjusted across the study groups by analysis of covariance (ANCOVA) (Online Tables III and IV). Furthermore, multiple linear regression analyses were performed to verify if a given endpoint is predicted by combination of independent variables, including group, duration of diabetes, fasting glycemia, age, gender, coronary artery disease (CAD), stroke, hypertension, body mass index (BMI), smoking and treatment (Online Table V). Since grouping factor, duration of diabetes and fasting glucose showed a high collinearity index (~2), each of the 3 independent variables was computed distinct from the other 2 in multiple linear regression analyses. The relationship between miR-155 and FOXO3a expression in CD34pos cells was calculated using the Spearman correlation coefficient. A p value <0.05 was considered significant. Stated n values represent biological replicates. Histomorphometry, immunohistochemistry, and immunofluorescence analyses A femoral bone fragment of at least 0.5 cm3 was fixed in 10% neutral buffered formalin, decalcified in 10% formic acid, and embedded in paraffin. Analyses were conducted on 3 micrometer-thick BM sections mounted on silane-coated slides (Dako, Denmark). Briefly, for histomorphometric studies, BM sections were stained with Hematoxylin-Eosin (HE) and images captured with a Nikon E800light microscope (Nikon, The Netherlands) were analyzed to assess the marrow composition using Image J, a Java-based processing program from the National Institutes of Health (NIH) (3 sections and 4-field/section for each patient). For immunohistochemistry, BM staining was conducted with primary antibodies followed by appropriate secondary antibody (Online Table VI). A Peroxidase/DAB commercial kit



(EnVision Detection Systems, Dako, Denmark) was used as revelation system. Counterstaining of nuclei was performed using Mayer's Hematoxylin (Sigma, St. Louis, USA). Images were captured with a Nikon E800light microscope as mentioned above. Immunofluorescence microscopy was performed by incubating BM sections with appropriately diluted primary antibodies overnight at 4°C, followed by secondary fluorescence-conjugated antibodies for 1 hour at room temperature. Image acquisition was carried out with a confocal laser microscope (Leica TCS-SP2, Leica Microsystems, Germany). In addition, percentage of apoptotic BM cells was measured as function of FITC-positive nuclei that bear DNA breaks labeled by TUNEL assay (Calbiochem, Germany). Fluorescence was visualized and captured using AXIO OBSERVER A1 microscope equipped with digital image processing software (AxioVision Imaging System), both from Zeiss (Germany). Cell isolation and culture MNCs were isolated from total BM single cell suspension by centrifugation on Histopaque-1077 density medium (Sigma, St. Louis, USA). CD34pos cells were separated from BM-MNCs by magnetic bead-assisted cell sorting (MACS, MiltenyiBiotec, Germany). CD34 purity was ~80% as confirmed by flow cytometry. Cells were then used for molecular biology studies either immediately after sorting or following exposure to high glucose (HG). To the latter aim, CD34pos cells were cultured for 24 hours in standard medium (EBM) (LONZA, Switzerland), 10% fetal bovine serum (FBS), human Stem Cell Factor (hSCF, 100ng/mL) (RD, USA), interleukin 3 (IL3, 20 ng/mL), and FMS Like Tyrosine Kinase 3 Ligand (Flt3-L, 100ng/mL) (both from Provitro Gmbh, Germany) containing 5 mM D-glucose (normal glucose, NG), or 25mM D-glucose (HG). Preservation of the antigenic profile after cell culture was confirmed by flow cytometry. miR-155 Transfection CD34pos were transfected with 50 nmol/L pre-miR-155 or negative control (a non-targeting sequence, also identified as scramble, SCR, throughout the manuscript) (all from Applied Biosystems) using GeneSilencer (Dharmacon) following the manufacturer’s instructions. Changes in relative miR-155 expression were measured by TaqMan PCR (vide infra). Colony Forming Unit Assay CD34pos cells were transduced with miR-155or scramble control (Applied Biosystem) as mentioned above and used in standard colony-forming unit (CFU) assays for myeloid and erythroid differentiation following the manufacturer’s instructions (Stem Cells Technologies). Colonies were harvested at day 14 and isolated cells were placed on glass slides for staining with specific markers. In particular, myeloperoxidase (MPO), a marker of granulocytes, was used to stain cells in CFU-GEMM colonies, CD68, a marker of macrophages, to stain cells from CFU-GM colonies, and glyphorin, a marker of erythrocutes to characterize cells from CFU-E colonies. Flow cytometry The following cell populations were identified within BM-MNCs and PB-MNCs as previously described in detail:1-5 HSCs (CD45dimCD34pos and CD45dimCD34posCD133pos), CD34posCD14posCD45dimKDRposCXCR4pos and CD34posCD14negCD45negKDRposCXCR4pos cells, Natural Killer cells (NKs, CD3negCD56posCD16pos), T-lymphocytes (CD45posCD3pos), B-lymphocytes (CD45posCD19pos), non-hematopoietic SCs (CXCR4posCD34posCD45neg and c-Kitpos cells) and endothelial cells (ECs, CD45negCD31posCD144pos). Cell cycle was analyzed on CD34pos cells fixed in 70% ethanol solution for 18 hours at -20°C. Cells were then washed twice with cold PBS, incubated in a 0.1% TritonX-100, 0.1% sodium

citrate, 50g/ml propidium iodide solution for 40min at 37°C in the dark, and analyzed by flow cytometry within 45 min. For each test, 1X105 to 5X106 total events were analyzed in a FACSCanto flow cytometer using the FACSDiva software (both from BD Biosciences, New Jersey, USA). RT-PCR analysis on sorted CD34pos BM-MNCs



RNA extraction and cDNA synthesis was performed using Power SYBR® Green Cells-to-CT™ Kit(Applied Biosystems, California, USA). PCR primers for gene expression were designed with Primer3 software. CDKN1A-p21 (sense: 5’-GACACCACTGGAGGGTGACT-3’; antisense: 5’-CAGGTCCACATGGTCTTCCT-3’); CDKN1B-p27kip1 (sense: 5’-GAGTGGCAAGAGGTGGAGAA-3’; antisense: 5’-GCGTGTCCTCAGAGTTAGCC-3’); FOXO3A (sense: 5’- ACAAACGGCTCACTCTGTCC-3’; antisense: 5’- ATTCTGGACCCGCATGAAT 3’); 18S (sense: 5’-CGCAGCTAGGAATAATGGAATAGG-3’; antisense: 5’-CATGGCCTCAGTTCCGAAA-3’). RNA reverse transcription and PCR to measure miR expression was performed using commercially available TaqManmiRNA reverse transcription kit and miR-specific primers, according the manufacturer’s instructions (Applied Biosystems, California, USA). Each Real time PCR reaction was performed in triplicate, and relative expression of mRNAs and miRs was calculated by the 2– Ct method 6 using 18S ribosomal RNA or the U6 small nucleolar RNA (snRU6) as endogenous control respectively. Supplemental References 1. Spinetti G, Fortunato O, Cordella D, Portararo P, Krankel N, Katare R, Sala-Newby GB, Richer C, Vincent MP, Alhenc-Gelas F, Tonolo G, Cherchi S, Emanueli C, Madeddu P. Tissue kallikrein is essential for invasive capacity of circulating proangiogenic cells. Circ Res. 2011;108(3):284-293. 2. Hirschi KK, Ingram DA, Yoder MC. Assessing identity, phenotype, and fate of endothelial progenitor cells. Arterioscler Thromb Vasc Biol. 2008;28(9):1584-1595. 3. Amadesi S, Reni C, Katare R, Meloni M, Oikawa A, Beltrami AP, Avolio E, Cesselli D, Fortunato O, Spinetti G, Ascione R, Cangiano E, Valgimigli M, Hunt SP, Emanueli C, Madeddu P. Role for substance p-based nociceptive signaling in progenitor cell activation and angiogenesis during ischemia in mice and in human subjects. Circulation. 2012;125(14):1774-1786, S1771-1719. 4. Schmidt-Lucke C, Fichtlscherer S, Aicher A, Tschope C, Schultheiss HP, Zeiher AM, Dimmeler S. Quantification of circulating endothelial progenitor cells using the modified ISHAGE protocol. PLoS One. 2010;5(11):e13790. 5. Sutherland DR, Anderson L, Keeney M, Nayar R, Chin-Yee I. The ISHAGE guidelines for CD34+ cell determination by flow cytometry. International Society of Hematotherapy and Graft Engineering. J Hematother. 1996;5(3):213-226. 6. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T) method. Nature protocols. 2008;3(6):1101-1108.



Online Table I: Characteristics of study subjects

Controls (n=49) T2D (n=10) T2D+CLI (n=23) Test p= Age, years 70 (42-85) 73 (55-78) 76 (37-88) ANOVA 0.14

Male, % 57 90 57 Chi-square 0.13

BMI, kg/m2 27 (20-36) 33 (24-38) 26 (21-35) ANOVA <0.01

Diabetes duration, years - 9 (1-15) 20 (14-36) Unpaired t test <0.001

Fasting glucose, mg/dL 91 (72-129) 122 (102-170) 182 (91-277) Kruskal-Wallis test <0.001

HbA1c , %Hb ND 6.6 (6-8.2) 7.4 (6.0-12.1) Unpaired t test 0.13

Smoking, % 11 0 65 Chi-square <0.001

Hypertension, % 47 90 96 Chi-square <0.001

Complications, % Coronary artery disease 14.8 20.0 56.5 Chi-square 0.001

Stroke 2.1 0 39.1 Chi-square <0.001

Retinopathy - 10 34.8 Fisher exact test 0.41

Neuropathy - 10 14 Fisher exact test 0.94

Nephropathy - 10 39.1 Fisher exact test 0.25

Laboratory tests

LDL, mg/dL 78 (21-190) 104 (19-126) ND Unpaired t test 0.74

HDL, mg/d L 43 (20-84) 48 (28-62) ND Unpaired t test 0.56

Triglycerides, mg/dL 107 (53-311) 119 (51-171) 125 (73-314) Kruskal-Wallis test 0.58

Creatinine, mg/dL 0.8 (0.5-1.6) 1.0 (0.7-2.1) 1.1 (0.6-2.6) Kruskal-Wallis test 0.01

Medications

Insulin, % - 11 90.4 Fisher exact test <0.001

Oral anti-diabetic drugs, % - 100 4.7 Fisher exact test <0.001

Glitazones, % - 33 0 Fisher exact test <0.05

Statin, % 4.3 40.0 52.3 Chi-square <0.001

Anti-hypertensive drugs, n. 1 (0-2) 2 (0-4) 2 (0-2) Chi-square <0.001

Quantitative data are expressed as median with minimum and maximum. T2D: Type 2 diabetic patients, CLI: Critical Limb Ischemia, CAD: Coronary

Artery Disease, ND: not determined



Online Table II. Samples distribution to different assays

Number Group Morphometry IHC FACS PCR

1 Control x x

2 Control x x

3 Control x x

4 Control x x x

5 Control x x x

6 Control x x x

7 Control x x x

8 Control x x x

9 Control x x x

10 Control x x x

11 Control x

12 Control x

13 Control x

14 Control x

15 Control x

16 Control x

17 Control x

18 Control x

19 Control x

20 Control x

21 Control x

22 Control x

23 Control x

24 Control x

25 Control x x

26 Control x x

27 Control x x

28 Control x x



29 Control x x

30 Control x

31 Control x

32 Control x

33 Control x

34 Control x

35 Control x

36 Control x

37 Control x

38 Control x

39 Control x

40 Control x

41 Control x

42 Control x

43 Control x

44 Control x

45 Control x

46 Control x

47 Control x

48 Control x

49 Control x


1 T2D x x x

2 T2D x x x

3 T2D x x x

4 T2D x x x x

5 T2D x x x x

6 T2D x x x x

7 T2D x x x x

8 T2D x x



9 T2D x x

10 T2D x


1 T2D+CLI x x x

2 T2D+CLI x x x

3 T2D+CLI x x x

4 T2D+CLI x x x

5 T2D+CLI x x x

6 T2D+CLI x x

7 T2D+CLI x x x

8 T2D+CLI x x x x

9 T2D+CLI x x x

10 T2D+CLI x x

11 T2D+CLI x

12 T2D+CLI x

13 T2D+CLI x

14 T2D+CLI x

15 T2D+CLI x

16 T2D+CLI x

17 T2D+CLI x

18 T2D+CLI x x

19 T2D+CLI x

20 T2D+CLI x

21 T2D+CLI x x

22 T2D+CLI x

23 T2D+CLI x



Online Table III: Analysis of covariance for histological and cellular endpoints.

Group mean values Pairwise Comparison, p=

Between-group p=

Dependent Variable C T2D T2D+CLI C vs. T2D

C vs. T2D+CLI

T2D vs.T2D+CLI

Co

vari

ates

: Age

, Gen

de

r, B

MI

His

tolo

gy/I

HC

Hematopoietic fraction 50.2±4.1 33.5±5.3 21.5±3.9 0.0729 0.0002 0.2807 <0.001

Fat fraction 38.1±3.5 60.7±4.5 76.1±3.4 0.0031 <0.0001 0.0466 <0.001

Capillaries 25.3±2.4 11.3±3.1 11.8±2.3 0.0075 0.0022 1.0000 0.001

Sinusoids 16.0±1.7 11.6±2.2 5.5±1.6 0.4160 0.0009 0.1246 0.001

Arterioles 23.4±2.4 19.2±3.1 7.7±2.3 0.9016 0.0004 0.0256 <0.001

CD34+CD45+ cells 62±11 60±14 18±12 1.0000 0.0449 0.1082 0.030

CD34-CD45+ cells 141±13 82±17 34±15 0.0571 0.0002 0.1477 <0.001

FAC

S

BM CD34+CD45dim cells 1.04±0.14 0.45±0.23 0.42±0.16 0.1477 0.0352 0.1477 0.023

PB CD34+CD45dim cells 0.052±0.016 0.035±0.021 0.047±0.017 1.0000 1.0000 1.0000 0.978

BM CD34+CD133+ cells 0.007±0.001 0.002±0.001 0.001±0.001 0.0851 0.0130 1.0000 0.012

PB CD34+CD133+ cells 0.004±0.001 0.002±0.001 0.001±0.001 1.0000 0.3228 1.0000 0.257

BM CD34+KDR+CD45dim cells 0.031±0.005 0.004±0.011 0.010±0.001 0.0606 0.1191 0.8904 0.025

PB CD34+KDR+CD45dim cells 0.007±0.001 0.002±0.001 0.001±0.006 0.0851 0.0130 0.8904 0.041

BM CD34+CD14+CD45dimKDR+CXCR4+ cells 0.051±0.011 0.006±0.008 0.024±0.010 0.2368 0.3393 1.0000 0.134

PB CD34+CD14+CD45dimKDR+CXCR4+ cells 0.019±0.005 0.006±0.001 0.0001±0.005 0.6655 0.0568 1.0000 0.054

Co

vari

ates

: Glit

azo

ne

, Sta

tin

, An

ti-

hyp

ert

en

sive

Dru

gs

His

tolo

gy/I

HC

Hematopoietic fraction 50.2±4.6 28.5±5.5 26.3±4.4 0.0319 0.0097 1.0000 0.007

Fat fraction 38.4±3.8 63.3±4.5 72.1±3.7 0.0027 0.0001 0.4909 <0.001

Capillaries 21.3±1.8 12.9±2.1 15.2±1.7 0.0316 0.1094 1.0000 0.025

Sinusoids 14.5±1.8 12.9±2.2 6.4±1.8 1.0000 0.0335 0.1225 0.025

Arterioles 23.7±2.8 19.8±3.4 7.5±2.8 1.0000 0.0050 0.0452 0.004

CD34+CD45+ cells 60±11 57±14 22±12 1.0000 0.1268 0.2131 0.077

CD34-CD45+ cells 144±15 74±19 36±14 0.0500 0.0010 0.4695 0.001

FAC

S

BM CD34+CD45dim cells 1.09±0.17 0.36±0.26 0.36±0.21 0.1309 0.0081 1.0000 0.005

PB CD34+CD45dim cells 0.043±0.016 0.051±0.017 0.051±0.016 1.0000 1.0000 1.0000 0.932

BM CD34+CD133+ cells 0.008±0.002 0.003±0.002 0.001±0.002 0.1895 0.0434 1.0000 0.042

PB CD34+CD133+ cells 0.004±0.002 0.002±0.002 0.002±0.002 1.0000 0.9637 1.0000 0.699







Co

vari

ate

: Cre

atin

ine

His

tolo

gy/I

HC

Hematopoietic fraction 50.4±3.9 32.7±4.8 20.9±4.0 0.0297 0.0001 0.2345 <0.001 Fat fraction 37.7±3.3 60.9±3.9 76.8±3.3 0.0006 <0.0001 0.0192 <0.001

Capillaries 25.9-±2.4 9.8±2.9 12.7±2.4 0.0012 0.0034 1.0000 0.001

Sinusoids 15.8±1.7 10.9±2.0 5.7±1.7 0.2320 0.0014 0.1986 0.002

Arterioles 23.2±2.3 18.4±2.8 7.9±2.3 0.6182 0.0006 0.0312 0.001

CD34+CD45+ cells 58±10 58±12 23±13 1.0000 0.1612 0.2066 0.109

CD34-CD45+ cells 134±13 83±17 42±17 0.0803 0.0019 0.3166 0.002

FAC

S

BM CD34+CD45dim cells 1.16±0.14 0.29±0.22 0.32±0.17 0.0089 0.0042 1.0000 0.001

PB CD34+CD45dim cells 0.051±0.014 0.039±0.016 0.047±0.018 1.0000 1.0000 1.0000 0.854

BM CD34+CD133+ cells 0.008±0.001 0.003±0.001 0.001±0.001 0.0573 0.0227 1.0000 0.015

PB CD34+CD133+ cells 0.004±0.001 0.002±0.001 0.001±0.002 1.0000 0.9637 1.0000 0.538





Co

vari

ates

: Pre

vio

us

Myo

card

ial

Infa

rcti

on

, Str

oke

, Hyp

ert

en

sio

n

His

tolo

gy/I

HC

Hematopoietic fraction 51.4±3.9 36.03±4.5 20.8±4.0 0.0497 0.0002 0.0795 <0.001

Fat fraction 38.1±3.3 57.4±3.8 76.6±3.4 0.0024 <0.0001 0.0051 <0.001

Capillaries 24.2±2.5 15.6±2.9 11.4±2.6 0.1033 0.0111 0.9693 0.010

Sinusoids 14.9±1.7 12.9±1.9 6.0±4.2 1.0000 0.0079 0.0582 0.008

Arterioles 24.7±2.3 21.2±2.7 5.9±2.4 1.0000 0.0001 0.0015 <0.001

CD34+CD45+ cells 65±9 51±12 18±12 1.0000 0.0248 0.1846 0.027

CD34-CD45+ cells 136±13 83±16 37±16 0.0534 0.0006 0.1839 0.001

FAC

S

BM CD34+CD45dim cells 1.34±0.17 0.37±0.22 0.32±0.20 0.0373 0.0379 1.0000 0.018

PB CD34+CD45dim cells 0.047±0.017 0.044±0.017 0.055±0.017 1.0000 1.0000 1.0000 0.905

BM CD34+CD133+ cells 0.007±0.001 0.003±0.001 0.002±0.001 0.0955 0.0398 1.0000 0.033

PB CD34+CD133+ cells 0.004±0.001 0.003±0.001 0.002±0.001 1.0000 1.0000 1.0000 0.491





Values are estimated marginal means ± Std. Error. Insulin, Oral Anti-diabetic Drugs, HDL, LDL and Triglycerides not included because the statistical software cannot solve equation.



Online Table IV: Effect of diabetes on the expression of miR-155, FOXO3a, p21 and p27kip1 before and after adjustment for background covariates by ANCOVA.

Covariate

Dependent Variable

p= Adjusted

p=

Age, Gender, BMI miR-155 0.019 0.013

FOXO3a 0.004 0.005

p21 0.001 0.006

p27 kip1 0.006 0.035

Glitazone, Statin, Anti-hypertensive Drugs

miR-155 0.019 0.050

FOXO3a 0.004 0.031

p21 0.001 0.030

p27 kip1 0.006 0.196

Creatinine miR-155 0.019 0.004

FOXO3a 0.004 0.010

p21 0.001 0.001

p27 kip1 0.006 0.004

Previous Myocardial Infarction, Stroke, Hypertension

miR-155 0.019 0.041

FOXO3a 0.004 0.017

p21 0.001 0.008

p27kip1 0.006 0.072



Online Table V: Multiple Regression Analysis

Dependent Variable Independent Variable

p= Independent Variable

p= Independent Variable

p=

His

tolo

gy/I

HC

Hematopoietic fraction Group <0.001 Diabetes duration 0.009

Fat fraction Group <0.001 Diabetes duration BMI

<0.001 0.040

Fasting glucose 0.017

Capillaries Group Hypertension

<0.001 <0.001

Diabetes duration Hypertension

<0.001 <0.001

Sinusoids Group Hypertension

0.005 0.048

Diabetes duration 0.004 Fasting glucose Hypertension

0.022 0.031

Arterioles Group <0.001 Diabetes duration <0.001 Fasting glucose 0.038

CD34+CD45+ cells Group 0.020

CD34-CD45+ cells Group <0.001 Diabetes duration 0.003 Fasting glucose 0.032

FAC

S

BM CD34+CD45dim cells Group <0.001 Diabetes duration 0.011 Fasting glucose 0.003

BM CD34+CD133+ cells Group 0.003 Diabetes duration 0.012 Fasting glucose 0.016

BM CD34+KDR+CD45dim cells Group 0.017 Diabetes duration 0.050 Fasting glucose 0.028

PB CD34+KDR+CD45dim cells Group 0.050 Diabetes duration 0.043

PB CD34+CD14+CD45dimKDR+CXCR4+ cells Group 0.049

Results indicate the independent variables that, alone or in linear combination with other independent variables, predict the dependent variable

under study



Online Table VI. Monoclonal and polyclonal antibodies used in this study

Primary antibody Dilution Incubation Time Source

Anti-CD34 (Code:M7165, mc) 1:10 O/N DAKO

Anti-CD45 (Code ab10559, pc) 1:50 60 min Abcam

Anti-CD31 (Code: M0823, mc) 1:10 60 min DAKO

Anti-alpha-SMA (Code: M0851, mc) 1:20 60 min DAKO

Anti-CD68 (mc) - 30 min Ventana Medical Systems

Anti-Myeloperoxidase (pc) - 30 min Ventana Medical Systems

Anti-Gycophoryn (mc) - 30 min Ventana Medical Systems

mc: monoclonal; pc: polyclonal ; O/N : overnight



Supplemental Figures and Figure legends

Online Figure I. Immunohistochemical assessment of apoptosis in CD34pos BM cells.

Representative microphotograph of human BM stained with anti-CD34 antibody in combination

with TUNEL reaction to show apoptosis.



Online Figure II. Flow cytometry analysis of c-kitposMNCs. Bar graphs showing the

percentage of cKitpos cells. *p<0.05 vs. controls, §p<0.05 vs. T2D. Controls, n=11; T2D, n=6 to 9;

T2D+CLI, n=8 to 9.



Online Figure III. Flow cytometry analysis of cell cycle in freshly-sorted CD34pos cells: (i) gating

strategy, (ii) control, (iii) T2D.



Online Figure IV. Effect of high glucose on cell counts and gene expression. A to E) Bar

graphs showing the effect of high glucose (25mM D-glucose, HG) on BM CD34pos cell counts

(A), levels of miR-155 (B), FOXO3a (C), CDKN1A/p21 (D) and CDKN1B/p27kip1 (E). Normal

glucose (5mM D-glucose, NG). *p<0.05 and **p<0.01 vs. NG. n=7 per group.



Online Figure V. The osmotic control mannitol does not influence the expression of miR-

155 and downstream targets in CD34pos-PCs from non-diabetic patients. Bar graphs of real

time PCR showing levels of miR-155 (A), FOXO3a (B), CDKN1A/p21 (C) and CDKN1B/p27kip1

(D) in the presence or absence of 25mM mannitol. N=2 donors assayed in duplicate.



Online Figure VI. miR-155 expression after transfection of CD34pos-PCs with pre-miR or

scramble (SCR). N=5 donors assayed in duplicate, ***p<0.01 vs. SCR.



Global Remodeling of the Vascular Stem Cell Niche in Bone Marrow of Diabetic Patients: Implication...

Documents

Transcript of Global Remodeling of the Vascular Stem Cell Niche in Bone Marrow of Diabetic Patients: Implication...