Reversal of hyperglycemia in diabetic mice by a marginal islet mass together with human blood outgrowth endothelial cells is independent of the delivery technique and blood clot-induced processes

Abstract

We recently reported that human blood outgrowth endothelial cells (BOEC) are supportive to reverse hyperglycemia in marginal islet mass-transplanted diabetic mice. In this report, we investigated whether the observed effect was evoked by islet packing in a blood clot prior to transplantation or could be mimicked by another method of islet/cell delivery. A marginal islet mass with or without BOEC was grafted underneath the kidney capsule of diabetic recipient mice via a (blood clot-independent) tubing system and compared with previous islet packing in a blood clot. The effect on metabolic outcome of both delivery techniques as well as the additive effect of BOEC was subsequently evaluated. Marginal islet mass transplantation via a tubing system required more islets per recipient than via a blood clot. Using the tubing method, transplantation of a marginal islet mass combined with 5 x 10⁵ BOEC resulted in reversal of hyperglycemia, improved glucose tolerance and increased kidney insulin content. The present study provides evidence that (1) previous packing in a blood clot results in more effective islet delivery compared with tubing; (2) BOEC exert a beneficial effect on marginal islet transplantation, independent of grafting technique and potential blood clot-induced processes. These data further support the use of BOEC in (pre-) clinical studies that aim to improve current islet transplantation protocols.

Introduction

Donor scarcity and early graft failure strongly compromise widespread use and long-term positive outcome of islet transplantation as a cure for brittle type 1 diabetes.¹ Development of abundant β-cell sources as well as alternative grafting sites and techniques to improve post-transplantation β-cell survival are therefore under thorough investigation.

In clinical islet transplantation, islets are infused through the portal vein. Micro-embolization of the islets in the portal circulation results in rapid activation of the complement and coagulation cascade with subsequent inflammation, collectively defined as instant blood-mediated inflammatory response (IBMIR).^2,3 IBMIR has been demonstrated as detrimental for islet function and identified, together with (sub-)acute ischemia and immune-rejection, as a major cause of early graft failure.^2,3

In a search to improve currently available islet transplantation protocols, we recently reported that human blood outgrowth endothelial cells (BOEC) promote β-cell survival, proliferation, and glycometabolic control when co-engrafted with a marginal islet graft in the renal subcapsular space.⁴ As the islet-cell mixture was entrapped in a blood clot prior to engraftment, thereby initiating IBMIR-like processes and since endothelial and endothelial colony forming cells have been demonstrated to (partially) prevent IBMIR,^5,6 it could not be excluded that the beneficial effect of BOEC was solely due to their differential effect on the deleterious effects of blood clot-evoked IBMIR.

In the current report, we compared islet delivery via a blood clot to the more commonly applied, blood clot-independent, tubing system⁷ and investigated whether BOEC also mediate a beneficial influence on marginal islet graft transplantation in the latter transplantation technique. By avoiding direct contact between the islets and recipient blood, the tubing technique limits graft damage through thrombotic reactions. Although islet transplantation using the blood clot technique has been described as early as 1986⁸ and used to prevent cell loss of small and loosely formed cell aggregates upon renal subcapsular transplantation,⁹ reports demonstrating the detrimental effects of coagulation processes on allo- and xenogeneic islet grafts^10,11 and inferior metabolic outcome¹² have rendered the technique obsolete, favoring tubing transplantation protocols. Its widespread use has made the tubing technique the current standard procedure for experimental subcapsular kidney islet transplantation.

As we previously reported a beneficial effect of BOEC on islet engraftment using the blood clot technique,⁴ we aimed to evaluate whether BOEC also favor the glycometabolic outcome of islet transplantation when co-engrafted via a tubing system. Demonstrating a positive effect of BOEC in the latter experimental condition further supports BOEC as valuable graft supporting cells for islet transplantation.

Results

Establishment of the marginal islet mass model in tubing-method transplantation

Our previous report demonstrated that 60 rat islets, entrapped in a blood clot and transplanted under the kidney capsule of diabetic NOD SCID mice were insufficient to normalize glycemia whereas co-engraftment with BOEC was able to do so.⁴ To evaluate whether BOEC also exert a beneficial effect on marginal islet transplantation without prior entrapment of the graft in a blood clot, the experimental setup of marginal islet mass transplantation was applied to transplantation via a tubing technique. Sixty, 75, or 150 islets were transplanted by inserting a polyethylene tubing with 0.58mm inner diameter (PE-50) underneath the kidney capsule of immune deficient mice with alloxan-induced diabetes. Follow up of glycemia for 28 d after transplantation (PTx) demonstrated that grafts of 60 (islet⁶⁰) or of 75 rat islets (islet⁷⁵) were not sufficient to restore normoglycemia in contrast to grafts of 150 rat islets (islet¹⁵⁰) (PTx day 28 glycemia: 25.89 ± 3.09 mmol/L, n = 4 islet⁶⁰ vs 25.37 ± 2.08 mmol/L, n = 4 islet⁷⁵ vs 5.31 ± 0.30 mmol/L, n = 3 islet¹⁵⁰; Fig. 1). The islet⁷⁵ condition was therefore adopted as the model of marginal islet graft transplantation via tubing. Interestingly, 75 islets transplanted in a blood clot were able to reverse diabetes while the same number of islets introduced via tubing was not (Fig. 2). This indicates that islet transplantation with prior entrapment in a blood clot is a more efficient delivery technique.

Figure 1. Transplantation via tubing of 75 rat islets does not reverse hyperglycemia of diabetic recipient mice and is therefore adopted as marginal islet mass model. One month follow up of 2 h fasting blood glucose levels of diabetic mice, transplanted with 150 (n = 4, squares), 75 (n = 4, open circles) or 60 (n = 4, closed circles) rat islets.

Figure 2. Transplantation of 75 rat islets via tubing does not reverse hyperglycemia of diabetic recipient mice in contrast to 75 islets transplanted in a blood clot. One month follow up of 2 h fasting blood glucose levels of diabetic mice transplanted with 75 rat islets via tubing (n = 4, open circles) or in a blood clot (n = 2, closed circles).

BOEC/marginal islet graft co-transplantation via tubing reduces hyperglycemia in diabetic mice

To investigate whether the positive influence of BOEC on the metabolic outcome of marginal islet mass-grafted diabetic mice depended on the presence of a blood clot, 75 rat islets, either alone (islet⁷⁵), or in combination with 5x10⁵ BOEC (islet^75+BOEC), were grafted in the renal subcapsular space of diabetic mice via PE-tubing. Two hours fasting glycemia was measured at several time points during the 28 d follow up period and revealed that the presence of BOEC decreased hyperglycemia (glycemia at day 28 PTx in islet⁷⁵ mice is 23.90 ± 2.79 mmol/L (n = 10) and in islet^75+BOEC mice 11.81 ± 4.71 mmol/L [n = 6]) (Fig. 3A). Statistically significant differences were reached from PTx day 14 on (** p < 0.01; * p < 0.05). At day 27 PTx, only 40% (4/10) of islet⁷⁵ mice showed an overnight fasting blood glucose level of < 10 mmol/L, while this was the case for 83% (5/6) of islet^75+BOEC mice. In this subset of normoglycemic mice, islet^75+BOEC animals still displayed improved glucose tolerance compared with islet⁷⁵ mice (glycemia 2 h after intra-peritoneal glucose bolus in islet⁷⁵ mice: 18.36 ± 5.27 mmol/L (n = 4) vs. islet^75+BOEC mice: 4.71 ± 1.87 mmol/L (n = 5), * p < 0.05) (Fig. 3B). Moreover islet^75+BOEC mice displayed a significantly higher kidney insulin content at the end of the 28 d follow up period, compared with islet⁷⁵ mice (439.60 ± 100.9 ng in islet^75+BOEC mice [n = 6] vs. 110.9 ± 17.95 ng in islet⁷⁵ mice [n = 8], p < 0.01) (Fig. 3C).

Figure 3. Co-transplantation of a marginal islet mass with BOEC via tubing improves the metabolic outcome of diabetic recipient mice. (A) 1 mo follow up of 2 h fasting blood glucose levels of diabetic mice transplanted with 75 rat islets via tubing without (n = 10, open circles) or with 5 x 10⁵ BOEC (n = 6, closed circles); (B) Intra-peritoneal glucose tolerance test performed at PTx day 27 in diabetic mice transplanted with 75 rat islets via tubing without (n = 4, open circles) or with 5 x 10⁵ BOEC (n = 5, closed circles) that displayed an overnight fasting blood glucose of < 10 mmol/L; (C) PTx day 28 kidney insulin content of diabetic mice transplanted with 75 islets via tubing without (n = 8, white bar) or with 5x10⁵ BOEC (n = 6, black bar). p < 0.05 (*), p < 0.01 (**).

These data therefore provide evidence that BOEC are beneficial to the function of a marginal islet graft, independent of potential influences on blood clot- related processes or of the grafting method. Of note, co-transplantation of BOEC with 75 rat islets via blood clot (islet^75+BOEC, in se sufficient to reverse diabetes; Fig. 2) did not induce hypoglycemia at any point during the follow up period of 28 d, thereby indicating that BOEC do not hyperactivate β cells in vivo (Fig. 4).

Figure 4. BOEC do not induce hypoglycemia when co-transplanted with an above-marginal islet mass in diabetic recipient mice. One month follow up of 2 h fasting blood glucose levels of diabetic mice transplanted via blood clot without (n = 2, open circles) or with 5 x 10⁵ BOEC (n = 2, closed circles).

Discussion

We previously demonstrated that transplantation of a marginal islet mass entrapped in a blood clot together with blood outgrowth endothelial cells (BOEC), reverses hyperglycemia in diabetic mice.⁴ Since the instant blood-mediated inflammatory response (IBMIR) is a known contributor to early graft failure in clinical islet transplantation^2,3 and endothelial cells can, at least partially, prevent blood coagulation processes,^5,6 we contemplated alteration of the blood clot-evoked coagulation cascade and its downstream phenomena as a potential mechanism behind the observed beneficial effect of BOEC on islet transplantation. While prevention of IBMIR by BOEC would per se be an important advantage in the context of intra-portal islet delivery, site-specific risks like liver hematoma and thrombosis¹⁵ combined with disadvantages like poor graft monitoring and accessibility currently inspire the search for alternative islet transplantation sites. In this context, BOEC should also act advantageous in extravascular islet transplantation protocols devoid of immediate coagulation processes. Therefore, islets and BOEC were co-delivered underneath the kidney capsule of diabetic recipient mice using a polyethylene tubing instead of a blood clot. Follow up of blood glucose and determination of kidney insulin content demonstrated that co-transplantation of BOEC with a marginal islet mass via tubing also proves beneficial for the glycometabolic outcome of diabetic recipient mice. Although we cannot exclude that the observed improvement in blood glucose control and increased insulin content is due to an increased insulin content per cell, our previous report⁴ illustrated that BOEC co-engraftment resulted in an increased β-cell volume via a reduction in β-cell death and stimulation of β-cell proliferation. While others have reported that endothelial (progenitor) cells of different origins can improve survival and function of transplanted islets,^16-18 our study is, to our knowledge, the first to illustrate the beneficial effect on islet transplantation of adult, human, peripheral blood derived endothelial cells. As we were previously able to mimic the observed beneficial glycometabolic effect with BOEC derived from a type 1 diabetic individual,⁴ we further confirmed the autologous transplantation potential of BOEC. Moreover, while some research groups reported an effect of endothelial (progenitor) cells on rapid graft revascularization,¹⁷ BOEC co-engraftment did not affect islet vessel density but rather resulted in an increased graft-vessel and β-cell volume at day 28 PTx. Although we cannot exclude that BOEC, like other primary human endothelial cells,¹⁶ do have a potential benefit on IBMIR-related islet damage, our previous and current report illustrates that BOEC provide other, presumably trophic, support to promote the outcome of islet transplantation and subsequently reverse hyperglycemia.

Moreover, the current report demonstrates that BOEC do not mediate uncontrolled insulin release from the co-transplanted islets as hypoglycemia was never observed during the entire follow-up period, even when BOEC were co-transplanted with an islet mass that was per se sufficient to restore normoglycemia in diabetic recipient animals.

Finally, while others reported a detrimental effect of the use of a blood clot on experimental islet transplantation,^10,11 we found that the islet number needed to define the marginal islet mass via tubing was higher than that via a blood clot. This indicates that, at least in our hands, blood clot transplantation results in a more efficient islet delivery and/or an improved islet survival/function upon engraftment. Indeed, in our experience, transplanting islets with tubing is inherently susceptible to islet loss. Frequently, islets are flushed out due to the back-pressure buildup underneath the kidney capsule upon infusion. While these recipient animals should be disqualified from data collection, we do not exclude that some degree of islet loss might occur unnoticed. In this case, the absolute number of effectively engrafted islets sub-numbers the original number of islets and might thereby be insufficient to normalize glycemia. While others have applied the blood clot technique for transplantation of substantially higher numbers of islets, using a significantly higher blood volume, thereby likely evoking a more pronounced coagulation and inflammatory reaction and negatively affecting islet survival and function,¹² we propose to not abandon the blood clot technique in the setting of marginal islet mass transplantation but rather to use the smallest blood clot volume possible. This approach should on the one hand promote a more efficient islet/cell delivery without significantly biasing the outcome upon engraftment.

In summary, the current report provides evidence that (1) previous packing into a blood clot results in a superior islet/cell delivery compared with the commonly applied tubing technique and (2) BOEC exert a beneficial influence on the glycometabolic outcome of marginal islet mass transplantation in diabetic recipient mice, independently of the applied grafting technique and by mechanisms that are at least partially different than inhibition of IBMIR-like processes. These data thereby further nourish studies that aim to explore the effect of BOEC on islet transplantation in (pre-)clinical models and/or alternative (extra-portal) delivery sites.

Methods

Animals

Nine week-old, male NOD SCID mice were obtained from Charles River Laboratories (L'Arbresle) and housed alone or in pairs. The ‘Principles of laboratory animal care’ (NIH publication no. 85–23, revised 1985) as well as specific national laws were followed. In addition, the institutional “Ethical Committee for Animal Use” of the Vrije Universiteit Brussel (VUB) approved the experimental plan. For induction of hyperglycemia, the mice were intravenously (iv) injected with alloxan monohydrate (ALX) (Sigma-Aldrich) at a concentration of 50 mg/kg bodyweight (BW) and only used for experiments when 2 h fasting blood glucose levels, 2 d after ALX injection, reached at least 19.5 mmol/L.

Islet isolation and culture

Islets were isolated as previously described¹³ from 6 to 10 week-old, male Wistar rats (Janvier), weighing 250‒300 g. After isolation, islets were cultured during 2 d in Ham’s F10, supplemented with 2% fetal calf serum (FCS) and 0.5% bovine serum albumin (BSA). One hour before transplantation, islets were washed twice, re-suspended in PBS and handpicked randomly. As we previously described,⁴ only medium-sized islets (150–250 μm) were used for transplantation after being hand-picked by an independent investigator who was unaware of the experimental conditions.

Cell derivation and culture

After giving informed consent, BOEC were isolated from the peripheral blood of a 31-y old male healthy volunteer (isolation was approved by the Ethical Committee of the University Hospital, Brussels), cultured as described previously¹⁴ and frozen in liquid nitrogen when they reached passage 6. One hour before transplantation, cells were thawed and re-suspended in PBS. For each recipient mouse, 5x10⁵ BOEC were mixed with islets. The islets or islets-BOEC mixture were centrifuged and washed once in 0.2 ml transplantation medium (BioWhittaker), supplemented with 2% male NOD SCID serum (Charles River Laboratories).

Transplantation and tissue harvest

Mice were sedated using 10 μl/g BW of 10 mg/ml ketamine + 0.1% xylazine, dissolved in physiological saline (0.9% NaCl). Blood clot method: 6 μl of tail tip blood from the recipient mouse was added to the islet or islet-BOEC pellet, stirred gently and left to clot during 4 min. The blood clot-containing graft was then placed as a whole under the left kidney capsule. Tubing method: a volume of 50 μl serum-supplemented transplantation medium containing the islets or islets-BOEC mixture was aspirated into a polyethylene tubing (PE-50, BD N.V., Erembodegem), connected to a micrometer pump. The graft-containing tubing was then centrifuged manually until the entire graft was collected at the bottom of the tubing. Subsequently, the tubing was inserted underneath the kidney capsule, and the graft was gently flushed out using the micrometer pump. Tail vein glycemia was measured between 10 and 12 AM, following 2 h of fasting. Intraperitoneal (i.p.) glucose tolerance tests were performed by injecting glucose (2 g/kg BW) after an overnight fast. For plasma C-peptide measurement, 0.250 ml blood was collected from the tail vein in tubes containing Na-EDTA (2.2 mg/ml blood) and aprotinin (0.14 mg/ml blood). Animals were killed by cervical dislocation. The transplanted kidney was harvested, snap-frozen in liquid nitrogen and stored at -70°C until processing. After thawing, kidneys were homogenized in 1 ml Azal extraction buffer (Bovine Serum Albumin [2.5 g] in a mixture of glacial acetic acid [113 ml] and water [887 ml]) with the mixer mill for 2 times 2 min at 30 Hz. Another 2 ml Azal was added to the mixture before spinning it down for 25 min at 12.000 RPM, 4 °C. The supernatant was complemented with Azal to a final volume of 3ml. The insulin content was determined by radio-immuno assay (RIA) using the rodent-insulin specific kit, according to manufacturer’s instructions (Linco Research).

Glossary

Abbreviations:

BOEC

human blood outgrowth endothelial cells

IBMIR

instant blood-mediated inflammatory response

NOD SCID

non-obese diabetic severe combined immunodeficient

iv

intravenous

ALX

alloxan monohydrate

BW

bodyweight

FCS

fetal calf serum

BSA

bovine serum albumin

PBS

phosphate-buffered saline

NaCl

sodium chloride

PE-50

polyethylene tubing

Na-EDTA

sodium-ethylenediaminetetraacetic acid

RIA

radio-immuno assay

PTx

days after transplantation

islet^x

graft consisting of X islets

mmol/L

millimolar per liter

islet^X+BOEC

graft consisting of X islets supplemented with 5x10⁵ BOEC

10.4161/isl.26778