New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy

ARTICLE IN PRESS

JID: ACTBIO

[m5G;November 5, 2019;1:18]

Acta Biomaterialia xxx (xxxx) xxx

Contents lists available at ScienceDirect

Acta Biomaterialia journal homepage: www.elsevier.com/locate/actbio

Full length article

New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy D. Cristea a,1, S. Krishtul b,1, P. Kuppusamy c, L. Baruch b, M. Machluf b, A. Blank a,∗ a

Schulich Faculty of Chemistry, Technion – Israel Institute of Technology, Haifa 32000, Israel Faculty of Biotechnology and Food Engineering, Technion – Israel Institute of Technology, Haifa, Israel c Departments of Radiology and Medicine, Geisel School of Medicine, Dartmouth College, Lebanon, NH, USA b

a r t i c l e

i n f o

Article history: Received 16 July 2019 Revised 6 October 2019 Accepted 23 October 2019 Available online xxx Keywords: Oxygen diffusion Cell microencapsulation Electron spin resonance imaging Extracellular matrix

a b s t r a c t Cell microencapsulation within biocompatible polymers is an established technology for immobilizing living cells that secrete therapeutic products. These can be transplanted into a desired site in the body for the controlled and continuous delivery of the therapeutic molecules. One of the most important properties of the material that makes up the microcapsule is its oxygen penetrability, which is critical for the cells’ survival. Oxygen reaches the cells inside the microcapsules via a diffusion process. The diffusion coefficient for the microcapsules’ gel material is commonly measured using bulk techniques, where the gel in a chamber is first flushed with nitrogen and the subsequent rate of oxygen diffusion back into it is measured by an oxygen electrode placed in the chamber. This technique does not address possible heterogeneities between microcapsules, and also cannot reveal O2 heterogeneity inside the microcapsule resulting from the living cells’ activity. Here we develop and demonstrate a proof of principle for a new approach to measuring and imaging the partial pressure of oxygen (pO2 ) inside a single microcapsule by means of high-resolution and high-sensitivity electron spin resonance (ESR). The proposed methodology makes use of biocompatible paramagnetic microparticulates intercalated inside the microcapsule during its preparation. The new ESR approach was used to measure the O2 diffusion properties of two types of gel materials (alginate and extracellular matrix – ECM), as well as to map a 3D image of the oxygen inside single microcapsules with living cells. Statement of Significance The technology of cell microencapsulation offers major advantages in the sustained delivery of therapeutic agents used for the treatment of various diseases ranging from diabetes to cancer. Despite the great advances made in this field, it still faces substantial challenges, preventing it from reaching the clinical practice. One of the primary challenges in developing cell microencapsulation systems is providing the cells with adequate supply of oxygen in the long term. Nevertheless, there is still no methodology good enough for measuring O2 distribution inside the microcapsule with sufficient accuracy and spatial resolution without affecting the microcapsule and/or the cells’ activity in it. In the present work, we introduce a novel magnetic resonance technique to address O2 availability within cell-entrapping microcapsules. For the first time O2 distribution can be accurately measured and imaged within a single microcapsule. This new technique may be an efficient tool in the development of more optimal microencapsulation systems in the future, thus bringing this promising field closer to clinical application. © 2019 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction Cell therapy has long been suggested as a promising solution for numerous pathological conditions and diseases that require the

∗

1

Corresponding author. E-mail address: [email protected] (A. Blank). These authors equally contributed to the work.

continuous and regulated delivery of a therapeutic factor. However, non-autologous transplantation (in which the cells are not sourced from the host itself) might result in immune rejection by the recipient, and consequently, in graft failure. Cell microencapsulation technology was developed primarily to overcome this barrier [1,2]. In this technology, cells that secrete therapeutic products are immobilized within the confines of a semipermeable membrane and transplanted to the desired site for the controlled and continuous delivery of the therapeutic molecule. The immunobarrier permits

https://doi.org/10.1016/j.actbio.2019.10.032 1742-7061/© 2019 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

JID: ACTBIO 2

ARTICLE IN PRESS

[m5G;November 5, 2019;1:18]

D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

inward diffusion of nutrients and oxygen and secretion of waste and the therapeutic product, but prevents access by the host’s immune system components. This technology has been vastly investigated for the past three decades as a treatment for a wide variety of diseases and dysfunctions such as diabetes, Alzheimer’s disease, liver failure, cancer, anemia, and others [3]. Cell microencapsulation is most commonly achieved using biocompatible polymers such as alginate. However, research published in recent years reveals the advantages of employing natural complex biomaterials such as a decellularized extracellular matrix (ECM) or isolated ECM proteins, which also endow the encapsulated cells with functional biological support [4–6]. One of the most important properties of the material comprising the microcapsule is its oxygen penetrability. This is because oxygen availability is considered to be a critical factor that limits cell microencapsulation technology [2,7]. Oxygen reaches the cells inside the microcapsules via a diffusion process. The diffusion coefficient for the microcapsules’ gel material is commonly measured using bulk techniques, where the gel in a chamber is first flushed with nitrogen and the subsequent rate of oxygen diffusion back into it is measured by an oxygen electrode placed in the chamber [8]. Similar techniques can be used also for chambers filled with actual microcapsules rather than just the gel material; in these cases, diffusion into microcapsules is monitored by measuring the changes in concentration in the surrounding liquid [9]. Such bulk techniques, however, cannot provide information regarding potential heterogeneity among microcapsules, and cannot provide data about possible oxygen gradients generated in the microcapsules due to limited diffusion or cell metabolism. To address potential bulk heterogeneity issues and possibly even provide high-resolution spatially-resolved data on the partial pressure of oxygen (pO2 ) in individual microcapsules, several methods can be implemented. These include (a) electrochemical methods based primarily on Clark-type oxygen electrodes [10,11], which have long been the main technique for measuring pO2 in liquid and gas phases in general; and (b) fluorescence and fluorescence quenching techniques [12–14], which have also been applied to the measurement of individual microcapsules [15,16]. This type of technique is based on the ability of oxygen to quench the optically excited state of luminescent oxygen-sensitive molecular or nanometric probes. While in principle such optical techniques render themselves useful also to imaging, we have failed to locate such an example in the literature, which only contains theoretical models for the 3D distribution of O2 in the microcapsule [17,18] that stress the need for in vitro data for their validation. Here we propose a method for mapping oxygen distribution in microcapsules that is based on an electron spin resonance (ESR) technique. This technique, called ESR oximetry, requires the incorporation of a suitable stable paramagnetic spin probe into the desirable sample. The spin-lattice and spin-spin relaxation times (T1 and T2 , respectively) of these spin probes are highly oxygensensitive. This is due to the interaction between the paramagnetic molecular oxygen and the incorporated spin probe. As O2 is paramagnetic, its interaction with the spin probe results in a spin exchange that leads to shorter relaxation times and a broadening of the linewidth of the spin probe’s ESR spectrum [19,20]. It should be noted that the ESR technique has been previously employed extensively in pO2 measurements and imaging in biological contexts. For example, it has been used to measure oxygen consumption and oxygenation in cells encapsulated in different biocompatible matrices, both in-vitro and in-vivo [21–24]. As for the exogenous spin probe, there are many possibilities ranging from soluble stable free radicals [25], which can or cannot penetrate the cell’s membrane [26], to solid paramagnetic microcrystals [27]. In the case of soluble charged spin probes, penetration of lipid-rich membranes is minimal unless there is a car-

rier for them; on the other hand, small neutral soluble spin probes can penetrate the membrane much more easily (but need to be re-charged inside the cell in order to prevent them from escaping back out [27]). In the context of microencapsulated cells, soluble spin probes can be placed in the medium solution and then made to homogenously fill the microcapsule by a diffusion process. However, such spin probes may be oxidized or metabolized quite fast, within a day or two at the most. In contrast, solid paramagnetic particulates are harder to incorporate homogeneously into the microcapsule but are very inert, have “universal” calibration curves, and can provide readings many weeks after their encapsulation [28,29]. The main limitation of ESR as a whole is its relatively low sensitivity compared to optical techniques, meaning that it usually requires a large number of paramagnetic molecules (spins) to obtain a measurable signal. As a consequence, common ESR-based imaging methodologies suffer from low imaging resolution (when compared to optical microscopy-based methods). The ESR technology’s sensitivity limitation seems to render it difficult to apply to small samples, such as microencapsulated cells, which are relevant to this work. Nevertheless, recently we have made efforts to alleviate this ESR handicap by means of a unique methodology we developed for high-sensitivity high-resolution pulsed ESR microimaging [30]. This approach was implemented in our study of oxygen concentration distribution near cyanobacteria cells, in which we incorporated a soluble trityl free radical spin probe into the solution media surrounding the cells [31]. In a later project, we took another step along this path and made use of our advanced imaging capability, in conjunction with a unique solid particulate spin probe distributed inside submillimeter-size cancer spheroids, to obtain high-resolution 3D pO2 mapping inside these tissuemodel microsystems [32]. This work was closely followed by another team that provided similar measurements showing hypoxia in cell spheroids using ESR spectroscopy (without imaging) in a bulk of many spheroids [33]. Here, we further the application of this methodology using ESR microimaging to study oxygen diffusion properties in alginate- and ECM-based microcapsules that entrap living cells. We first use subsecond time-resolved ESR to measure the diffusion coefficient of O2 in/out of a single empty microcapsule, without any living cells, thereby learning about the basic diffusion properties of these 3D gel structures. Subsequently, we employ 3D ESR microimaging to learn about oxygen distribution in single microcapsules holding living cells inside them. In this way, we provide a proof of principle for the use of this methodology to measure spatial oxygen availability within the microcapsule. 2. Materials and methods 2.1. Cell culture Bone marrow-derived human mesenchymal stem cells (MSCs) were purchased from Lonza AG (Basel, Switzerland). The cells were cultured in α -MEM (Sigma-Aldrich, St. Louis, MO, USA), supplemented with 10% fetal bovine serum (FBS), 1% Pen-Strep solution, 0.4% Fungizone® (Biological Industries, Israel), and 5 ng/ml basic fibroblast growth factor (bFGF, Peprotech, Rehovot, Israel). Cells were maintained at 37 °C in a humidified incubator with 5% CO2 . 2.2. ECM preparation Porcine pancreatic ECM was decellularized and liquefied as described in a previous paper [4]. Briefly, pancreases removed from healthy commercial pigs were decellularized employing alternating hyper-/hypotonic NaCl solutions, trypsin-EDTA treatment, and Triton X-100 washing. The decellularized ECM was then lyophilized,

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

JID: ACTBIO

ARTICLE IN PRESS

[m5G;November 5, 2019;1:18]

D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

3

crushed into a particulate form, and subsequently solubilized through pepsin digestion. 2.3. Spin probe In this work we made use of a spin probe known as LiNc-BuO [34]. It comes in solid form and is not soluble in aqueous media. The cytotoxicity and lack of immunogenic response of the LiNcBuO radical in solid microparticulate form was verified in previous in-vitro and in-vivo work [35]. As noted above, ESR can provide information about the spin-spin relaxation time (T2 ) of the paramagnetic states, which is inversely proportional to the spectral linewidth of the signal (LW∼1/π T2 ). This capability and property can be used to accurately quantify oxygen concentration in the vicinity of the paramagnetic probe because oxygen is known to have a dramatic effect on the LW of paramagnetic samples. Fig. S1 shows how the signal of the LiNc-BuO spin probe is broadened by the presence of oxygen. The calibration of linewidth vs. pO2 is the same, regardless of the radical concentration or the solution’s properties, given a specific batch of LiNc-BuO [36]. The assynthesized radical particulates undergo a few cycles of sonication in distilled water. The resulting particulates are smaller than 2 μm and are then entrapped inside the microcapsules during the latter’s preparation procedure (see below). 2.4. Microcapsule production Alginate- and ECM-based microcapsules were produced as we described in a previous paper [4], albeit with some modifications. Prior to encapsulation, the radical particulates were immersed in a serum-free medium and sonicated at 40% amplitude for 30 s using a VibraCell VCX750 sonicator (Sonics & Materials Inc., Newtown, CT, USA). The particulates (1 mM) were then mixed with sodium alginate (1.6%, PronovaTM UP-MVG, NovaMatrix, Norway) for alginate-based microcapsules or with sodium alginate (1.6%) and solubilized pancreatic ECM (2.6 mg/ml) for ECM-based microcapsules, and sonicated once more for 2 min in a 40-kHz bath sonicator (UC-02 Ultrasonic Cleaner, Lab Companion, Seoul, Korea). All sonication procedures were carried out in ice to prevent the sample from overheating. MSCs were then added to the alginate-ECMLiNc-BuO mixture at a ratio of 1.5 × 106 cells/ml, and the resulting suspension was sprayed into a calcium chloride crosslinking solution (1.52% (w/v) CaCl2 , 0.31% (w/v) HEPES) using an electrostatic bead generator. Following a 20-min incubation period, the microcapsules were coated with poly-L-lysine (PLL, 28.2 kDa, 0.06% (w/v) in 0.84% (w/v) NaCl) for 10 min. For the ECM-based microcapsules, additional steps of ECM gelation and alginate removal were applied. Thus, these microcapsules were incubated at 37 °C for 30 min to enable the self-assembly of the 3D fibrous network within the capsules, and then alginate was removed through suspension in a sodium citrate solution (1.5% (w/v), in 0.31% (w/v) HEPES). All the resulting cell-laden microcapsules were cultured as described for MSCs until further use. The final microcapsule diameter was between 500 and 600 μm with accordingly ∼100–170 cells per microcapsule. Fig. 1 shows typical microcapsules with the cells and the encapsulated particulates homogenously distributed inside them. 2.5. Encapsulated cell viability The viability of encapsulated MSCs was evaluated through staining with fluorescein diacetate (FDA, 10 μg/ml, Sigma) and propidium iodide (PI, 2 μg/ml, Sigma) for 10 min in the dark, followed by 3 washes with PBS (pH 7.4). The stained cells were visualized under a fluorescence microscope (Eclipse TE20 0 0-E, Nikon, Netherlands).

Fig. 1. Microcapsules with the entrapped LiNc-BuO particulates. Scale bar: 200 μm.

2.6. Sample preparation for ESR measurements For the purpose of performing ESR measurements, a single microcapsule is inserted into a custom-designed sample holder (Fig. 2) that fits our miniature ESR resonator and imaging probe system. The sample holder is prepared from two types of materials: PEEK and Rexolite®. The former is used to create a sealed environment where O2 and water from the medium cannot diffuse in or out, while the latter is useful to enable O2 diffusion through the sample holder while maintaining the inside compartment wet and avoiding dehydration. Once the microcapsule is inserted into the sample holder with its associated medium, it is covered in a thin layer of paraffin oil to avoid dehydration. The sample holder can then be sealed with a specially designed glass cover made using lithography techniques [37]. The seal is adhered to the sample holder using UV glue (NOA 61 by Norland, USA). The sample holder can be sealed completely or partially (leaving it open to air but minimizing dehydration), depending on the nature of the experiments to be performed. Complete sealing can also be achieved inside a glove box with controlled O2 content (e.g., 4% O2 , 96% N2 in some of the experiments). 2.7. ESR measurements ESR measurements were used to quantify the partial pressure of oxygen in the microcapsules. ESR experiments were carried out using a homemade pulsed ESR microimaging system operating at X-band frequency (∼9.5 GHz). Full details of this imaging system are provided in a previous publication [31]. In this work we performed two types of measurements: a. Integral ESR measurements employing single-pulse free induction decay (FID) to measure the signal from the entire sample (pulse length of 30 ns); and b. Imaging experiments in which the ESR signal is obtained in a spatiallyresolved manner. This means that in order to obtain an image, the system used a combination of microwave pulses and pulsed magnetic field gradients in all 3 axes (x, y, and z). A single-pulse FID signal was measured with the pulsed magnetic field gradients (lasting about 200 ns) induced between the microwave pulse and the signal acquisition period. The pulsed gradient magnitude reached a peak value of ∼10.5 T/m, corresponding to a spatial resolution of ∼28 μm [38]. Gradients were induced over the sample using miniature coils located near it (schematically shown in Fig. 2(c)). The gradients spatially encoded the position of the radi-

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

JID: ACTBIO 4

ARTICLE IN PRESS

[m5G;November 5, 2019;1:18]

D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

Fig. 2. Sample holder for the microcapsule. (a) General 3D view of the sample holder with the microcapsule inside it. (b) Cross section of the sample holder with the microcapsule. (c) Sample holder placed inside the dielectric ESR resonator and surrounded by the imaging gradient coils (shown in a schematic manner – in practice there are more gradients coils to facilitate full 3D imaging).

2.9. Statistical analysis The software package MatlabTM ver 2019a was applied to all statistical analyses and plots. The mean, standard deviation, and Student’s t-test for confidence levels were employed in assessing oxygen diffusion times in/out of the microcapsules. 3. Results 3.1. Measuring oxygen diffusion into blank microcapsules

Fig. 3. Model of the sample holder with a single microcapsule used in the finite element analysis (COMSOL software). Three domains can be identified: (1) air (or a controlled atmosphere with a specific oxygen content); (2) microcapsule that can optionally contain living cells (manifested through the O2 consumption rate defined for this domain); and (3) cell cultivation medium solution.

cal microparticulates in the phase of their ESR signal (see details in [31,32]). After proper processing, the phase of the signal recorded for various values of gradient strengths in x, y, and z made it possible to obtain a full 3D image of the radical particulates, which provided both the concentration of the particulates in space and their spatially-resolved linewidth. This linewidth information was then used to extract the local pO2 using the calibration curve shown in Fig. S1. During the experiments, we controlled the temperature and moisture around the sample holder by flowing air or a specific gas mixture in its immediate vicinity.

2.8. Simulation of oxygen diffusion and consumption Throughout the paper we compare our experimental results and analyze the experimental data with the aid of a finite element simulation tool (COMSOL Multiphysics®, COMSOL AB, Stockholm, Sweden) to account for oxygen diffusion and oxygen consumption rates in our sample. For this purpose, we use the computational sample model shown in Fig. 3. The parameters used in the simulation are provided in Table 1. The boundary conditions enforced varied according to the specific experiment being carried out (i.e., sample holder open or closed).

This section focuses on quantifying the diffusion properties of two different microcapsule types: the commonly used alginate-PLL microcapsules and the bioactive ECM-based microcapsules, which we have previously introduced [4]. The diffusion of oxygen in/out of the microcapsule was measured in the following way: blank microcapsules (with no cells) were produced using the procedure described above. They contained small (< 2 μm in size) particulates of the LiNc-BuO radical. The microcapsule in the sample holder was measured in different ambient conditions and atmospheric contents. Accordingly, in order to quantify the diffusion of oxygen into the microcapsule as compared to oxygen diffusion in regular aqueous media, we carried out the following test protocol: (a) We flushed the sample holder with moisture-saturated Ar gas for a period of ∼30 min, until the Ar diffused into all parts of the sample chamber. This was verified by constantly monitoring the ESR signal’s magnitude (which is inversely proportional to the linewidth) and waiting until it reached its maximum amplitude. Fig. 4 shows a typical kinetics graph of the ESR signal for Ar gas flowing into the microcapsule. (b) We flushed the sample holder with moisture-saturated regular airflow and monitored the kinetics of the decrease in the ESR signal as O2 diffused back into the microcapsule. Fig. 5 shows an example of such time-dependent ESR signal. This kinetics is much faster than that of Ar flushing and is primarily influenced by the diffusion properties of O2 in the cell medium and microcapsule media. The same experiment was repeated for several microcapsules of each type. The results were compared to measurements of radical particulates placed at the bottom of the sample holder without any aqueous media, and to samples of radical particulates placed at the bottom of the sample holder and covered with the same medium and oil combination that covers the microcapsule. Table 2 summarizes these results:

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

ARTICLE IN PRESS

JID: ACTBIO

[m5G;November 5, 2019;1:18]

D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

5

Table 1 Parameters used in the COMSOL oxygen diffusion simulation. The values of [O2 _Air], D_O2 _Air, D_O2 _Med, and α are taken from [39]. The nominal O2 consumption rate and the T_death value were found by fitting the experimental data to simulation results (see Fig. 9 below). The value of O2 _Cons_Max appears in ref. [39] as the maximum O2 consumption rate found for living cells in general. Parameter

Symbol

Value

Units

O2 concentration in air O2 diffusion coefficient in air O2 diffusion coefficient in medium Nominal O2 consumption rate Maximum O2 consumption rate Solubility coefficient Exponential death time of cells

[O2 _Air] D_O2 _Air D_O2 _Med O2 _Cons_Nom O2 _Cons_Max

9.375 7.6 × 10−5 2.69 × 10−9 15 100 0.024 35,000

mol/m3 m2 /s m2 /s amol/cell/s amol/cell/s Unitless s

α

T_death

Fig. 4. Typical kinetic curve showing the increase in ESR signal as Ar is flowed into the sample chamber. The time Tout represents the characteristic sigmoidal function rise time fitted to the experimental data ESR_Signal = A/(1 + e−t/Tout ) + B. The ESR signal is obtained by a 1-pulse FID sequence with the following parameters: microwave π /2 pulse of 30 ns and acquisitions time of 1 μs, starting 80 ns after the end of the microwave pulse.

Fig. 5. Typical kinetic curve showing the decrease in ESR signal as air (with O2 ) flows into the sample chamber that was flushed previously with Ar. The time Tin represents the characteristic exponential fall time fitted to the experimental data ESR_Signal = Ae−t/Tin + B. Table 2 Comparison of O2 Tin (time of O2 to diffuse into the sample holder) for various sample preparations. Sample type Particulates Particulates Particulates Particulates

without any medium in medium + oil layer in alginate microcapsule + medium + oil layer in ECM microcapsule + medium + oil layer

Tin [s]

Standard deviation (STD) [s]

95% confidence interval [s]

3 111.7 42.9 45.3

1 29 14 11

– 111.7 ± 22.4 42.9 ± 8.0 45.3 ± 6.7

Number of microcapsules measured (N) 2 10 15 13

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

JID: ACTBIO 6

ARTICLE IN PRESS

[m5G;November 5, 2019;1:18]

D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

Fig. 6. (Left) Results of finite element simulation of O2 diffusion into the sample chamber with aqueous medium in it. This is a single snapshot of the time-domain simulation taken 80 s after its initiation (see also Supplementary Movie 1). The oxygen concentration is shown to decrease from top to bottom, as expected, and as time passes it reaches the steady-state concentration of the partial pressure of normal air all the way down at the bottom of the sample holder. (Right) Calculated ESR signal as a function of time for different positions of the LiNc-BuO particulates with respect to the upper surface of the cup sample holder. The calculation uses the time-domain simulation results as well as the known variations of the ESR signal of the particulates with varying pO2 .

The results clearly show that O2 diffuses easily into the microcapsule matrix and that there is no apparent difference between alginate and ECM materials. Moreover, at first glance, the diffusion properties seem to be even better for the microcapsules’ materials than in medium-only cases. However, this is due to the fact that the particulates in microcapsule samples are closer to the upper surface of the sample holder (near the medium/gas boundary) than in the case of the medium-only samples (where the particulates are at the bottom of the sample holder). This is also verified by a diffusion simulation executed using our sample chamber geometry and the known diffusion properties of O2 in air and in water medium. Fig. 6 and Supplementary Movie 1 show the results of this simulation (prepared using the COMSOL Multiphysics® package). The result obtained by calculating and summing up the values of the ESR signal vs. time for several depths inside the sample holder is Tin ∼38 s, which is very similar to the result we observed in our measurements with the microcapsules. The simulation also shows that as particulates are placed more deeply inside the sample holder, Tin increases significantly. 3.2. Measuring and mapping pO2 in microcapsules containing living cells While oxygen diffusion through gel media is an important parameter for the proper growth and survival of cells inside the polymeric matrix, it is not the only factor affecting this result. The living cells themselves can also influence the quantity of oxygen found in the microcapsules, both because they consume oxygen directly and because, in a large enough concentration, they can affect the diffusion of oxygen through the microcapsule. This may result in oxygen gradients that could limit the availability of oxygen for cells located in the inner parts of the microcapsule (resulting in a hypoxic core) [7,40]. Thus, measuring the oxygen concentration from within the microcapsule that holds intact living cells in it, and ideally mapping it in 3D, are actions relevant to the subject matter of this paper. To that end, microcapsules with living cells and LiNc-BuO microparticulates were prepared according to the procedure described above. Our primary concern was to assess whether the LiNc-BuO microparticulates harm the co-encapsulated cells. Therefore, cell viability was addressed with and without the

LiNc-BuO microparticulates 8 weeks into encapsulation. Typical results provided in Fig. 7 indicate that the effect of the LiNc-BuO particulates on the cells’ viability is negligible (Fig. 7(a) and (b)). Furthermore, no change in the cell phenotype could be observed when staining the cells for their typical markers CD29 (Fig. 7(c) and (d)) and CD90 (Fig. 7(e) and (f)). As noted in Section 2.4, the typical cell concentration in the microcapsules is 1.5 × 106 cells/ml, meaning that a typical microcapsule with a diameter of 600 μm contains ∼170 cells at most. Previous studies have reported that the oxygen consumption rate for such MSCs has rather large variations in the range of ∼25– 100 amol/cell s [41,42]. When using these numbers in our oxygen diffusion simulation, we find that the steady-state pO2 gradients expected when the sample is open to the atmosphere, even at the maximum reported O2 consumption rate, are very small (Fig. 8). Thus, we cannot possibly expect to measure any respiration effects in our sample holder for a single microcapsule when it is open to air. It is thus clear that in order to measure significant variations from the normal atmospheric pO2 leading to the consumption rate characteristic of our samples, as well as to verify the integrity of our cells during the measurement procedure, we would have to seal the sample. The typical result of such an experiment that refers to pO2 as a function of time (based on the linewidth of the radical signal, measured over a 24-h period for a sealed sample), is shown in Fig. 9. The slow but steady change in O2 concentration due to cellular activity is clearly measurable. Superimposed over the experimental data is the simulated graph, with two fitting parameters - the nominal O2 computation rate and the exponential decay time of the cell’s population, (both fitted parameters are provided in Table 1 above). It is evident that the cells are viable and consume a nominal amount of oxygen for at least 8 h, even though they reside in a sealed miniature chamber. At longer times, O2 consumption slowly levels off due to the death of the cells (simulated by an exponential decrease in cell population using “T_death”, see Table 1, due to the measurement conditions). The small discrepancies between the measured and simulated plots can be attributed to the crudeness of our cell survival model, whose population is characterized by a simple exponential decrease in time. Nevertheless, the model is good enough to capture the essence of cell viability over time.

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

JID: ACTBIO

ARTICLE IN PRESS

[m5G;November 5, 2019;1:18]

D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

7

Fig. 7. Typical fluorescein. diacetate (FDA)/ propidium iodide (PI) viability analyses performed 8 weeks following encapsulation with (a) or without (b) 1 mM (∼1.3 mg/ml) LiNc-BuO particulates. Green – living cells, red – dead cells. Typical MSC markers immunostaining of MSC encapsulated with (c, e) or without (d, f) 1 mM LiNc-BuO particulates 8 weeks following encapsulation. Blue-nuclei (DAPI), Green – CD29 (c, d) and CD90 (e, f). Scale bars: 50 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 8. (a) Calculated steady-state oxygen profile along a line running from top to bottom at the center of the sample holder (as marked on the right by a light blue line), for the maximum possible O2 consumption rate of the cells in the microcapsule. (b) The results of the same calculation shown for an entire 2D cross section of the sample holder, with pO2 color scale ranging from 0 to 160 mmHg (i.e., atmospheric O2 levels). It is clear that in this scale the gradients are negligible. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

In order to provide further support for these findings and to test our capability to acquire spatially-resolved pO2 data, we carried out high-resolution ESR microimaging experiments to map out the pO2 inside the sample holder in 3D. These measurements can provide the 3D spatially resolved spectral amplitude and shape linewidth of the radical inside the sample and translate them into a 3D signal amplitude and pO2 maps. Imaging experiments were conducted with the solid LiNc-BuO microparticulates mentioned above, which were co-encapsulated with the cells. These microparticulates remain inside the microcapsule and do not exit to the outside medium. In addition, they offer relatively large variations in linewidth according to the changes in pO2 (Fig. S1). This is good

for accurate pO2 monitoring, but also causes the linewidth to be too broad near a normal atmospheric pO2 , which makes the signal drop a bit fast in our pulsed ESR imaging experiments. Fig. 10 shows typical imaging results measured for a microcapsule co-entrapping living cells and the LiNc-BuO radical. The imaging acquisition time was 45 min. In this case, the microcapsule was sealed in a glove box with an atmosphere of 4%O2 /96%N2 (corresponding to a pO2 level of 30 mmHg). The activity of the cells during the measurements was verified by monitoring via ESR the slow overall reduction in pO2 in the entire sample holder, as well as by inspecting the cells immediately after the measurement using the fluorescein diacetate test. As expected, even though the

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

JID: ACTBIO 8

ARTICLE IN PRESS

[m5G;November 5, 2019;1:18]

D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

Fig. 9. Oxygen concentration as a function of time for a sealed sample with a single microcapsule. (Red) Measured data; (blue) simulated data. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article)

cells were active, we cannot observe a statistically significant pO2 gradient. The data in the pO2 image corresponds to a mean value of 31.3 mmHg and a standard deviation of 4.2 mmHg. These variations are most probably due to the accuracy of our pO2 imaging procedure (which can be affected by magnetic field instabilities and noise in the acquired data).

4. Discussion 4.1. Measuring oxygen diffusion into blank microcapsules Based on our measurement results, we can deduce that the diffusion properties of O2 in the microcapsule gel are practically indistinguishable from those in the aqueous cell medium. There are also no apparent differences between the diffusion properties of O2 in alginate- and in ECM-based microcapsules. Such conclusions are in accordance with previous observations carried out on gel bulk material. For example, Berber and co-workers found that the diffusion coefficient of O2 in alginate gel and water at 37 °C is ∼2.54 × 10−9 m2 /s [8], which is very similar to the coefficient found in pure water of ∼2.6 × 10−9 m2 /s [43]. Similar observations are described also in the references cited in that article. The addition of salts and proteins, like those found in common cell medium, is known to reduce the diffusivity of O2 but only by a few percentage points at the most [39]. In contrast to Berber’s observations, Zhang, Ma, et al. [9] found that the diffusion coefficient of O2 is significantly dependent on the size of the microcapsule, decreasing from ∼2.1 × 10−9 m2 /s to ∼0.17 × 10−9 m2 /s as the microcapsule’s diameter is reduced from 1800 to 450 μm (the microcapsule’s density was increased from 1.013 g ml−1 at the larger diameter to 1.034 g ml−1 at the smaller diameter). These somewhat surprising findings, measured in bulk using multiple microcapsules, emphasize again the need to explore additional methodologies, such as the one provided here, to verify or disprove such measurements.

4.2. Measuring and mapping pO2 in microcapsules containing living cells The measurements of single microcapsules with living cells revealed a nominal O2 consumption rate of 15 amol/cell/s. This is in accordance with previous experiments carried out in bulk samples of many cells, that obtained values of ∼27 amol/cell/s (which may actually be as low as 3 amol/cell/s, depending on incubator conditions) [41]. Imaging results of oxygen distribution in the microcapsule did not reveal any appreciable gradient due to the combination of low O2 consumption rate per cell, sparse cell population, and good diffusivity of O2 in the gel material – all in accordance with the simulated data. It should be noted that in our experiments only ∼0.1% of the volume is occupied by the cells, and thus the present results are valid for the case of dilute singlecell dispersion. Under such conditions, O2 gradients are expected to appear only inside the cell in micron-scale lengths and are not expected to prevail outside the cell (see for example [44]). Supplementary Fig. S2 shows calculated data for such a case. Most probably, O2 diffusion will be different in more dense cellular environments. Accordingly, significant hypoxic conditions can be expected for MSC cells in gel microcapsules only if their concentration is significantly (∼5 to 10 times) larger, and/or the supply of ambient O2 is limited. Thus, one must be careful when designing the microcapsule in terms of gel material properties and acceptable cell concentrations. Our measurement approach should make it possible to see such effects, and we are currently experimenting with more dense cell systems to that end. 4.3. Comparison between different methodologies for measuring and imaging O2 in cell microcapsules We now return to the electrochemical and optical-based methods, briefly described in the Introduction, and compare them with the ESR method presented in this work in the context of pO2 mapping in live cell microcapsules (see also Table 3). (a) Electrochem-

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

ARTICLE IN PRESS

JID: ACTBIO

[m5G;November 5, 2019;1:18]

D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

9

Fig. 10. ESR microimaging data for microcapsule co-encapsulating living cells and LiNc-BuO microparticulates. The top 2D images are a subset of the complete 3D imaging data set of the entire microcapsule (shown at the bottom). Image resolution is ∼28 μm. (Left) Amplitude image showing the normalized amplitude of the radical distributed inside the microcapsule. (Right) pO2 image showing O2 distribution inside the microcapsule in units of mmHg. The white spots correspond to areas where the signal-tonoise-ratio was not high enough and/or the error in the lineshape fitting was too large to be included in the image. Table 3 Summary comparison of different methodologies to image pO2 inside single microcapsules with living cells. Method Electrochemical Luminescence lifetime-based ESR microscopy

Accuracy √√√ √√(2) √√

Lack of O2 consumption – √(3) √√√

Spatial resolution √√(1) √√√(4) √

Lack of invasiveness – √(5) √(5)

(1) Resolution depends on the size of the glass electrode’s tip (usually at least 10 μm), but 3D imaging is not possible. (2) See remarks above about potential accuracy limitation at low pO2 levels. (3) O2 is not directly consumed by the fluorophore, but optical quenching can result in single O2 , which is reactive. (4) While in principle 3D optical resolution can be in the sub-micron scale, we failed to locate an example of 3D image of pO2 in live cell microcapsules. (5) The method is minimally invasive since it requires the incorporation of an exogenous optical/ESR probe.

ical measurement with small glass electrodes is arguably considered to be the “gold standard” for pO2 measurements in biological media [45]. However, while in many cases it has proved to be very useful, its methodology is associated with a number of limitations, especially in relation to measuring microencapsulated cells. These include oxygen consumption by the electrode during measurement, which can alter the oxygen concentration in the sample and lead to oxygen gradients, even if there are no cells in a microcapsule [46]; puncturing of capsule by the microelectrode tip, which may damage the cells; and possible compression of the cells ahead of the tip [47]. Furthermore, this technique offers only point-like measurements and not 3D repetitive non-destructive imaging. (b) Optical-based methodologies are

very attractive due to the ever-increasing availability of specially designed O2 sensing molecules or nanoparticles, and to accessibility to advanced optical microscopes with high spatial resolution equipped with luminescence lifetime measurement capabilities [48]. However, in practical terms, as noted in the Introduction, we could not locate an optical study that presents pO2 images in live cell microcapsules. This might be due to certain limitations of optical techniques, such as the potential to induce cell damage as a result of high-power optical excitation and the generation of highly reactive excited single O2 states due to O2 interaction with the optical luminescent tracer [49]. It should be noted that although at low concentrations of oxygen (< 10 mmHg) the luminescent signal becomes larger, the accuracy of pO2 measurements

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

JID: ACTBIO 10

ARTICLE IN PRESS

[m5G;November 5, 2019;1:18]

D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

can decrease. This is primarily because O2 becomes less dominant in the quenching process and luminescence lifetime can be greatly affected by the specific bio-environment and the presence of other quenchers that commonly exist in cellular media, such as proteins with aromatic amino acid residues [50,51], nucleic acids (purines and pyrimidines), and many others [52]. The use of optical methods is also of limited applicability in medium-size samples (100–1000 μm, typical of cell microcapsules) which may be optically opaque in their inner areas, although in recent years twophoton methods have proved capable of resolving some of these issues [53]. The main advantages of the ESR-based oximetry method are: (a) lack of O2 consumption, meaning that low O2 levels can be reliably quantified [54]; (b) ease and accuracy of calibration, which does not depend on environmental conditions and medium content [20]; (c) the selectivity of the measurement regarding O2 quenching, since there are practically no other paramagnetic molecules in the tissue culture that can affect the ESR properties (T1 , T2 ) of the spin probe; and (d) ESR can make use of biocompatible micronscale solid spin probes as O2 reporters (see below) whose physical entrapment within the microcapsules prevents their leakage and allows for long-term follow-up. These characteristics of ESR oximetry appear to be especially useful for measuring O2 in the context of cell microcapsules.

5. Conclusions and future prospects We present here a new methodology that tests O2 diffusion properties in gel microcapsules used for entrapping live cells. Our approach, based on ESR spectroscopy and ESR microimaging technology, uses solid biocompatible paramagnetic microparticulates entrapped in microcapsules to report on the local pO2 through the strong effect O2 has on their ESR linewidth. Using this approach, we show that the diffusion coefficient of O2 in two types of gel materials (alginate and ECM), at the single microcapsule level, is practically indistinguishable from that of the cell medium itself. We also show that our ESR imaging procedure can be applied to a single microcapsule containing live cells without causing appreciable adverse effects to the cells, for a measurement period of up to ∼8 h. With this capability, we show ways to record cell activity in a single microcapsule and to extract oxygen consumption rates per cell. In addition, we make use of ESR microimaging to provide the first-of-its-kind spatially resolved data for pO2 in a single microcapsule containing living cells. Due to the relatively low cell concentration and fast diffusion of O2 , we are unable to detect any appreciable pO2 gradient. However, these measurements provide proof of principle for the accuracy and precision of the spatially resolved pO2 data. Future work may involve further methodological developments and engage in more specific applications to meet the needs of the cell microencapsulation community. Thus, for example, we may conduct studies of diffusion properties using different cell concentrations, microcapsule dimensions, and material densities. Moreover, in order to improve the accuracy of the results of such measurements, we may use our COMSOL simulation data more extensively, possibly integrating the simulated ESR signal from the entire microcapsule and not only from a few specific locations. Another path we intend to explore involves the use of a watersoluble trityl spin probe. Such work may take place using the Oxo63 trityl, the Finland trityl [55], or possibly newer derivatives of trityl, which allow obtaining pO2 , pH, and phosphate buffer concentrations simultaneously [56–58]. A tool like this may grant researchers a huge advantage regarding the development of microencapsulation systems and other micro-scale biomedical devices.

Acknowledgments The authors wish to acknowledge the Israeli Ministry of Science and Technology (Grant No. 3-14572). Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.actbio.2019.10.032. References [1] G.J. Lim, S. Zare, M. Van Dyke, A. Atala, Cell microencapsulation, in: J.L. Pedraz, G. Orive (Eds.), Therapeutic Applications of Cell Microencapsulation, Springer, New York, NY, 2010, pp. 126–136. [2] E.C. Opara, Springer Science+Business Media, Cell Microencapsulation: Methods and Protocols, Humana Press, New York, 2016. [3] J.L. Pedraz, G. Orive, Therapeutic applications of cell microencapsulation, Springer Science+Business Media, Landes Bioscience, New York, Austin, Tex., 2010. [4] D. Chaimov, L. Baruch, S. Krishtul, I. Meivar-Levy, S. Ferber, M. Machluf, Innovative encapsulation platform based on pancreatic extracellular matrix achieve substantial insulin delivery, J. Controlled Release 257 (2017) 91–101. [5] S. Krishtul, L. Baruch, M. Machluf, Processed tissue–derived extracellular matrices: tailored platforms empowering diverse therapeutic applications, Adv. Funct. Mater. 1900386 (2019) 1–26, doi:10.10 02/adfm.20190 0386. [6] L.A. Llacua, M.M. Faas, P. de Vos, Extracellular matrix molecules and their potential contribution to the function of transplanted pancreatic islets, Diabetologia 61 (6) (2018) 1261–1272. [7] J. Malda, T.J. Klein, Z. Upton, The roles of hypoxia in the in vitro engineering of tissues, Tissue Eng. 13 (9) (2007) 2153–2162. [8] U. Mehmetoglu, S. Ates, R. Berber, Oxygen diffusivity in calcium alginate gel beads containing Gluconobacter suboxydans, Artif. Cell Blood Sub. 24 (2) (1996) 91–106. [9] W. Zhao, Y. Zhang, Y. Liu, M.Q. Tan, W.T. Yu, H.G. Xie, Y. Ma, G.W. Sun, G.J. Lv, S. Zhao, X.J. Ma, Oxygen diffusivity in alginate/chitosan microcapsules, J. Chem. Technol. Biotechnol. 88 (3) (2013) 449–455. [10] I.A. Silver, The oxygen micro-electrode, in: Oxygen Transport to Tissue, Springer, 1973, pp. 7–15. [11] W.F. Mueller-Klieser, R.M. Sutherland, Oxygen tensions in multicell spheroids of two cell lines, Br. J. Cancer 45 (2) (1982) 256–264. [12] N. Opitz, D.W. Lübbers, Increased resolution power in pO2 analysis at lower pO2 levels via sensitivity enhanced optical pO2 sensors (pO2 optodes) using fluorescence dyes, in: Oxygen Transport to Tissue—VI, Springer, 1984, pp. 261–267. [13] A. Raza, H.E. Colley, E. Baggaley, I.V. Sazanovich, N.H. Green, J.A. Weinstein, S.W. Botchway, S. MacNeil, J.W. Haycock, Oxygen mapping of melanoma spheroids using small molecule platinum probe and phosphorescence lifetime imaging microscopy, Sci. Rep. 7 (1) (2017) 10743. [14] J.M. Vanderkooi, G. Maniara, T.J. Green, D.F. Wilson, An optical method for measurement of dioxygen concentration based upon quenching of phosphorescence, J. Biol. Chem. 262 (12) (1987) 5476–5482. [15] W.Y. Chen, M. Lisowski, G. Khalil, I.R. Sweet, A.Q. Shen, Microencapsulated 3-Dimensional sensor for the measurement of oxygen in single isolated pancreatic islets, PLoS One 7 (3) (2012) 1–10. [16] J. Weidling, S. Sameni, J.R.T. Lakey, E. Botvinick, Method measuring oxygen tension and transport within subcutaneous devices, J. Biomed. Opt. 19 (8) (2014) 1–5. [17] A.S. Johnson, R.J. Fisher, G.C. Weir, C.K. Colton, Oxygen consumption and diffusion in assemblages of respiring spheres: performance enhancement of a bioartificial pancreas, Chem. Eng. Sci. 64 (22) (2009) 4470–4487. [18] R. Cao, E. Avgoustiniatos, K. Papas, P. de Vos, J.R.T. Lakey, Mathematical predictions of oxygen availability in micro- and macro-encapsulated human and porcine pancreatic islets, J. Biomed. Mater. Res. Part B: Appl. Biomater. (2019) 1–10, doi:10.1002/jbm.b.34393. [19] J.S. Hyde, W.K. Subczynski, Spin-label oximetry, in: Spin Labeling, Springer, 1989, pp. 399–425. [20] W.K. Subczynski, H.M. Swartz, EPR oximetry in biological and model samples, in: Biomedical EPR, Part A: Free Radicals, Metals, Medicine, and Physiology, Springer, 2005, pp. 229–282. [21] S. Veriter, J. Mergen, R.M. Goebbels, N. Aouassar, C. Gregoire, B. Jordan, P. Leveque, B. Gallez, P. Gianello, D. Dufrane, Vivo selection of biocompatible alginates for islet encapsulation and subcutaneous transplantation, Tissue Eng. Pt. A 16 (5) (2010) 1503–1513. [22] S. Veriter, N. Aouassar, P.Y. Adnet, M.S. Paridaens, C. Stuckman, B. Jordan, O. Karroum, B. Gallez, P. Gianello, D. Dufrane, The impact of hyperglycemia and the presence of encapsulated islets on oxygenation within a bioartificial pancreas in the presence of mesenchymal stem cells in a diabetic Wistar rat model, Biomaterials 32 (26) (2011) 5945–5956. [23] S. Veriter, P. Gianello, Y. Igarashi, G. Beaurin, A. Ghyselinck, N. Aouassar, B. Jordan, B. Gallez, D. Dufrane, Improvement of subcutaneous bioartificial pancreas vascularization and function by coencapsulation of pig islets and mesenchymal stem cells in primates, Cell Transpl. 23 (11) (2014) 1349–1364.

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032

JID: ACTBIO

ARTICLE IN PRESS D. Cristea, S. Krishtul and P. Kuppusamy et al. / Acta Biomaterialia xxx (xxxx) xxx

[24] A. Lafosse, C. Desmet, N. Aouassar, W. Andre, M.S. Hanet, C. Beauloye, R. Vanwijck, H.A. Poirel, B. Gallez, D. Dufrane, Autologous adipose stromal cells seeded onto a human collagen matrix for dermal regeneration in chronic wounds: clinical proof of concept, Plast. Reconstr. Surg. 136 (2) (2015) 279–295. [25] H.J. Halpern, Stable soluble paramagnetic compounds, in: Vivo EPR (ESR), Springer, 2003, pp. 201–232. [26] Y.P. Liu, F.A. Villamena, J. Sun, T.Y. Wang, J.L. Zweier, Esterified trityl radicals as intracellular oxygen probes, Free Radic. Biol. Med. 46 (7) (2009) 876–883. [27] J.F. Dunn, H.M. Swartz, In vivo electron paramagnetic resonance oximetry with particulate materials, Methods 30 (2) (2003) 159–166. [28] R.P. Pandian, N.L. Parinandi, G. Ilangovan, J.L. Zweier, P. Kuppusamy, Novel particulate spin probe for targeted determination of oxygen in cells and tissues, Am. J. Physiol.-Heart Circ. Physiol. 292 (6) (2007) H3207. [29] R.P. Pandian, N.L. Parinandi, G. Ilangovan, J.L. Zweier, P. Kuppusamy, Novel particulate spin probe for targeted determination of oxygen in cells and tissues, Free Radic. Biol. Med. 35 (9) (2003) 1138–1148. [30] L. Shtirberg, Y. Twig, E. Dikarov, R. Halevy, M. Levit, A. Blank, High-sensitivity Q-band electron spin resonance imaging system with submicron resolution, Rev. Sci. Instrum. 82 (4) (2011) 043708. [31] R. Halevy, V. Tormyshev, A. Blank, Microimaging of oxygen concentration near live photosynthetic cells by electron spin resonance, Biophys. J. 99 (3) (2010) 971–978. [32] M. Hashem, M. Weiler-Sagie, P. Kuppusamy, G. Neufeld, M. Neeman, A. Blank, Electron spin resonance microscopic imaging of oxygen concentration in cancer spheroids, J. Magn. Reson. 256 (2015) 77–85. [33] L.M. Langan, N.J.F. Dodd, S.F. Owen, W.M. Purcell, S.K. Jackson, A.N. Jha, Direct measurements of oxygen gradients in spheroid culture system using electron parametric resonance oximetry, PLoS One 11 (8) (2016) 1–13. [34] R.P. Pandian, N.P. Raju, J.C. Gallucci, P.M. Woodward, A.J. Epstein, P. Kuppusamy, A new tetragonal crystalline polymorph of lithium octa-n-butoxy-naphthalocyanine (LiNc-BuO) radical: structural, magnetic and oxygen-sensing properties, Chem. Mater. 22 (23) (2010) 6254–6262. [35] S.M. Chacko, M. Khan, M.L. Kuppusamy, R.P. Pandian, S. Varadharaj, K. Selvendiran, A. Bratasz, B.K. Rivera, P. Kuppusamy, Myocardial oxygenation and functional recovery in infarct rat hearts transplanted with mesenchymal stem cells, Am. J. Physiol.-Heart Circ. Physiol. 296 (5) (2009) H1263–H1273. [36] M.M. Kmiec, H. Hou, M. Lakshmi Kuppusamy, T.M. Drews, A.M. Prabhat, S.V. Petryakov, E. Demidenko, P.E. Schaner, J.C. Buckey, A. Blank, P. Kuppusamy, Transcutaneous oxygen measurement in humans using a paramagnetic skin adhesive film, Magn. Reson. Med. 81 (2) (2019) 781–794. [37] R. Halevy, Y. Talmon, A. Blank, Photolithographic production of glass sample holders for improved sensitivity and resolution in ESR microscopy, Appl. Magn. Reson. 31 (3–4) (2007) 591–598. [38] A. Blank, J.H. Freed, ESR microscopy and nanoscopy with "Induction" detection, Isr. J. Chem. 46 (4) (2006) 423–438. [39] T.L. Place, F.E. Domann, A.J. Case, Limitations of oxygen delivery to cells in culture: an underappreciated problem in basic and translational research, Free Radic. Biol. Med. 113 (2017) 311–322. [40] U. Barkai, A. Rotem, P. de Vos, Survival of encapsulated islets: more than a membrane story, World J. Transpl. 6 (1) (2016) 69–90. [41] G. Pattappa, H.K. Heywood, J.D. De Bruijn, D.A. Lee, The metabolism of human mesenchymal stem cells during proliferation and differentiation, J. Cell. Physiol. 226 (10) (2011) 2562–2570. [42] B.A. Wagner, S. Venkataraman, G.R. Buettner, The rate of oxygen utilization by cells, Free Radic. Biol. Med. 51 (3) (2011) 700–712.

[m5G;November 5, 2019;1:18] 11

[43] P. Han, D.M. Bartels, Temperature dependence of oxygen diffusion in H2 O and D2 O, J. Phys. Chem. 100 (13) (1996) 5597–5602. [44] H. Kurokawa, H. Ito, M. Inoue, K. Tabata, Y. Sato, K. Yamagata, S. Kizaka-Kondoh, T. Kadonosono, S. Yano, M. Inoue, T. Kamachi, High resolution imaging of intracellular oxygen concentration by phosphorescence lifetime, Sci. Rep. 5 (2015) 1–13. [45] H.B. Stone, J.M. Brown, T.L. Phillips, R.M. Sutherland, Oxygen in human tumors: correlations between methods of measurement and response to therapy: summary of a workshop held November 19–20, 1992, at the National Cancer Institute, Bethesda, Maryland, Radiat. Res. 136 (3) (1993) 422–434. [46] W. Muller, A. Winnefeld, O. Kohls, T. Scheper, W. Zimelka, H. Baumgartl, Real and pseudo oxygen gradients in ca-alginate beads monitored during polarographic pO2 -measurements using Pt-needle microelectrodes, Biotechnol. Bioeng. 44 (5) (1994) 617–625. [47] J. Klinowski, S.E. Korsner, C.P. Winlove, Problems associated with the micropolarographic measurement of the arterial wall pO2 , Cardiovasc. Res. 16 (8) (1982) 448–456. [48] D.B. Papkovsky, R.I. Dmitriev, Quenched-Phosphorescence Detection of Molecular Oxygen: Applications in Life Sciences, Royal Society of Chemistry, Cambridge, UK, 2018. [49] S.M. Borisov, Fundamentals of quenched phosphorescence O2 sensing and rational design of sensor materials, in: D.B. Papkovsky, R.I. Dmitriev (Eds.), Quenched-Phosphorescence Detection of Molecular Oxygen: Applications in Life Sciences, 2018. [50] A. Fürstenberg, O. Kel, J. Gradinaru, T.R. Ward, D. Emery, G. Bollot, J. Mareda, E. Vauthey, Site-dependent excited-state dynamics of a fluorescent probe bound to avidin and streptavidin, ChemPhysChem 10 (9–10) (2009) 1517–1532. [51] H. Chen, S.S. Ahsan, M.E.B. Santiago-Berrios, H.D. Abruña, W.W. Webb, Mechanisms of quenching of Alexa fluorophores by natural amino acids, J. Am. Chem. Soc. 132 (21) (2010) 7244–7245. [52] J.R. Lakowicz, Quenching of fluorescence, in: Principles of Fluorescence Spectroscopy, Springer, 1983, pp. 257–301. [53] S. Sakadzic, E. Roussakis, M.A. Yaseen, E.T. Mandeville, V.J. Srinivasan, K. Arai, S. Ruvinskaya, A. Devor, E.H. Lo, S.A. Vinogradov, D.A. Boas, Two-photon high-resolution measurement of partial pressure of oxygen in cerebral vasculature and tissue, Nat. Methods 7 (9) (2010) 755–U125. [54] T. Presley, P. Kuppusamy, J.L. Zweier, G. Ilangovan, Electron paramagnetic resonance oximetry as a quantitative method to measure cellular respiration: a consideration of oxygen diffusion interference, Biophys. J. 91 (12) (2006) 4623–4631. [55] J.H. Ardenkjaer-Larsen, I. Laursen, I. Leunbach, G. Ehnholm, L.G. Wistrand, J.S. Petersson, K. Golman, EPR and DNP properties of certain novel single electron contrast agents intended for oximetric imaging, J. Magn. Reson. 133 (1) (1998) 1–12. [56] A.A. Bobko, T.D. Eubank, B. Driesschaert, I. Dhimitruka, J. Evans, R. Mohammad, E.E. Tchekneva, M.M. Dikov, V.V. Khramtsov, Interstitial inorganic phosphate as a tumor microenvironment marker for tumor progression, Sci. Rep. 7 (2017) 1–12. [57] A.A. Bobko, T.D. Eubank, I. Dhimitruka, J.L. Zweier, V.V. Khramtsov, Multifunctional assessment of tissue pO2 , PH and inorganic phosphate (Pi) using in vivo EPR and phosphanated trityl probe: interstitial Pi as a new prognostic factor in tumorigenesis, Free Radic. Biol. Med. 76 (2014) S121–S122. [58] A.A. Bobko, I. Dhimitruka, J.L. Zweier, V.V. Khramtsov, Fourier transform epr spectroscopy of trityl radicals for multifunctional assessment of chemical microenvironment, Angew. Chem. Int. Ed. 53 (10) (2014) 2735–2738.

Please cite this article as: D. Cristea, S. Krishtul and P. Kuppusamy et al., New approach to measuring oxygen diffusion and consumption in encapsulated living cells, based on electron spin resonance microscopy, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.10.032