1. Introduction

Diabetes mellitus (DM) is a group of metabolic disorders characterized by a high blood glucose level over a prolonged period and threatening human health in epidemic spreading. DM is mainly classified into type 1 DM (T1DM) and type 2 DM (T2DM), where the dysfunction and death of beta cells and the decrease of insulin sensitivity/secretion, respectively, are considered as major causalities [1–6]. Beta cells, a type of the endocrine cells in the pancreatic islets, regulate blood glucose level by secreting insulin, and their loss or dysfunction is related with diabetes progression [7,8]. Therefore, it is important to maintain the survival and function of pancreatic beta cells for daily glucose homeostasis, and the monitoring of islets is important for the prevention of diabetes related complications. However, islet monitoring has remained a clinical challenge due to their anatomical location within the endocrine pancreas [9]. We developed a technology platform transplanting pancreatic islets into the anterior chamber of the eye (ACE), which could be used as the surrogate of pancreatic islet monitoring in situ or the treatment of diabetes in replacement therapy [10–12]. This method has been tested not only in the small animal models but also in primates [13–15], and efforts for clinical applications are ongoing [16]. Optical coherence microscopy (OCM) was previously used for the non-invasive monitoring of pancreatic islets transplanted into the ACE of non-obese diabetic (NOD) mice as well as for detecting lesions in the pancreas such as cancer [17–19]. OCM is a 3D imaging technique based on light scattering and depth-resolved detection for the rapid and non-invasive visualization of tissue micro-structures [20–22]. OCM visualizes vasculature by analyzing the phase fluctuation of reflected light with blood flow [23–25]. OCM visualized islet grafts in the ACE as highly scattering and vascularized round objects. In the progression of insulitis by autoimmune attack, OCM detected various changes in the islet grafts non-invasively. OCM could be used as a monitoring method of ACE transplanted islets for early detection and intervention, and further study is needed for verification.

In this study, OCM was applied to streptozotocin (STZ)-induced diabetic mouse models for testing the feasibility of early detecting beta cell death or dysfunction in the ACE transplanted islet grafts. STZ is a pancreatic beta cell specific cytotoxin which induces oxidative stress and then cell death [26]. A custom extended-focus optical coherence microscopy (xfOCM) was developed and used to visualize the micro-structure and vasculature of islet grafts in the ACE. Bright-field (BF) imaging and glucose measurement were conducted together with xfOCM imaging as the reference. For the cellular examination of islet grafts, confocal microscopy of ex vivo grafts with fluorescent staining was conducted. Quantitative analysis of longitudinal xfOCM images was performed.

2. Method

2.1 Experimental animal care and islet transplantation

All experimental procedures were approved by the Pohang University of Science and Technology Institutional Animal Care and Use Committee (POSTECH-2017-0084, POSTECH IACUC). 12 8-week-old C57BL/6J mice in total were used in this study and were maintained in cages with free access to water and food under a 12-h light/dark cycle. Pancreatic islets were isolated by collagenase digestion (1 mg/ml collagenase P; Roche Diagnostics, Indianapolis, IN) and subsequently handpicked under a stereomicroscope. Isolated islets were cultured in RPMI 1640 medium (Gibco, Carlsbad, CA) supplemented with 10% FBS, 100 U/ml penicillin, and 100µg/ml streptomycin at 37°C in a humidified atmosphere of 5% CO₂ for one day before transplantation. Recipient mice were anesthetized by a gas mixture of 1.5%/vol isoflurane (Terrel, Piramal) and medical grade oxygen, and placed on a heating pad. Their heads were restrained by a head holder (SGM-3, Narishige) and eyeballs were positioned facing upward and gently held by a universal solid joint (UST-2, Narishige). The cornea was punctured with a 26G 1/2” needle and the isolated islets were introduced into the anterior chamber of eye (ACE) by a pulled glass micropipette (1.5mm outer diameter, World Precision Instruments, USA). 4 weeks after transplantation, the transplanted islets were imaged by xfOCM, in vivo.

2.2 Extended-focus optical coherence microscopy (xfOCM)

A custom xfOCM was developed for high-resolution 3D imaging in the extended depth range, and the simplified schematic is shown in Fig. 1. The xfOCM system was a spectral domain OCM using a broadband light source and a custom spectrometer and was implemented by using Bessel beam illumination and Gaussian beam detection [27,28]. The light source was a high-power super-luminescence diode (SLD-371-HP3, SUPERLUM) with the center wavelength at approximately 840 nm and the bandwidth of approximately 50 nm in full width at half maximum (FWHM) intensity. Light from the source was split into the sample and reference arms with a fiber coupler in 75:25 ratio (TW850R3A2, Thorlabs). In the sample arm, collimated beam was generated by a collimator (HPUCO-13A-850-S-5AC-UNL, OZ optics) and converted to radial zero-order Bessel beam by an axicon lens with 170° apex angle (AX255-B, Thorlabs). The focal plane of Bessel beam was relayed to the object plane by using three pairs of lens combinations (L1-L6) in 4-f configuration. L1-L5 (60mm, 150mm, 100mm, 50mm, 150mm in focal length) were standard achromatic lenses (AC254/AC508, Thorlabs) and the ring beam was expanded in diameter from 4.7mm to 9.4mm by passing L1-L5. L6 was a 10x objective lens (UPlanFL 10x, EF: 18mm, OLYMPUS). A x-y galvano scanning mirror (GVS012, Thorlabs) was placed in between L3 and L4. Back scattered light from the specimen was collected back by the objective lens, relayed backward in the same path as the illumination beam down to the scanner pair, and de-scanned after reflection on the scanner pair. The scattered light was separated from the illumination beam via reflection on a right-angled prism (aluminum coated and 5 mm leg, Edmund optics) and coupled into a fiber via a collimator (HPUCO-13A-850-S-7.5AC-UNL, OZ optics). Therefore, scattered light from the specimen was collected as the Gaussian beam. The Bessel beam illumination field had 0.2 NA. Detection field with Gaussian distribution had 0.05 NA to avoid specular reflection. Scattered light from the specimen was combined with reference light, which was transmitted through the reference arm with an optical delay path for the matching of optical path lengths, using a non-polarizing beam splitter (BS038, Thorlabs). In addition to the optical delay, light in the reference arm was attenuated with a neutral density filter (NDC-100C-4, Thorlabs) and dispersed by going through a prism pair (N-SF11 uncoated prism, Edmund optics). The prism pair was for matching the amount of light dispersion of the scattered light from the sample in the sample arm. Combined light in the fiber was delivered to the custom spectrometer and collimated by a collimator (HPUCO-13A-900-S-30AC-UNL, OZ optics), and spectrally dispersed by using a volume phase holographic grating (VPHG, WP-HD1800/840-35X45, Wasatch Photonics). The dispersed light transmitted by a lens (1:1.4 85mm AS IF ΜMC, Samyang), and then collected as a line at a line CCD camera (EV71YEM4CL2014-BA9, Teledyne e2v). Depth scans (A-scans) with the line camera were acquired at 40kHz. Imaging volume was 0.42 mm × 0.42 mm × 0.86 mm in the x, y and z directions corresponding to 500 pixels × 500 pixels × 375 pixels, respectively. The depth of field was 0.86 mm in the air. The lateral resolution was 3.1 µm, measured with a USAF-1951 resolution target (R3L3S1N, Thorlabs). The measured axial resolution was 6.4 ± 0.2 µm.

 

Fig. 1. A system configuration of custom xfOCM. Bessel-beam illumination and Gaussian-beam detection paths are depicted in gray and red colors, respectively. Optical delay part in the reference arm and a spectrometer is presented in black and red dashed line boxes, respectively. FC: fiber coupler, COL: collimator, AP: angled prism, SM: scanning mirror, BS: non-polarizing beam splitter, VPHG: volume phase holographic grating, L1-L5: achromatic lens, L6: 10x objective lens.

Download Full Size | PDF

2.3 Experiment design

For the xfOCM characterization study of STZ induced islet damage, 6 islet grafts in four mice were used. The islet grafts were imaged daily starting from before STZ injection and until day 5 after the injection. Two islet grafts were extracted at the end of the imaging period, labeled, and imaged with confocal microscopy for cellular examination, and the confocal images of damaged islet grafts were compared with the ones of normal islet grafts obtained from two healthy mice. For the xfOCM study of early damage detection, 11 islet grafts in six mice were imaged in two separate sessions: before the injection and in 12 hours after the injection. After the imaging, 10 islet grafts were extracted and imaged with confocal microscopy. BF imaging and glucose level measurement was performed together with xfOCM, immediately before the xfOCM imaging.

2.4 In vivo imaging protocols with a custom xfOCM

150 mg/kg of STZ was administered to mice intraperitoneally to induce beta cell death. In vivo imaging of islet grafts in the ACE was conducted under gas anesthesia with the inhalation of a gas mixture of 1.5%/vol isoflurane and medical grade oxygen through an anesthesia machine (VETIA, J&TEC). Anesthetized mice were placed on the heating pad for hypothermia prevention, and their heads were restrained by the head-holder. The eye was positioned facing upwards and held with the universal solid joint, and covered with lubricant eye gel (GenTeal, Alcon, Fort Worth, TX) and a coverslip. xfOCM imaging was conducted by acquiring 3D cross-sectional images with lateral scanning. For vasculature imaging, cross-sectional images at the same lateral positions were acquired for 10 times, and the repeated images were processed by using complex differential variance (CDV) algorithm to exploit both intensity and phase changes from blood flow [29]. Total imaging time per volume was 60s. En-face images were produced by using the ImageJ software (1.53a, National Institutes of Health).

2.5 Glucose measurement protocols

Mice were fasted overnight for 16 hours (6pm to 10am next morning), and blood glucose level was measured by using a glucometer with test strips (Accu-Chek Active, Roche, Mannheim, Germany). A small amount of blood was obtained by the tail-tip amputation of the fasted mice after cleaning the tail with 70% ethanol and was dropped to the tip of the test strip, which was inserted in the glucometer. The glucose level was read. Bleeding spot was pressed with a clean gauze to stop bleeding before returning mice to the animal cage. STZ injection was conducted after the glucose measurement. Non-fasting glucose level after STZ injection was measured in the same procedure as above just before xfOCM imaging, which was conducted longitudinally.

2.6 Immunostaining and confocal imaging of islet grafts

Mice were euthanized after the completion of in vivo imaging and glucose measurement by increasing the concentration of isoflurane, and islet graft-bearing eyeballs were surgically taken out. The iris with islet grafts was carefully removed from the eyeball, fixed in 4% paraformaldehyde overnight at 4 °C, and permeabilized with phosphate buffered saline (PBS) containing 0.3% triton x-100, and 5% bovine serum albumin. The islet graft was incubated with insulin and glucagon antibodies (Abcam ab7842, Sigma G2654) overnight and washed with PBS containing 0.1% Tween 20 (PBST) for 5 times. Secondary antibodies conjugated with Alexa488 or Alexa594 (Thermo fisher scientific) were given to the sample for 3 hours and the sample was again washed with PBS for 5 times. The sample was mounted in anti-fade mounting medium and analyzed with a commercial confocal microscopy system (SP-5, Leica).

2.7 Quantitative analysis and statistics

Statistical analysis was performed with the xfOCM images of 11 islet grafts in six mice before and in 12 hours after STZ injection. Light scattering of the islet graft was measured by analyzing en-face sectional xfOCM intensity images. Islet regions in the image were segmented, and intensity histograms were generated. Both median and skewness values were calculated from the intensity histogram and presented as boxplot with 2 sigma whiskers including 95% in normal distribution. Note that vascular regions in the en-face image were excluded in the generation of intensity histogram to analyze light scattering in non-vascular regions only. Vascular regions could be identified from the corresponding xfOCM angiography image. Vascular volume in the islet was calculated by counting the number of pixels in the vascular region of 3D xfOCM angiography images. Relative vascular volume was calculated by dividing the volume after STZ injection with the one before injection. All image processing and statistical analysis were conducted in MATLAB (Mathworks). Welch’s t-test was used to assess the difference in the quantitative results between the groups with unequal variances. All statistical tests were performed by two-tailed tests, and p-values < 0.05 were considered as the criteria for statistical significance. Statistical significance was presented by the number of asterisks depending on the p-value; *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

3. Results

3.1 xfOCM characterization of ACE transplanted islets after STZ-induced damage

Islets from donor C57BL/6 mice were isolated and transplanted into the ACE of the recipient C57BL/6 mice. In 4 weeks after transplantation, islet grafts were imaged longitudinally with xfOCM. The imaging started from a few days before STZ injection and continued daily until 5 days after the injection, because the islet damage induced by STZ was typically observed within 3 days after the injection. For comparison between intact islet grafts and seriously damaged ones by STZ, representative images of an islet graft before and on day 3 after STZ injection are presented in Fig. 2. Both BF and xfOCM images of the same islet graft before and on day 3 after STZ injection, and blood glucose levels of the mouse model at the two time points are shown in Fig. 2(a, b) and 2(c), respectively. Confocal image of the same islet graft prepared on day 5 after STZ injection (on the final day of the monitoring period) was presented together with that of the normal islet graft to show cellular changes induced by STZ in Fig. 2(d, e). The images before STZ injection showed the characteristics of normal islet grafts. BF image showed the entire ACE and an islet residing on the iris in Fig. 2(a1). Magnified BF image of the islet graft, which was presented as an inset, showed an opaque white and round object. xfOCM intensity and angiography images of the same islet were presented in both the cross-sectional (x-z) and en-face (x-y) planes in Fig. 2(b1). En-face images were acquired approximately in the central plane of the islet graft before and after STZ injection. The central plane was found to be the one with maximum en-face area. The islet graft appeared elliptical in the cross-sectional image due to image distortion with anisotropic image scales. The intact islet graft was round in morphology and highly scattering across the section in the intensity image. Strong light scattering in islet grafts was owing to zinc-insulin crystals in the pancreatic beta cells as verified in previous reports [17,30,31]. xfOCM intensity image showed relatively low scattering in blood vessels. The reduced scattering in blood vessels was owing to light absorption and speckle washout with blood flow [32,33]. Blood vessels within the islet graft were visualized more clearly in the angiography images. Large and small vessels occupied in the center and periphery of the islet graft in the en-face image, respectively. Vessels occupied more than 50% of the sectional area in the islet graft before STZ injection.

 

Fig. 2. Representative bright-field (BF) and xfOCM images of an islet graft in the ACE before and after STZ-induced damage. (a1-a2) BF images of the islet before and 3 days after STZ injection, respectively. Magnified BF islet images were presented as insets. (b1-b2) xfOCM images of the same islet graft before and 3 days after STZ injection, respectively. Yellow dashed lines in the x-z cross-sectional images indicated the depth of x-y en-face images. (c) Blood glucose levels of mice (n = 4) at the two different time points. (d-e) Representative confocal fluorescence images of a control islet graft and the one taken from the mice euthanized on day 5 after STZ injection, respectively. Confocal image of the islet graft on day 5 after STZ injection was the same one observed in the longitudinal BF and xfOCM images. Insulin (red) and glucagon (green) were visualized by immunofluorescence labeling. Scale bars indicate 100 µm.

Download Full Size | PDF

The xfOCM and BF images of the same islet graft on day 3 after STZ injection showed various changes. BF image showed the islet graft less white and more translucent on day 3 after STZ injection than the one before the injection in Fig. 2(a2). The magnified BF image showed morphological changes as well. The islet graft, which was round and smooth before damage, became less smooth on day 3 after STZ treatment. 3D xfOCM images showed the changes clearly via optical sectioning in Fig. 2(b2). The en-face sectional xfOCM intensity image showed that the islet graft had significantly low light scattering and rough boundary compared to the one before STZ injection. The lowered light scattering in the islet graft could be attributed to the destruction of beta cells by STZ. The angiography image in the central plane showed reduced vasculature compared to the one before STZ injection. Vasculature in the islet graft was approximately 22% in the section in Fig. 2(b2). Both BF and xfOCM images on day 3 after STZ injection showed the changes of the islet graft associated with the beta cell destruction. The mouse models had a fasting blood glucose level of approximately 100 mg/dl just before the STZ injection, and their non-fasting glucose level on day 3 after STZ injection was over 400mg/dl. The glucose level increase indicated STZ induced diabetes with significant pancreatic beta cell death. This was also consistent with xfOCM image results. Confocal image of a damaged islet graft after immunostaining in Fig. 2(e) showed significant pancreatic beta cell death compared with the one of an intact islet graft in Fig. 2(d). Confocal images visualized both beta cells and alpha cells in the islet grafts via insulin and glucagon labeling in red and green colors, respectively. Confocal image of the intact islet showed dense cell distribution in the islet graft, mainly beta cells and some alpha cells. On the other hand, confocal image of the damaged islet graft showed low insulin fluorescence, as well as sparse and irregular cell distribution on the surface.

3.2 xfOCM observation of early-period STZ-induced damage in ACE transplanted islets

xfOCM detection of damaged islet grafts on day 3 after STZ injection motivated us to monitor islet grafts earlier for testing the feasibility of early detection. Sensitive detection of islet damage is essential for the early diagnosis of diabetes and prompt treatment [34–36]. xfOCM monitoring was conducted 12 hours after STZ injection and representative BF and xfOCM images of islet grafts taken from two individual mice are presented in Fig. 3. Blood glucose measurement was conducted together with the imaging, and the results are shown in Fig. 3(e). Confocal image of an islet graft in 12 hours after STZ injection presented together with the one of an intact islet graft for comparison in Fig. 3(f, g).

 

Fig. 3. Early detection of islet damage caused by STZ using xfOCM. (a-b) and (c-d) BF and xfOCM images of two representative islet grafts in the ACE before and in 12 hours after STZ injection. (a1-a2) and (c1-c2) BF images of the two islet grafts at the two time points. (b1-b2) and (d1-d2) xfOCM images of the two islet grafts. (e) Glucose levels in experimental mice (n = 6) at the two time points. (f-g) Confocal fluorescence images of islets before and in 12 hours after STZ injection. The islet in (g) was the same one shown in (a2) and (b2), and it was marked with asterisks. The confocal images showed cells in the islet with insulin (red) and glucagon (green) labeling. Scale bars indicate 100 µm.

Download Full Size | PDF

BF images of the islet grafts in 12 hours after STZ injection in Fig. 3(a2) and 3(c2) were not much different from the ones before the injection in Fig. 3(a1) and 3(c1). On the other hand, xfOCM images showed some changes in the islet grafts. xfOCM image of the islet graft in Fig. 3(b2) showed a low intensity region inside the islet marked with a cyan arrowhead and some irregular surface marked with a white arrowhead. Because this low intensity region in the intensity image was not co-registered with vasculature in the angiography image, the local intensity decrease could be due to beta cell death. The irregular boundary was also found in the seriously damaged islet on day 3 after STZ injection in Fig. 2(b2). Similar features were found in xfOCM image of the other islet graft in Fig. 3(d2). xfOCM intensity images showed low scattering regions on the islet graft boundary, marked with cyan arrowheads. xfOCM angiography images showed vasculature decrease in both the islet grafts in 12 hours after STZ injection in Fig. 3(b2) and 3(d2). Reduction of vascular area was consistent with the feature shown in the severely damaged islet graft on day 3 after STZ injection in Fig. 2(b2). In the blood glucose level, non-fasting glucose levels of the mice in 12 hours after STZ injection were not much different from the fasting glucose levels before the injection in Fig. 3(e). Confocal image of the islet graft in 12 hours after STZ injection showed damage of insulin labeled beta cells in red color and irregular islet surface in Fig. 3(g), compared to the one before STZ injection in Fig. 3(f). The maintenance of glucose levels in 12 hours after STZ injection could be explained by some remaining functional beta cells and/or the burst release of insulin by the death of beta cells after STZ injection.

3.3 Quantitative analysis of longitudinal xfOCM islet images in the early period of STZ-induced damage

xfOCM could detect changes in the islet grafts in the ACE caused by STZ as early as 12 hours post-injection. The changes found by xfOCM were analyzed quantitatively and the results were shown in Fig. 4. Post-processing of xfCOM images was conducted for quantitative analysis and details explained in Method. Representative post-processed xfOCM images of an islet graft before and in 12 hours after STZ injection in the en-face planes and their intensity histograms are shown in Fig. 4(a). Statistical results from the xfOCM images of 11 islet grafts before and in 12 hours after STZ injection were presented as median and skewness values of the intensity histograms in Fig. 4(b, c), respectively. The relative change of vascular volume in the islet grafts between the two time points was presented as a bar graph in Fig. 4(d).

 

Fig. 4. Quantitative analysis of longitudinal xfOCM images of islet grafts in the early period of islet damage induced by STZ. (a) Representative xfOCM images and intensity histograms of an islet graft before and in 12 hours after STZ injection. Cyan arrow heads in the xfOCM image in 12 hours after the injection marked local reduced scattering regions. (b-c) Median intensity and skewness values of the intensity histograms from 11 islet grafts at the two different time points. (d) Ratio of vascular volume in the islet grafts between the two time points. Statistical significances were presented by number of asterisks depending on the p-value: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

Download Full Size | PDF

xfOCM images showed internal structure and vasculature of the islet graft in the en-face plane at approximately 40 µm deep from the surface. Intensity histograms of the two xfOCM images showed the changes of intensity distribution. The intensity histogram, which was approximately symmetric before STZ injection, became asymmetric with some stretch in the negative direction in 12 hours after STZ injection. The median intensity value decreased as well. The skewness of intensity histogram increased from -0.15 ± 0.20 to -0.45 ± 0.18 in the negative direction and the median intensity decrease from 2710 ± 110 to 2550 ± 120. The changes in the intensity histogram would reflect local intensity decrease found in the xfOCM image in 12 hours after STZ injection. Vascular change in the islet grafts was analyzed as the ratio of vascular volumes from the angiographic xfOCM images before and in 12 hours after STZ injection. The ratio of vascular volume after STZ injection to the one before STZ injection was 0.74 ± 0.26.

4. Discussion and conclusion

xfOCM is a non-invasive high-speed 3D imaging technique with an extended depth of field (DOF) while maintaining a relatively high lateral resolution by using Bessel beam illumination and Gaussian beam detection. Its application to intraocular transplantation models would be beneficial for non-invasive and close observation of pancreatic islets with 3D imaging. xfOCM was used for the longitudinal monitoring of islet grafts in the ACE of STZ-treated diabetic mice for the early detection of islet damage. BF imaging and glucose level measurement were conducted together with xfOCM imaging for comparison. Longitudinal xfOCM monitoring showed the more sensitive detection of beta cell destruction in the islet caused by STZ than BF imaging and glucose measurement. The structural and vascular changes in the islet were observed by xfOCM imaging. Local light scattering reduction and irregular surface morphology in the islet grafts were detected, and these changes could be associated with the damage of highly scattering beta cells. The reduction of vasculature was also observed within the islet graft in the early period after STZ injection. The cause of vasculature reduction is unclear, but it could be associated with beta cell death, because beta cells are known to have physical contact with blood vessels in the islet, and beta cells secrete vascular endothelial growth factor-A (VEGF-A) that helps blood vessel formation [37–39]. The detection of islet damage by xfOCM was confirmed by cellular examination via immunostaining and confocal microscopy. Beta cell death and structural destruction in the islet grafts were identified. Although xfOCM did not have the image resolution and contrast to visualize individual cells, it could detect various changes in islet grafts such as the changes of light scattering, morphology, and vasculature. On the other hand, BF imaging was less sensitive than xfOCM to the local changes because the changes inside the islet were not observable. Blood glucose measurement was not able to detect beta cell death caused by STZ in the early period, because blood glucose levels could be collectively affected by food intake, corresponding insulin secretion and/or burst release of insulin in the process of beta cell death. Longitudinal xfOCM images of islet grafts during STZ-induced beta cell death were quantitatively analyzed in terms of image intensity and vasculature. The intensity changes were analyzed by using intensity histograms. The decrease of median intensity and the increase of skewness in the negative direction were observed. These changes were associated with local intensity reduction in the image. Vasculature decrease was analyzed by measuring vascular volume within the islet grafts by using angiography images, and the vascular volume change could be used as another parameter for the sensitive detection of beta cell death.

In this study, we treated mice with STZ to induce beta cell death in both pancreatic islets in situ and islet grafts in the ACE. Although STZ induces beta cell death in a very short period, which could be different from the gradual destruction of pancreatic beta cells in diabetes progression, it is an appropriate experimental model because it induces the specific beta cell death, and inflammation is also accompanied by apoptosis and necrosis in a similar way as diabetes progression [26,40,41]. However, for the application of xfOCM to the intraocular islet grafts in the clinic, verification of this technique in large animals such as monkeys would be necessary owing to differences between species. In addition, it was difficult to conduct the experiment at a short time interval because frequent anesthesia could induce changes in the body metabolism and immune system. In the future, changes at shorter time points will be assessed by conducting the experiment at different time schedules with different mouse groups.

In conclusion, non-invasive xfOCM was applied to the longitudinal imaging of islet grafts in the ACE of STZ-treated diabetic mice. xfOCM detected various changes in islet grafts associated with beta cell death caused by STZ more sensitively than BF imaging and glucose measurement. The changes were local decrease of light scattering, surface roughening, and vasculature decrease, and they were quantified. The study results showed that xfOCM could be used as a non-invasive and sensitive monitoring method of pancreatic islets transplanted in the ACE of animal models and has a potential for clinical applications.