The goal of this protocol is to continuously monitor the dynamics of the human pancreatic islet engraftment process and the contributing host versus donor cells. This is accomplished by transplanting human islets into the anterior chamber of the eye (ACE) of an NOD.(Cg)-Gt(ROSA)26Sortm4–Rag2-/-mouse recipient followed by repeated 2-photon imaging.
Imaging beta cells is a key step towards understanding islet transplantation. Although different imaging platforms for the recording of beta cell biology have been developed and utilized in vivo, they are limited in terms of allowing single cell resolution and continuous longitudinal recordings. Because of the transparency of the cornea, the anterior chamber of the eye (ACE) in mice is well suited to study human and mouse pancreatic islet cell biology. Here is a description of how this approach can be used to perform continuous longitudinal recordings of grafting and revascularization of individual human islet grafts. Human islet grafts are inserted into the ACE, using NOD.(Cg)-Gt(ROSA)26Sortm4–Rag2-/-mice as recipients. This allows for the investigation of the expansion of recipient versus donor cells and the contribution of recipient cells in promoting the encapsulation and vascularization of the graft. Further, a step-by-step approach for image analysis and quantification of the islet volume or segmented vasculature and islet capsule forming recipient cells is outlined.
Diabetes mellitus describes a group of metabolic diseases characterized by elevated levels of blood glucose as results of insufficient insulin production from loss or dysfunction of pancreatic islet beta cells, often accompanied by insulin resistance. Type 1 (T1D) and type 2 diabetes (T2D) are complex diseases in which the progressive dysfunction of the beta cells causes disease development. T1D is precipitated by an autoimmune attack on the beta cells, while T2D is considered to be driven by metabolic factors, albeit with increasing evidence of low-grade systemic inflammation1. Transplantation of donor human islets, particularly to T1D patients, offers the potential for providing physiological glycemic control. However, a shortage of tissue donors and poor islet engraftment has prevented islet transplantation to become a mainstream therapeutic option. A substantial proportion of the functional islet graft is lost in the immediate posttransplantation period (24–48 h) due to the hypoxic, inflammatory, immunogenic host environment2,3. To evaluate the efficiency of intervention methods for the improvement of islet survival, continuous monitoring of such transplantations is necessary.
In vivo techniques to image and track the fate of transplanted human pancreatic islets after transplantation still remains a challenge for diabetes research4,5. To date, noninvasive imaging techniques, including positron emission tomography (PET), magnetic resonance imaging (MRI), or ultrasound (US) show potential for the quantification and functional evaluation of transplanted islets in experimental conditions5. However, given the small islet sizes, quantitative measurements by those modalities suffer from insufficient resolution. The anterior chamber of the eye (ACE) as a transplantation site for observation is a promising noninvasive imaging solution offering effectively higher spatial resolution and frequent monitoring over long time periods6. This method has been successfully exploited to study mouse islet biology (reviewed in Yang et al.7), autoimmune immune responses8, as well as human islet grafting9,10.
Here the ACE transplantation method is combined with a 2-photon imaging approach to investigate the dynamics of the human pancreatic islet engraftment process by continuous and repeated recordings on individual islet grafts for up to 10 months after transplantation. The multiphoton imaging properties of greater imaging depths and reduced overall photobleaching and photo damage overcome the imaging limitations of confocal microscopy11. Quantification of fluorescent imaging involves several stages, including islet sample preparation, islet transplantation, image acquisition, image filtering to remove islet noise or background, segmentation, quantification, and data analysis. The most challenging step is usually partitioning or segmenting an image into multiple parts or regions. This could involve separating signal from background noise, or clustering regions of voxels based on similarities in color or shape to detect and label voxels of a 3D volume that represents islet vasculature, for example. Once segmented, statistics such as object volume sizes are typically straightforward to extract. Provided is a method for the quantification and extraction of the imaging data, such as segmentation and data visualization. Particular attention is paid to the removal of autofluorescence in human islets and distinction between islet vasculature and islet capsule forming recipient cells.
The Regional Ethics Committee in Lund, Sweden, approved the study according to the Act Concerning the Ethical Review of Research Involving Humans. Animal experiments were performed in strict accordance with the Swedish ethics of animal experiments and approved by the ethics committees of Malmö and Lund. 6 to 8-week-old immunodeficient NOD.(Cg)-Gt(ROSA)26Sortm4–Rag2-/- (NOD.ROSA-tomato.Rag2-/-) recipient mice were used as recipients for transplantation of human islets10.
1. Islet preparation for transplantation
2. Preparation of transplantation equipment and surgery table
NOTE: All surgical tools should be autoclaved, and the surgery table and instruments disinfected with 70% alcohol.
3. Anesthesia and positioning of recipient mice for surgery
NOTE: All animals were bred and maintained in a pathogen-free environment at the animal facilities at Lund University.
4. Transplantation procedure
NOTE: This method has been previously described for the transplantation of mouse islets6. A slightly modified procedure is presented here.
5. Imaging of implanted human islets by 2-photon microscopy
NOTE: Taking overview images of the eye using a fluorescence stereoscopic microscope (Figure 2a–c) 4–5 days after transplantation prior to 2-photon imaging is recommended to localize the islets of interest. Avoid restraining the eye too tightly this early after transplantation. Use 2-photon imaging 6–7 days posttransplantation.
6. Imaging of implanted human islets by confocal microscopy
NOTE: The total volume, morphology, and plasticity of transplanted islets can be assessed by monitoring the in vivo scattering signal in a separate scan (i.e., separate track) by detection of laser backscatter light10.
7. Image analysis
NOTE: Commercial software (see Table of Materials) was used for this step.
Non-labeled human islets were transplanted into the ACE of 8-week-old female NOD.(Cg)-Gt(ROSA)26Sortm4–Rag2-/-(NOD.ROSA-tomato.Rag2−/−) recipient mice. To prevent human tissue rejection, immunodeficient Rag2 knockout mice were chosen as recipients. In these transgenic mice, all cells and tissues expressed a membrane-targeted tomato fluorescence protein (mT) that allows clear identification of the recipient and the donor tissue. Repeated imaging of islet grafts by high-resolution 2-photon microscopy could identify mT+ recipient cells involved in the engraftment process and their dynamic migration pattern.
In general, the transplantation setup of human islets into the ACE of the recipient mice (Figure 1a) was similar to syngeneic mouse islet transplantations with regard to islet shape and size (Figure 1b,c). Figure 1d,e shows a typical transplantation with a detailed schematic showing the lateral incision site (Figure 1f) and dispersion of islets into the eye chamber (Figure 1g) and a representative outcome of injected mouse islets (Figure 1h) or human islets (Figure 1i). The grafting rate of human islets was generally lower (~30%) than that of mouse islets (50–70%). A possible reason for this discrepancy is that the availability of human islets is unpredictable, usually with only a 1 day notice, and also that the obtained islets vary in donor age, donor BMI, purity (45%–86%), and postisolation culture time (2–5 days). In contrast, islet isolations from mice can be easily planned. These were isolated from 6–8-week-old (i.e., young adult) mice with high purity and short, overnight culture times. These discrepancies between mouse and human islet grafts should be considered when interpreting the results.
In vivo fluorescence imaging of human islet grafts was compromised by significant interference from tissue autofluorescence (Figure 4). The visualization of the revascularization process of ACE-transplanted human islet grafts by imaging agents using traditional wavelengths in the visible spectrum was hindered by a low signal-to-noise ratio and high level of natural islet autofluorescence, illustrated by images of a λ-scan in the spectral range of 500–700 nm (Figure 4a) or by the sensitive nondescanned PMT detector with filter for detection of dextran TR (610–675 nm) (Figure 4b). This high background was not observed in mouse islet grafts (Figure 4c). The NIR emitting imaging agent Angiosense 680 showed a visibly higher signal-to-background ratio due to decreases in background noise in the NIR range 690–730 nm (Figure 4d, Figure 3a). The additional recording of an “unused” channel (i.e., 500–550 nm) could be used in a later image processing step to remove remaining background noise caused by natural islet autofluorescence. The 2-photon/confocal imaging setup (Figure 2d–f) is similar to previously described ones6, save for the use of 900 nm 2-photon excitation and fluorescence detection of Angiosense 680, autofluorescence, and tomato channels with nondescanned PMTs 1-3 (Figure 2d). It is advisable to measure the laser output power reaching the sample for each microscope setup (Figure 2e). Further, it is recommended to initially image the grafted islets with stereomicroscopy (Figure 2a–c), which will help select islets of interest and avoid 2-photon microscopy images of an unsuccessful transplantation.
The purpose of extracting quantitative data from many images was to remove potential bias in data selection and to obtain the statistical power necessary to detect a legitimate effect when comparing experiments. Quantification from fluorescent imaging of pancreatic islets involved several steps, each of which may have influenced the results in another. Here, interactive imaging software was used. It contains features allowing visualization of volume images and objects and identification of objects according to their morphology or intensity. Figure 3 illustrates the different steps in image segmentation. The removal of autofluorescence by subtraction of the “autofluorescence” channel improved the signal-to-noise ratio (Figure 3a,b). The islet boundary called “islet mask” (Figure 3c) and corresponding channels (Figure 3d) was defined manually. The segmentation of the tomato signal into the islet vasculature and islet capsule (Figure 3f,g) and final surface rendering (Figure 3e,h,i) was used to extract quantitative data.
Figure 5 shows a representative longitudinal imaging session of the same human islet graft at 2 weeks, 2 months, 5 months, and 8 months posttransplantation into the ACE of a NOD.ROSA-tomato.Rag2-/- recipient mouse. Illustrated are maximum intensity projections (MIP) images of originally recorded RAW data (Figure 5a), processed images after islet autofluorescence removal (Figure 5b–e, f–i), and segmented islet objects including capsule (Figure 5j,m) or islet vasculature forming mT+ recipient cells (red, Figure 5n–q) and total islet vasculature (green, Figure 5n–q).
Figure 1: Transplantation of pancreatic islets into the anterior chamber of the eye.
(a) Transplantation setup showing anesthetized mouse fixed in stereotaxic head holder and the exposed eye (inset) next to the prepared Hamilton syringe fixed to the table and the stereomicroscope ready to pick mouse islets (b) or human islets (c). (d) Waiting position of the eye cannula loaded with islets. (e) Transplantation in process, and (f) schematic drawing of a single lateral incision used to carefully lift up the cornea with the tip of the eye cannula and dispense islets into the anterior chamber (g). Image of the eye immediately after injection of mouse islets (h) or human islets (i). Scale bar = 500 µm. Please click here to view a larger version of this figure.
Figure 2: Imaging of ACE-implanted pancreatic islet grafts.
(a) Experimental fluorescent stereo microscope imaging setup. Widefield (b) or fluorescent image (c) of B6.ROSA-tomato islet grafts. (d) Simplified scheme of the emission light path. LBF: Laser-blocking filter (main beam splitter); LP: Long pass; BP: Band pass; PMT: Photomultiplier. (e) Laser output power depends strongly on the wavelength. Diagram shows actual laser output power in Watts (W) related to the power level of the laser beam (%), measured for the microscope setup. Circle: 800 nm, square: 900 nm, triangle: 1,000 nm. (f) Experimental imaging setup showing the recipient mouse with ACE-grafted islets (insert) mounted onto the motorized stage a commercial microscope. Scale bar = 100 µm Please click here to view a larger version of this figure.
Figure 3: Image segmentation of a human islet grafted in the ACE of a NOD.ROSA-tomato.Rag2-/- recipient mouse.
(a) Original recording: Shown is a 3D rendering of an image stack of a human islet with merged channels or optical sections of split channels Angiosense 680 (ch 1), autofluorescence (ch 2), and tomato (ch 3). (b) 3D rendering of image stack with merged channels or optical sections with split channels “Vasculature” (ch 4) and “Tomato(all)” (ch 5) after autofluorescence removal. (c) Islet mask (yellow). (d) 3D rendering of image stack with merged or split channels “islet vasculature” (ch 6, white), “islet tomato” (ch 7, red). (e) Surface object “islet vasculature” created from ch 6. (f) 3D rendering of image stack with merged channels “islet vasculature” (ch 6, green) and “islet tomato vasculature” (ch 8, red). (g) 3D rendering of “tomato capsule” (ch 9) after channel subtraction ch 7–ch 8 or optical section with merged channels “tomato capsule” (ch 9, white), islet tomato vasculature (ch 8, red), and islet vasculature (ch 6, green). (h) Surface rendering of tomato capsule (created from ch 9) and islet tomato vasculature (created from ch 8). (i) Islet segmentation from human islet backscatter signal showing the selection of “Region of interest”(left) and the segmented surface object (middle) compared to the channel intensity (right). (j) Retrieval of data from the statistics tab. Scale bar = 50 µm. Please click here to view a larger version of this figure.
Figure 4: Autofluorescence of human islets.
Human islets (a,b,d) or B6 mouse islets (c) were transplanted into the ACE of NOD.Rag2-/- or B6 recipient mice and injected with dextran TR (a–c) or Angiosense 680 imaging agent (d) for the visualization of blood vessels. (a) λ-scan of human islet graft at the range from 500–700 nm with 2-photon excitation at 900 nm. Maximum image projection (MIP) image of human (b) or mouse islet graft (c) excited at 900 nm and detected with the TR filter (610–675 nm) of the NDD. (d) MIP of a human islet graft excited at 900 nm and detection with the Angiosense 680 filter (690–730 nm) of the NDD. Scale bar = 50 µm. Please click here to view a larger version of this figure.
Figure 5: Transplanted human islets are progressively revascularized and encapsulated by mT expressing cells of recipient origin.
Human islets transplanted into the anterior eye chamber of NOD. ROSA-tomato.Rag2−/− recipient mice, were imaged repeatedly for up to 8 months. Angiosense 680 was injected for the visualization of vasculature. (a) Maximum intensity projections (MIP) of original recorded RAW data of split (vasculature, autofluorescence, mT) or merged channels and backscatter light signal at 5 months posttransplantation. MIPS of total vasculature alone (b–e) or merged with membrane-targeted tomato fluorescence (mT) (f–i) after islet autoflourescence removal (green, a). 3D renderings of segmented islet tomato capsule (j–m) and islet tomato vasculature (red) or total islet vasculature (green) (n–q) at indicated time points posttransplantation. Scale bar = 50 µm. Please click here to view a larger version of this figure.
A method is presented to study the human pancreatic islet cell grafting process by observing the involvement of recipient and donor tissue. After a minimal invasive surgery implanting human islets into the anterior chamber of an immunodeficient mouse eye, the mouse recovers quickly within minutes after surgery. The procedure is performed on one eye. Generally, from 5–7 days postimplantation onwards the cornea is sufficiently healed to perform intravital imaging.
In this protocol, the quality of the human islet grafts is critical. Human islet quality and thereby transplantation outcome can vary depending on the donor's age, BMI, and isolation process, as well as time of islet culture prior to and after arrival. Human islets processed from organ donor pancreases approved for clinical transplantation and research purposes were used with consent of the donors. They were obtained after a process of pancreas digestion and islet purification described elsewhere12. Similar to isolated murine islets, these islets undergo a number of cellular assaults such as ischemia, mechanical stress, loss of basement proteins, and partial disruption of intra-islet endothelial cells (ECs) during the enzymatic digestion step14,15. Despite constantly improving culture conditions and recovery of large numbers of high-quality human islets, the time to transplantation remains a critical factor. During the first days of culture, human islets show a significant loss of intra-islet ECs and upon six days of culture only small endothelial structures in the islet core can be detected16. This loss of intra-islet endothelial cells could constitute a critical factor in the reduced grafting of human islets, because donor islet ECs are important players in the revascularization process17.
Maintaining sterility during the transplantation procedure is critical to avoid eye infections in the immunocompromised NOD.ROSA-tomato.Rag2-/- mice. Typically, transplantation procedures are performed under clean conditions using gloves, lab coat, head cover, and mouth mask, but not inside a biosafety cabinet. All used solutions are sterile filtered, and syringes, cannula, tubing, and gauze are rinsed in 70% ethanol. Wake-up cages are autoclaved. While full sterility cannot be guaranteed because of manual handling of the mouse during the procedure, islet contamination or eye infections have not been issues with this procedure.
Stable and adequate anesthesia is an important and critical factor for reducing eye movements and drift during low framerate data acquisition, and effectiveness can vary between different anesthetic agents18. Under light, general isoflurane anesthesia, the eye occasionally moves in a slow rolling fashion and increased depth of anesthesia (from 1–2%) can generally reduce eye movements. The clear advantage of using inhalable anesthesia is its quick effect and recovery time. It should be pointed out, however, that the use of isoflurane may alter insulin secretion19.
Image analysis and segmentation, or the partitioning of an image into multiple regions, are challenging and subject to potential bias in data selection20. Segmentation aims to identify objects like the “islet vasculature” for quantification. Here, methods based on intensity-threshold were applied to select regions (i.e., “absolute intensity”) or intensity differences to find edges (i.e., “background subtraction”). Currently, there are no universal solutions for segmentation in fluorescent microscopy. One approach is to experiment with commercial software that supports a range of methods. If thresholding fails because of background intensity variation, then small changes in image capture settings may improve segmentation results.
A limitation of the method lies in the fact that the human graft is more rapidly replaced by recipient cells and, in fact, completely reconstituted by mouse recipient endothelial cells. This would preclude studies of human cell interactions if the aim is to study adoptively transferred human immune cells migrating into the islet parenchyma via interactions with human endothelial cells. However, both in mouse and human islet grafts, similar revascularization occurs from the recipient and follows the different anatomical 3D plan specific for the species.
The success rates of beta cell replacement therapy have continued to improve. Still, various challenges in evaluating the efficiency of islet grafting and survival in vivo remain unresolved due to the lack of appropriate intravital imaging technologies. The anterior eye chamber is a useful transplantation site, supporting the repeated and long-term in vivo imaging of islet cells for the study of islet morphology, vascularization patterns, beta cell function, and beta cell death at a cellular resolution.
The protocol for studying the grafting process of human islet transplants in the ACE transplantation site reported here allows for longitudinal monitoring of human islets grafts with considerably higher resolution than that obtained with other alternative longitudinal platforms such MRI21,22, PET23, SPECT24, or Bioluminescence25.
The authors have nothing to disclose.
This study was supported by the Swedish Research Council, Strategic Research Area Exodiab, Dnr 2009-1039, the Swedish Foundation for Strategic Research Dnr IRC15-0067 to LUDC-IRC, the Royal Physiographic Society in Lund, Diabetesförbundet and Barndiabetesförbundet.
Anasthesia machine, e.g. Anaesthesia Unit U-400 | Agnthos | 8323001 | used for isofluran anasthesia during surgery and imaging |
-induction chamber 1.4 L | Agnthos | 8329002 | connect via tubing to U-400 |
-gas routing switch | Agnthos | 8433005 | connect via tubing to U-400 |
AngioSense 680 EX | Percin Elmer | NEV10054EX | imaging agent for injection, used to image blood vessels in human islet grafts |
Aspirator tubes assemblies | Sigma | A5177-5EA | connect with pulled capillary pipettes for manual islet picking |
Buprenorphine (Temgesic) 0.3mg/ml | Schering-Plough Europé | 64022 | fluid, for pain relief |
Capillary pipettes | VWR | 321242C | used together with Aspirator tubes assemblies |
Dextran-Texas Red (TR), 70kDa | Invitrogen | D1830 | imaging agent for injection |
Eye cannula, blunt end , 25 G | BVI Visitec/BD | BD585107 | custom made from Tapered Hydrode lineator [Blumenthal], dimensions: 0.5 x 22mm (25G x 7/8in) (45⁰), tip tapered to 30 G (0.3mm) |
Eye gel | Novartis | Viscotears, contains Carbomer 2 mg/g | |
Hamilton syringe 0.5 ml, Model 1750 TPLT | Hamilton | 81242 | Plunger type gas-tight syringe for islet injection |
Head holder | |||
-Head holding adapter | Narishige | SG-4N-S | assemled onto metal plate |
-gas mask | Narishige | GM-4-S | |
-UST-2 Solid Universal Joint | Narishige | UST-2 | assemled onto metal plate |
-custom made metal plate for head-holder assembly | |||
-Dumont #5, straight | Agnthos | 0207-5TI-PS or 0208-5-PS | attached to UST-2 (custom made) |
Heating pad, custom made | taped to the stereotaxic platform | ||
Human islet culture media | |||
-CMRL 1066 | ICN Biomedicals | cell culture media for human islets | |
-HEPES | GIBCO BRL | ||
-L-glutamin | GIBCO BRL | ||
-Gentamycin | GIBCO BRL | ||
-Fungizone | GIBCO BRL | ||
-Ciproxfloxacin | Bayer healthcare AG | ||
-Nicotinamide | Sigma | ||
Image analysis software | Bitplane | Imaris 9 | |
Image Aquisition software | Zeiss | ZEN 2010 | |
Infrared lamp | VWR | 1010364937 | used to keep animals warm in the wake-up cage |
Isoflurane Isoflo | Abott Scandinavia/Apotek | fluid, for anesthesia | |
Needle 25 G (0.5 x 16mm), orange | BD | 10442204 | used as scalpel |
Petri dishes, 90mm | VWR | 391-0440 | |
2-Photon/confocal microscope | |||
-LSM7 MP upright microscope | Zeiss | ||
-Ti:Sapphire laser Tsunami | Spectra-Physics, Mai Tai | ||
-long distance water-dipping lens 20x/NA1.0 | Zeiss | ||
-ET710/40m (Angiosense 680) | Chroma | 288003 | |
-ET645/65m-2p (TR) | Chroma | NC528423 | |
-ET525/50 (GFP) | Chroma | ||
-ET610/75 (tomato) | Chroma | ||
-main beam splitter T680lpxxr | Chroma | T680lpxxr | Dichroic mirror to transmit 690 nm and above and reflect 440 to 650 nm size 25.5 x 36 x 1 mm |
Polythene tubing (0.38mm ID, 1.09 mm OD) | Smiths Medical Danmark | 800/100/120 | to connect with Hamilton syringe and eye canula |
Stereomicroscope | Nikon | Model SMZ645, for islet picking | |
Stereomicroscope (Flourescence) | for islet graft imaging | ||
-AZ100 Multizoom | Nikon | wide field and long distance | |
-AZ Plan Apo 1x | Nikon | ||
-AZ Plan Apo 4x | Nikon | ||
-AZ-FL Epiflourescence with C-LHGFI HG lamp | Nikon | ||
-HG Manual New Intensilight | Nikon | ||
-Epi-FL Filter Block TEXAS RED | Nikon | contains EX540-580, DM595 and BA600-660 | |
-Epi-FL Filter Block G-2A | Nikon | (EX510-560, DM575 and BA590) | |
-Epi-FL Filter Block B-2A | Nikon | (EX450-490, DM505 and BA520) | |
-DS-Fi1 Colour Digital Camera (5MP) | Nikon | ||
Syringe 1-ml, Omnitix | Braun | 9161406V | for Buprenorphine injection, used with 27 G needle |
Surgical tape | 3M |