Here we use bioluminescent, X-ray, and positron-emission tomography/computed tomography imaging to study how inhibiting mTOR activity impacts bone marrow-engrafted myeloma tumors in a xenograft model. This allows for physiologically relevant, non-invasive, and multimodal analyses of the anti-myeloma effect of therapies targeting bone marrow-engrafted myeloma tumors in vivo.
Multiple myeloma (MM) tumors engraft in the bone marrow (BM) and their survival and progression are dependent upon complex molecular and cellular interactions that exist within this microenvironment. Yet the BM microenvironment cannot be easily replicated in vitro, which potentially limits the physiologic relevance of many in vitro and ex vivo experimental models. These issues can be overcome by utilizing a xenograft model in which luciferase (LUC)-transfected 8226 MM cells will specifically engraft in the mouse skeleton. When these mice are given the appropriate substrate, D-luciferin, the effects of therapy on tumor growth and survival can be analyzed by measuring changes in the bioluminescent images (BLI) produced by the tumors in vivo. This BLI data combined with positronic-emission tomography/computational tomography (PET/CT) analysis using the metabolic marker 2-deoxy-2-(18F)fluoro-D-glucose (18F-FDG) is used to monitor changes in tumor metabolism over time. These imaging platforms allow for multiple noninvasive measurements within the tumor/BM microenvironment.
MM is an incurable disease made up of malignant plasma B-cells that infiltrate the BM and cause bone destruction, anemia, renal impairment, and infection. MM makes up 10% – 15% of all hematological malignancies1 and is the most frequent cancer to involve the skeleton2. The development of MM stems from the oncogenic transformation of long-lived plasma cells that are established in the germinal centers of lymphoid tissues before eventually homing to the BM3. The BM is characterized by highly heterogeneous niches; including diverse and critical cellular components, regions of low pO2 (hypoxia), extensive vascularization, complex extracellular matrices, and cytokine and growth factor networks, all of which contribute to MM tumorgenesis4. Thus, the development of a disseminated MM xenograft model characterized by tumors that are strictly engrafted in the BM would be a very powerful and clinically relevant tool to study MM pathology in vivo5,6. However, numerous technical hurdles can limit the effectiveness of most xenograft models, making them costly and difficult to apply. This includes problems associated with consistent and reproducible tumor engraftment within the BM niche, a prolonged time to tumor development, and limitations in the ability to directly observe and measure changes of tumor growth/survival without having to sacrifice mice during the course of the experiment7,8.
This protocol uses a modified xenograft model that was initially developed by Miyakawa et al.9, in which an intravenous (IV) challenge with myeloma cells results in "disseminated" tumors that consistently and reproducibly engraft in the BM of NOD/SCID/IL-2γ(null) (NOG) mice10. The in situ visualization of these tumors is achieved by the stable transfection of the 8226 human MM cell line with a LUC allele and serially measuring the changes in the BLI produced by these engrafted tumor cells6. Importantly, this model can be expanded to utilize various other LUC-expressing human MM cell lines (e.g., U266 and OPM2) with a similar propensity to specifically engraft in the skeleton of NOG mice. The identification of the tumors by bioluminescent imaging of the mice is followed by measuring the uptake of radiopharmaceutical probes (such as 18F-FDG) by PET/CT. Together, this allows for additional characterization of critical biochemical pathways (i.e., alterations in metabolism, changes in hypoxia, and the induction of apoptosis) within the tumor/BM microenvironment. The major strengths of this model can be highlighted by the availability of a wide range of radiolabeled, bioluminescent and fluorescent probes and markers that can be used to study MM progression and pathology in vivo.
All animal procedures described below were approved by the Institutional Animal Care and Use Committee (IACUC) of the Greater Los Angeles VA Healthcare system and were performed under sterile and pathogen-free conditions.
1. Preparation of Luciferase-expressing 8226 Cells (8226-LUC)
2. Orthotopic Xenograft Model
3. Measurement of BLI In Vivo
Note: Typically, measurable BLI from engrafted tumors in mice can be observed first between 10 – 20 days postchallenge, but this may need to be experimentally determined.
4. Treatment of the Mice with Targeted Therapy
5. PET/CT Analysis Using 18F-FDG to Measure Changes in Tumor Metabolism
In initial pilot studies, IV injections of 8226-LUC cells into NOD/SCID mice did not develop BM-engrafted MM tumors, although squamous MM tumors were easily formed (100% success rate). In contrast, IV challenges with 8226 cells into NOG mice generated (within 15 – 25 days) tumors in the skeleton (and only rarely formed tumors in non-skeletal tissue, such as the liver or spleen). Since tumors in the skeleton could not be visually confirmed by physically examining the animals, other methods had to be considered to confirm successful tumor engraftment and the location of the tumors within the BM (Figure 1). Similar to the reports by Miyakawa et al.9, human IgG, produced by 8226 cells, could be detected in the serum of challenged NOG mice using ELISA (data not shown), although these data only confirmed engraftment and not the location or extent of the tumors. Serial imaging of multiple animals shows the distribution of BM-engrafted MM tumors (Figure 4). The BLI produced by these engrafted tumor cells can then be serially and non-invasively measured to assess changes in tumor growth and survival in mice treated with the mTOR inhibitor temsirolimus (Figure 5). Finally, PET/CT analysis for the uptake of 18F-FDG in the tumors was used to demonstrate temsirolimus-mediated changes in glucose metabolism and that this correlated with changes in tumor growth (Figure 6). However, by stably transfecting 8226 cells with a luciferase allele, in conjunction with an optical imaging/X-ray analysis, the exact location and distribution of MM tumors in the mouse skeleton could be quickly and non-invasively determined.
Figure 1: Confirmation of the engraftment of 8226-LUC cells in the mouse skeleton. (A) NOG mice were challenged with 5 x 106 8226-LUC2 cells IV (mouse on the left) or with saline (mouse on the right). After 15 days, the mice were given an IP injection of D-luciferin substrate and imaged using a small-animal imaging system. (B) The mice were sacrificed, the femurs were collected, and the BM cellular exudate was harvested. The expression of human CD45 was determined by flow cytometry using a PE-conjugated anti-human CD45 antibody. (C) This panel shows an immunohistochemistry of serial sections of mouse femurs stained for hypoxia using pimonidazole (top panels) or human CD45 (bottom panels) in mice challenged with 8226 (right panels) or saline (control; left panels). The asterisk indicates the location of a large blood vessel in the serial sections. The scale bar = 100 µm. Please click here to view a larger version of this figure.
Figure 2: Anesthetized NOG mice showing positioning in the imaging system prior to measuring the BLI signal from bone marrow-engrafted MM tumors. Please click here to view a larger version of this figure.
Figure 3: Set-up of the optical imaging and X-ray analysis of the bioluminescent signal collected from NOG mice. Once the animals have been successfully challenged and the engraftment of tumors in the BM has been confirmed, the mice can be randomized into treatment groups. At various times during the course of the experiment, the animals can be monitored for changes in the BLI. Because the procedure is noninvasive, the frequency of monitoring could be altered to reflect the expected growth of the tumors, as well as how quickly the cancer cells will respond to therapy. Please click here to view a larger version of this figure.
Figure 4: Serial imaging of multiple animals showing the distribution of bone marrow-engrafted MM tumors. The regions of interest (ROI) for each mouse are identified. Please click here to view a larger version of this figure.
Figure 5: Changes in the BLI of engrafted tumors in NOG mice during an anti-MM therapeutic regimen. (A) This panel shows a change in the average radiance (p/s/cm2/ser) in temsirolimus (20 mg/kg mouse)-treated NOG mice challenged IV with 5 x 106 cells. Once the mice were observed to be positive for engrafted tumors, they were randomized into various treatment groups. The mice were given five IP injections daily, followed by 2 days of rest and, then, an additional five IP injections of temsirolimus (the bars on the X-axis indicate the days of treatment) or vehicle control. On various days, the mice were given IP injections of "in vivo-grade" D-luciferin, and the BLI was measured using an optical imaging platform. The values represent the average radiance (p/s/cm2/ser) ± a 95% confidence interval. (B) This Kaplan-Meier plot shows the change in percentage of the surviving mice over time. Mice that had reached the endpoint criteria (i.e., a loss of > 10% body weight or hind-limb paralysis) were humanely euthanized. (C) This panel shows representative images of mice taken on days 28, 35, and 40, showing changes in the LUC activity. Please click here to view a larger version of this figure.
Figure 6: Serial imaging and relative changes in the 18F-FDG uptake by PET/CT. (A) This panel shows serial imaging of the same mouse for bioluminescence (optical imaging/X-ray) imaging and 18F-FDG uptake by PET/CT. BLI was measured in anesthetized mice, and the PET/CT imaging of the same mouse was performed 24 h later. The mouse fasted overnight and, on the day of the study, received an IV injection of 50 µCi 18F-FDG. The mice were maintained under anesthesia during the probe uptake incubation (60 min) and, then, were analyzed for 18F-FDG uptake by PET/CT analysis. Tumors (T1 and T2) are indicated.H = heart, K = kidney, and B = bladder. (B) This panel shows relative changes (in a percentage of untreated control tumors) in the 18F-FDG uptake in control mice (no treatment) or temsirolimus (20 mg/kg mouse)-treated mice (measured at days 22, 35, and 40). The data is presented as the mean (n = 5 mice) ± 1 S.D. Please click here to view a larger version of this figure.
Despite a variety of preclinical xenograft models of MM6,9,11,12,13, the ability to study the tumor/BM interactions within the BM microenvironment remains difficult14. The techniques described here allow for the rapid and reproducible engraftment of 8226-LUC tumors cells in the skeleton of NOG mice.
The critical steps in this protocol involved the establishment of luciferase-expressing MM cell lines and the verification of a stable expression of luciferase in vitro15. Once the cell lines have been established, NOG mice are challenged by IV injection, and the engraftment of tumors to the skeleton is confirmed by measuring the BLI activity in vivo (typically within 10 – 20 days postchallenge) (Figure 1A). The use of NOG mice is important because other immunodeficient strains (such as NOD/SCID) do not typically form BM-engrafted tumors. The relatively high success rate of implantation (> 90%) and the short interval to observe a positive BLI signal makes this model ideal for a high-throughput and longitudinal analysis of anti-MM therapeutic modalities (Figure 4). Another very important aspect of this model is that the BM-engrafted MM cell lines maintain their morphological and immunologic features of the parent cell lines (e.g.,8226, U266); they have consistent and reproducible growth patterns and exhibit characteristics of patient diseases, such as increasing serum human IgG paraprotein levels and the formation of bone lesions.
The advantage of this strategy is the utilization of these powerful imaging techniques to repeatedly study biochemical and molecular components of the tumor/BM microenvironment over the course of the disease progression and in response to anti-tumor therapies. In this experiment, we found that targeting mTOR activity in mice with BM-engrafted myeloma tumors resulted in a decrease in bioluminescent measurements that could be observed over time. Furthermore, we found that changes in the uptake of the 18F-FDG probe (Figure 6B) in these same tumors were correlated to changes in the bioluminescent signal (Figure 5B). Another critical component of this procedure is that multiple biochemical variables can be measured within the tumor over time. This is especially important because tumor progression is a dynamic process characterized by changes in tumor physiology and microenvironment characteristics. Thus, this procedure, because of its non-invasive nature and the relatively short half-life of the probes, allows for multiple longitudinal analyses of changes in the tumor response to therapy.
After mastering the technique, the future applications of the technique present a great deal of flexibility. For example, the use of other bioluminescent or fluorescent tags (e.g., fluorescently tagged antibodies) could also be utilized, as could various other 18F-labeled probes and radiopharmaceutical compounds to study specific biochemical pathways or processes in the tumor/BM microenvironment. Because of the short half-life of many of these radionucleotides, multiple serial measurements could be made during the course of a specific experiment to study drug uptake, distribution, kinetics, and decay. Using this xenograft model, the ability of MM to utilize anaerobic glycolysis and other metabolic pathways (i.e., fatty acid synthesis) by using other probes (e.g.,18F-fluoro-2deoxyglucose [18F-FDG] and 18F-fluoro-4-thia-palmitate [FTP])16,17, as well as the ability to measure the anti-proliferative effects of treatment using another widely used probe (i.e., 3'-deoxy-3'-[18F]fluorothymidine [18F-FLT]) can be included in the experimental designs. In addition, it may be possible to directly measure cell death by using 18F-Annexin B1 to measure apoptosis in vivo18. Many of these compounds are commercially available.
The authors have nothing to disclose.
This work was supported by a VA MERIT grant1I01BX001532 from the United States Department of Veterans Affairs Biomedical Laboratory Research and Development Service (BLRDS) to P.F., and E.C. acknowledges support from the VA Clinical Science R&D Service (Merit Award I01CX001388) and VA Rehabilitation R&D Service (Merit Award I01RX002604). Further support came from a UCLA Faculty Seed Grant to J.K. These contents do not necessarily represent the views of the US Department of Veterans Affairs or the US Government.
8226 human myeloma cell line | ATCC | CCL-155 | |
NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ Mice (NOG) | Jackson Labs | 5557 | |
VivoGlo Luciferin substrate | Promega | P1041 | |
Hypoxyprobe-1Kit | HPL | HP1-100 | |
PE-CD45 (clone H130) | BD Biosciences | 555483 | Used for flow cytometry to identify human CD45+ tumor cells in BM exudate |
rabbit anti-human CD45 (clone D3F8Q) | Cell Signaling Technology | 70527 | Primary antibody used for Immunohistochemistry of excised bone |
Goat Anti-rabbit IgG (HRP conjugated) | ABCAM | ab205718 | Seconday antibody used for Immunohistochemistry of excised bone |
Dual-Luciferase Reporter Assay System | Promega | E1910 | |
pGL4.5 Luciferase Reporter Vector | Promega | E1310 | |
IVIS Lumina XRMS In Vivo Imaging System | Perkin Elmer | ||
Sofie G8 PET/CT Imaging System | Perkin Elmer |