We present a protocol to generate a minimally invasive orthotopic pancreatic cancer model by ultrasound-guided injection of human pancreatic cancer cells and the subsequent monitoring of tumor growth in vivo by ultrasound imaging.
Pancreatic cancer (PCa) represents one of the deadliest cancer types worldwide. The reasons for PCa malignancy mainly rely on its intrinsic malignant behavior and high resistance to therapeutic treatments. Indeed, despite many efforts, both standard chemotherapy and innovative target therapies have substantially failed when moved from preclinical evaluation to the clinical setting. In this scenario, the development of preclinical mouse models better mimicking in vivo characteristics of PCa is urgently needed to test newly developed drugs. The present protocol describes a method to generate a mouse model of PCa, represented by an orthotopic xenograft obtained by ultrasound-guided injection of human pancreatic tumor cells. Using such a reliable and minimally invasive protocol, we also provide evidence of in vivo engraftment and development of tumor masses, which can be monitored by ultrasound (US) imaging. A noteworthy aspect of the PCa model described here is the slow development of the tumor masses over time, which allows precise identification of the starting point for pharmacological treatments and better monitoring of the effects of therapeutic interventions. Moreover, the technique described here is an example of implementation of the 3Rs principles since it minimizes pain and suffering and directly improves the welfare of animals in research.
PCa, and its most common form, the Pancreatic Ductal Adeno Carcinoma (PDAC), is one of the most common causes of cancer-related death with a 1-year survival rate lower than 20% and a 5-year survival rate of 8%, regardless of the stage1,2. The disease is almost always fatal, and its incidence is forecasted to continuously grow in the next years, unlike other cancer types, whose incidence is declining3. Factors such as late cancer detection, the tendency of rapid progression, and lack of specific therapies lead to a poor prognosis of PCa4. Great advances in cancer research have been obtained, thanks to the development of more accurate preclinical mouse models. The models have provided appropriate insights to the understanding of the molecular mechanism underlying cancer and to the development of new treatments5. These advances poorly apply to PCa, which, despite great recent efforts, remains resistant to current chemotherapeutic therapies1. For these reasons, the development of novel approaches to improve patients' prospects is mandatory.
Over the years, many PCa mouse models have been developed, including xenografts, which are the most widely used models nowadays5. Xenograft models are classified as subcutaneous heterotopic and orthotopic, depending on the location of the implanted tumor cells. Subcutaneous heterotopic xenografts are easier and cheaper to accomplish but miss certain characteristic features of PCa (i.e., the peculiar tumor microenvironment, characterized by the accumulation of fibrotic tissue, hypoxia, acidity, and angiogenesis)6,7. This explains why subcutaneous xenografts often fail to provide robust data for therapeutic treatments leading to failures when translated to the clinical setting8. On the other hand, orthotopic xenografts resemble the tumor microenvironment more closely, leading to better mimicking of the natural development of the disease. In addition, orthotopic xenografts are more suitable for studying the metastatic process and the invasive features of PCa, which almost do not occur in subcutaneous models9. Overall, orthotopic xenograft mouse models are nowadays preferred to perform preclinical drug testing9,10. Orthotopic xenografts usually rely on surgical procedures to implant either cells or very small tumor tissue pieces into the pancreas. Indeed, several papers based on surgical models of PCa have been published in the last few decades11. However, the quality and the outcome of the surgical procedure for the establishment of an orthotopic tumor model strongly depend on the technical skill of the operator. Another key point for a successful orthotopic PCa xenograft for a translational clinical approach is the possibility to establish localized disease with predictable growth kinetics.
To address these issues, here we describe an innovative procedure to produce an orthotopic PCa xenograft, exploiting ultrasound (US)-guided injection of human PCa cells into the tail of the pancreas in immunodeficient mice. This procedure generates a reliable PCa mouse model. The tumor growth is followed in vivo by US imaging.
The present protocol received approval from the Italian Ministry of Health with the authorization number 843/2020-PR. In order to ensure aseptic conditions, the animals were maintained inside the barrier room of the research animal vivarium (Ce.S.A.L.) of the University of Florence. All procedures were performed in the same space where the mice were housed at the LIGeMA facility of the University of Florence (Italy).
1. Cell preparation
2. Mouse preparation for ultrasound-guided injection (US-GI)
NOTE: Following steps were performed under sterile conditions. The entire procedure of US-guided injection, from the beginning of the anesthesia until the mouse is removed from the animal platform, takes around 10-12 min plus 5 min for complete mouse recovery.
3. Injection of PANC1 cells in the pancreas by US-GI method
4. 3D US imaging for monitoring pancreatic tumors in mice
NOTE: The evaluation of tumor development was performed starting 8 days after the cell injection, using the same instrument used for the US-guided injection (listed in Table of Materials). Hence, some procedures, such as the system ignition (step 2.2.), anesthesia (steps 2.3. – 2.6.), and mouse placement on the animal platform (step 2.7.), fully match what was described above in the protocol.
Following the protocol described above, mice were first anesthetized in an isoflurane chamber, and placed on the animal platform (Figure 1A). The pancreas was visualized with ultrasound imaging (Figure 1B). A 50 µL Hamilton syringe was loaded with 1 x 106 PANC1 cells suspended in 20 µL of PBS and placed on the needle holder (Figure 1C). The optimal angle between the syringe and the US transducer was 45° (Figure 1D). Using the micromanipulator, the syringe needle was inserted into the pancreas and its trajectory was observed on the US imager display (Figure 1E). A 20 µL bolus of tumor cells was then injected into the pancreas (Figure 1F) and after 10 s the needle was retracted. Relevant actions to prevent the recoil of cells and their subsequent leakage into the peritoneal cavity were pursued, such as applying a constant pressure during the inoculation, pausing after cell injection, and using a thin needle with a 30° bevel angle. After the US-guided injection, mice's weight was monitored every day to assess any signs of stress due to the procedure. No significant change in the weight of injected mice was observed.
3D US imaging was then applied to monitor tumor cell engraftment and tumor development. The first image acquisition was performed on day 8 after cell injection, followed by weekly acquisitions (for a total of 6 acquisitions). During imaging, animals were placed on the right flank onto the warmed (37 °C) platform and anesthetized by inhalation of isoflurane gas (induction dose at 4% and maintenance dose at 2%) (Figure 2). The 3D acquisition was performed in B-mode using the 3D motor that allows the transducer to scan the abdomen in various sections, perpendicular to its axis. A series of 2D images were obtained, which were then assembled by the analysis software, reconstructing the 3D anatomical image of the organ (3D re-rendering). To perform 3D imaging, it was necessary to synchronize the mouse breath phases with the acquisition of the 3D scan. The reproducibility of the protocol was confirmed by the presence of the tumor mass, a hypo-echogenic structure (represented with dark colors) in the tail of the pancreas, starting at day 8, in 16 out of 20 (80%) animals (Figure 3A). In four mice (20% of animals), the development of tumor mass was observed 2 weeks after the injection. This latency could be traced back to a sub-optimal procedure of needle insertion into the pancreas, which caused leakage of cells into the peritoneum.
Figure 3B shows the mean tumor volume (n = 16) evaluated by US imaging of the US-guided injection mouse model. The day after the last US acquisition, mice were euthanized first by CO2 inhalation and then by cervical dislocation. The abdominal cavity was examined and the tumor masses were explanted for histologic analysis (Figure 3C–E).
Figure 1: US-guided method for the establishment of an orthotopic PCa mouse model. (A) The mouse is placed on the right flank on a special support heated at 37 °C. (B) The US transducer is located transversely to the animal body at the level of the spleen and the pancreas is visualized in B-Mode. The flank view of the mouse abdomen shows the pancreas (1) between the spleen (2) and the left kidney (3). Scale bar: 2 mm. (C) The Hamilton syringe is placed on the appropriate holder. (D) In this situation, the transducer (2) and the syringe (1) are positioned forming an angle of 45°. (E) The needle of the syringe (1) is inserted in the tail of the pancreas, just below the spleen. Scale bar: 2 mm. (F) A bolus containing 1 x 106 PANC1 cells in 20 µL of PBS is injected into the tail of the pancreas, using a Hamilton syringe with a 28 G needle. Scale bar: 2 mm. Please click here to view a larger version of this figure.
Figure 2: Imaging workstation used for monitoring pancreatic tumors in mice. (A) Workplace for US imaging. (1) 55 MHz transducer; (2) Mouse handling table; (3) 3D Motor; (4) Height control knob of US transducer. (B) The mouse is placed on its right flank onto the handling table with its snout in the anesthetic tube (1), the skin is stretched, and the US gel is applied on the right flank. (C) The transducer is lowered to touch the mouse skin and positioned transversely to the animal body. Please click here to view a larger version of this figure.
Figure 3: Development of orthotopic PCa xenograft in the mouse pancreas. (A) 2D US image acquisition of the tumor mass developed after echo-guided cell injection. The development of the tumor is monitored with US imaging after 8 days of cell injection, followed by once-a-week acquisition until day 44, for a total of 6 acquisitions. The tumor is represented by a hypo-echogenic structure in the pancreatic parenchyma. The hypo-echogenic structure is echographically represented by an area with echoes of reduced intensity, compared to the surrounding parenchyma or a neighboring structure. Scale bar: 2 mm. (B) Time course of the tumor volume of orthotopic PDAC derived from PANC1 cells. (C) US imaging acquisition of a tumor mass at day 43 after cell injection. Scale bar: 2 mm. 3D re-rendering of the tumor mass is shown in the inset. (D) After imaging, the animals are euthanized, and necropsy examination is performed. (1) Tumor mass; (2) A small portion of a healthy pancreas; (3) Spleen; (4) Stomach; (5) Liver. Scale bar: 5 mm. (E) Hematoxylin and eosin staining of the tumor mass at 4x magnification. Scale bar: 200 µm. On the right upper side of the panel, inset of the tumor mass in which pancreatic acinar cells are still visible; 40x magnification. Scale bar: 100 µm. Please click here to view a larger version of this figure.
Although the use of US imaging is widespread in the clinic, tumor development in many preclinical mouse models is usually described using bioluminescent imaging11. The latter is an indirect way to evaluate tumor engraftment and expansion and it also does not provide a reliable tumor growth kinetics. In the present study, we have applied US imaging for performing cell injection as well as for monitoring tumor development. The protocol we have described and the results we have provided represent a relevant breakthrough in PCa research.
The US-guided method described here for the injection of pancreatic tumor cells is minimally invasive and allows to overcome many drawbacks caused by the surgical procedure, which is usually applied to establish an orthotopic PCa mouse model. Although indications on mortality rate or side effects due to the surgical procedure are missing in the literature, our experience suggests that the surgical method produces higher postoperative mortality in a low percentage of cases (around 10%), with frequent animal suffering, requiring a certain amount of recovery time12. In addition, the stitches applied after surgery prevent the correct application of imaging procedures. This results in the acquisition of US images with very poor quality, leading to inaccurate data extrapolation caused by artifacts, such as reverberation and shadows regions. Taking all these facts into consideration, a method that does not need a surgical procedure for the establishment of an orthotopic mouse model is preferred and strongly recommended for preclinical research purposes. Furthermore, being minimally invasive, the recovery time from cell inoculation is faster and the animal suffering is minimal in the US-guided method. A great advantage is that the US-guided method is compatible with the use of isoflurane anesthesia which allows faster induction and recovery, relatively less sparing effect on cardiovascular function and cerebral blood flow autoregulation, compared to other methods of anesthesia (i.e., ketamine/xylazine solution). Overall, the negligible metabolism of isoflurane makes it particularly useful in anesthetic management13. Indeed, roughly 5 min after the end of the procedure, animals regain sufficient consciousness to maintain sternal recumbency. In addition, mice are pre-medicated with an analgesic (carprofen) to minimize pain, and no post-operative treatments are necessary due to the rapid and complete recovery.
Nevertheless, there are some crucial troubleshooting steps in the procedure that must be considered. The number of cells to be inoculated must not exceed 1 x 106 cells and the volume of cell inoculum must not be greater than 20 µL to avoid the risk of cell leakage out of the pancreas. Cell injection should be performed immediately after their detachment, to avoid a decrease in cell vitality. After cell injection, it is necessary to wait for at least 10 s before removing the needle to avoid the leakage of cells out of the pancreas.
One of the most critical aspects of the present protocol concerns the positioning of the mouse on the animal platform. The mouse must be lying on its right side onto the animal handling table and the skin must be stretched. If the skin is not properly stretched during the injection, the needle will struggle to pierce and enter the pancreas, with the risk of cells spilling out of the pancreas. The ultrasound-guided injection to produce a PCa mouse model was first described by Huynh et al. in 201114. Here, we focused on the methodology suitable to produce such a model, describing in detail all the steps to make it reproducible. Furthermore, compared to Huynh et al.14, we have introduced some innovations mainly related to the advancements in US technology that occurred in the last 5 years. In the present protocol, the use of a Hamilton syringe with a stiff needle and a bevel angle of 30° is strongly recommended to obtain better accuracy during cell injection and to minimize cell leakage. Finally, during the US-guided injection, mice were secured on its right side onto the animal platform, while in Huynh et al.14 mice were placed in dorsal recumbency. The side position of the mouse allows to visualize the tail of the pancreas, just below the spleen, which acts as a visual reference (Figure 1B) during the cell injection, ensuring the reproducibility of the technique.
As with other techniques, there are also some limitations with the US-guided procedure. The greatest limitation is that the injection can be performed only in the tail of the pancreas because the body and the head are covered by other organs and are difficult to reach with the needle. Furthermore, another limitation associated with the use of US imaging to monitor tumor growth is the inability to visualize the whole tumor mass when it becomes too large, after 6 weeks of inoculation (Figure 3C-E). Although, such large volumes do not mimic the clinical course of the disease.
Another advantage of the PCa model obtained with the protocol described here is the slow cell engraftment and growth of the tumor masses (Figure 3B). This is ideal for precise identification of the starting point for pharmacological intervention and better monitoring of the effects of therapeutic interventions over time.
Finally, the technique described is an example of implementation of the 3Rs principles involving the refinement of technique and development of procedure which minimizes pain, suffering, distress, or lasting harm, directly improving the welfare of the animals in research.
Future applications of the PCa model obtained by the US-guided injection include studies aimed at the development of appropriate therapeutic treatments or treatment combinations. In fact, the use of such innovative in vivo techniques could be coupled with the use of small molecules, e.g., antibodies, antibody fragments, and antibody-drug conjugates (ADC), with a theragnostic approach. The latter could be used alone, for therapeutic purposes, as our group has demonstrated in our recent work15 or after a proper fluorophore conjugation, with the goal of monitoring tumor growth.
The authors have nothing to disclose.
This work was supported by Associazione Italiana per la Ricerca sul Cancro (AIRC, grant no. 15627, IG 21510, and IG 19766) to AA, PRIN Italian Ministry of University and Research (MIUR). Leveraging basic knowledge of ion channel network in cancer for innovative therapeutic strategies (LIONESS) 20174TB8KW to AA, pHioniC: European Union's Horizon 2020 grant No 813834 to AA. CD was supported by a AIRC fellowship for Italy ID 24020.
100 mm Petri dish | Sarstedt, Germany | P5856 | |
3D-Mode package | Visualsonics Fujifilm, Italy | Includes the 3D Motor; necessary for volumetric imaging | |
Aquasonic 100, Sonypack 5 lt Ultrasound Transmission Gel | PARKER LABORATORIES, INC. | 150 | Gel for ultrasound |
Athymic Mice (Nude-Foxn1nu) | ENVIGO, Italy | 69 | 20 females, 8 weeks old, Athymic Nude-Foxn1nu, 20-22 g body weight |
CO2 Incubator Function Line | Heraeus Instruments, Germany | BB16-ICN2 | |
Display of ECG, Respiration Waveform and body temperature | Visualsonics Fujifilm, Italy | 11426 | |
DMEM (Dulbecco’s Modified Eagle Medium) | Euroclone Spa, Italy | ECM0101L | |
DPBS (Dulbecco’s Phosphate Buffered Saline) | Euroclone Spa, Italy | ECB4004L | |
Eppendorf (1.5mL) | Sarstedt, Germany | 72.690.001 | |
FBS (Fetal Bovine Serum) | Euroclone Spa, Italy | ECS0170L | |
Hamilton Needle Pointstyle 4, lenght 30 mm, 28 Gauge | Permax S.r.l., Italy | 7803-02 | |
Hamilton Syringe 705RM 50 µL | Permax S.r.l., Italy | 7637-01 | |
Isoflo (250 mL) | Ecuphar | 7081219 | |
L-glutamine 100X | Euroclone Spa, Italy | ECB3000D | |
Mouse Handling table II | Visualsonics Fujifilm, Italy | 50249 | |
MX550D: 55 MHz MX Series Transducer | Visualsonics Fujifilm, Italy | 51069 | Ultrasound Transducers |
Oxygen/isofluorane mixer | Angelo Franceschini S.r.l. | LFY-I-5A | |
PANC1 cell line | American Type Culture Collection (ATCC), USA | CRL-1469 | |
Rimadyl (carprofen) | Pfizer | 11319 | 20 mL, injection solution |
Trypsin-EDTA 1X in PBS | Euroclone Spa, Italy | ECB3052D | |
Vet ointment for eyes, Systane nighttime | Alcon | 509/28555-1 | |
Vevo Compact Dual Anesthesia System (Tabletop Version) | Visualsonics Fujifilm, Italy | VS-12055 | complete with gas chamber |
Vevo Imaging Station 2 | Visualsonics Fujifilm, Italy | VS-11983 | Imaging WorkStation 1 plus Imaging Station Extension with injection mount |
Vevo Lab | Visualsonics Fujifilm, Italy | VS-20034 | Data Analysis Software |
Vevo LAZR-X Photoacoustic Imaging System | Visualsonics Fujifilm, Italy | VS-20054 | Includes analytic software package for B-mode |
Vevo Photoacoustic Enclosure | Visualsonics Fujifilm, Italy | 53157 |