Detection of Lung Tumor Progression in Mice by Ultrasound Imaging

Summary

This protocol describes the steps taken to induce KRAS lung tumors in mice as well as the quantification of formed tumors by ultrasound imaging. Small tumors are visualized in early timepoints as B-lines. At later timepoints, relative tumor volume measurements are achieved by the measurement tool in the ultrasound software.

Abstract

With ~1.6 million victims per year, lung cancer contributes tremendously to the worldwide burden of cancer. Lung cancer is partly driven by genetic alterations in oncogenes such as the KRAS oncogene, which constitutes ~25% of lung cancer cases. The difficulty in therapeutically targeting KRAS-driven lung cancer partly stems from having poor models that can mimic the progression of the disease in the lab. We describe a method that permits the relative quantification of primary KRAS lung tumors in a Cre-inducible LSL-KRAS G12D mouse model via ultrasound imaging. This method relies on brightness (B)-mode acquisition of the lung parenchyma. Tumors that are initially formed in this model are visualized as B-lines and can be quantified by counting the number of B-lines present in the acquired images. These would represent the relative tumor number formed on the surface of the mouse lung. As the formed tumors develop with time, they are perceived as deep clefts within the lung parenchyma. Since the circumference of the formed tumor is well-defined, calculating the relative tumor volume is achieved by measuring the length and width of the tumor and applying them in the formula used for tumor caliper measurements. Ultrasound imaging is a non-invasive, fast and user-friendly technique that is often used for tumor quantifications in mice. Although artifacts may appear when obtaining ultrasound images, it has been shown that this imaging technique is more advantageous for tumor quantifications in mice compared to other imaging techniques such as computed tomography (CT) imaging and bioluminescence imaging (BLI). Researchers can investigate novel therapeutic targets using this technique by comparing lung tumor initiation and progression between different groups of mice.

Introduction

As the leading cause of cancer-related deaths worldwide, lung cancer remains refractory to treatments, mainly due to lack of relevant pre-clinical models that can recapitulate the disease in the lab¹. Around 25% of lung cancer cases are due to mutations in the KRAS oncogene². KRAS-driven lung cancer is often associated with poor prognosis and low response to therapy, highlighting the importance of further studies in this disease².

We optimized a method that allows the relative evaluation of lung tumor growth in real time in KRAS lung cancer-induced immune-competent mice. We use Lox-Stop-Lox KRAS G12D (LSL-KRAS G12D) mice in which the KRAS G12D oncogene can be expressed by Cre lentiviral vectors^3,4. These vectors are driven by carbonic anhydrase 2, allowing the viral infection to take place specifically in alveolar epithelial cells⁵. In addition, to accelerate the initiation and progression of lung tumors, the lentiviral construct also expresses P53 shRNA from an U6/H1 promoter (the lentiviral construct herein will be referred to as Ca2Cre-shp53)⁶. The biological relevance of this method lies in the natural course of lung tumor development in mice as opposed to xenografts of non-orthotopic tumors in mice. An obstacle using the orthotopic method is monitoring lung tumor growth without sacrificing the mouse. To overcome this limitation, we optimized ultrasound imaging to permit the analysis of lung tumor progression in two-dimensional (2D) mode in this mouse model. Initiating tumors at 7 weeks post-infection are reflected as B-lines in ultrasound images, which can be counted, but will not reflect the exact number of tumors present on the lung. B-lines are characterized by laser-like vertical white lines arising from the pleural line in the lung parenchyma^7,8. Large tumors can be visualized after 18 weeks of infection. The relative volume of these tumors is quantified by 2D measurements done on ultrasound.

This method is optimal for researchers investigating the effect of pharmacological drugs on lung tumor growth in the LSL-KRAS G12D mouse model. In addition, lung tumor progression can be compared between mice with different genetic lineages, to examine the importance of the presence or absence of certain genes/proteins on the development of lung tumor volume.

Protocol

Animal studies were performed in accordance with the Institutional Animal Care and Use Committee (IACUC) of McGill University and procedures were approved by the Animal Welfare Committee of McGill University (animal use protocol # 2009-5754).

1. Generation of CA2Cre-shp53 Lentiviral Titre

NOTE: The following protocol is the same as that described in Xia et al.⁶, with minor modifications.

1. Preparation of lentivirus (for 15 cm x 10 cm dishes)
   1. On day 1, plate healthy HEK293T cells (7.5 x 10⁶ cells per 10 cm dish) with 10 mL of Dulbecco's modified Eagle medium) (DMEM), 10% fetal bovine serum (FBS), and 1% pen/strep. Culture in a 37 °C, 5% CO₂ incubator.
   2. Prepare the mix for calcium-phosphate transfection (mixture for 15 plates). Prepare tube A containing 225 µg (15 µg/plate) of the lentiviral vector (Ca2Cre-shp53), 75 µg (5 µg/plate) of PsPAX2 plasmid (containing HIV-1 gag and HIV-1 pol genes), 75 µg (5 µg/plate) of pMD2.G plasmid (containing VSV-G gene) for packaging, a final concentration of 0.15 M of CaCl₂ and fill the tube with distilled H₂O up to 3.75 mL. Prepare tube B containing 3.75 mL of 2x HEPES-buffered saline (HBS; 50 mM HEPES, pH 7.05, 280 mM NaCl, 10 mM KCl, 1.5 mM Na₂HPO₄·2H₂O, and 12 mM D-dextrose).
   3. Vortex tube A and add it dropwise to tube B under continuous vortexing.
      NOTE: The total volume will be 7.5 mL.
   4. Incubate at room temperature in the dark for 20−30 min.
   5. Approximately 9 h after plating the cells, add 500 µL of the transfection mixture dropwise into the cell medium (7.5 mL/15 dishes: 0.5 mL per dish).
   6. Swirl gently each dish to mix, and incubate all dishes at 37 °C, 5% CO₂.
   7. On day 2, 12−18 h after transfection, replace the media with 10 mL of antibiotic-free reduced serum media (Table of Materials) per 10 cm dish. Place dishes back in the 37 °C, 5% CO₂ incubator.
   8. On day 3, collect the media containing lentiviruses expressing Ca2Cre-shp53 and filter through 0.45 μm filters. Replenish the media with fresh antibiotic-free reduced serum media.
      NOTE: The collected media can be kept at 4 °C for no more than 3 days.
   9. On day 4, collect, for a second time, the media containing the lentivirus and filter through a 0.45 µm filter. Keep at 4 °C (for no more than 3 days).
   10. Combine the virus supernatants from steps 1.1.8 and 1.1.9. Concentrate them through centrifugal filter units (Table of Materials) by centrifuging at 1,372 x g for 30 min at 4 °C. Repeat the process until all the collected media has passed through the columns.
       NOTE: After each centrifugation, 100−200 μL of concentrated filtrate can be harvested.
   11. Collect and mix the concentrated filtrates in a 15 mL ice-cold tube. Mix the concentrated lentiviruses well and then aliquot (e.g., 100 μL/ tube). Store at -80 °C.
2. Lentiviral titration
   NOTE: Immortalized mouse embryonic fibroblasts (MEFs) expressing a loxP-flanked allele of green fluorescent protein (GFP) are used in this protocol for the quantification of viral titer. However, any cell line with a loxP-GFP allele should be suitable for this step.
   1. Culture cells expressing a loxP-flanked allele of GFP with DMEM, 10% FBS and 1% pen/strep at 37 °C, 5% CO₂.
   2. Plate 2 x 10⁵ cells in two 50 mm wells of a 6-well plate.
      NOTE: The cells of one well will be used for lentiviral infection while that of the other will be used as a negative control.
   3. The next day, replenish the cells with 2 mL of DMEM, 10% FBS, and 1% pen/strep 2 h before lentiviral infection.
   4. Add 20 µL of the CA2Cre-shp53 lentivirus (the volume may vary when needed) in the lentiviral infection well.
   5. After 3 days in culture, determine the frequency of the Cre-induced GFP-positive cells by flow cytometry, as shown in Figure 1. Wash cells with phosphate-buffered saline (PBS), detach by trypsinization and collect cells by centrifugation at 112 x g.
   6. Wash the cells twice with PBS. Finally, suspend the cells in 100 µL of PBS. Determine the percentage of GFP positive cells by a cell analyzer (Table of Materials).
   7. Calculate the lentivirus infectious/functional titer using the following formula:
      
      where F is the frequency of GFP-positive cells and calculated by subtracting the frequency of GFP -positive cells after infection (Fi) from the frequency of GFP-positive cells (background) prior to infection (Fc) as shown in Figure 1 (F = Fi-Fc), Cn is the number of plated cells (2 x 10⁵), V is the volume of the inoculum (mL), and DF is the virus dilution factor.
      NOTE: The infectious/functional concentration = ~2 x 10⁶ TU/mL.

2. Intratracheal Intubation of Lentiviruses in LSL-KRAS^G12D Mice

NOTE: The method of intratracheal intubation was used as described in the published protocol by Vandivort et al.⁹. In this protocol, mice LSL-KRAS^G12D mice in C57BL/6 background are used at age 6−8 weeks. A home-made working procedure board is used as described in Vandivort et al.⁹. The board is positioned in front of the experimenter in a convenient workspace (approximately 1 m²).

1. Prepare a spirometer by removing the plunger of a 1 mL syringe and loading 60 µL of PBS into the syringe.
2. Attach a 22 G catheter tip into the syringe and set aside.
3. Anaesthetize the mice by intraperitoneal injection of a 1 µL/g of mouse body weight of ketamine (50 mg/mL)/xylazine (5 mg/mL)/acepromazine (1 mg/mL) cocktail. Ensure proper sedation of the mouse by a low respiratory rate (1 breath every 2 s).
4. Aspirate 20 µL of CA2Cre-shp53 lentivirus into a pipettor and set aside.
5. Position the sedated mouse on the working procedure board by hooking its upper incisors into the thread of the board.
   NOTE: The mouse's dorsum should be flat against the platform.
6. Tape the caudal portion of the thoracic cavity to the platform to ensure alignment of the mouse during the procedure.
7. Adjust a gooseneck illuminator between 80−100% intensity and place the light 1−2 cm from the skin surface.
8. From behind the platform, draw the tongue out of the oral cavity of the mouse using sterile forceps.
9. While securing the tongue, insert a depressor into the oral cavity of the mouse then release the tongue.
10. Position the gooseneck on the main stem bronchi to illuminate trachea.
    NOTE: The trachea may be visible through the action of respiration, causing fluctuation of the light.
11. When the trachea is clearly viewed, insert the spirometer prepared (syringe with PBS and catheter) into the tracheal path.
12. Remove the depressor and observe the rise and fall of PBS in the syringe with each breath.
    NOTE: This is an indicator that the catheter is properly positioned into the trachea.
13. Remove the syringe containing PBS while maintaining the catheter inside the trachea as the previous position.
14. Deposit the 20 µL CA2Cre-shp53 lentivirus into the center of the catheter.
15. While keeping the catheter in place, inject 300 µL of air into the catheter using an empty syringe to ensure proper distribution of lentivirus in the lungs.
16. Keep the catheter in place and reinsert the spirometer into the catheter.
    NOTE: The rise and fall of the PBS bubble will ensure that the procedure is successful.
17. Remove the catheter and tape. Place the animal in a warm and dry place until it is revived.

3. Ultrasound Imaging of Lung Tumors in Mice

NOTE: Ultrasound imaging was performed after 7 and 18 weeks of lentiviral intubation using the system listed in Table of Materials; however, any model can be used for the analysis.

1. One day prior to imaging, remove fur from the chest area of the intubated mouse.
   NOTE: Mouse should be sedated during this step by placing them in an induction chamber of 3% isoflurane and 2L/minute O₂.
2. On the day of imaging, set up the workspace as shown in Figure 2. Turn on the heating pump for ultrasound gel and the temperature monitor.
3. Set up a warm 33 °C incubator to place the mice in post-imaging.
4. Place the three-dimensional (3D) motor (Figure 2F) on the integrated rail system.
5. Make sure that the 3D motor and the transducer mounting system are securely in place.
6. Connect a preferred transducer (frequency: 40 MHz; Figure 2E and Table of Materials) for tumor measurements perpendicular to the 3D motor.
7. Start a new study on the ultrasound software.
   1. Select Study Browser, then select New at the bottom of the screen.
   2. Select New Study, a new window will appear that enables the addition of a Study Name in addition to further information about the study, i.e., date of study, name of researcher, etc.
   3. Fill in the information in Series Name, i.e., Animal ID, Strain, Weight, Date of Birth, etc.
   4. Select Done, the program will change for B-mode.
8. Put a heating lamp in a convenient position above the animal platform.
9. Place the mouse in the induction chamber (3.5% isoflurane).
   NOTE: Proper anesthetization is confirmed by unconsciousness of the mouse, and slower respiratory rate of around 1 breath per 2 seconds.
10. When the mouse is sedated, change the connection of the anesthetic machine to be directed towards the animal platform, reduce isoflurane to 2.5%.
11. Place the mouse on the animal platform in decubitus ventral, with its oral cavity directed towards the anesthetics tube.
12. Apply lubricant on the eyes of the mouse.
13. Place the mouse in decubitus dorsal and tape its hands and feet firmly onto the animal platform.
14. Apply a small layer of ultrasound gel on the mouse chest.
15. Lower the acquisition probe using the height-control knob to touch the surface of the mouse chest. Position the probe such that the heart of the mouse is approximately centered.
16. Use the micro-knobs to acquire images of the whole chest, from both extremities, in the transverse orientation ideally gathering 500 frames per mouse (number of frames may vary depending on personal choice).
17. Once imaging is done, remove the gel from the chest of the mouse and place the mouse in the warm incubator.

4. 2D Analysis of Ultrasound Images

1. After opening acquired frames on the ultrasound software, scan the frames for tumors.
2. For small initiating tumors, count the number of B-lines periodically every 10 frames for the full length of the 500 frames acquired.
   Therefore, B-lines are counted in a total of 50 images, each image is separated by 10 frames. B-lines are characterized by longitudinal white straight lines fully traversing the screen.
3. For 2D measurements of large tumors, select the linear tool and measure the width and length of the tumor present.
4. To calculate the volume of the tumors, use the following formula:
   
   where L and W are the length and width of the tumor, respectively.

Representative Results

After obtaining a lentiviral infectious titer of ~2 x 10⁶ TU/mL (Figure 1), the Ca2Cre-shp53 lentivirus was intratracheally injected when LSL-KRAS G12D mice reached an appropriate age (6−8 weeks)⁹. Ultrasound imaging was performed after 7 weeks of infection upon initiation of tumors (Figure 3B). Imaging was done at 7 weeks in order to include the various types of precursor lesions that occur in the LSL-KRAS^G12D mouse model, ranging from hyperplasia to adenoma³ as shown in Figure 4A after 8 weeks of infection. Imaging earlier than 6 weeks will not ensure the inclusion of adenoma-forming tumors in the analysis³. These precursor lesions are visualized as B-lines in ultrasound and can be counted by eye since they can be identified as narrow beams of white light vertically traversing the screen and are the result of a strong reflection of ultrasound wave (Figure 3B)¹⁰. B-lines represent small tumors that are on the surface of the lungs; tumors being initiated in deeper structures of the lung such as close to the trachea will not be reflected as B-lines. The number of tumors being initiated on the pleural surface can be quantified by counting the B-lines, since they are too small for volume measurements. We acquired 500 frames spanning the full lung area in 5 mice. B-lines were counted every 10 frames; thus, they were counted in a total of 50 images per mouse that span the full 500 frames acquired. The scatter plot in Figure 3D shows the sum of B-lines counted in every 10 images (100 frames) for 5 mice. The number of B-lines per mouse is then summed up to include B-lines present in the full lung area. The mean quantity of B-lines representing tumors in the five mice was 141.6 ± 41.52 (Figure 3E). Figure 3A shows an ultrasound image acquired of a non-infected mouse lung, for comparison.

According to Jackson et al.³, 16 weeks post-intubation includes adenoma as well as adenocarcinoma, which are tumor types of interest in KRAS lung cancer. To ensure inclusion of these types of tumors when analyzing volume measurements via ultrasound in our model, we imaged the lungs of mice at 18 weeks post-infection. In order to quantify the volume of the large tumors via ultrasound, we used 2D analysis on the ultrasound software. Large tumors appear as deep clefts interrupting the pleural surface as shown in the image of Figure 3C. When a tumor is spotted, we measure the length (L) and width (W) of the tumor, using the measurement tool of the ultrasound software. Once obtained, L and W are applied in the conventional formula used for calculating tumor volume (protocol step 4.4). We analyzed the tumor volume of 10 mice, as shown in Figure 3F, where the number of tumors formed in each mouse ranged from 5 to 26 tumors. The volumes of all tumors formed per mouse are then summed up to represent the tumor volume per mouse. For this analysis, we assume that the tumors formed in the lungs of mice are epileptic in shape, thus the width of the tumor is considered as perpendicular to the imaging plane. Relative volume quantification can be done when linear size of tumors reaches 0.3 mm (either length or width). After analyzing the lung parenchyma of 10 mice, the average tumor volume obtained was 31.8 ± 6.61 mm³ (Figure 3G). Hematoxylin and Eosin (H&E) staining analysis on lung sections confirmed the formation of large tumors at 20 weeks post-infection (Figure 4B) and confirmed formation of adenoma and adenocarcinoma. It should be noted that as tumor growth progresses, different tumors might merge together to form large ones, as shown in Figure 4B. Although the ultrasound software permits 3D analysis of tumors, this type of analysis proved difficult in our model, mainly due to the high dynamic movement of the primary tumors formed that is consistent with lung sliding (movement of the pleural line with respiration) as well as cardiac beats¹¹. We reasoned that 2D analysis would provide consistency in relative tumor volume measurements, especially when comparing lung tumor progression between different groups of mice.

Figure 1: Flow cytometry analysis of MEFs with a GFP LoxP-allele. Representative flow cytometry images showing frequency of GFP-positive MEFs that are non-infected (Fc) (A) and infected (Fi) with the Ca2Cre-shp53 lentivirus (B). Please click here to view a larger version of this figure.

Figure 2: Workspace of integrated rail system of ultrasound imaging. (A) X-axis micro-knob. (B) Y-axis micro-knob. (C) Animal platform. (D) Height control knob of scan head-probe. (E) Transducer. (F) 3D motor. (G) Heating lamp. (H) Anesthetics tube. Please click here to view a larger version of this figure.

Figure 3: Representative ultrasound images of mouse lungs before and after lung tumor induction. (A) Image of a non-infected mouse lung section. Red arrows point to ribs (R) and the pleural surface (PS) of the lung. (B) Image of a mouse lung 7 weeks after tumor induction by intratracheal intubation of lentivirus. Arrows indicate B-lines reflecting small tumors on the lung surface. (C) Image of a mouse lung section 18 weeks after tumor induction. Blue lines indicate measurements of the length (L: mm) and width (W: mm) of the tumor shown. (D) Scatter plot showing B-line quantification in 10 images spanning 500 frames (each image is separated by 10 frames). (E) Quantification of relative tumor number in lungs at 7 weeks post-induction. (F) Scatter plot showing the relative volume of the different tumors formed in the lungs of 10 mice at 18 weeks post-infection. (G) Quantification of lung tumor volume in mice (n = 10). Please click here to view a larger version of this figure.

Figure 4: Representative H&E staining of LSL-KRAS G12D lung sections. (A) Lung section at 8 weeks post-infection at 500 µm magnification. Arrows point at precursor lesions formed at this timepoint, which are reflected as B-lines in ultrasound. Black arrows point at magnified lesions at 60 µm. (B) Lung section at 20 weeks post-infection at 2 mm magnification. Arrowheads of the same color show different tumors that have merged together. Arrows show magnified tumors at 60 µm. Please click here to view a larger version of this figure.

Discussion

We demonstrate a method that can assess lung tumor growth in the Cre-inducible LSL-KRAS G12D mouse model by ultrasound. This method can be used for evaluating the effect of pharmacological inhibitors on lung tumor growth. It can also be used to compare lung tumor growth between mice of different genetic backgrounds. Using this technique does not require specialized computational skills, however, it is important to be systematic in the number of frames used for analysis to allow for proper comparison if the method is used for comparing different groups of mice.

The intratracheal intubation of ~2 x 10⁶ TU/mL Ca2Cre-shp53 in LSL-KRAS G12D mice led to the formation of precursor lesions after 7 weeks of infection. These appear as longitudinal white lines on sonographic images and are called B-lines (Figure 3B)¹². The average number of B-lines per mouse was 141.6 ± 41.52 (Figure 3E). We chose to visualize initiating tumors at 7 weeks post-infection since this timepoint would include various types of precursor lesions; atypical adenomatous hyperplasia, epithelial hyperplasia and adenoma³. The formation of precursor lesions on the lung was verified by H&E staining after 8 weeks of lentiviral infection (Figure 4A). After 18 weeks of viral intubation, the tumors are large enough to allow 2D volume measurements in the ultrasound software (Figure 3C). Analysis of ultrasound images was done at the 18-week timepoint in order to include the quantification of well-developed adenoma and adenocarcinoma³. The formation of these types of tumors was confirmed at 20 weeks post-infection by H&E staining analysis (Figure 4B). 3D volume measurements are also possible by the ultrasound software but require high level of expertise. The titre described in this protocol is optimal for the formation of lung tumors that can be properly visualized and quantified in 2D.

Other imaging techniques can also be used to monitor lung tumor growth in mice, such as CT imaging and BLI¹³. However, CT imaging requires delivering a radiation dose to the tumor, which might affect tumor growth upon repeated assays¹³. Moreover, it requires the use of contrast agents for accurate tumor measurements¹³. BLI is an important imaging technique that is often used for quantifying tumor proliferation; however, its main limitation is its dependence on metabolic activities and the presence of ATP and O₂¹³. The advantage of ultrasound imaging lies in it being a non-invasive, non-irradiating, speedy and inexpensive technique for acquiring images of the lung parenchyma¹³. Ultrasound has been previously used to monitor growth of orthotopic human tumors xenografted in lung and pancreas of mice^13,14. Raes et al. have been successful in longitudinally monitoring the growth of a single human lung tumor implanted near the posterior diaphragmic surface in mice via ultrasound¹³. The difference in the LSL-KRAS G12D mouse model is having spontaneous development of lung tumors that can be monitored from initiation to progression and this development can be compared between mice groups. Also, there are numerous tumors being developed on the lung of the LSL-KRAS G12D mouse model, that eventually grow, and might even merge together to form large tumors (Figure 4B). This urged the need to use a method that can monitor lung tumor initiation and progression in this mouse model. We optimized the 2D brightness (B) mode analysis to allow quantitative assessment of lung tumors developed in this mouse model.

A possible limitation with ultrasound imaging is the presence of false positives¹⁵. These are seen when a highly reflective surface is present (diaphragm or liver) that interrupts the trajectory of the beam and might create a virtual object mimicking a true object¹⁵. Such images could be false tumors but can be bypassed upon proper examination; a large tumor is usually dynamic whereas a virtual object would be static.

Ultrasound imaging for the LSL-KRAS mouse model is ideal due to the non-invasive nature of this technique. The method of analysis used in our lab to monitor lung tumor progression in the LSL-KRAS mouse model is fast and user-friendly.

Materials

Name	Company	Catalog Number	Comments
0.45 μm Acrodisc Syringe Filters	Pall Corporation	PN 4614
100-mm Cell Cultre Plate	CELLSTAR	664 160
6-well Cell Culture Plate	CELLSTAR	657 160
Amicon Ultra - 15 Centrifugal Filter Units	Merck Millipore Ltd.	UFC910024
BD LSR-Fortessa	BD Biosciences	649225B 3024
CA2Cre-shp53 lentiviral vector			From Dr. I Verma Laboratory
DMEM	Multicell	319-005-CL
FBS	Multicell	80450
LSL-KRASG12D mouse	JAX Mice	8179
MX550S; Centre Transmit: 40 MHz	FUJIFILM VisualSonics	51070
OptiMEM	gibco	11058-021
Pen/strep	Multicell	450-201-EL
pMD2.G	Addgene	12259
PsPAX2	Addgene	12260
VEVO-3100	FUJIFILM VisualSonics	51072-50