In vivo Bioluminescent Imaging of Mammary Tumors Using IVIS Spectrum



Mammary tumor cells expressing luciferase are implanted subcutaneously in mice and visualized using optical imaging to monitor tumor growth and development non-invasively in a longitudinal study.

Cite this Article

Copy Citation | Download Citations

Lim, E., Modi, K. D., Kim, J. In vivo Bioluminescent Imaging of Mammary Tumors Using IVIS Spectrum. J. Vis. Exp. (26), e1210, doi:10.3791/1210 (2009).


4T1 mouse mammary tumor cells can be implanted sub-cutaneously in nu/nu mice to form palpable tumors in 15 to 20 days. This xenograft tumor model system is valuable for the pre-clinical in vivo evaluation of putative antitumor compounds.

The 4T1 cell line has been engineered to constitutively express the firefly luciferase gene (luc2). When mice carrying 4T1-luc2 tumors are injected with Luciferin the tumors emit a visual light signal that can be monitored using a sensitive optical imaging system like the IVIS Spectrum. The photon flux from the tumor is proportional to the number of light emitting cells and the signal can be measured to monitor tumor growth and development. IVIS is calibrated to enable absolute quantitation of the bioluminescent signal and longitudinal studies can be performed over many months and over several orders of signal magnitude without compromising the quantitative result.

Tumor growth can be monitored for several days by bioluminescence before the tumor size becomes palpable or measurable by traditional physical means. This rapid monitoring can provide insight into early events in tumor development or lead to shorter experimental procedures.

Tumor cell death and necrosis due to hypoxia or drug treatment is indicated early by a reduction in the bioluminescent signal. This cell death might not be accompanied by a reduction in tumor size as measured by physical means. The ability to see early events in tumor necrosis has significant impact on the selection and development of therapeutic agents.

Quantitative imaging of tumor growth using IVIS provides precise quantitation and accelerates the experimental process to generate results.


A wide variety of luciferase expressing cancer cell lines can be used for pre-clinical research in mouse models.

These cells are provided as a pathogen-free frozen culture, which will readily grow in standard media with no need for selection markers.

For our experiment, we’ll use the 4T1-luc2 murine mammary tumor cell line, which expresses the luciferase gene that serves as an optical indicator of gene expression or tumorgenesis in vivo. We will use luciferase expression to track growth of the primary tumor non-invasively, but they can also be used to locate and monitor metastatic lesions.

Luciferase activity must be verified before injection, and in order to do this, a 90% confluent flask is harvested by trypsinization and then counted.

50,000 cells are then dispensed in a single well of a microtiter plate and serial dilutions are performed. Luciferin is added to the wells at 150ug per ml and incubated for 2 minutes. The microplate can be imaged in IVIS or a luminescent plate reader to determine expression levels. This cell line expresses up to 6500 photons per second per cell, but any expression level above 500 photons per second per cell can be imaged successfully in vivo.

Now that we have cells of optimal activity, we can proceed to the subcutaneous injection step.

In order to facilitate optimal detection of the tumor, we are using an athymic immunocompromised nude mouse strain. Prior to injection, animals are anesthetized to effect for deep anesthesia.

Next we will inject up to 250,000 cells in 100ul PBS are injected subcutaneously into the flank. Load the cells in a 1 ml syringe and attach a 26 gauge needle. Lift the skin gently with forceps to make a “tent” and inject the cells at the base. The newly injected cells can be imaged immediately. In each imaging session a total of 150mg of Luciferin per kg body weight is then administered via two injections into the peritoneal cavity.

In this study, animals are imaged 10 minutes after Luciferin injection to ensure consistent photon flux. We’ll show you this in the next step. Luciferin kinetics can vary from day to day. In this particular model, measuring 10 minutes after Luciferin injection gave a result within a range of 15% variability.

For our experiment we’ll use the IVIS Spectrum in vivo imaging system uses a back-thinned charge coupled device cooled to -90°C to achieve maximum sensitivity. To support absolute quantitation, the system measures dark charge during down-time and runs a self-calibration during initialization. To start imaging, initialize the IVIS system (one click) and set the imaging parameters for the experiment.

Select field of view for the number of animals being imaged. Up to 5 animals can be maintained in the instrument using the integral anesthetic manifold. The stage is at a constant 37°C to maintain body temperature in the animals. An EKG port is provided to monitor animal health during stressful procedures.

In Living Image software, exposure time, f-stop and pixel binning can be optimized based on the expression level of the cell line. These settings can be changed at any time during an experiment without impacting the quantitative result.

IVIS acquires a photographic image of the animal under white light and a quantitative bioluminescent or fluorescent signal, which is overlaid on the image.

The bioluminescent signal is expressed in photons per second and displayed as an intensity map. The image display is adjusted to provide optimal contrast and resolution in the image without affecting quantitation.

Luminescence from the cells can be measured at the site of injection using a region of interest tool. Measurement data are displayed in the table together with all experimental parameters relating to the image capture which can be saved or exported for analysis

Multiple images can be acquired and compared in longitudinal studies covering seconds or months depending on the nature of the experiment. We will measure photon flux from the tumor at time zero and monitor for 4 weeks with imaging at bi-weekly intervals.

Photon flux from the tumor is proportional to the number of live cells expressing luciferase so bioluminescence correlates directly with tumor size.

At 5 days post implantation the tumor is not yet palpable but the cells can be quantified through bioluminescence and the tumor is seen to be actively growing. At this stage the bioluminescence signal is much stronger. The exposure time, f-stop and pixel binning can be adjusted so that the image is clear and the camera does not saturate. IVIS automatically compensates for the changes in light collection so these measurements can be compared to those collected earlier and later in the experiment.

At 7 days post implantation tumors are palpable for the first time and bioluminescence measurement has already generated 7 days of data.

At 28 days post implantation, tumors are becoming necrotic and cells begin to die. Tumor size estimated by caliper measurement does not change appreciably, but luminescence from the tumor will decrease indicating cell death.

Caliper measurements and bioluminescence measurement can be continued until a humane endpoint is reached. Tumor necrosis due to hypoxia or treatment regimes will be indicated by reduced bioluminescence even if they do not reduce the tumor mass.


I would like to assure that all experiments performed to produce the Journal of Visualized Experiments article were performed in compliance with the Animal Protocol 060 approved by the Institutional Animal Care and Use Committee (IACUC) at Caliper Life Sciences, Alameda CA” Jae Kim, Ph.D., IACUC Chair, Caliper Life Sciences.



  1. I  can't see any vedio, why?

    Posted by: Jack M.
    April 30, 2009 - 8:57 AM
  2. Jack, sorry you are having problems.  Please check that you have Flash installed.  If you still have problems, please send me an email at and I'll help you out.

    Posted by: Anonymous
    April 30, 2009 - 10:37 AM
  3. 1.why did you give ² injections of the ip luciferin?
    ².Do you have correlation in the caliper measurements compared to the BLI measurements?

    Posted by: Tomi O.
    June 22, 2009 - 7:18 AM
  4. Hi Tomi,

    1) Two ip injections of luciferin are given in case of a missed injection, although a partial dose may not give you a satisfactory image (in which case you will need to re-image the animal). A single injection of luciferin can be administered and the animal may be conscious or unconscious. Select a method and use it consistently throughout your study to avoid any variability.

    ²) Yes, there is correlation between bioluminescence and caliper measurements in the 4T1-luc² subcutaneous model.

    Posted by: Ed L.
    June 26, 2009 - 5:29 PM
  5. Microparticles can be stained fluorescently or a labeled compound in a microparticle can be traced in vivo as long as either is sufficiently bright and the signal dŒsn&#x²019;t interfere with that from the labeled cells. I don&#x²019;t know what type of microparticle you are using but there are commercially available staining kits depending on the type.
    Yes, you can inject non-labeled cells subcutaneously and visualize the resulting tumor with a fluorescently-labeled antibody.

    Posted by: Ed L.
    January 8, 2010 - 7:46 PM
  6. Is it possible to trace microparticles loaded with a drug in vivo ? If positive how should I label them? Also would it be possible to visualize in the way you showed in video antibodies directed to non-labeled tumoral cells?



    Posted by: Helio T.
    August 19, 2009 - 2:18 PM
  7. How can I use this methods to quantify apoptosis?

    Posted by: Silvio M.
    December 9, 2009 - 1:05 AM
  8. Miranda,

    Here is a nice publication covering apoptosis:
    Clinical Cancer Research ²008;14(8) April 15, ²008

    Posted by: Ed L.
    January 8, 2010 - 7:48 PM
  9. How can I use this methods to quantify apoptosis?

    Posted by: Silvio M.
    December 9, 2009 - 1:06 AM
  10. New in this. I have a question:
    The signal intensity change over time, in order to give a good pic. I guess people use different settings at different time point. The picture thus are not comparable? How could I take the pictures over time under different settings but still comparable?

    Thanks a lot.


    Posted by: Anonymous
    December 28, 2009 - 12:35 PM
  11. As the tumor increases in size, the light emission will increase and camera settings may need to be adjusted so as not to get a saturated image (which cannot be accurately quantified). When performing a longitudinal study, it would be ideal to have all the images with the same scale and binning. You may not always be able to use the same scale for the entire study but you can still get a good overall picture of disease progression. If you select the same units throughout (ie &#x²01C;photons/second/cm²/sr&#x²01D;), the images and quantification will be comparable. In the case where different binning was used to capture the image, the Living Image software allows you to change the binning on any image after it was taken.

    Posted by: Ed L.
    January 8, 2010 - 7:50 PM
  12. So may I ask if you have any suggest Binning for all the models.

    Posted by: Chenpan N.
    May 10, 2019 - 1:35 AM
  13. Is there a reason to nu/nu mice vs. normal Balb/C mice ? 4T1 is a syngeneic xenograft model.

    Posted by: Anonymous
    January 27, 2010 - 2:00 PM
  14. Is there a reason to use nu/nu mice vs. normal Balb/C mice ? 4T1 is a syngeneic xenograft model.

    Posted by: Anonymous
    January 27, 2010 - 2:00 PM
  15. Nu/nu mice would be preferable if you are using fluorescence. Fur will auto-fluoresce and may mask your signal (you can shave or use a depilatory in this case). Otherwise, we could have also used a Balb/C mouse with the 4T1-luc² in the video.

    Posted by: Anonymous
    January 27, 2010 - 4:07 PM
  16. hello Is there any mice model for the research of glioma? Our lab has just begin the study in this area.We really need some help here.

    Posted by: Anonymous
    April 20, 2010 - 9:09 AM
  17. Hi Aaron,

    There is a very good video demonstrating this procedure
    You can use U87-MG-luc cells.

    Posted by: Anonymous
    September 21, 2010 - 3:14 PM
  18. In the video you've mentioned monitoring of metastatic lesions.
    What settings do I need to enter in order to see the lesions (in my case lymph node mets) with the primary tumor in one hole body shot?

    Posted by: Anonymous
    September 20, 2010 - 6:33 AM
  19. Hi Rachel,

    The primary tumor will be very bright while the lymph node metastases will be dim. You won't be able to image everything in a single, whole body image. In order to image the mets, you will need to shield the primary tumor and try imaging with large binning for 3 minutes as a start. Adjust your exposure time as needed and you can also change the FOV to bring the animal closer to the camera.

    Posted by: Anonymous
    September 21, 2010 - 1:01 PM
  20. Is is possible to use this method for imaging intra-abdominal metastases? Are there any specific tricks to doing so (other than placing the mouse on its back) and imaging for very dim signals?

    Posted by: Anonymous
    June 7, 2011 - 5:13 PM
  21. Dr. Holtz,
    Using the exact same method, you can image intra-abdominal metastases. You may want to use brighter cells to increase your chances to detect small metastases.

    Posted by: Anonymous
    June 7, 2011 - 6:08 PM
  22. Dear Mr. Lim,

    I'm new to the IVIS technology and I have some problems understanding the absolute quantitation of the obtained data. Prior to s/c injection of Luc²-cancer cells, cel expression levels were measured via dilution series in 96 well plate. I understand that after using the averaging tool one receives a certain value expressed in photons/sec/cell (for instance 700 p/s/cell). After injection of these cells, several bioluminescent pictures were taken in function of time. When ROI analyses is performed and data is displayed in Radiance mode, one obtaines a value in photons/sec/cm²/sr (for instance 7x10E7 p/s/cm²/sr) . How dŒs one now quantify the amount of cells ? In the example between brackets, would this be a cell number of 100000 cells in a tissue surface area of 1cm² ?

    Kind regards

    Posted by: Anonymous
    September 20, 2011 - 4:08 PM
  23. Hi Bram,

    The well plate values cannot be used directly to quantitate the in vivo image. Imaging the cells on the well plate gives you the light output in photons/sec and you can further calculate the photons/sec/cell to ensure that the cells are within the parameters in the product sheet. For in vivo imaging, a specified number of cells are injected into the animal and imaged. The light output is in units of photons/sec/cm^²/sr. This is the light that is detected at the surface of the animal per square centimeter per steradian. The bioluminescent signal will increase as the tumor increases in volume. Quantitating the in vivo signal needs to be approached with caution: photon intensity decreases with increase in tissue depth and optical properties of tissues may affect scattering/absorption of light (i.e. necrosis).


    Posted by: Anonymous
    October 12, 2011 - 7:39 PM
  24. what is the type of 96 -microwell plate do you use?what is the lotnumber of it ?

    Posted by: Anonymous
    October 8, 2011 - 10:31 PM
  25. Hi Lily,

    We use 96-well plates with black walls and clear bottom (polystyrene). The particular plate is Costar #3603.


    Posted by: Anonymous
    October 12, 2011 - 7:42 PM
  26. Hi,
    I have been trying the same procedure for LUC B16F10 cells..but I am not getting any signal in vitro... I tried changing the procedure for luminescence but not getting any results. i have tries seeding different number of cells in 1² well plate, though cannot see any signal. What could be the possible reason for the same? and can i just use PBS+luciferin for the imaging instead of the complete media?


    Posted by: Anonymous
    February 20, 2012 - 12:04 PM
  27. Hi Krishna,

    Are your B16F10 cells luc or luc²? The luc version has a light output of 15photons/second/cell whereas the luc² version is brighter at 450ph/sec/cell. In contrast, 4T1-luc² is about 6500ph/sec/cell. What is your exposure time when you image your cells? You may need to use a longer exposure time and a larger binning to see a signal. Hope this helps.


    Posted by: Anonymous
    March 6, 2012 - 3:14 PM
  28. Hi,

    our group is working on adoptive T cell therapy and is successfully using the IVIS ²00 for tracing T cells and measuring tumor size.
    But since I'm new I do have some questions no one could properly answer :-)

    We always inject the substrate (Luciferin od CŒlenterazine) i.v.
    Is there a difference to s.c. or i.p. injection?

    And I was wondering if one could also detect resting tumor cells in the bone marrow? Or is the luc² signal (of let say 1-6 cells) to weak to shine thru the bones?



    Posted by: Peter M.
    June 15, 2012 - 7:23 AM
  29. hi, i am in confusion and would like to know from you that,
    is bioluminescence signal in photos/sec/cm² just shown of tumors ? why it is not shown of other live cells of mouse other than tumor? is luciferin is selective to tumors or what would be the reasons?pls

    Posted by: arjun t.
    November 7, 2012 - 7:13 PM
  30. Could you share any data about group size, ie how much variability would I expect with subcutaneous tumors? Also, how does the presence of hair affect the measurement and reproducibility?

    Posted by: e w.
    June 24, 2014 - 2:19 AM

Post a Question / Comment / Request

You must be signed in to post a comment. Please or create an account.

Usage Statistics