The ability to noninvasively quantitate tumor burden from conditional (Cre/loxP-dependent) mouse cancer models would greatly increase their range of useful applications. We now report the generation of a reporter mouse that enables visualization of spontaneous tumor development from pre-existing conditional mouse tumor models via in vivo bioluminescence imaging. We demonstrate that bioluminescence can be “switched-on” in a Cre-dependent manner in every organ analyzed, and that this gives rise to between a 4 and 6-log increase in light emission per mg of wet tissue weight. Furthermore, we highlight the utility of this reporter by showing that it can be used as a sensitive means to measure spontaneous Kras2v12-induced lung tumorigenesis in a pre-existing mouse model of non-small cell lung cancer. Taken together, our results suggest that this reporter may be combined with a wide-range of other Cre/loxP tumor mouse models, irrespective of their tissue specificity and render them immediately amenable to longitudinal monitoring of tumor growth and therapeutic response with a noninvasive in vivo imaging approach.

The recent widespread application of Cre/loxP technology has resulted in a new generation of conditional mouse cancer models that can better recapitulate many salient features of human tumorigenesis (1). To date these models have proven most useful for furthering our understanding of in vivo tumor etiology (2, 3). However, like other spontaneous tumor mouse models (4, 5, 6, 7), they also possess great potential to become an improved means with which to evaluate novel preclinical cancer treatments and to elucidate key genetic factors that underlie chemoresistance.

Despite comprising an improved tool for modeling cancer, the stochastic and often nonvisible nature of tumorigenesis associated with conditional tumor models has thus far limited their widespread application for such purposes. These characteristic features complicate attempts to correlate tumor burden with onset of treatment and conventional tumor latency-based estimates are prone to a high degree of error. Furthermore, as autopsy is often the most effective way to evaluate treatment efficacy, longitudinal studies examining the dynamics of tumor response and relapse to treatment become problematic. Consequently, large treatment cohorts of mice are required to generate statistically meaningful data.

In vivo bioluminescence imaging could be used to obviate these particular limitations of conditional tumor models, as it has been shown to comprise a reliable, sensitive, and rapid modality with which to noninvasively quantify tumor mass at deep-tissue sites in mice (8, 9, 10). This imaging modality is also well-suited for the purposes of preclinical therapy evaluation, because it has been extensively validated in multiple xenograft-based systems as an accurate means to noninvasively determine cell viability in vivo and to measure tumor cell kill in response to therapy (8).4 To image spontaneously arising tumor burden from existing Cre/loxP models via this approach, however, one needs a strategy to introduce tumor- or tissue-specific luciferase expression into the mouse model. To limit the need to generate, optimize, and validate novel independent transgenic luciferase mouse strains for each individual Cre/loxP model, we now report the generation of a ubiquitously expressing conditional luciferase reporter mouse that can be used to render a wide range of Cre/loxP mouse tumor models suitable for bioluminescence imaging.

Generation of the LucRep Mouse.

The LucRep transgenic mouse strain was generated by pronuclear microinjection of the LucRep transgene (depicted in Fig. 1) into fertilized FVB/N oocytes as described previously (11).

Genotyping of transgenic mice was determined by PCR. The primer sequences and PCR reaction conditions to detect Luciferase (3), Cre (10), and Kras2v12(12) alleles were as described previously.

Imaging Protocol.

Bioluminescence was measured noninvasively using the IVIS imaging system (Xenogen Corp., Alameda, CA). All of the images were taken 10 min after i.p. injection of luciferin (225 mg/kg−1; Xenogen Corp.) as a 60-s acquisition, binning 10, unless otherwise stated in the text. During image acquisition, mice were sedated continuously via inhalation of ∼3% isoflurane (Abbott Laboratories Ltd., Kent, United Kingdom). Ex vivo bioluminescence imaging of isolated organs was performed immediately after euthanasia of the animals by CO2, 10 min after i.p. injection of luciferin as described. Dissected organs were placed on a sheet of black plastic and imaged by IVIS; strong bioluminescent signals remained detectable >30 min after dissection. Image analysis and bioluminescent quantification was performed using Living Image software (Xenogen Corp.).

Induction of Lung Tumors.

Approximately 106 or 107 particles of AdCre5 were introduced directly into the lungs of compound LucRep/conditional Kras2v12 mice via intratracheal intubation as described previously (12).

Because tumorigenesis is initiated by either Cre-mediated “knock-out” or “knock-in” of tumor-associated gene function in conditional (Cre/loxP-dependent) cancer models, we reasoned that the generation of a generally expressed conditional luciferase transgene would give rise to tumor-specific light emission after introduction into pre-existing conditional tumor models. Accordingly, we generated a small series of transgenic founders via pronuclear microinjection of the conditional luciferase reporter (LucRep) construct as depicted in Fig. 1.

To determine the likely utility of these transgenic founders, we crossed each one with a ubiquitously expressing Cre transgenic line (ACM-Cre; Ref. 14) and measured the extent of bioluminescence from the resultant F1 generation by IVIS (Xenogen Corp.). This approach clearly demonstrated that one transgenic line, hereby denoted as the LucRep line, was capable of Cre-dependent luciferase expression (Fig. 2 A). Southern analysis on DNA derived from the tails of LucRep and LucRep/ACM-Cre compound mice showed that the nonrecombined LucRep allele had integrated as a long concatamer (>20 copies) at one locus, which was subsequently reduced to single copy after Cre mediated recombination (data not shown).

We next measured the difference in bioluminescence from individual “switched” and “nonswitched” tissues by sacrificing and dissecting three LucRep-only and three LucRep/ACM-Cre compound mice after administration of luciferin. The amount of light emitted from each exposed organ was then quantified by IVIS.

This experiment showed that bioluminescence could be “switched on” in a Cre-dependent manner in every LucRep organ analyzed and gave rise to between a 4 and 6-log increase in light emission per mg of wet tissue weight (Fig. 2 B). Furthermore, the extent of light emission from switched LucRep tissue was comparable with that of another bioluminescent mouse tumor model (Pomc05) that we have used previously to noninvasively determine in vivo tumor growth dynamics and response to therapy from a deep tissue site (10).

The amount of background luminescence arising from individual nonrecombined LucRep organs was essentially negligible in all of the cases except for the heart, liver, and spleen, which were seen to emit very low levels of light. The extent of this background luminescence is unlikely to significantly affect the sensitivity of most LucRep imaging applications, however, as light emission was only weakly detectable, mainly around the thorax of intact mice, at sensitive image acquisition settings (60 s, binning 10). We next crossed the LucRep mouse with a conditional oncogenic Kras2 transgenic mouse to demonstrate that the LucRep allele could be used to image spontaneous tumor development in vivo. We have shown previously that multiple NSCLC-like lung adenocarcinomas develop in conditional Kras2v12 mice after adenoviral-mediated delivery of Cre recombinase (AdCre) to the lungs (12). Accordingly, several compound LucRep/conditional Kras2v12 mice were imaged by IVIS at weekly intervals after the introduction of ∼106 AdCre particles into their lungs via intubation. The result of this experiment showed that the LucRep allele could enable noninvasive visualization of conditional Kras2v12-induced lung tumorigenesis by IVIS (Fig. 3,A). Direct imaging and histological processing of the lungs from this mouse additionally emphasized the sensitivity of this imaging approach, because the light detected from the mouse depicted in Fig. 3,A was shown to originate from a single lesion measuring between 1 and 2 mm in diameter (see Fig. 3, B and C).

Longitudinal tumor growth from several AdCre-infected LucRep/conditional Kras2v12 compound mice was also followed over time, after the introduction of ∼107 particles of AdCre to their lungs (Fig. 4, A and B). Measured bioluminescence arose slightly faster in this experimental cohort, and ex vivo analysis of the lungs indicated the presence of as many as 40 independent lesions of varying size. Ex vivo analysis of these lungs also indicated that the majority of lung tumors induced in this model should be suitable for bioluminescence imaging, because we observed only one macroscopically visible lesion from a total of 20 distinct lesions that did not glow.

The stochastic nature of spontaneous tumor onset associated with this mouse model of NSCLC is clearly illustrated by the graph depicted in Fig. 4 B and, thus, highlights the usefulness afforded by noninvasive quantification, especially for the purposes of standardizing tumor burden within experimental cohorts of mice. More animals will need to be included in this experiment before a definitive statement can be made regarding the individual growth kinetics of Kras2v12-induced lung adenocarcinomas. However, there is little doubt that the LucRep allele will facilitate this work in the future.

To date, with one exception (10), the advantages associated with noninvasive bioluminescence imaging of tumor burden in mice have been confined to xenograft-based models. The ability to noninvasively quantify tumor burden in living conditional tumor model mice, however, will ultimately lead to the development of more accurate models of human cancer that are better suited for evaluating and optimizing preclinical cancer therapy.

A similar conditional reporter transgene strategy to ours has been used previously, and because of the ubiquitous and conditional nature of reporter transgene expression, now comprises the most commonly used approach by the research community to test and validate novel Cre transgenic strategies (13). Because this reporter mouse is lacZ-based, however, its utility is effectively limited to studies on tissues postmortem. We now demonstrate that a ubiquitously expressed conditional luciferase transgene can be used as a sensitive means to noninvasively visualize cells that have undergone Cre-mediated recombination in living mice. We also show that this reporter can be used to noninvasively follow tumor growth over time via sequential measurement of tumor cell light emission, which illustrates that expression of the reporter is stable and that it can be used effectively to measure tumor burden from deep tissue sites like the lung. Furthermore, because of the ubiquitous nature of conditional luciferase transgene expression, we propose that the LucRep mouse will enable longitudinal analyses of spontaneous tumor development and response to therapy for a wide variety of conditional tumor models in the future.

A condition that must be met for such a purpose is that all of the conditional alleles present in a particular cell are switched, resulting in cells that carry one or more oncogenic lesions together with an activated luciferase reporter. This criterion can be met under two circumstances. First, when the luciferase reporter switches more efficiently than the conditional alleles that initiate tumor growth upon recombination. Second, when conditions are created in which all of the conditional alleles switch efficiently in the cell.

In the first situation, luciferase will also switch on in some cells in which no oncogenic lesions have been induced. Because these cells do not clonally expand, however, tumor growth can still be monitored effectively, as they will not provide an appreciable contribution to signal. We have not tested whether the LucRep reporter switches more effectively than the various conditional tumor suppressor and oncogene alleles. This issue is complicated by the fact that different alleles show large differences in Cre-mediated recombination (14), making the utility of the reporter dependent on the alleles with which it is combined. Switching efficiency is also dependent on the expression level of Cre recombinase in the cell. Therefore, in conditions where partial switching of conditional oncogenic lesions occur, the frequency of resultant tumors suitable for bioluminescence imaging will need to be determined empirically. We have found that these issues also apply to the Rosa26-LacZ reporter, which switches less well than several conditional alleles studied by us (14).

The second situation requires a relatively high level of Cre recombinase in the cell, which can be achieved using somatic gene transfer with vectors that deliver high transient expression of Cre recombinase. We have found that AdCre is particularly suited for this purpose, as the majority of infected cells appear to switch all of the floxed alleles (12). Our data are in agreement with this, because to date we have observed only one macroscopically visible tumor ex vivo without concomitant luciferase expression. This implies that in the context of our mouse model of NSCLC, >90% of Kras2v12-induced tumors should be amenable to noninvasive imaging via the LucRep allele. In more generalized terms, we believe that the LucRep reporter mouse will be most useful for imaging conditional tumorigenesis when Cre expression is introduced via the germ line in a tissue/cell type restricted manner or when introduced somatically via adenoviral or lentiviral expression vectors. It should be noted, however, that the extent of background bioluminescence may become significant when LucRep is used in conjunction with conditional tumor models with low tumor penetrance (i.e., in situations where many more normal cells switch relative to resultant tumorigenesis). If considered in terms of wet weight, a background of 1 mg switched normal tissue would be unlikely to affect the visualization of a 10-mg tumor, whereas 500 mg of switched normal tissue would. This issue did not apply to our model of NSCLC, because we switch only a very small fraction of the cells. As a result we observed no appreciable difference in bioluminescence before or several weeks after AdCre infection. Currently the LucRep reporter mouse lacks the spatial resolution of the Rosa26-LacZ reporter, because an antibody has yet to be developed against luciferase that can be used for immunohistochemical detection in paraffin-embedded tissue. One can envisage, however, that the future development of such an antibody will additionally increase the value of the LucRep mouse as a reporter, because it will then enable the detection of Cre recombinant cells in both living and fixed tissues.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

Supported by a grant of the Dutch Cancer Foundation (K. W. F.).

4

D. Jenkins et al. Bioluminescent imaging (BLI) to improve and refine traditional murine models of tumor growth and metastasis. Clin. Exp. Metastasis, in press, 2003.

5

The abbreviations used are: AdCre, adenovirus Cre; NSCLC, non-small cell lung carcinoma; Kras2, Kirsten-Ras.

Fig. 1.

Generation of the LucRep mouse. Schematic depicting the conditional transgene strategy used to ensure Cre-dependent luciferase transgene expression. The β-actin promoter drives ubiquitous transgene expression, but in the absence of Cre-mediated recombination, transcription terminates at the “floxed” polyadenylation acid sequence before the luciferase open reading frame. Cre-mediated recombination results in the specific deletion of the floxed polyadenylation sequence and consequently luciferase transgene expression, driven by the same β-actin promoter.

Fig. 1.

Generation of the LucRep mouse. Schematic depicting the conditional transgene strategy used to ensure Cre-dependent luciferase transgene expression. The β-actin promoter drives ubiquitous transgene expression, but in the absence of Cre-mediated recombination, transcription terminates at the “floxed” polyadenylation acid sequence before the luciferase open reading frame. Cre-mediated recombination results in the specific deletion of the floxed polyadenylation sequence and consequently luciferase transgene expression, driven by the same β-actin promoter.

Close modal
Fig. 2.

Characterization of the LucRep mouse. A, IVIS images of two transgenic LucRep littermates resulting from a LucRep/ACM-Cre cross. The mouse in the top panel is LucRep-positive/Cre-negative and does not luminesce, whereas the mouse in the bottom panel is LucRep-positive/Cre-positive and appears to luminesce brightly from all areas of the body (image taken 10 min after i.p. administration of luciferin; 1 s acquisition, binning 2). B, graph depicting the amount of light emitted from switched and nonswitched LucRep organs, expressed as relative light units per mg wet tissue weight per min. The Pomc05 bar on the far right of the graph corresponds to the amount of light emitted from another luminescent model that has been used previously by us to image spontaneous tumorigenesis (see Ref. 8).

Fig. 2.

Characterization of the LucRep mouse. A, IVIS images of two transgenic LucRep littermates resulting from a LucRep/ACM-Cre cross. The mouse in the top panel is LucRep-positive/Cre-negative and does not luminesce, whereas the mouse in the bottom panel is LucRep-positive/Cre-positive and appears to luminesce brightly from all areas of the body (image taken 10 min after i.p. administration of luciferin; 1 s acquisition, binning 2). B, graph depicting the amount of light emitted from switched and nonswitched LucRep organs, expressed as relative light units per mg wet tissue weight per min. The Pomc05 bar on the far right of the graph corresponds to the amount of light emitted from another luminescent model that has been used previously by us to image spontaneous tumorigenesis (see Ref. 8).

Close modal
Fig. 3.

Bioluminescence imaging of spontaneous lung tumorigenesis arising from LucRep/conditional Kras2v12 compound mice. A, IVIS image of a compound LucRep/conditional Kras2v12 mouse 13 weeks after AdCre intubation shows a bright focal region of luminescence originating from the thorax. (Image taken 10 min after i.p. administration of luciferin; 60 s acquisition, binning 10). B, IVIS image of the lungs dissected from the mouse depicted in A also shows a single origin of light. (Image taken ∼20 min after i.p. administration of luciferin; 60 s acquisition, binning 10). C, the same lungs after H&E processing (at ×2.5 and ×10 magnification) showing that the light detected in A and B originated from a single lesion measuring between 1 and 2 mm in diameter.

Fig. 3.

Bioluminescence imaging of spontaneous lung tumorigenesis arising from LucRep/conditional Kras2v12 compound mice. A, IVIS image of a compound LucRep/conditional Kras2v12 mouse 13 weeks after AdCre intubation shows a bright focal region of luminescence originating from the thorax. (Image taken 10 min after i.p. administration of luciferin; 60 s acquisition, binning 10). B, IVIS image of the lungs dissected from the mouse depicted in A also shows a single origin of light. (Image taken ∼20 min after i.p. administration of luciferin; 60 s acquisition, binning 10). C, the same lungs after H&E processing (at ×2.5 and ×10 magnification) showing that the light detected in A and B originated from a single lesion measuring between 1 and 2 mm in diameter.

Close modal
Fig. 4.

Longitudinal measurement of Kras2v12-induced lung tumor growth from individual mice. A, three sequential images taken at 2-week intervals of mouse 36948; a representative mouse developing multiple Kras2v12-induced lung tumors (images taken 10 min after i.p. administration of luciferin; 60 s acquisition, binning 10). The number of weeks annotating each individual image refers to the time post-AdCre infection. B, graph depicting the longitudinal measurements of Kras2v12-induced lung tumor growth from individual mice. Light emission is expressed as corrected relative light units (i.e., measured light post-AdCre infection minus measured light pre-AdCre infection).

Fig. 4.

Longitudinal measurement of Kras2v12-induced lung tumor growth from individual mice. A, three sequential images taken at 2-week intervals of mouse 36948; a representative mouse developing multiple Kras2v12-induced lung tumors (images taken 10 min after i.p. administration of luciferin; 60 s acquisition, binning 10). The number of weeks annotating each individual image refers to the time post-AdCre infection. B, graph depicting the longitudinal measurements of Kras2v12-induced lung tumor growth from individual mice. Light emission is expressed as corrected relative light units (i.e., measured light post-AdCre infection minus measured light pre-AdCre infection).

Close modal
1
Jonkers J., Berns A. Conditional mouse models of sporadic cancer.
Nat. Rev. Cancer
,
2
:
251
-265,  
2002
.
2
Jonkers J., Meuwissen R., van der Gulden H., Peterse H., van der Valk M., Berns A. Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model for breast cancer.
Nat. Genet.
,
29
:
418
-425,  
2001
.
3
Marino S., Vooijs M., van Der Gulden H., Jonkers J., Berns A. Induction of medulloblastomas in p53-null mutant mice by somatic inactivation of Rb in the external granular layer cells of the cerebellum.
Genes Dev.
,
14
:
994
-1004,  
2000
.
4
Roberts R. B., Min L., Washington M. K., Olsen S. J., Settle S. H., Coffey R. J., Threadgill D. W. Importance of epidermal growth factor receptor signaling in establishment of adenomas and maintenance of carcinomas during intestinal tumorigenesis.
Proc. Natl. Acad. Sci. USA
,
99
:
1521
-1526,  
2002
.
5
Bergers G., Javaherian K., Lo K. M., Folkman J., Hanahan D. Effects of angiogenesis inhibitors on multistage carcinogenesis in mice.
Science (Wash. DC)
,
284
:
808
-812,  
1999
.
6
Schmitt C. A., Rosenthal C. T., Lowe S. W. Genetic analysis of chemoresistance in primary murine lymphomas.
Nat. Med.
,
6
:
1029
-1035,  
2000
.
7
Schmitt C. A., Fridman J. S., Yang M., Lee S., Baranov E., Hoffman R. M., Lowe S. W. A senescence program controlled by p53 and p16INK4a contributes to the outcome of cancer therapy.
Cell
,
109
:
335
-346,  
2002
.
8
Rehemtulla A., Stegman L. D., Cardozo S. J., Gupta S., Hall D. E., Contag C. H., Ross B. D. Rapid and quantitative assessment of cancer treatment response using in vivo bioluminescence imaging.
Neoplasia
,
2
:
491
-495,  
2000
.
9
Adams J. Y., Johnson M., Sato M., Berger F., Gambhir S. S., Carey M., Iruela-Arispe M. L., Wu L. Visualization of advanced human prostate cancer lesions in living mice by a targeted gene transfer vector and optical imaging.
Nat. Med.
,
8
:
891
-897,  
2002
.
10
Vooijs M., Jonkers J., Lyons S., Berns A. Noninvasive imaging of spontaneous retinoblastoma pathway-dependent tumors in mice.
Cancer Res.
,
62
:
1862
-1867,  
2002
.
11
Hogan B., Beddington R., Costantini F., Lacy E. .
Manipulating the Mouse Embryo, A Laboratory Manual
, Ed. 2
219
-252, Cold Spring Harbor Laboratory Press Cold Spring Harbor, New York Section E 
1994
.
12
Meuwissen R., Linn S. C., van der Valk M., Mooi W. J., Berns A. Mouse model for lung tumorigenesis through Cre/lox controlled sporadic activation of the K-Ras oncogene.
Oncogene
,
20
:
6551
-6558,  
2001
.
13
Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain.
Nat. Genet.
,
21
:
70
-71,  
1999
.
14
Vooijs M., Jonkers J., Berns A. A highly efficient ligand-regulated Cre recombinase mouse line shows that LoxP recombination is position dependent.
EMBO Rep.
,
2
:
292
-297,  
2001
.