Use of the ODD-luciferase transgene for the non-invasive imaging of spontaneous tumors in mice

Abstract

Background: In humans, imaging of tumors provides rapid, accurate assessment of tumor growth and location. In laboratory animals, however, the imaging of spontaneously occurring tumors continues to pose many technical and logistical problems. Recently a mouse model was generated in which a chimeric protein consisting of HIF-1α oxygen-dependent degradation domain (ODD) fused to luciferase was ubiquitously expressed in all tissues. Hypoxic stress leads to the accumulation of ODD-luciferase in the tissues of this mouse model which can be identified by non-invasive bioluminescence measurement. Since solid tumors often contain hypoxic regions, we performed proof-of-principle experiments testing whether this transgenic mouse model may be used as a universal platform for non-invasive imaging analysis of spontaneous solid tumors.

Methods and materials: ODD-luciferase transgenic mice were bred with MMTV-neu/beclin1+/- mice. Upon injection of luciferin, bioluminescent background of normal tissues in the transgenic mice and bioluminescent signals from spontaneously mammary carcinomas were measured non-invasively with an IVIS Spectrum imaging station. Tumor volumes were measured manually and the histology of tumor tissues was analyzed.

Conclusion: Our results show that spontaneous mammary tumors in ODD-luciferase transgenic mice generate substantial bioluminescent signals, which are clearly discernable from background tissue luminescence. Moreover, we demonstrate a strong quantitative correlation between the bioluminescent tumor contour and the volume of palpable tumors. We further demonstrate that shrinkage of the volume of spontaneous tumors in response to chemotherapeutic treatment can be determined quantitatively using this system. Finally, we show that the growth and development of spontaneous tumors can be monitored longitudinally over several weeks. Thus, our results suggest that this model could potentially provide a practical, reliable, and cost-effective non-invasive quantitative method for imaging spontaneous solid tumors in mice.

Figure 1. The effect of hypoxia on the HIF1-α degradation pathway and the scheme of the use of the ODD-Luciferase transgene for bioluminescent imaging of spontaneous tumors.

(A) HIF prolyl hydroxylase depends upon the substrate oxygen for hydroxylation of the ODD domain in constitutively expressed HIF1-α protein in normoxic cells, ultimately leading to the degradation of HIF1-α via the ubiquitin-proteasome pathway. The expression of the ODD-Luciferase transgene is identically regulated, leading to the accumulation of the ODD-Luciferase protein under hypoxic conditions. (B) A schematic of the generation of mammary carcinoma-prone MMTV-neu/ODD-Luc/Beclin1 +/− mice.

Figure 2. Evaluation of baseline bioluminescent signals in MMTV-neu/ODD-Luc/Beclin1 +/+ or +/− mice.

(A) Representative dorsal, ventral, right lateral, and left lateral views of two female mice expressing the ODD-Luc and MMTV-neu transgenes. Each image is individually scaled for radiance according to scale bars shown. (B) The images in panel (A) are adjusted to conform to the same radiance scale, showing the relative intensity of signaling among all images.

Figure 3. Bioluminescent images of spontaneous tumors.

(A) Photographic and luminescent images of two female mice, one bearing a subcutaneous tumor above the right scapula (top panels) and another bearing a subcutaneous tumor on the right abdomen/inguinal region (bottom panels) with photographs of the dissected tumor masses. White circles denote the approximate boundaries of the tumor masses in this view. Note the underlying strong signal generated by the right kidney (arrow) intensified due to decreased subcutaneous and adipose tissue mass. (B) Longitudinal tracking of subcutaneous tumor development in a female mouse.

Figure 4. Histological analysis of mammary carcinomas and the expression of ODD-luciferase.

(A) Representative subcutaneous tumors on the left cervical and right abdominal aspects of a female MMTV-neu/ODD-Luc/Beclin1 +/− mouse. White circles denote approximate boundaries of tumor masses. (B) Hematoxylin and Eosin stained paraffin-embedded tumor tissue sections at 200× (top panels) and 1000× (bottom panels) magnification. White arrows denote the presence of mitotic cells in the viable peripheral tumor tissue. (C) Immunofluorescent analysis of two tumor tissue samples with anti-luciferase antibody. Left panels show immunofluorescent signaling representing the presence of luciferase. Center panels are DAPI stained representing the nuclei of viable cells. Right panels are merged fluorescein/DAPI images. NC: necrotic core (region without DAPI staining). Histograms illustrate the fluorescein signal intensity in relation to the distance from local vasculature.

Figure 5. Correlation of tumor bioluminescence signal and tumor volume.

(A) Manually calculated tumor area compared to computer-obtained tumor bioluminescent contour by view (left lateral, or LL, versus dorsal, or D) of subcutaneous abdominal tumor in a female MMTV-neu/ODD-Luc/Beclin1 +/− mouse over time. (B) Manually calculated tumor area compared to computer-obtained tumor bioluminescent contour by view (right lateral, or RL, versus dorsal, or D) of subcutaneous cervical tumor in the same female MMTV-neu/ODD-Luc/Beclin1 +/− mouse over time. (C) Computer-obtained radiance (photons/second/cm²/steradian) in both cervical and abdominal tumors over time, by view (right or left lateral and dorsal). (D). Statistical analyses of results from (A) and (B) to determine the correlation between the tumor size measured manually and by bioluminescent signal contour. Pearson coefficients and P values are shown.

Figure 6. Tracking the regression of spontaneous tumors during drug treatment.

(A) Photographic and bioluminescent images of a shrinking palpable tumor over a 14- day drug treatment as described in materials and methods. White arrow in photographic view indicates location of tumor. The red circles in bioluminescent view indicate region of interest. (B) Relative tumor volume expressed as a percent of initial tumor from manually measuring tumor volume (▪) vs. software calculated (ROI) tumor volume (◊) based on bioluminescence signal contour over the course of drug treatment as described in materials and methods. Pearson coefficient (ρ) and P value are shown.

Figure 7. Longitudinally tracking impalpable spontaneous tumors.

Left panels: Bioluminescent images of a subcutaneous tumor of the left cranial thoracic mammary gland developing in an agouti female MMTV-neu/ODD-Luc/Beclin1 +/+ mouse over an eight week period. Arrow indicates tumor signal at beginning and end of imaging period. Only this region persistently shows hypoxic signal throughout the imaging period (the hypoxic signals are expected near the thyroid and the extremities). Note the decrease in background signal observed in agouti mice. This is due to attenuation of the luciferase signal by the darker hair. Right panels: Photograph of external tumor appearance at day 67 of imaging and dissected tumor mass. Arrows denote tumor location.